Attachments B, C

PDF 10 - att B C.pdf

2015 National Survey on Drug Use and Health (NSDUH)

Attachments B, C

OMB: 0930-0110

⚠️ Notice: This form may be outdated. More recent filings and information on OMB 0930-0110 can be found here:

Document [pdf]

Download: pdf | pdf

2015 NSDUH, Supporting Statement
Attachment B – Questionnaire Field Test (QFT)
Final Report

NATIONAL SURVEY ON DRUG
USE AND HEALTH: 2012
QUESTIONNAIRE FIELD TEST
FINAL REPORT

Substance Abuse and Mental Health Services Administration
Center for Behavioral Health Statistics and Quality
Rockville, Maryland 20857
March 19, 2014

NATIONAL SURVEY ON DRUG
USE AND HEALTH: 2012
QUESTIONNAIRE FIELD TEST
FINAL REPORT
Deliverable 27: Field Test Protocol
Contract No. HHSS283201000003C
RTI Project No. 0212800.001.102.003.008.005
RTI Authors:
Doug Currivan
Patrick Chen
Teresa Davis
Becky Granger
Ilona Johnson
Larry Kroutil
Patty LeBaron
Gretchen McHenry

Project Director:
Allison McKamey
Katie Morton
Susan Myers
Victoria Scott
Bonnie Shook-Sa
Stephanie Terrey
Kevin Wang
Hilary Zelko

David Hunter
SAMHSA Project Officer:
Peter Tice

For questions about this report, please email [email protected].
Prepared for Substance Abuse and Mental Health Services Administration,
Rockville, Maryland
Prepared by RTI International, Research Triangle Park, North Carolina
March 19, 2014
Recommended Citation: Center for Behavioral Health Statistics and Quality.
(2014). National Survey on Drug Use and Health: 2012 Questionnaire Field
Test Final Report. Substance Abuse and Mental Health Services
Administration, Rockville, MD.

Acknowledgments
This report would not be possible without the guidance and input of staff from the Center for
Behavioral Health Statistics and Quality. In particular, Jonaki Bose and Dicy Painter provided
useful comments. At RTI International (a trade name of Research Triangle Institute), Debbie
Bond, Valerie Garner, and Richard Straw provided report production assistance, and Dave Heller
and Chris Stringer contributed to revisions and additions to the report.

Table of Contents
Chapter

Page

List of Tables .................................................................................................................................. ix
List of Exhibits............................................................................................................................ xxix

Background and Goals .........................................................................................................1

Study Design, Field Preparations, and Data Collection Procedures ....................................5
2.1
Overview of the Study Design, Field Preparations, and Data Collection ................5
2.2
Study Design ............................................................................................................5
2.2.1 Target Population .........................................................................................5
2.2.2 Selection of State Sampling Regions and Segments....................................5
2.2.3 Selection of Dwelling Units .........................................................................7
2.2.4 Age Group Allocations ................................................................................8
2.2.5 Selection of Persons .....................................................................................8
2.3
Field Preparations ....................................................................................................9
2.3.1 Preparing Field Equipment ..........................................................................9
2.3.2 Staffing .......................................................................................................12
2.3.3 Training Procedures ...................................................................................13
2.4
Data Collection Procedures....................................................................................15
2.4.1 Questionnaire and Protocol Changes for the 2012 QFT ............................15
2.4.2 Contacting Dwelling Units ........................................................................19
2.4.3 Dwelling Unit Screening............................................................................20
2.4.4 Interview Administration ...........................................................................21
2.4.5 Controlled Access Procedures ...................................................................22
2.4.6 Refusal Conversion Procedures .................................................................23
2.4.7 Data Collection Management and Quality Control ...................................23
2.4.8 Problems Encountered ...............................................................................25

Processing and Analysis of the 2012 Questionnaire Field Test Data and 2011 and
2012 Comparison Data ......................................................................................................29
3.1
Overview of Data Processing and Analysis Approach ..........................................29
3.2
Defining Usable Cases ...........................................................................................29
3.2.1 Overview of Defining Usable Cases ..........................................................29
3.2.2 Usable Case Definitions .............................................................................29
3.3
Editing and Coding Procedures .............................................................................30
3.3.1 Overview of Editing and Coding Procedures ............................................30
3.3.2 Coding of "OTHER, Specify" Data ...........................................................31
3.3.3 General Editing Principles .........................................................................31
3.3.4 Special Editing Situations ..........................................................................33
3.4
Imputation Procedures ...........................................................................................41
3.4.1 Overview of Imputation Procedures ..........................................................41
3.4.2 Imputation Methodology ...........................................................................42
3.5
Weighting Procedures ............................................................................................45
3.5.1 Overview of Weighting Procedures ...........................................................45
3.5.2 Weighting Procedures ................................................................................46
iii

Table of Contents (continued)
Chapter

3.6

3.7

Page
3.5.3 Distribution of QFT Analysis Weights ......................................................52
3.5.4 Creation of Variance Estimation Strata and Replicates .............................52
Data File Preparation .............................................................................................53
3.6.1 QFT Data File ............................................................................................53
3.6.2 2011 Comparison Data File .......................................................................53
3.6.3 2012 Comparison Data File .......................................................................54
Data Analysis Issues ..............................................................................................54
3.7.1 Primary Analytic Goals..............................................................................54
3.7.2 Comparison with Current NSDUH Data ...................................................55
3.7.3 Comparisons with Other Survey Data .......................................................56

Data Collection Outcomes and Data Quality Assessment .................................................59
4.1
Overview of Data Collection and Data Quality Outcomes ....................................59
4.2
Unit Response Rates ..............................................................................................59
4.2.1 Screening Response Rates (SRRs) and Number of Visits for
Completed and Noncompleted Screenings ................................................59
4.2.2 Interview Response Rates (IRRs) and Number of Visits for
Completed and Noncompleted Screenings ................................................61
4.3
Imputation Rates for Common 2011 Comparison Data, 2012 Quarters 3
and 4 Comparison Data, and QFT Variables .........................................................65
4.4
Missing Data Rates for New or Revised QFT Items and Comparisons of
Missing Data Rates for Moved QFT Items with 2011 and 2012 Quarters 3
and 4 Comparison Data ..........................................................................................71
4.4.1 Missing Data Rates for New, Revised, or Moved Items in the QFT
Questionnaire .............................................................................................71
4.4.2 Missing Data Rates for Items Moved in the QFT Questionnaire for
the QFT Data, 2011 Comparison Data, and 2012 Quarters 3 and 4
Comparison Data .......................................................................................71
4.5
Interview Timing Results.......................................................................................84
4.5.1 Central Tendency Statistics for Overall and Module Timing
Results for the 2011 and 2012 Quarters 3 and 4 Comparison Data
and the 2012 QFT Data ..............................................................................84
4.5.2 Selected Detailed Interview Timing Data for the 2012
Questionnaire Field Test and the 2011 and 2012 Quarter 3 and 4
Comparison Data .....................................................................................125
4.5.3 Timing Data for High and Low Reports of Numbers of
Prescription Drugs Used or Misused in the Past Year in the QFT
Sample......................................................................................................149
4.6
Other Data Quality Indicators ..............................................................................168
4.6.1 Overview of Other Data Quality Indicators .............................................168
4.6.2 Triggering of Inconsistency Flags in Core Drug Use Data......................169
4.6.3 Responding to Lead Questions for "OTHER, Specify" Data ..................170

Table of Contents (continued)
Chapter

Page
4.6.4 Triggering of Hard Errors Involving Ages at First Prescription
Drug Misuse .............................................................................................172
4.6.5 Triggering of Specific Consistency Checks in the Prescription
Drug Modules ..........................................................................................173
4.6.6 Patterned Responses in the Core Drug Questions for the
Comparison Data .....................................................................................174
4.6.7 Patterned Responses in the Drug Use Questions for the QFT Data ........174
4.6.8 Issues to Consider for the Dress Rehearsal ..............................................178

Assessments of the Redesigned Protocol .........................................................................181
5.1
Overview of QFT Protocol Assessment ..............................................................181
5.2
Summary of Results from Field Observations of QFT Field Interviewers ..........181
5.3
QFT Field Interviewer Debriefing Results ..........................................................185
5.4
QFT Equipment Surveys......................................................................................198
5.4.1 Purpose and Development of the Equipment Surveys .............................198
5.4.2 Procedures for Conducting the Equipment Surveys ................................198
5.4.3 Summary and Discussion of Results from the Equipment Surveys ........199
5.4.4 FI Comments on the Tablet, Screening Program, and Tablet
Accessories ..............................................................................................201
5.4.5 FI Feedback on the QFT Handbook.........................................................203
5.5
Focus Groups with QFT Field Interviewers ........................................................203
5.5.1 Purpose of the Focus Groups ...................................................................203
5.5.2 Sites and Participants ...............................................................................204
5.5.3 Focus Group Protocol and Procedures.....................................................204
5.5.4 Focus Group Results by Topic .................................................................205

QFT Estimates Compared with NSDUH Estimates: Substance Use Items Other
than Methamphetamine and Prescription Drugs ..............................................................211
6.1
Overview of QFT Estimates Compared with NSDUH Estimates for
Substance Use Items Other than Methamphetamine and Prescription
Drugs ....................................................................................................................211
6.2
Marijuana, Cocaine, and Heroin ..........................................................................211
6.3
Hallucinogens and Inhalants ................................................................................212
6.3.1 Hallucinogens ..........................................................................................212
6.3.2 Inhalants ...................................................................................................215
6.4
Illicit Drug Summary Measures ...........................................................................217
6.4.1 Any Illicit Drug ........................................................................................219
6.4.2 Illicit Drugs Other than Marijuana ...........................................................219
6.5
Tobacco ................................................................................................................221
6.5.1 Cigarettes .................................................................................................222
6.5.2 Smokeless Tobacco..................................................................................222
6.6
Alcohol .................................................................................................................224
6.6.1 Any Alcohol Use......................................................................................224

Table of Contents (continued)
Chapter

Page
6.6.2 Past Month Binge Alcohol Use................................................................225

QFT Estimates Compared with Current NSDUH Estimates: Methamphetamine
and Prescription Drug Items ............................................................................................227
7.1
Overview of QFT Estimates Compared with NSDUH Estimates for
Methamphetamine and Prescription Drug Items .................................................227
7.2
Estimates for Methamphetamine Items ...............................................................228
7.3
Estimates for Prescription Drug Items .................................................................229
7.3.1 Any Prescription Psychotherapeutic Drug ...............................................230
7.3.2 Pain Relievers ..........................................................................................231
7.3.3 Tranquilizers ............................................................................................232
7.3.4 Sedatives ..................................................................................................232
7.3.5 Stimulants ................................................................................................233
7.4
Effects of Methamphetamine and Prescription Drugs on Illicit Drug Use
Estimates ..............................................................................................................236
7.5
Methamphetamine, Prescription Drug, and Illicit Drug Estimation Issues
to Consider for the 2013 Dress Rehearsal and 2015 Redesign ............................237
7.5.1 Methamphetamine....................................................................................237
7.5.2 Prescription Drugs ...................................................................................239
7.5.3 Illicit Drugs ..............................................................................................240

QFT Estimates Compared with NSDUH Estimates: Noncore Items...............................241
8.1
Overview of QFT Estimates Compared with NSDUH Estimates for
Noncore Items ......................................................................................................241
8.2
Estimates for Substance Dependence and Abuse ................................................241
8.3
Estimates for Needle Use Items ...........................................................................245
8.4
Comparisons of Medical Marijuana Reports by State in Reference to
Current State Laws ...............................................................................................247
8.5
Estimates for Noncore Demographic and Household Items ................................248
8.6
Estimates for Selected Items Potentially Subject to Context Effects Due to
Questionnaire Redesign .......................................................................................250
8.7
Estimates for New, Revised, and Moved Items in the QFT Instrument ..............252
8.8
Comparison of the Distribution of Relationships for Proxy Respondents
and Estimates for Selected Items Based on Proxy Report Status ........................254

Selected QFT Estimates Compared with Other Survey Estimates ..................................257
9.1
Overview of Selected QFT Estimates Compared with Other Survey Data .........257
9.2
Estimates for Prescription Drug Misuse ..............................................................257
9.2.1 NAMCS and NHAMCS ..........................................................................258
9.2.2 Prescription Drug Use and Misuse in the QFT and Prescription
Drug Mentions in NAMCS and NHAMCS .............................................259
9.2.3 Monitoring the Future ..............................................................................271
9.2.4 Prescription Drug Misuse in the QFT and Monitoring the Future...........272
9.3
Estimates for Selected Health and Demographic Items .......................................274
vi

Table of Contents (continued)
Chapter
9.4

10.

Page
Estimates for Additional Demographic and Household Items ............................277
9.4.1 Received Income and Participation in Government Assistance
Programs ..................................................................................................278
9.4.2 Health Insurance Coverage ......................................................................279
9.4.3 Family Income .........................................................................................280
9.4.4 Employment Status and Unemployment Rates........................................281
9.4.5 Education .................................................................................................282

Summary and Implications ..............................................................................................285
10.1 Data Collection Outcomes and Data Quality Assessment ...................................285
10.2 Assessments of the Redesigned Protocol .............................................................287
10.3 QFT Estimates Compared with NSDUH Estimates: Substance Use Items
Other than Methamphetamine and Prescription Drugs ........................................289
10.4 QFT Estimates Compared with NSDUH Estimates: Methamphetamine
and Prescription Drug Items ................................................................................290
10.5 QFT Estimates Compared with NSDUH Estimates: Noncore Items...................291
10.5.1 Substance Dependence and Abuse...........................................................291
10.5.2 Needle Use ...............................................................................................292
10.5.3 Medical Marijuana ...................................................................................292
10.5.4 Demographic and Household Items .........................................................293
10.5.5 Selected Items Potentially Subject to Context Effects .............................293
10.6 Selected QFT Estimates Compared with Other Survey Estimates ......................294
10.7 Summary of QFT Questionnaire Items Identified as Needing
Reexamination in the DR Analysis ......................................................................296

References ....................................................................................................................................299

vii

List of Appendices
Appendices

Page

Redesigned NSDUH Questionnaire and Redesigned Contact Materials for the
2015 Partial Redesign ..................................................................................................... A-1

Questionnaire Field Test Screening and Interview Response Rates, by Sample
Release and Age Group and for Each State .....................................................................B-1

Missing Data Rates for New, Moved, or Revised Items in the 2012 Questionnaire
Field Instrument ...............................................................................................................C-1

QFT Field Observation Materials – Screening Checklist, QFT Field Observation
Interview Checklist, and Field Observer Reference Sheet ............................................. D-1

QFT Field Interviewer Debriefing Questions .................................................................. E-1

Complete Results from the QFT New Equipment User Satisfaction Survey .................. F-1

Moderator's Guide for QFT Focus Groups with Field Interviewers ............................... G-1

Selected Notes on Analysis Variables for the QFT ........................................................ H-1

Detailed Tables for Core Substance Use Items Other than Methamphetamine and
Prescription Drugs in the 2011 and 2012 Comparison Data and the QFT ....................... I-1

Detailed Tables for Methamphetamine and Prescription Drug Items in the 2011
and 2012 Comparison Data and the QFT ......................................................................... J-1

Detailed Tables for Noncore Estimates in the 2011 and 2012 Comparison Data
and the QFT .................................................................................................................... K-1

Detailed Tables for Prescription Drug Use and Misuse in the 2012 Questionnaire
Field Test and Data from Sources Other than NSDUH ................................................... L-1

Estimates for New Items in the 2012 Questionnaire Field Test That Were
Included in the 2013 NSDUH Main Study Questionnaire..............................................M-1

Moved Demographic and Household Items in the 2012 Questionnaire Field Test:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data ........................................................................................ N-1

Estimates and Standard Errors for New, Moved, or Revised Items in the 2012
Questionnaire Field Test among Persons Aged 12 or Older........................................... O-1

Proxy Reports from the QFT and the Comparison Samples............................................ P-1

Protocol Changes Considered for the Dress Rehearsal and Whether the Changes
Will Be Implemented for the Dress Rehearsal ............................................................... Q-1

2012 Questionnaire Field Test—Investigation of Data Quality Issues for Items
Moved from CAPI to ACASI ..........................................................................................R-1

viii

List of Tables
Table

Page

2.1

Number of 2012 Questionnaire Field Test State Sampling Regions and Sample
Sizes, by State ......................................................................................................................6

2.2

Summary of the 2012 Questionnaire Field Test Sample Results ........................................8

2.3

Tenure Distribution of 2012 Quarters 3 to 4 Main Study Field Interviewers
Compared with 2012 Questionnaire Field Test Field Interviewers ...................................13

2.4

Questionnaire Field Test Field Interviewer Training Program ..........................................14

2.5

Questionnaire Field Test Items with Programming Logic Errors ......................................26

3.1

Imputed Variables ..............................................................................................................43

3.2

Weight Distribution of QFT Analysis Weights .................................................................52

3.3

Data Files Created for the 2012 Questionnaire Field Test Analyses .................................54

4.1

Screenings, Interviews, and Response Rates for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test ....................................60

4.2

Number of Visits Made for Completed Screenings for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test ....................................62

4.3

Number of Visits Made for Noncompleted Screenings for the 2011 Main Study,
2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test ...........................62

4.4

Interview Response Rates, by Age, for the 2011 Main Study, 2012 Quarters 3 and
4 Main Study, and 2012 Questionnaire Field Test (QFT) .................................................63

4.5

Number of Visits Made for Completed Interviews for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test ....................................64

4.6

Number of Visits Made for Noncompleted Interviews for the 2011 Main Study,
2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test ...........................64

4.7a

Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3
and 4 Main Study, and 2012 Questionnaire Field Test: Substance Use Variables ............66

4.7b

Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3
and 4 Main Study, and 2012 Questionnaire Field Test: Selected Demographic and
Socioeconomic Variables...................................................................................................67

4.7c

Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3
and 4 Main Study, and 2012 Questionnaire Field Test: Health Insurance Variables ........68

4.7d

Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3
and 4 Main Study, and 2012 Questionnaire Field Test: Income Variables .......................69

4.8

Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test,
2011 Comparison, 2012 Comparison, and Questionnaire Field Test Data........................74

List of Tables (continued)
Table

Page

4.9a

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 12 or Older) .................................................................................87

4.9b

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 12 to 17).......................................................................................90

4.9c

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 18 to 25).......................................................................................93

4.9d

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 26 to 49).......................................................................................96

4.9e

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 50 to 64).......................................................................................99

4.9f

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes
(All Respondents Aged 65+) ...........................................................................................102

4.9g

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 12 or Older) ..............................................................................107

4.9h

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 12 to 17) ...................................................................................110

4.9i

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 18 to 25) ...................................................................................113

4.9j

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 26 to 49) ...................................................................................116

4.9k

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 50 to 64) ...................................................................................119

4.9l

Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4
2012 Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative
Gate Respondents Aged 65+) ..........................................................................................122

List of Tables (continued)
Table

Page

4.10a Unweighted Overall Interview Timing Data for the Questionnaire Field Test
Protocol in Minutes, in Total and by Age Groups: All QFT Respondents ......................127
4.10b Unweighted Overall Interview Timing Data for the 2011 Comparison Protocol in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents ...................128
4.10c Unweighted Overall Interview Timing Data for the 2012 Comparison Protocol in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents ...................129
4.10d Unweighted Overall Interview Timing Data for the QFT Tobacco Module in
Minutes, in Total and by Age Groups: All QFT Respondents Answering
LEADCIG ........................................................................................................................130
4.10e Unweighted Overall Interview Timing Data for the 2011 Tobacco Module in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents
Answering LEADCIG .....................................................................................................131
4.10f Unweighted Overall Interview Timing Data for the 2012 Tobacco Module in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents ...................132
4.10g Unweighted Overall Interview Timing Data for the QFT Pain Relievers Screener
in Minutes, in Total and by Age Groups: All Respondents .............................................133
4.10h Unweighted Overall Interview Timing Data for the QFT Tranquilizer Screener in
Minutes, in Total and by Age Groups: All Respondents .................................................134
4.10i Unweighted Overall Interview Timing Data for the QFT Stimulant Screener in
Minutes, in Total and by Age Groups: All Respondents .................................................135
4.10j Unweighted Overall Interview Timing Data for the QFT Sedative Screener in
Minutes, in Total and by Age Groups: All Respondents .................................................136
4.10k Unweighted Overall Interview Timing Data for the Pain Reliever Module in
Minutes, in Total and by Age Groups: All QFT Respondents ........................................137
4.10l Unweighted Overall Interview Timing Data for the Pain Reliever Module in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents ...................138
4.10m Unweighted Overall Interview Timing Data for the Pain Reliever Module in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents ...................139
4.10n Unweighted Overall Interview Timing Data for the Tranquilizer Module in
Minutes, in Total and by Age Groups: All QFT Respondents ........................................140
4.10o Unweighted Overall Interview Timing Data for the Tranquilizer Module in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents ...................141
4.10p Unweighted Overall Interview Timing Data for the Tranquilizer Module in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents ...................142

List of Tables (continued)
Table

Page

4.10q Unweighted Overall Interview Timing Data for the Stimulants Module in
Minutes, in Total and by Age Groups: All QFT Respondents ........................................143
4.10r Unweighted Overall Interview Timing Data for the Stimulants Module in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents ...................144
4.10s Unweighted Overall Interview Timing Data for the Stimulants Module in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents ...................145
4.10t Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes,
in Total and by Age Groups: All QFT Respondents........................................................146
4.10u Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes,
in Total and by Age Groups: All 2011 Comparison Respondents...................................147
4.10v Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes,
in Total and by Age Groups: All 2012 Comparison Respondents...................................148
4.11a Overall Interview Timing Data for the QFT Pain Relievers Screener in Minutes,
in Total and by Age Groups for Respondents Reporting Extreme High Numbers
of Prescription Pain Relievers Used in the Past Year ......................................................153
4.11b Overall Interview Timing Data for the QFT Pain Relievers Screener in Minutes,
in Total and by Age Groups for Respondents Reporting Lower Numbers of
Prescription Pain Relievers Used in the Past Year ..........................................................154
4.11c Overall Interview Timing Data for the QFT Pain Relievers Screener and Main
Module in Minutes, in Total and by Age Groups for Respondents Reporting
Extreme High Numbers of Prescription Pain Relievers Misused in the Past Year .........155
4.11d Overall Interview Timing Data for the QFT Pain Relievers Screener and Main
Module in Minutes, in Total and by Age Groups for Respondents Reporting
Lower Numbers of Prescription Pain Relievers Misused in the Past Year......................156
4.11e Overall Interview Timing Data for All QFT Prescription Drug Screeners in
Minutes, in Total and by Age Groups for Respondents Reporting Extreme High
Numbers of Prescription Drugs Used in the Past Year ....................................................157
4.11f Overall Interview Timing Data for All QFT Prescription Drug Screeners in
Minutes, in Total and by Age Groups for Respondents Reporting Lower Numbers
of Prescription Drugs Used in the Past Year ...................................................................158
4.11g Overall Interview Timing Data for All QFT Prescription Drug Screeners and
Main Modules in Minutes, in Total and by Age Groups for Respondents
Reporting Extreme High Numbers of Prescription Drugs Misused in the Past Year ......159
4.11h Overall Interview Timing Data for All QFT Prescription Drug Screeners and
Main Modules in Minutes, in Total and by Age Groups for Respondents
Reporting Lower Numbers of Prescription Drugs Misused in the Past Year ..................160

xii

List of Tables (continued)
Table

Page

4.11i Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Pain Relievers Used in the Past Year ..........................................................161
4.11j Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Pain Relievers Used in the Past Year ...............................................................................162
4.11k Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Pain Relievers Misused in the Past Year .....................................................163
4.11l Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Pain Relievers Misused in the Past Year .........................................................................164
4.11m Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Drugs Used in the Past Year ........................................................................165
4.11n Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Drugs Used in the Past Year ............................................................................................166
4.11o Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Drugs Misused in the Past Year ..................................................................167
4.11p Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total
and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Drugs Misused in the Past Year .......................................................................................168
5.1

Screening Errors Specific to the Questionnaire Field Test ..............................................182

5.2

Interview Errors Specific to the Questionnaire Field Test...............................................182

5.3

Screening Respondent Recall of Lead Letter, by Screening Respondent Age ................186

5.4

Screening Respondent Recall of Lead Letter, by Dwelling Unit Interview Status .........186

5.5

Recall of Lead Letter among Screening Respondents Aged 18 to 25, by Dwelling
Unit Interview Status .......................................................................................................187

5.6

Recall of Lead Letter among Screening Respondents Aged 26 to 49, by Dwelling
Unit Interview Status .......................................................................................................187

5.7

Recall of Lead Letter among Screening Respondents Aged 50 to 64, by Dwelling
Unit Interview Status .......................................................................................................187

xiii

List of Tables (continued)
Table

Page

5.8

Recall of Lead Letter among Screening Respondents Aged 65 or Older, by
Dwelling Unit Interview Status .......................................................................................188

5.9

Interview Status, by Recall of Lead Letter and Screening Respondent Age ...................188

5.10

Screening Respondent Comments on Lead Letter, by Screening Respondent Age ........189

5.11

Screening Respondent Comments on Lead Letter, by Dwelling Unit Interview
Status ................................................................................................................................189

5.12

Timing of Providing Q&A Brochure ...............................................................................190

5.13

Comments on Q&A Brochure .........................................................................................190

5.14

Comments on Q&A Brochure, by Timing of Providing Brochure ..................................190

5.15

Respondent Comments on the Interview Being Too Long ..............................................190

5.16

Respondent Comments on the Interview Being Too Long, by Interview
Respondent Age ...............................................................................................................191

5.17

Respondent Comments on the Interview Being Too Long, by Interview
Respondent Education .....................................................................................................191

5.18

Classification of Open-Ended Comments on Prescription Drug Questions ....................191

5.19

Interview Respondent Questions or Comments on Prescription Drug Questions ...........192

5.20

Interview Respondent Questions or Comments on Prescription Drug Questions,
by Interview Respondent Age ..........................................................................................192

5.21

Interview Respondent Questions or Comments on Prescription Drug Questions,
by Interview Respondent Education ................................................................................192

5.22

Any Interview Respondent Questions or Comments on On-Screen Calendars ...............193

5.23

Any Interview Respondent Questions or Comments on On-Screen Calendars, by
Interview Respondent Age ...............................................................................................193

5.24

Any Interview Respondent Questions or Comments on On-Screen Calendars, by
Interview Respondent Education .....................................................................................193

5.25

Types of Interview Respondent Questions or Comments on On-Screen Calendars .......193

5.26

Interview Respondent Troubles with Other Questions ....................................................194

5.27

Proxy Used for Income and Health Insurance Questions ................................................194

5.28

Proxy Used for Income and Health Insurance Questions, by Interview Respondent
Age ...................................................................................................................................194

5.29

Interview Respondent Concerns about Revealing Answers to Proxy Respondent..........195

5.30

Interview Respondent Questions or Comments about Proxy Interview ..........................195

xiv

List of Tables (continued)
Table

Page

5.31

Problems with Proxy on ACASI Tutorial ........................................................................195

5.32

Types of Problems with Proxy on ACASI Tutorial .........................................................195

5.33

Problems with Proxy Use of ACASI to Answer Income and Health Insurance
Questions..........................................................................................................................195

5.34

Types of Problems with Proxy Use of ACASI to Answer Income and Health
Insurance Questions .........................................................................................................196

5.35

Interviews Conducted at Respondent's Home for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test (QFT) .......................196

5.36

Interview Location Not at Respondent's Home for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test (QFT) .......................196

5.37

Field Interviewer (FI) Evaluation of Interview Privacy in Respondent's Home for
the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire
Field Test (QFT) ..............................................................................................................197

5.38

Field Interviewer (FI) Reports of Others Present during Interview for the 2011
Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field
Test (QFT) .......................................................................................................................197

5.39

Field Interviewer Opinions on Use of the Tablet before Questionnaire Field Test
(QFT) Data Collection and after QFT Data Collection ...................................................200

5.40

Field Interviewer (FI) Expectations on Referencing the Questionnaire Field Test
(QFT) Handbook before QFT Data Collection and FI Need to Reference the QFT
Handbook after QFT Data Collection ..............................................................................200

5.41

Sites and Number of Participants for QFT Focus Groups ...............................................204

5.42

QFT FIs' "Wish List" for Modifications to Tablet Functions ..........................................207

6.1

Substances Included in Definitions of Illicit Drugs and Illicit Drugs Other than
Marijuana .........................................................................................................................218

8.1

Current State of Residence without a Medical Marijuana Law in Effect and
Current or Former Bordering States with Medical Marijuana Laws in Effect for
Eight QFT Respondents Reporting Medical Use of Marijuana .......................................247

9.1

Comparison of Summary Data for Pain Relievers from the 2012 NSDUH
Questionnaire Field Test and the 2010 National Ambulatory Medical Survey and
the 2010 National Hospital Ambulatory Medical Care Survey .......................................266

9.2

Comparison of Summary Data for Tranquilizers and Sedatives from the 2012
NSDUH Questionnaire Field Test and the 2010 National Ambulatory Medical
Survey and the 2010 National Hospital Ambulatory Medical Care Survey ....................267

List of Tables (continued)
Table

Page

9.3

Comparison of Summary Data for Stimulants from the 2012 NSDUH
Questionnaire Field Test and the 2010 National Ambulatory Medical Survey and
the 2010 National Hospital Ambulatory Medical Care Survey .......................................269

10.1

Questionnaire Items Identified from the QFT Analysis as Needing Reexamination
in the DR Analysis ...........................................................................................................297

B-1

2012 Questionnaire Field Test Weighted Screening and Interview Response
Rates, by Sample Release and Age Group.......................................................................B-1

B-2

2012 Questionnaire Field Test Unweighted Screening and Interview Response
Rates, by Sample Release and Age Group.......................................................................B-2

B-3

2012 Questionnaire Field Test Weighted Screening and Interview Response
Rates, by State ..................................................................................................................B-3

B-4

2012 Questionnaire Field Test Unweighted Screening and Interview Response
Rates, by State ..................................................................................................................B-4

C-1

Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire
Field Test among Persons Aged 12 or Older ...................................................................C-1

I-1

Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime
among Persons Aged 12 or Older: Percentages, Differences, and Standard Error
of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test .......................................................................................................................... I-1

I-2

Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime
among Persons Aged 12 to 17: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... I-2

I-3

Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime
among Persons Aged 18 to 25: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... I-3

I-4

Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime
among Persons Aged 26 or Older: Percentages, Differences, and Standard Error
of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test .......................................................................................................................... I-4

I-5

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 12 or Older: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................... I-5

xvi

List of Tables (continued)
Table

Page

I-6

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 12 to 17: Percentages, Differences, and Standard Error
of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test .......................................................................................................................... I-6

I-7

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 18 to 25: Percentages, Differences, and Standard Error
of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test .......................................................................................................................... I-7

I-8

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 26 or Older: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................... I-8

I-9

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 or Older: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................... I-9

I-10

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 to 17: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. I-10

I-11

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 18 to 25: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. I-11

I-12

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 26 or Older: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. I-12

I-13

Specific Hallucinogen Use in Lifetime, by Age Group: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ........................................................................................ I-13

I-14

Specific Inhalant Use in Lifetime, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. I-14

I-15

Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................. I-15

xvii

List of Tables (continued)
Table

Page

I-16

Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age
Group and Gender: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test .................... I-16

I-17

Lifetime Use of Felt-Tip Pens, Computer Cleaners, or Other Inhalants, by Age
Group and Past Year Use of Inhalants according to Types of Inhalants Used in
Lifetime among Persons Aged 12 or Older: Percentages, 2012 Questionnaire
Field Test ........................................................................................................................ I-17

I-18 Use of Hallucinogens in Lifetime among Persons Aged 12 or Older with or
without Noncore Hallucinogen Data, by Age Group: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ........................................................................................ I-18
I-19 Use of Hallucinogens in the Past Year among Persons Aged 12 or Older with or
without Noncore Hallucinogen Data, by Age Group: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ........................................................................................ I-19
I-20 Use of Hallucinogens in the Past Month among Persons Aged 12 or Older with or
without Noncore Hallucinogen Data, by Age Group: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ........................................................................................ I-20
J-1

Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons
Aged 12 or Older: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ...................... J-1

J-2

Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons
Aged 12 to 17: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................... J-2

J-3

Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons
Aged 18 to 25: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................... J-3

J-4

Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons
Aged 26 or Older: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ...................... J-4

J-5

Misuse of Prescription Drugs or Methamphetamine in the Past Year among
Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... J-5

xviii

List of Tables (continued)
Table

Page

J-6

Misuse of Prescription Drugs or Methamphetamine in the Past Year among
Persons Aged 12 to 17: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... J-6

J-7

Misuse of Prescription Drugs or Methamphetamine in the Past Year among
Persons Aged 18 to 25: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... J-7

J-8

Misuse of Prescription Drugs or Methamphetamine in the Past Year among
Persons Aged 26 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... J-8

J-9

Misuse of Prescription Drugs or Methamphetamine in the Past Month among
Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................... J-9

J-10

Misuse of Prescription Drugs or Methamphetamine in the Past Month among
Persons Aged 12 to 17: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. J-10

J-11

Misuse of Prescription Drugs or Methamphetamine in the Past Month among
Persons Aged 18 to 25: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. J-11

J-12

Misuse of Prescription Drugs or Methamphetamine in the Past Month among
Persons Aged 26 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. J-12

J-13 Misuse of Stimulants in Lifetime among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-13
J-14 Misuse of Stimulants in the Past Year among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-14

xix

List of Tables (continued)
Table

Page

J-15 Misuse of Stimulants in the Past Month among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-15
J-16 Misuse of Sedatives in Lifetime among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-16
J-17

Misuse of Sedatives in the Past Year among Persons Aged 12 or Older with or
without Noncore Ambien® Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-17

J-18 Misuse of Sedatives in the Past Month among Persons Aged 12 or Older with or
without Noncore Ambien® Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test ................................................................................................. J-18
K-1

Substance Dependence or Abuse in the Past Year among Persons Aged 12 or
Older, by Survey Protocol: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. K-1

K-2

Substance Dependence or Abuse in the Past Year among Persons Aged 12 to 17,
by Survey Protocol: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test .................... K-2

K-3

Substance Dependence or Abuse in the Past Year among Persons Aged 18 to 25,
by Survey Protocol: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test .................... K-3

K-4

Substance Dependence or Abuse in the Past Year among Persons Aged 26 or
Older, by Survey Protocol: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. K-4

K-5

Substance Use with a Needle in Lifetime, Past Year, and Past Month among
Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test .................................................................................................................................. K-5

K-6

Demographic, Socioeconomic, and Household Characteristics among Persons
Aged 12 or Older: Percentages, Chi-Square Test Statistic, and P Value, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................. K-6

List of Tables (continued)
Table

Page

K-7

Demographic, Socioeconomic, and Household Characteristics among Persons
Aged 12 to 17: Percentages, Chi-Square Test Statistic, and P Value, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................. K-7

K-8

Demographic, Socioeconomic, and Household Characteristics among Persons
Aged 18 to 25: Percentages, Chi-Square Test Statistic, and P Value, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................. K-8

K-9

Demographic, Socioeconomic, and Household Characteristics among Persons
Aged 26 or Older: Percentages, Chi-Square Test Statistic, and P Value, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ............................. K-9

K-10 Demographic and Geographic Characteristics among Persons Aged 12 or Older:
Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test .......................................................... K-10
K-11 Geographic Characteristics among Persons Aged 12 to 17: Percentages, ChiSquare Test Statistic, and P Value, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ...................................................................................... K-11
K-12 Demographic and Geographic Characteristics among Persons Aged 18 to 25:
Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test .......................................................... K-12
K-13 Demographic and Geographic Characteristics among Persons Aged 26 or Older:
Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test .......................................................... K-13
K-14 Perceived Great Risk of Harm Associated with Substance Use among Persons
Aged 12 or Older: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test .................. K-14
K-15 Number of Years Since Last Use for Selected Substances among Lifetime Users
Aged 12 to 49: Averages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ........................... K-15
K-16 Received Substance Use Treatment in Lifetime and Past Year and Types of Past
Year Substance Use Treatment among Persons Aged 12 or Older: Percentages,
Differences, and Standard Error of Differences, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test .......................................................... K-16
K-17 Adult Mental Health Treatment in the Past Year and Type of Facility Where
Received Treatment among Persons Aged 18 or Older: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and
2012 Questionnaire Field Test ...................................................................................... K-17

xxi

List of Tables (continued)
Table

Page

K-18 Youth Mental Health Treatment in the Past Year and Number of Nights Received
Treatment among Persons Aged 12 to 17: Percentages, Chi-Square Test Statistic,
and P Value, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test ................................................................................................................................ K-18
K-19 Selected Mental Health Measures among Persons Aged 18 or Older: Percentages,
Differences, and Standard Error of Differences, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test .......................................................... K-19
K-20 Adolescent Depression Characteristics among Persons Aged 12 to 17:
Percentages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test ................................................. K-20
K-21 Arrested and Booked in Lifetime and Past Year for Breaking the Law among
Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field
Test ................................................................................................................................ K-21
L-1

Comparison of Data for Pain Relievers from the 2012 NSDUH Questionnaire
Field Test and the 2010 National Ambulatory Medical Survey and 2010 National
Hospital Ambulatory Medical Survey ............................................................................. L-1

L-2

Comparison of Data for Tranquilizers and Sedatives from the 2012 NSDUH
Questionnaire Field Test and the 2010 National Ambulatory Medical Survey and
2010 National Hospital Ambulatory Medical Survey ..................................................... L-4

L-3

Comparison of Data for Stimulants from the 2012 NSDUH Questionnaire Field
Test and the 2010 National Ambulatory Medical Survey and 2010 National
Hospital Ambulatory Medical Survey ............................................................................. L-7

L-4

NSDUH Questionnaire Field Test and Monitoring the Future Comparisons for
Past Year Misuse among Adolescents ............................................................................. L-9

L-5

NSDUH Questionnaire Field Test and Monitoring the Future Comparisons for
Past Year Misuse among Young Adults Aged 19 to 24 ................................................ L-10

L-6

Selected Characteristics among Persons Aged 12 or Older: Percentages and
Standard Errors, 2012 Questionnaire Field Test and 2011 National Health
Interview Survey ............................................................................................................ L-11

L-7

2011 NHIS and 2009-2010 NHANES Height Statistics among Persons Aged 16
or Older for Comparison with the 2012 Questionnaire Field Test ................................ L-13

L-8

2011 NHIS and 2009-2010 NHANES Weight Statistics among Persons Aged 16
or Older for Comparison with the 2012 Questionnaire Field Test ................................ L-14

L-9

Received Income and Program Participation among Persons Aged 12 or Older:
Percentages and Totals for 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and Other Surveys ................................................................ L-15
xxii

List of Tables (continued)
Table

Page

L-10 Received Income and Program Participation among Persons Aged 12 to 17:
Percentages and Totals for 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and Other Surveys ................................................................ L-16
L-11 Received Income and Program Participation among Persons Aged 18 to 25:
Percentages and Totals for 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and Other Surveys ................................................................ L-17
L-12 Received Income and Program Participation among Persons Aged 26 or Older:
Percentages and Totals for 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and Other Surveys ................................................................ L-18
L-13 Health Insurance Coverage among Persons Aged 12 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ................................................................................... L-19
L-14 Health Insurance Coverage among Persons Aged 12 to 17: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ................................................................................... L-20
L-15 Health Insurance Coverage among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ................................................................................... L-21
L-16 Health Insurance Coverage among Persons Aged 26 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ................................................................................... L-22
L-17 Income among Persons Aged 12 or Older: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS ..................................................................................................................... L-23
L-18 Income among Persons Aged 12 to 17: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS ..................................................................................................................... L-24
L-19 Income among Persons Aged 18 to 25: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS Data............................................................................................................. L-25
L-20 Income among Persons Aged 26 or Older: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
NHIS Data ...................................................................................................................... L-26
L-21 Levels of Current Employment among Persons Aged 18 or Older: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire
Field Test, and CPS Data ............................................................................................... L-27

xxiii

List of Tables (continued)
Table

Page

L-22 Levels of Current Employment among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field
Test, and CPS Data ........................................................................................................ L-28
L-23 Levels of Current Employment among Persons Aged 26 or Older: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire
Field Test, and CPS Data ............................................................................................... L-29
L-24 Unemployment Rates among Persons Aged 18 or Older, by Age Group:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and CPS Data........................................................................ L-30
L-25 Levels of Education among Persons Aged 18 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS ..................................................................................................................... L-31
L-26 Levels of Education among Persons Aged 18 to 25: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS ..................................................................................................................... L-32
L-27 Levels of Education among Persons Aged 26 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS ..................................................................................................................... L-33
M-1

Estimates and Standard Errors for New Items in the 2012 Questionnaire Field
Test That Were Included in the 2013 NSDUH Main Study Questionnaire among
Persons Aged 12 or Older ............................................................................................... M-1

N-1

O-1

Estimates and Standard Errors for New, Moved, or Revised Items in the 2012
Questionnaire Field Test among Persons Aged 12 or Older........................................... O-1

P-1

Distribution of Respondent Relationship with Proxy among Persons Aged 12 or
Older Who Obtained a Proxy, by Age Group: Percentages, and Standard Errors,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ..................... P-1

P-2

Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test .................................................... P-2

P-3

Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test ............................................................. P-4

xxiv

List of Tables (continued)
Table

Page

P-4

Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 18 or Older: Percentages and Standard Errors, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test .................................................... P-6

R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality
Issues Observed ...............................................................................................................R-4

R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues
in the 2012 Questionnaire Field Test and Item Missingness Rates for these Items
in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison
Sample............................................................................................................................R-11

R-3

Item Missingness Rates for Moved Items with Observed Data Quality Issues in
the 2012 Questionnaire Field Test and Item Missingness Rates for these Items in
the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample .....R-23

R-4

Distribution of "Don't Know" and "Refused" Item Response Rates for Moved
Items with Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for These Items in the 2011 Comparison Sample and the
2012 Quarters 3 and 4 Comparison Sample ..................................................................R-29

R-5

Received Income and Program Participation among Persons Aged 12 or Older:
Percentages and Totals for 2011 Comparison, 2012 Comparison,
2012 Questionnaire Field Test, and Other Surveys .......................................................R-40

R-6

Received Income and Program Participation among Persons Aged 12 to 17:
Percentages and Totals for 2011 Comparison, 2012 Comparison,
2012 Questionnaire Field Test, and Other Surveys .......................................................R-41

R-7

Received Income and Program Participation among Persons Aged 18 to 25:
Percentages and Totals for 2011 Comparison, 2012 Comparison,
2012 Questionnaire Field Test, and Other Surveys .......................................................R-42

R-8

Received Income and Program Participation among Persons Aged 26 or Older:
Percentages and Totals for 2011 Comparison, 2012 Comparison,
2012 Questionnaire Field Test, and Other Surveys .......................................................R-43

R-9

Health Insurance Coverage among Persons Aged 12 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ...................................................................................R-44

R-10 Health Insurance Coverage among Persons Aged 12 to 17: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ...................................................................................R-45
R-11 Health Insurance Coverage among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ...................................................................................R-46

xxv

List of Tables (continued)
Table

Page

R-12 Health Insurance Coverage among Persons Aged 26 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, Questionnaire Field Test,
2011 ACS, and 2011 NHIS Data ...................................................................................R-47
R-13 Income among Persons Aged 12 or Older: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS .....................................................................................................................R-48
R-14 Income among Persons Aged 12 to 17: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS .....................................................................................................................R-49
R-15 Income among Persons Aged 18 to 25: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
2011 NHIS Data.............................................................................................................R-50
R-16 Income among Persons Aged 26 or Older: Percentages and Standard Errors, 2011
Comparison Data, 2012 Comparison Data, 2012 Questionnaire Field Test, and
NHIS Data ......................................................................................................................R-51
R-17 Levels of Current Employment among Persons Aged 18 or Older: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire
Field Test, and CPS Data ...............................................................................................R-52
R-18 Levels of Current Employment among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field
Test, and CPS Data ........................................................................................................R-53
R-19 Levels of Current Employment among Persons Aged 26 or Older: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire
Field Test, and CPS Data ...............................................................................................R-54
R-20 Unemployment Rates among Persons Aged 18 or Older, by Age Group:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, 2012
Questionnaire Field Test, and CPS Data........................................................................R-55
R-21 Levels of Education among Persons Aged 18 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS .....................................................................................................................R-56
R-22 Levels of Education among Persons Aged 18 to 25: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS .....................................................................................................................R-57
R-23 Levels of Education among Persons Aged 26 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and
2011 NHIS .....................................................................................................................R-58

xxvi

List of Tables (continued)
Table

Page

R-24 Distribution of Respondent Relationship with Proxy among Persons Aged 12 or
Older Who Obtained a Proxy, by Age Group: Percentages, and Standard Errors,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test ...................R-63
R-25 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test ..................................................R-64
R-26 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test ...........................................................R-66
R-27 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among
Persons Aged 18 or Older: Percentages and Standard Errors, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test ..................................................R-68

xxvii

xxviii

List of Exhibits
Exhibit

Page

2.1

Questionnaire Field Test Field Interviewer Training Agenda ...........................................16

3.1

Collapsing Imputation Classes: Race.................................................................................44

3.2

Collapsing Imputation Classes: Race and Gender .............................................................45

5.1

Field Interviewer (FI) Experience with Touch Screen Devices before
Questionnaire Field Test (QFT) Training ........................................................................201

xxix

xxx

1. Background and Goals
This report describes the data collection and analytic methods and results of the 2012
Questionnaire Field Test (QFT) for the National Survey on Drug Use and Health (NSDUH),
including comparisons of selected QFT estimates with current and comparable NSDUH data and
other data sources. Sponsored by the Substance Abuse and Mental Health Services
Administration (SAMHSA), NSDUH is a national survey of the U.S. civilian,
noninstitutionalized population aged 12 or older. The annual conduct of NSDUH is paramount in
meeting a critical objective of SAMHSA's mission to maintain current data on the prevalence of
substance use in the United States.
In order to continue producing data that accurately reflect current conditions, SAMHSA's
Center for Behavioral Health Statistics and Quality (CBHSQ) must update NSDUH periodically
to reflect changing substance use and mental health issues. CBHSQ is planning to implement
changes related to a partial NSDUH redesign. These changes include use of a new sample design
in 2014 and a limited update to the interview questionnaire in 2015. The new sample design will
allow for continued national, State, and substate-level estimation comparable with estimation
from previous surveys. The sample design's improved efficiency will result in significant cost
savings. The primary change to the questionnaire is an updated set of prescription drug modules,
which will include current prescription drugs and incorporate a new questionnaire structure.
Other planned changes to the questionnaire include a revised health module that contains new
questions about drug and alcohol screening by primary care physicians. These changes will seek
to achieve three main goals: (1) to revise the questionnaire to address changing policy and
research data needs, (2) to modify the survey methodology to improve the quality of estimates
and the efficiency of data collection and processing, and (3) to maintain trends in core substance
use estimates1 across survey years. The 2012 QFT is meant to test the revisions to the
questionnaire and protocols.
The NSDUH questionnaire used in the 2012 QFT was revised to improve some of the
questions that cause known or suspected problems with data from the current questionnaire.
New content that addresses current data needs has also been added. Revisions designed to reduce
errors associated with usability problems in the design and layout of the computer-assisted
interviewing (CAI) instrument have been added. These changes include revising the prescription
drug modules, the front-end demographics, the binge drinking definition for women, the special
drugs module, and the back-end demographics section, as well as including a new
methamphetamine module. In addition, materials that describe the survey to respondents have
been revised. These materials include the NSDUH lead letter that is mailed to respondents prior
to their being contacted by an interviewer and the "question & answer" (Q&A) brochure that
interviewers provide to respondents. Section 2.4.1 provides a complete and detailed list of the
questionnaire and protocol changes that were implemented for the 2012 QFT. In addition,
Appendix A shows the changes to the NSDUH questionnaire modules in interview sequence and
provides copies of the redesigned lead letter and Q&A brochure that were used in the 2012 QFT
and are planned for main study implementation in the 2015 survey year.
1

Drugs defined as core substance use items in NSDUH include tobacco, alcohol, marijuana, cocaine, crack
cocaine, heroin, hallucinogens, inhalants, pain relievers, tranquilizers, stimulants, and sedatives.

To inform the questionnaire and protocol for the 2012 QFT, pretesting activities were
conducted. Revised questions were tested with 80 respondents across two phases of cognitive
interviewing. The cognitive interviews tested updated modules for pain relievers, tranquilizers,
stimulants, and sedatives. Questions about drugs that are newly available on the market were
added, and questions about drugs that are no longer commercially available were deleted. A new
definition of misuse of prescription drugs and respondent understanding of a number of new
questions and modules were also tested. In addition, focus groups were conducted in five
metropolitan areas in the United States to obtain feedback from diverse members of the target
population on alternative versions of the NSDUH lead letter and Q&A brochure, including
12 focus groups in English and 5 in Spanish (Currivan et al., 2009).
The primary goal of the field test is to measure the total effect on NSDUH estimates from
all changes to the protocol planned for the 2015 redesign, using multiple indicators. The field
test provides data to attempt to address the following research questions to the extent that sample
sizes allow:
1. To what extent do the planned changes in the protocol influence data quality as
measured by unit nonresponse, item nonresponse, imputation rates, and other
indicators of data quality?
2. To what extent does the redesigned protocol influence the overall timing of the full
interview, the section timing for revised modules, and the screener timing, including
the new field observation questions?
3. What measurable implications, if any, for the general feasibility of the redesigned
protocol were obtained from field observations, field interviewer (FI) debriefing
items, equipment surveys, or focus groups with QFT interviewers?
3a. What feedback from FIs or respondents is received on the redesigned prescription
drug questions on issues such as the ability to understand the questions,
repetitiveness of questions, and ease of interpreting the electronic drug images?
3b. What FI or proxy respondent feedback is received on the new audio computerassisted self-interviewing (ACASI) tutorial for proxy respondents?
3c. What FI and/or respondent feedback is received on any other new aspects of the
redesigned protocol elsewhere in the interview?
4. To what extent are the planned changes in the protocol associated with any increases
or decreases in the reporting of core substance use, methamphetamine, prescription
drugs, or noncore items?2
4a. To what extent are the planned changes in the protocol associated with any
differences in the reporting of core substance use across important demographic
subgroups, especially age groups?

The core consists of initial demographic items (which are interviewer-administered) and self-administered
questions pertaining to the use of tobacco, alcohol, marijuana, cocaine, crack cocaine, heroin, hallucinogens,
inhalants, pain relievers, tranquilizers, stimulants, and sedatives. Noncore items in the NSDUH questionnaire
include substance dependence or abuse, injection drug use, and various demographic and household items.

4b. To what extent do the planned changes to the prescription drug questions appear
to affect the reporting of the misuse of prescription drugs?
4c. To what extent do the planned changes in the protocol appear to be associated
with any differences in reporting for noncore survey items?
This report provides information on how the 2012 QFT was conducted and the results of
this field test. Chapter 2 describes the study design, field preparations, and data collection
procedures. Chapter 3 describes procedures for defining usable cases, editing, imputation,
weighting, data file preparation, and data analysis issues for the 2012 QFT data and the two
NSDUH datasets that were used to compare with the QFT data. This chapter also discusses key
analytic issues, especially comparisons of the 2012 QFT data with the 2012 quarters 3 and 4
NSDUH main study data and the 2011 NSDUH main study data. Chapter 4 details the data
collection outcomes, including screenings and interviews completed, screening and interview
response rates, overall interview timing, selected section timings, imputation rates, item
missingness rates, and other data quality indicators. Chapter 5 describes data collected from
QFT interviewers through multiple methods—including field observations of interviewers, field
debriefing questions completed by interviewers, two equipment surveys, and three focus
groups—to address the general performance of the redesigned protocol. Chapter 6 presents
comparisons of the 2012 QFT core substance use estimates, excluding methamphetamine and
prescription drug items, with 2011 NSDUH and 2012 quarters 3 and 4 NSDUH main study
estimates. Chapter 7 presents comparisons of QFT estimates for methamphetamine and
prescription drugs with 2011 NSDUH and 2012 quarters 3 and 4 NSDUH main study estimates.
Chapter 8 examines QFT estimates for selected noncore items compared with 2011 NSDUH and
2012 quarters 3 and 4 NSDUH main study estimates for these items. Chapter 9 compares
selected QFT estimates with relevant data from other sources, including the National
Ambulatory Medical Care Survey (NAMCS), National Hospital Ambulatory Medical Care
Survey (NHAMCS), Monitoring the Future (MTF), and the National Health Interview Survey
(NHIS). Finally, Chapter 10 summarizes the key findings in the report and presents the
implications of these findings for the partially redesigned NSDUH protocol.

2. Study Design, Field Preparations, and
Data Collection Procedures
2.1

Overview of the Study Design, Field Preparations, and Data Collection

This chapter provides details of the design and implementation of the 2012 Questionnaire
Field Test (QFT). Section 2.2 describes the study design, including the sample design and
selection procedures. Section 2.3 addresses preparations made for data collection, including
preparing the field equipment, selecting the field interviewers (FIs), and training the FIs and field
supervisors (FSs). Section 2.4 describes all of the data collection procedures followed in
implementing the 2012 QFT.

2.2

Study Design

This section describes the target population represented by the QFT, procedures for
selecting sampling regions and segments, selection of dwelling units, allocation of respondents
across age groups, and selection of persons to be respondents for the interviews.
2.2.1

Target Population

Similar to the main study of the National Survey on Drug Use and Health (NSDUH), the
respondent universe for the QFT was the civilian, noninstitutionalized population aged 12 or
older. In order to control costs, persons residing in Alaska and Hawaii, as well as persons who
were not able to complete the interview in English, were excluded from the QFT. Therefore, the
sample is representative of members of the noninstitutionalized population aged 12 or older in
the contiguous United States who are able to complete the interview in English.
2.2.2

Selection of State Sampling Regions and Segments

NSDUH is designed to yield 67,500 interviews from 7,200 segments each calendar year
(Morton, Martin, Shook-Sa, Chromy, & Hirsch, 2012). Thus, an estimated 213 segments were
needed to yield approximately 2,000 completed interviews. To make this sample representative
of the target population, a probability proportional to size (PPS) sample of 213 (of 876) State
sampling (SS) regions was selected. This design maximized the efficiency (i.e., increased the
precision) of the QFT estimates by reducing variation in the weights. In addition, this design had
the benefit of placing the sample in heavily populated areas where a sufficient mix of FIs with
various experience levels would be expected to meet staffing goals. As shown in Table 2.1, a
large portion of the sample was selected from the eight largest States (i.e., California, Florida,
Illinois, Michigan, New York, Ohio, Pennsylvania, and Texas).
Within each selected SS region, a sample of dwelling units was drawn from the segment
that was retired from use in quarter 1 of the 2012 NSDUH. If an insufficient number of dwelling
units remained in a segment, or if significant access problems were expected in a segment, the
segment was replaced with the quarter 4 2011 retired segment in the same SS region. A total of
6 segments were replaced because they had fewer than 10 dwelling units remaining, and a total
of 7 segments were replaced due to anticipated access problems in the segments.
5

Table 2.1 Number of 2012 Questionnaire Field Test State Sampling Regions and Sample Sizes, by
State

State
CA
TX
NY
FL
IL
PA
OH
MI
GA
NC
NJ
VA
MA
WA
IN
AZ
TN
MO
WI
MD
MN
CO
AL
SC
KY
LA
OR
OK
CT
IA
MS
AR
KS
NV
UT
NM

Population
Rank
(12 or Older)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Current Design
3,600
3,600
3,600
3,600
3,600
3,600
3,600
3,600
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900

NSDUH SS
Regions
48
48
48
48
48
48
48
48
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12

Number of
QFT SS
Regions/
Segments
(PPS)
23
14
11
18
10
10
7
9
6
5
6
6
4
5
6
4
4
2
4
3
4
6
4
3
3
5
1
5
5
0
0
0
2
2
6
1

2012 QFT
Respondents
170
146
105
169
72
121
103
86
60
50
52
83
33
46
63
14
51
16
38
32
36
33
45
31
28
66
8
40
41
0
0
0
19
33
63
4
(continued)

Table 2.1 Number of 2012 Questionnaire Field Test State Sampling Regions and Sample Sizes, by
State (continued)
Population
Rank
(12 or Older)
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Total

State
WV
NE
ID
ME
NH
HI
RI
MT
DE
SD
AK
VT
ND
DC
WY

Current Design
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
67,500

NSDUH SS
Regions
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
900

Number of QFT
SS Regions/
Segments (PPS)
2
3
0
2
1
0
0
1
0
0
0
0
0
0
0
213

2012 QFT
Respondents
23
25
0
12
11
0
0
16
0
0
0
0
0
0
0
2,044

PPS = probability proportional to size; QFT = Questionnaire Field Test; SS = State sampling.

2.2.3

Selection of Dwelling Units

Dwelling units that were not selected for the 2011 and 2012 main studies were eligible
for selection in the QFT. A sufficient number of dwelling units was drawn to account for the
lower sample yield resulting from conducting interviews in English only. The starting sample
size and the sample allocation across the segments were determined based on anticipated
eligibility, nonresponse, and the person-level sample selection procedures. Similar to the main
study, a small reserve sample (20 percent) of dwelling units from each segment was selected, and
the total sample was partitioned into four probability subsamples within each segment:
105 percent and three 5 percent partitions, for a total of 120 percent. Although the majority of the
sample (105/120) was released at the beginning of the QFT data collection period, having the
additional sample partitions allowed for greater flexibility in controlling the sample size and
provided the ability to ensure that data collection goals were attained within the field period.
Two additional 5 percent partitions were released in all but six States3 after 4 weeks of data
collection.
A total of 5,358 dwelling units were sampled and yielded 2,044 completed interviews as
shown in Table 2.2. The half-open interval procedure for missed dwelling units was
implemented during the QFT, but it is not scheduled to be implemented in the 2014 or 2015
NSDUHs.
3

Additional sample was not released in the following States: Connecticut, New Mexico, Oregon, South
Carolina, Virginia, and Utah.

Table 2.2 Summary of the 2012 Questionnaire Field Test Sample Results
Statistic
State Sampling (SS) Regions
Segments
Selected Dwelling Units
Eligible Dwelling Units
Completed Screening Interviews
Selected Persons
Eligible Persons1
Completed Interviews

Total
213
213
5,358
4,623
3,837
2,823
2,760
2,044

Rate
N/A
N/A
N/A
0.86
0.83
0.98
0.74

N/A = not applicable.
1
These are selected persons who were eligible for the QFT (excluding final language barriers).

2.2.4

Age Group Allocations

The respondent sample was allocated to the three major age groups in the following
proportions: 25 percent aged 12 to 17, 25 percent aged 18 to 25, and 50 percent aged 26 or older.
Within the 26 or older age group, 15 percent of the sample was allocated to persons aged 26 to
34, 20 percent of the sample was allocated to persons aged 35 to 49, and 15 percent was
allocated to persons aged 50 or older. This sample allocation matched the planned allocation for
the 2015 NSDUH partial redesign. One implication of the respondent sample allocation by age
groups is a potential impact on QFT response rates. Retaining more of the 26 or older adults
identified in households to complete interviews had a negative effect on unweighted interview
response rates. As shown in Table 4.4 in Chapter 4, both the weighted and unweighted response
rates for persons younger than 26 were higher than the response rates for persons aged 26 or
older. The unweighted interview response rate for the QFT sample was 72.41 percent compared
with 76.52 percent for the 2011 main study comparison sample and 79.31 percent for the 2012
quarters 3 and 4 main study comparison sample (see Table 4.4 in Chapter 4). Weighted
interview response rates are not affected by the change in age allocation. Although a smaller
proportion of 12 to 17 year olds were selected, this age group continued to drive the number of
dwelling units needed (i.e., relative to the total population in this age group, the age group
continued to be sampled at the highest rate). Thus, fewer dwelling units were needed to yield the
desired sample than would be needed under the current sample design.
2.2.5

Selection of Persons

After dwelling units were selected within each QFT segment, an FI visited each selected
dwelling unit to obtain a roster of all persons residing in the dwelling unit. This roster
information was used to select 0, 1, or 2 persons for the survey. Sampling rates were preset by
segment and age group. Roster information was entered directly into the electronic screening
program, which automatically implemented this stage of selection based on the segment and age
group sampling parameters. As indicated in Table 2.2, 2,823 people were selected from within
3,837 screened and eligible dwelling units, which yielded 2,044 completed interviews.

2.3

Field Preparations

The primary QFT field preparation activities are presented in this section, including
programming tablets, laptops, and field support systems for data collection; selecting FIs to
conduct the data collection; and developing and implementing the FI training program, materials,
and procedures.
2.3.1

Preparing Field Equipment

As part of a larger effort to evaluate data collection equipment options to be deployed for
the 2015 NSDUH survey year and beyond, the Substance Abuse and Mental Health Services
Administration (SAMHSA) and RTI International4 adopted a phased equipment evaluation
process beginning in the fall of 2011. This process will conclude with final selection of data
collection hardware in 2014. The first and second evaluation phases of this process were
conducted in late 2011 and early 2012. These phases focused on determining whether to pursue a
"one-device" approach in which a single convertible laptop would be used to conduct both
screening and interviews or a "two-device" approach in which a small tablet computer would be
used for screenings and a conventional laptop for interviews. Results from those evaluations
revealed that NSDUH FIs strongly preferred a "two-device" approach. As a result, SAMHSA
and RTI determined that further evaluation phases would focus on tablets running Google's
Android operating system (OS) for screening and laptops running Microsoft's Windows OS for
interviewing. Although NSDUH's technical team initially investigated the possibility of using
Apple devices running iOS, they were ruled out in the early phase because of software
development challenges and higher hardware costs.
Another outcome of the first two evaluation phases was that NSDUH FIs strongly
preferred the Samsung Galaxy Tab 7.0", the smallest and lightest of all devices assessed, as a
potential device to be used for household screenings. For this reason, SAMHSA decided that the
third evaluation phase would consist of field testing the Samsung Galaxy Tab 7.0" as part of the
2012 QFT. All QFT FIs used the tablet for screening QFT cases and completed two equipment
surveys to provide structured feedback about their experiences. (See Section 5.4 in Chapter 5 for
results of the equipment surveys.) Additional feedback about the tablet was gathered during three
FI focus group sessions held at the end of QFT data collection. (See Section 5.5 in Chapter 5 for
results of the focus groups.) Because the existing NSDUH screening software is implemented on
the Windows Mobile platform, a substantial development effort was required to create not only a
new screening program that could run on the Android OS, but also new transmission software
that would enable transmission of data from the tablet and the laptop.
New interview hardware was not field tested during the QFT, partly because it was
desirable to use the same equipment to enable comparisons of the redesigned QFT questionnaire
to the current NSDUH questionnaire and to minimize the risk of software bugs that might
compromise the ability to make these comparisons. Although new laptops were not used, all
QFT FIs received from the existing fleet a second laptop that was configured with the new QFT
questionnaire and transmission program.

RTI International is a trade name of Research Triangle Institute.

Substantial modifications were needed for a variety of supporting systems central to the
supervision and monitoring of NSDUH data collection. These systems include the Web-based
case management system (CMS) that enables supervisors to assign, transfer, and monitor cases;
the reporting systems used for tracking FI performance and costs; and the verification systems
used for data quality. Development work for these supporting systems proceeded in parallel with
work on the screening and interview software for the 2012 QFT.
2.3.1.1

Programming Tablets for Screenings

The current NSDUH screening software is a .NET compact framework program that runs
on Microsoft's Windows Mobile OS. This software steps FIs through a sequence of rostering and
demographic screening questions. The software also performs randomized selections of potential
respondents, based on age, as dictated by an embedded statistical sampling algorithm. Because
the tablet selected for the QFT uses the Android OS, a new Java-based screening program had to
be developed. The software development team chose to develop this as a native Android "app,"
using freely available and open source development tools. The primary development goal was to
replicate the functionality and user interface of the iPAQ program as much as possible in order to
take advantage of FIs' existing knowledge of the program and minimize the need for extensive
training. As a result, the starting point for development was the iPAQ screening software and the
QFT screening specifications. In addition to the standard screening questions and functionality,
these specifications included the addition of a series of interview debriefing questions
(previously embedded at the end of the computer-assisted interviewing [CAI] questionnaire) that
would be displayed once the FI entered the final "interview completed" code. Two features in the
iPAQ screening program—the integrated calendar and the call distribution—were not
implemented in the QFT screening program because of time constraints in the QFT development
schedule. These two features will be developed for the 2013 Dress Rehearsal (DR) version of the
screening program. Finally, new transmission software was developed to enable a connection
between the tablet and laptop and the transmission of screening data back to RTI.
The screening software was built following RTI's standard Software Development Life
Cycle (SDLC). Internal unit testing proceeded in parallel with software development and was
performed by the involved programmers, with external testing provided by unassociated
members of the programming staff and also the second tier of support from the NSDUH Help
team. Test results were communicated among the team using email and other direct
communication. When the iterative process of development, change, and internal testing had
sufficiently proven the prototype, the new screening software was passed to RTI's iTeam for
internal acceptance testing. Iteration again was allowed to occur as needed. Again, email was the
primary tool used to communicate and track progress during this phase. Once RTI's iTeam
accepted the screening software, the software was sent to SAMHSA for acceptance testing. Once
the SAMHSA team confirmed their acceptance test via email, RTI proceeded to integrate the
new screening software into the master configuration for the QFT.
2.3.1.2

Programming Laptops for Interviews

The current NSDUH CAI questionnaire is developed in Blaise, an industry standard
survey programming software, and deployed on Gateway laptops running the Windows XP OS.
As mentioned above, the SAMHSA and RTI teams decided to use existing laptops from the

current fleet of equipment for the QFT interviews. For this reason, no changes were needed in
the software to accommodate a new OS, and the starting point for development was the existing
CAI instrument. However, substantial changes to the CAI questionnaire were made for the QFT,
requiring an extensive programming effort. A complete list of changes to the CAI questionnaire
is provided in Section 2.4.1. A summary of the major changes includes the following:
•

addition of new questions and rewording of existing questions or changes to response
categories,

•

significant reordering of questions in various modules,

•

transitioning interviewer-administered questions into the self-interview portion of the
questionnaire,

•

addition of pop-up question help with accompanying audio, and

•

addition of an electronic calendar and electronic pill images.

As with the screening program, the software was built following the standard SDLC.
Internal unit testing proceeded in parallel with software development and was performed by the
involved programmers, with external testing provided by unassociated members of the
programming staff and with the second tier of support from the NSDUH Help team. Test results
were communicated among the team using email and other direct communication. When the
iterative process of development, change, and internal testing had sufficiently proven the
prototype, the new CAI software was passed to the RTI iTeam for internal acceptance testing.
Because of the magnitude of changes in the questionnaire, an additional set of staff was recruited
to test changes in the questionnaire across two phases of additional testing. Email was the
primary tool used to communicate and track progress during this phase. Once RTI's iTeam
accepted the CAI software, the software was sent to SAMHSA for acceptance testing. Once the
SAMHSA team confirmed their acceptance via email, RTI integrated the new interview software
into the master laptop configuration for the QFT. After this integration occurred, a final round of
integration testing was performed by the programming team.
2.3.1.3

Programming Field Support Systems

QFT data were collected from a national sample of households across the continental
United States from September 1, 2012, through November 3, 2012, concurrent with the 2012
quarters 3 and 4 of the main study. Therefore, data had to be collected, processed, and managed
separately from the 2012 quarters 3 and 4 main study data. This effort required numerous
modifications to existing support systems primarily used by RTI and NSDUH FSs. New pages
were added to the Web-based CMS to allow FSs to assign, transfer, and monitor QFT cases
separately from the main study. The NSDUH reporting system was changed to include a new set
of production, expense, and data quality reports for the QFT. Modifications to NSDUH databases
and data processing systems were required to accommodate CAI questionnaire changes that
involved a multitude of new Blaise variables and to ensure that data transmitted to and from the
field were appropriately identified and stored separately from main study data. Finally, a number
of changes were needed in the verification system, including development of a separate
computer-assisted telephone interviewing (CATI) questionnaire for telephone verifiers and new
functionality on the data quality intranet to support monitoring and tracking of verification data.

RTI employed the same iterative process of development and testing used for the
screening and interview software to change these systems. However, because these are internal
systems used primarily by RTI and FSs and exist largely for the automation and streamlining of
internal project operations, testing of functionality was primarily the responsibility of the
programming team. New functionality was developed and implemented on a development site,
pointed at back-end development databases. Testing was completed by members of the
programming team, with the second tier of support from the NSDUH Help team and in some
cases members of NSDUH's operations and data quality teams. Upon completion of testing,
these systems were released to the production environment, and the Web programming team
continued to monitor and support their operation.
2.3.2

Staffing

The field management team and structure for the QFT were identical to those used for the
main study. All of the FIs selected for the QFT also worked on the 2012 quarters 3 and 4 main
study data collection, which overlapped with the QFT field period. FIs were chosen for the QFT
data collection based on several factors. Initial consideration of FIs was determined by proximity
to QFT segments. Field managers analyzed the QFT sample distribution to determine which FIs
would be strategic choices for consideration. Location, however, was not the only determining
factor.
Length of service on NSDUH was also an important selection criterion for QFT FIs. The
goal for the QFT interviewing team was to have a mix of veteran and newer FIs working on the
QFT data collection effort that was similar to the distribution for FIs working in quarters 3 and 4
of the main study. FIs who had attended the January 2012 new-to-project (NTP) training session
or who had attended an earlier NTP session were eligible for selection for the QFT data
collection. Tenure information was gathered for the proposed cohort of QFT FIs, and the
distribution of their length of service was similar to the main study, with slightly more
experienced FIs working on the QFT. Table 2.3 shows the distribution of 2012 QFT FIs by
tenure level compared with the 2012 quarters 3 and 4 main study FIs collecting data at the same
time.
Proximity to sample segments and experience level were balanced with each FI's
previous data quality and cost efficiency results, availability, and dependability to take on the
additional QFT work from September 1 through November 30, 2012. A group of alternates was
also recruited as replacements in case there was any attrition among the initially selected group
of FIs. In total, 159 FIs successfully completed the QFT FI training and were prepared to conduct
QFT data collection (see Section 2.3.3).

Table 2.3 Tenure Distribution of 2012 Quarters 3 to 4 Main Study Field Interviewers Compared
with 2012 Questionnaire Field Test Field Interviewers
Number of Quarters Worked
on NSDUH Since 2005
0-4
5-8
9 - 12
13 - 16
17 - 20
21 - 24
25 - 28
≥ 29

Total

2.3.3

2012 Quarters 3 and 4 NSDUH
Field Interviewers
Count
Percent
216
27.5
107
13.6
54
6.9
53
6.7
55
7.0
36
4.6
44
5.6
221
28.1
786
100.0

2012 Questionnaire Field Test
Field Interviewers
Count
Percent
13
8.2
26
16.4
19
11.9
9
5.7
14
8.8
8
5.0
12
7.5
58
36.5
159
100.0

Training Procedures
2.3.3.1

Training Materials

Using a master list of needed supplies, all training materials were prepared and ordered
(if necessary) in preparation for QFT training activities. A detailed, near-verbatim training guide
was prepared for each member of the training team. Along with the training guide, numerous
printed materials were also developed:
•

QFT FI handbook that contained protocols and procedures for conducting work on
the QFT;

•

training workbook that contained necessary exercises, screening and interviewing
mock scripts, and additional instructions;

•

quality control forms specifically for the various training cases;

•

interview incentive receipts for use during the practice interviews;

•

showcard booklets for training and use during subsequent fieldwork;

•

supplies to be used during the course of training, including the lead letter, study
description, and question & answer (Q&A) brochure;

•

administrative forms providing site-specific details for proper travel reimbursement;
and

•

evaluation forms used by trainers when observing FIs in class.

Additionally, PowerPoint slides were developed to accompany the various training guide
sections, providing illustrations of the item under discussion or summarizing the main points
conveyed in the guide.
As part of the QFT training plan, the electronic multimedia, interactive training
application, referred to as iLearning (which stands for independent learning), was used. Using
13

iLearning allowed FIs to complete a QFT iLearning course at their own pace and review portions
of the course again as needed. The QFT iLearning course consisted of visual slides with text and
graphics, an audio component providing important information and instructions, a training video,
interactive practice exercises, and an assessment portion to ensure the FI's comprehension of the
QFT material presented. Upon completion of the course and transmission to RTI, the course
assessment results were posted to the CMS for field management staff review. The QFT
iLearning course was completed by all FIs selected for the QFT and prior to attendance at an inperson QFT FI training session. All 163 QFT FIs scheduled to attend the in-person QFT FI
training sessions successfully completed and passed the QFT iLearning course. (See Section
2.3.3.3 for more details on the number of FIs who actually completed the QFT FI training
sessions.)
2.3.3.2

Train-the-Trainers Session

To prepare trainers and instruct all project management staff—including FSs, regional
supervisors (RSs), and regional directors (RDs), as well as other NSDUH team members—in the
procedures for the QFT, a Train-the-Trainers (TTT) session was held in Raleigh, North Carolina,
on August 8 and 9, 2012. A 1-day management meeting was held the day prior to the TTT
session on August 7, 2012, to bring all staff together for discussions on key field management
topics.
The TTT session was led by members of the instrumentation team who reviewed all
portions of the QFT training guide and materials and the logistics for the QFT and instruction on
the equipment being used. Following the review of the QFT FI training, each RD led a special
QFT management session for his or her RSs and FSs to provide instructions and answer
questions related to managing the QFT fieldwork.
2.3.3.3

Field Interviewer Training Sessions

Training sessions for QFT FIs were held in two locations—Cincinnati, Ohio, and
Baltimore, Maryland—with each site hosting two separate training sessions. Session A was held
on August 25 and 26, 2012. Session B took place on August 28 and 29, 2012. Of the 163 QFT
FIs scheduled to attend the in-person QFT FI training, three FIs were unable to attend the
training and participate in the QFT prior to conducting the sessions. Of the 160 QFT FIs who
attended the QFT FI training sessions, 159 FIs successfully completed the training. One FI
demonstrated significant performance issues during the QFT training session and, therefore, did
not successfully complete the training. This FI was excused from the QFT data collection, and
the cases originally assigned to this FI were reassigned to another FI. Table 2.4 summarizes the
results of the QFT FI training sessions.
Table 2.4 Questionnaire Field Test Field Interviewer Training Program

QFT FI Training Session
Session A (August 25 and 26, 2012)
Session B (August 28 and 29, 2012)
Total QFT FIs Completing Training

Cincinnati, Ohio,
FIs Trained
51
48
99

Baltimore,
Maryland,
FIs Trained
36
24
60

Total
87
72
159

The QFT FI training program included an initial self-study component (completed at
home prior to training) in which FIs read the QFT FI handbook and completed the QFT
iLearning course. During the 2-day in-person classroom training, FIs had hands-on practice with
the QFT equipment, programs, and QFT-specific procedures. The 2-day QFT FI training agenda
is provided in Exhibit 2.1.
Day 1

Training classes began with an introduction of the QFT and the FI responsibilities on the
study. The next topic on the QFT equipment provided instruction in the use of the laptop
computer hardware and the basics of the tablet hardware and software, including the screening
program. FIs learned about locating and contacting respondents, completed a group walkthrough of a QFT screening, and were able to practice effectively answering respondent
questions. Then FIs were introduced to the QFT interview conventions and completed a group
walk-through of a QFT interview. The FI debriefing questions were covered, as well as
additional tips for answering QFT-related respondent questions and dealing with nonresponse.
The late afternoon was spent completing two paired mock interviews to gain more practice with
the overall QFT process. During these mock interviews, FIs were observed by trainers and were
given constructive feedback on their performance and understanding. This was also a time when
retraining could take place and FIs could ask questions. All FIs were invited to attend an evening
FI laboratory session for additional practice or assistance. FIs completed a QFT screening and
interview exercise for homework during the evening as well.
Day 2

Day 2 included instruction on the transmission process and how to troubleshoot problems
with the equipment. The homework from the previous evening was reviewed. FIs completed an
actual transmission during this session to ensure everything was working properly and to pick up
their assigned QFT cases. Then FIs completed two more paired mock interviews while trainers
observed, and they received feedback from their trainers. At the end of the training day,
administrative tasks were reviewed, including reporting to their FS, how to record time and
expenses, and tips on organization. During a session wrap-up, key procedures and protocols of
the QFT were reviewed and FI questions were answered. FIs also completed the first installment
of the FI feedback survey.

2.4

Data Collection Procedures

This section describes the data collection procedures for the QFT, including contacting
and screening sample dwelling units (SDUs), interview administration, controlled access and
refusal conversion procedures, data collection management and quality control, and problems
encountered.
2.4.1

Questionnaire and Protocol Changes for the 2012 QFT

The 2012 QFT data collection involved the following changes to the 2012 NSDUH
questionnaire and protocol:
•

The response categories in the highest education completed question were revised.
15

Exhibit 2.1

Questionnaire Field Test Field Interviewer Training Agenda

DAY 1
(1) Introduction to the QFT [30 minutes]

Introductions & Training Agenda

QFT Overview

QFT FI Responsibilities
9:30
(2) Introduction to the QFT Equipment [45 minutes]

Reviewing the Equipment Assignment and Receipt Form (EARF)

Tablet Hardware

Laptop Hardware

Getting Started on the Tablet

Equipment Care & Maintenance
10:15
Break
10:30
(3) Administering the QFT Screening [1 hour, 30 minutes]

Locating & Contacting Respondents

Screening Procedures

QFT Screening - Group Walk-Through

Answering Respondent Questions & Nonresponse

QFT Paired Screening Exercises
12:00
Lunch
1:00
(4) Administering the QFT Interview [2 hours]

Interview Materials & Procedures

QFT Interview - Group Walk-Through

FI Debriefing Questions - Interview

Answering Respondent Questions & Nonresponse
3:00
Break
3:15
(5) QFT Paired Mocks 1 & 2 [1 hour, 45 minutes]

Review of QFT Process

Paired Mocks 1 & 2

Review of Paired Mocks 1 & 2

Individual Feedback

Day 1 Questions & Wrap-Up
5:00
Adjourn
6:00 – 8:00
Field Interviewer Lab
Homework Exercise
DAY 2
9:00
(6) Transmission & Troubleshooting [45 minutes]

Review of Homework Exercise

Answer FI Questions from Day 1

Transmission Procedures (including Actual Transmission)

Troubleshooting & Technical Support
9:45
(7) QFT Paired Mocks 3 & 4 [2 hours]

Paired Mocks 3 & 4
10:30
Break
10:45
(7) QFT Paired Mocks 3 & 4 (continued)

Review of Paired Mocks 3 & 4

Individual Feedback
12:00
Lunch
1:00
(8) Administrative Tasks [45 minutes]

Reporting to Field Supervisor (FS)

Recording Time & Expenses

Organization
1:45
(9) Session Wrap-Up [45 minutes]

Review of Key Procedures & Protocols

Day 2 Questions

FI Feedback
2:30
Adjourn
9:00

•

The reference date calendar was converted to a computerized application that
appeared on-screen.

•

Variables in the audio computer-assisted self-interviewing (ACASI) tutorial section
were combined and streamlined.

•

Smokeless tobacco sections were combined into one section.

•

The definition of binge drinking was changed to four or more drinks for female
respondents.

•

Questions currently included in the special drugs module for hallucinogens, such as
ketamine, tryptamines (dimethyltryptamine [DMT], alpha-methyltryptamine [AMT],
5-MeO-DIPT [N, N-diisopropyl-5-methoxytryptamine], also known as "Foxy"), and
Salvia divinorum, were moved to the core hallucinogens module.

•

New inhalants questions for markers and air duster were added.

•

A new methamphetamine module was added.

•

The definition, approach, and terminology for measuring the misuse of prescription
drugs were all revised.

•

Modules were added asking respondents about any use of pain relievers, tranquilizers,
stimulants, and sedatives as opposed to just nonmedical use.

•

The focus of the prescription drug modules was on a 12-month reference period
rather than the lifetime reference period used in the current questionnaire.

•

Electronic pill images of prescription drugs replaced the current showcard versions.

•

Discontinued prescription drugs were removed.

•

Prescription drugs currently included elsewhere in the questionnaire were added to
the appropriate prescription drug module.

•

Questions about use of cough or cold medicines just to get high were moved to the
beginning of the special drugs module.

•

The special drugs module questions about needle use were reworded, and questions
about use of prescription stimulants with a needle were moved to the prescription
stimulants module.

•

The stimulant questions were revised to reflect separate methamphetamine and
prescription stimulant modules.

•

The marijuana marketing module was removed.

•

The prior substance use module was revised to remove prescription drug questions, to
revise methamphetamine questions to refer to the stand-alone question, and to drop
questions about which drug was used first.

•

The health care module was revised and expanded.

•

Questions about how many times the respondent moved in the past 5 years were
removed from the social environment and youth experiences modules.

•

Questions about prescription drugs were removed from the questions about using
drugs with alcohol in the consumption of alcohol module and moved to the
appropriate prescription drug modules.

•

Questions about drinking four or more drinks on an occasion that were asked of
females in the consumption of alcohol module were dropped.

•

Questions about disability status and how well the respondent speaks English were
added to the ACASI section of the questionnaire in the back-end demographics.

•

New questions about family members currently serving in the U.S. military were
added to the back-end demographics.

•

Industry and occupation questions were removed.

•

Marital status was moved from the core demographics to the back-end demographics.

•

The education, employment, health insurance, and income questions were all moved
to the ACASI portion of the interview. In addition, the top response category for
income was revised.

•

Questions about step, foster, adoptive, or foster relationships in the household roster
were removed

•

A new module introduced proxy respondents to the ACASI.

•

Questions about cellular telephones and landlines were revised. Two new questions
were added, and the previous questions were removed.

•

New FI debriefing questions were added and administered via a new screening
device, a tablet computer with a 7-inch screen size. These questions had previously
been completed by FIs on their laptop computers at the end of the CAI protocol, after
all other questions had been completed.

•

New contact materials, including a redesigned version of the lead letter and Q&A
brochure, were used.

Some of the questionnaire changes were implemented earlier than in the 2015 survey
year. A few select changes made to the QFT questionnaire were also adopted for the 2013 survey
year. These changes include the following items:
•

Two new response categories were added to the race question. The response options
now include (a) Guamanian or Chamorro and (b) Samoan.

•

New questions were added to ask about serving in the reserve components in the
military. The current questions were edited for consistency.

•

Questions about use of medical marijuana were added to the blunts module.

•

New questions were added to the health care module that ask about height, weight,
and the discussions one has had with a doctor about substance use and abuse in the
past year.

•

The Mental Health Surveillance Study (MHSS) questions were eliminated because no
MHSS recruitment occurred as part of the QFT, and the MHSS was discontinued in
2013.5

Each of these features of the QFT data collection represents a difference from how the FIs
administered the main study data collection in 2011 and 2012.
2.4.2

Contacting Dwelling Units

A few procedural changes were implemented during the QFT that differed from the 2012
main study. When contacting respondents, FIs referred to RTI International (or RTI) and the U.S.
Department of Health and Human Services (DHHS), as opposed to Research Triangle Institute
and the U.S. Public Health Service. These updates were reflected in all field materials used for
the QFT, including the lead letter, study description, Q&A brochure, "Sorry I Missed You"
(SIMY) card, Spanish card, interview appointment card, summary of the questionnaire, "Who
Uses the Data?" sheet, RTI/SAMHSA fact sheet, and the door person letters. Because the QFT
interviews were conducted in English only, Spanish versions of materials were not provided for
the QFT. To help FIs distinguish QFT materials from main study materials, the majority of the
QFT materials were printed on gray paper and had the QFT version number (v. QFT 9.12) in the
lower right corner.
2.4.2.1

Lead Letters

Similar to the main study, prior to an FI's arrival at an SDU, a lead letter was mailed to
the address briefly explaining the study and requesting the resident's cooperation (see
Appendix A). This letter was printed on DHHS letterhead with the signature of DHHS' national
study director and RTI's national field director. Upon arrival at the SDU, the FI referred the
respondent to this letter and answered any questions. If the respondent had no knowledge of the
lead letter, the FI provided another copy, explained that one was previously sent, and then
answered any further questions.
The lead letter was modified for the QFT with redesigned content and format changes to
the FI ID and letterhead. The "United States Public Health Service" reference was replaced with
the "U.S. Department of Health and Human Services" in the letter. Additionally, the letters were
preaddressed to include the county, parish, or district name as part of the address and salutation.
These changes were based on the Contact Materials Redesign Study, which included 12 English
focus groups and five Spanish focus groups in five metropolitan areas in the United States
(Currivan et al., 2009).
2.4.2.2

Introduction, Study Description, and Informed Consent

When in-person contact was made with an adult resident of the SDU, the QFT FIs
followed the same introductory and informed consent scripts and procedures for the screening as
the main study, with one exception. The "U.S. Department of Health and Human Services" was
identified as the sponsor of the study and "RTI International" was used instead of "Research
5

Appendix M provides estimates for new or revised items in the QFT questionnaire that were added to the
2013 main study questionnaire.

Triangle Institute" in the study introduction script. These same wording changes were made to
the study description, in addition to updating it with Peter Tice's signature at the bottom as the
current NSDUH Project Officer. All other informed consent procedures remained the same for
the QFT, including handing a study description to the respondent.
2.4.2.3

Callbacks

QFT FIs followed similar guidelines for callbacks as the main study, including
appropriate use of SIMY cards, unable to contact (UTC) letters, and appointment cards. These
materials were utilized by FIs in the same manner as the main study. If no one was at home
during the initial visit to the SDU, the FI left a SIMY card to inform the resident(s) that the FI
planned to make another callback at a later date/time. If the FI was unable to contact anyone at
the SDU after repeated attempts, the FS sent a UTC letter. Appointment cards were used to
remind respondents when the FI would return to complete the interview.
For the main study, except in the case of adamant refusals, FIs attempted to make at least
four callbacks (in addition to the initial call) to each SDU in order to complete the screening
process and complete an interview, if yielded. These contacts were made at different hours on
different days of the week to increase the likelihood of completing the screening. These same
guidelines were followed as best as possible for the QFT, but the more widely dispersed sample
and the limited number of QFT FIs available to travel longer distances resulted in less flexibility
for assignments and fewer staff for remote segments. For the main study, FSs were able to
generate more effective callbacks by strategically assigning and transferring cases based on FI
availability and experience.
For the QFT, FIs made five or more contacts to each dwelling unit with the exception of
language barrier cases, physically or mentally incompetent cases, or refusal cases. QFT data
collection ended on November 3, 2012, which was approximately a 2-month data collection
period as opposed to the 3-month data collection period on the main study and originally planned
for the QFT. Although the QFT did exceed the nationwide goal of 2,000 completed interviews,
the QFT experienced lower response rates than the main study. (See Section 4.2.1 and Table 4.1
in Chapter 4 for a comparison of response rates between the QFT and the two main study
comparison samples.) The lower response rates are mainly a result of the limited number of QFT
FIs available for assignments and the transfer of cases. However, the response rates may have
been higher if FIs had made additional callbacks to convert refusals and reach the UTC
respondents over another month of data collection.
2.4.3

Dwelling Unit Screening

QFT procedures for screening at a dwelling unit were similar to those used on the main
study. The most significant change was that all screenings were completed on the tablet, as
opposed to the iPAQ (see Section 2.3.1 for more information on the new equipment). The
introduction and informed consent scripts incorporated the changes specified above. The
information gathered from the respondent during the screening was the same as what is collected
in the main study.

After the interview respondent selections were made (codes 30, 31, and 32), the FI was
prompted by the tablet to complete debriefing questions. The questions were not read out loud to
the respondent; rather, the FI completed them on his or her own after leaving the SDU. In the
case of an on-the-spot interview, the FI answered the questions while setting up the laptop or
during the ACASI section of the interview. These post-screening debriefing questions ask about
the respondents' recollections and reactions to the lead letter (see Appendix E).
2.4.4

Interview Administration

Upon selection, FIs attempted to complete the QFT interview using many similar
techniques as in the main study. However, FIs were trained to answer common respondent
questions based on the QFT procedures. For example, as discussed previously, FIs used the QFT
naming conventions of "RTI International" and the "U.S. Department of Health and Human
Services" rather than "Research Triangle Institute" and the "U.S. Public Health Service." To
describe the types of questions asked, the FI provided the respondent with the QFT version of the
summary of the questionnaire, but FIs were instructed to never tell respondents that they were
part of a questionnaire field test or provide specific sample size information. Also different from
the main study, interviews for the QFT were only conducted in English. No interviews were
conducted in Spanish. Therefore, if an FI encountered a household or respondent unable to
complete the screener or interview in English, the FI thanked the respondent for his or her time
and coded out the case appropriately.
2.4.4.1

Informed Consent and Getting Started

Prior to beginning a QFT interview, FIs obtained informed consent by following the same
informed consent procedures as used in the main study. This included reading the QFT version
of the appropriate introduction and informed consent scripts from the QFT showcard booklet
before the interview began. These scripts were modified for the QFT to ensure that respondents
were accurately informed about the study. Specifically, the informed consent statement states
that the individual respondent will represent thousands of others. Because the representativeness
of each respondent differs in the QFT sample, the sample size information was removed from the
script. In addition, the reference to the "U.S. Public Health Service" in the introduction and
informed consent scripts for respondents aged 18 or older was replaced with the "U.S.
Department of Health and Human Services." Finally, as part of the informed consent, FIs
provided the QFT study description if they had not already done so. Respondents were never
informed that the interview was part of a questionnaire field test.
2.4.4.2

Computer-Assisted Interviews

FIs began the interview with the front-end computer-assisted personal interviewing
(CAPI) section, which contained demographic questions similar to those on the main study with
a few key differences. New questions were added regarding the respondent's prior military
service, two new categories were added to the race question ("Guamanian or Chamorro" and
"Samoan"), and response categories were adjusted in the education-level question. As in the
main study interview, the FI introduced the respondent to the computer prior to the respondent
completing the practice session and ACASI section on his or her own. As noted in Section 2.4.1,

there were several key changes to the ACASI portion of the interview for the QFT, including the
electronic reference calendar and on-screen pill cards.
Following the ACASI section of the interview, the FI took the computer back and asked
the household roster questions. Following these questions, the FI inquired about the use of a
proxy for the health insurance and income questions. For the QFT, a second ACASI section
administered the health insurance and income questions. If a proxy was used, the FI introduced
the proxy to the computer prior to the proxy completing a short practice session and the health
insurance and income questions on his or her own. However, if the respondent answered the
questions or the proxy had previously used the computer, there was no additional practice
session. The industry and occupation questions and MHSS recruitment screens were removed
from the QFT interview. In addition, the number of showcards was reduced because many of the
questions previously requiring showcards were moved to the ACASI portion of the interview for
the QFT, allowing respondents to view answer choices on-screen.
2.4.4.3

End of Interview Procedures

QFT quality control forms were completed in the same manner as on the main study.
Minor changes were made to the verification screen, including removing the word "home" in the
telephone number reference to match the wording on the QFT quality control form and asking
respondents to enter their current address. Text was added that told the respondent to return the
form in the sealed envelope to the FI.
Respondents received a $30 incentive for completing the interview following the same
procedures used on the main study. At this point, if not given earlier, the FI provided the
respondent with the QFT version of the Q&A brochure (see Appendix A). QFT certificates of
participation were also available for youth respondents and were presented in the same way as on
in the main study.
Finally, the FI debriefing questions were removed from the end of the interview because
these questions were answered in the tablet upon entering a code of 70 for the completed
interview. This change allowed the FIs to answer the questions after leaving the household and
reduce the length of time in the respondent's home. The questions were answered by the FIs
based on the interview and any comments the respondent may have offered.
2.4.5

Controlled Access Procedures

Controlled access was treated in much the same way for the QFT as for the main study.
When controlled access situations were encountered, controlled access packets were requested
by the FS. The QFT controlled access packets reflected the differences in the naming
conventions implemented for the QFT. To gain access in difficult situations, FSs also transferred
cases between QFT FIs. If those attempts failed, "Call-Me" letters were sent directly to a selected
household. These letters informed residents that an FI had been trying to contact them and asked
that they contact an FS by telephone.

2.4.6

Refusal Conversion Procedures

Refusal conversion procedures followed for the QFT were similar to those used for the
main study. If a potential respondent refused, the FI attempted to address the respondent's
concerns and was trained to accept the refusal in a positive manner, thereby avoiding the
possibility of creating an adversarial relationship and precluding future opportunities for
conversion. A refusal letter was then sent by the FS. The refusal letter was tailored to the specific
concerns expressed by the potential respondent and asked him or her to reconsider participation.
Based on the refusal situation, an in-person conversion was generally attempted by the original
FI or another QFT FI available nearby or on travel assignment. However, in some FS regions,
another QFT FI was not available nearby or on travel assignment due to the small number of
cases remaining in the area.
2.4.7

Data Collection Management and Quality Control

FIs and field management staff worked strategically to balance quality, cost, and
production goals for the QFT, just as they do for the main study. The case management tools,
features, and reports used by the management team to monitor fieldwork for the main study were
adapted for use during the QFT.
2.4.7.1

Web-Based Case Management Reports

The Web-based Case Management System (CMS) housed a QFT reports' page that
mirrored the main study reports' pages. The structure of the reports remained the same for the
QFT. The following daily reports were available for case management on the QFT: daily FS and
State response rate report, daily status reports, edited address reports, duplicate address reports,
and recruit reports. The following weekly reports were also available on the CMS: executive
summary report, data quality summary report, missing screening data report, record of calls
(ROC) time discrepancies, and the interview length report. These reports were the same as the
main study reports except that QFT data were used.
2.4.7.2

Field Interviewer Observation Procedures

In conjunction with QFT data collection, field observations of QFT FIs were conducted
by RTI and SAMHSA staff members. Groups of four FIs were chosen for field observations in
each of five metropolitan areas: Detroit, Michigan; Miami, Florida; Denver, Colorado; New
York City, New York; and Chicago, Illinois. SAMHSA staff also observed an additional five FIs
in North Carolina, Maryland, Virginia, and Pennsylvania. An observation was considered
complete only after a full interview was observed. An observation where only screenings or
partial interviews took place was not considered complete.
To keep travel costs to a minimum, FIs were chosen for QFT field observations based on
location and proximity to RTI and SAMHSA observation staff. FIs were observed in nine States
total, centered on metropolitan areas. Observers used the QFT field observation screening
checklist and the QFT field observation interviewing checklist to document their observations.
A field observer reference sheet and a field observer task list were used to help maintain
consistency in planning observation assignments and interacting with FIs and respondents (see
Appendix D). Observers were asked to ensure that a field observation FI instruction sheet was
23

sent to each FI prior to the FI's arrival in the field. The QFT housing unit (HU) and group
quarters unit (GQU) scripts and CAI specifications for the front-end and back-end CAPI
questions were provided to observers for their use during the observations. These materials were
developed specifically for the QFT data collection effort based on similar materials used for the
main study field observation process.
Observers were asked to transfer information from paper field observation screening
checklists and field observation interviewing checklists to spreadsheets designed specifically for
the QFT field observations. The field observation manager then used the spreadsheets to process
the results of the field observation, which included issuing any appropriate disciplinary action,
creating a retraining plan to address any observed errors, and sending any comments about the
performance of the questionnaire, equipment, or materials to the appropriate RTI staff member.
The same standardized retraining process was used for the QFT field observations as is
used for the main study field observations. After the field observation manager reviewed each
observation form, for each FI who had errors reported on his or her observation, a member of the
NSDUH operations team completed a document referred to as the FI retraining template. This
template indicates the errors the FI made, the type of retraining required, and the dates by which
the retraining must be completed. The FS used this form to provide standardized feedback and
retraining (as scripted on the template) and issued any appropriate disciplinary action as directed
by the field observation manager.
2.4.7.3

Verification of Completed Cases

Of the 2,044 completed QFT interviews, 16 QFT quality control forms were not returned.
Of the 2,029 that were returned to RTI, 1,859 came back with a status of OK (indicating no
problems), 167 came back with problems, and 3 respondents refused to complete the form.
Two types of changes were made to the verification scripts for the QFT:
•

minor change due to changes in the QFT protocol: for example, referencing a tablet
instead of an iPAQ, providing a different computer tutorial question as an example to
the respondent, and saying "U.S. Department of Health and Human Services" and
"RTI"; and

•

changes designed to improve falsification detection: having the respondent provide
some household roster (number of people who are male and female) and address
(street number and name) information. On the main study, respondents simply
confirmed the information is correct after it is provided. This change was also made
for the 2013 main study verification scripts.

Of the completed QFT interviews, 901 cases were selected for telephone verification. No
problems were found with 435 cases, 184 cases did have problems, 227 cases were unable to be
contacted, and 55 cases had other issues. Of the completed QFT screenings, 913 cases were
selected for telephone verification. No problems were found for 397 of the cases, 161 cases did
have problems, 252 cases were unable to be contacted, and 103 cases had other issues. Problem
cases were those that verified with errors, such as items the respondent did not remember the FI
performing, the respondent reported that this was not the correct phone number for that address,
24

or if the respondent said that he or she was not given the $30 incentive. Cases with "other issues"
were considered unresolvable and included situations in which the telephone interviewer was
never able to speak with the respondent, someone answered the phone but refused or hung up, or
an initial problem was reported but callback verification staff were not able to recontact the
respondent to confirm the issue. Staff on the callback verification team recontacted respondents
when a problem was reported and more information was needed to confirm or clarify the
situation because, during the initial call, the verification script was read verbatim by the
telephone verifiers.
2.4.8

Problems Encountered
2.4.8.1

CAI Questionnaire Issues

Several minor inconsistencies in the CAI program were uncovered, either during data
collection or during analysis. Most notably, a routing issue in the hallucinogens module caused
14 cases to be routed incorrectly for questions LS05, LS11, and LS17. This logic was included in
the specifications correctly, but it was not added to the program. If a respondent reported having
used lysergic acid diethylamide (LSD) in question LS01a or LSREF1, or reported using
phencyclidine (PCP) in question LS01b or LSREF2, or reported using Ecstasy in question LS01f
or LSREF3, and reported "YES" to any of the new questions (Salvia divinorum,
DMT/AMT/"Foxy," or ketamine), he or she was not routed to question LS05, LS11, or LS17 as
indicated in the specifications. Four respondents were incorrectly routed out of the LSD use
questions as a result. A final value for LSD recency was imputed for these cases. An additional
10 cases incorrectly skipped the Ecstasy use questions, and those respondents have unknown
Ecstasy recency. These errors did not cause a significant shift in the QFT prevalence estimates
for LSD, Ecstasy, or any other hallucinogen. The data that are not available for these cases are
initiation data for LSD and Ecstasy. However, initiation data were not analyzed as part of this
QFT report. Overall, the impact of the routing logic issue for these 14 cases is minimal.
A second routing inconsistency occurred for question HLTH29, which asks respondents
if they had cancer during the past 12 months. If a respondent indicated his or her current age as
the age of first cancer diagnosis in any of the preceding health questions, HLTH29 should have
been skipped. This logic was correctly indicated in the specifications, but it was not included in
the CAI program. HLTH29 was not skipped during the QFT, and respondents were asked for
redundant information. This routing error was corrected for the 2013 DR and 2015 redesign and
did not cause a loss of unique data for any case.
Additionally, some programming logic incorrectly remained in the QFT CAI from the
test questionnaire used in the two phases of cognitive interviewing conducted during QFT
pretesting. This logic affected two questions. Respondents who reported receiving the
prescription drug that they misused for free from a friend or relative were asked two follow-up
questions. The first question asked the respondent to specify how that friend or relative got the
prescription drugs (e.g., question PRY42BSP). If the respondent answered, "He or she got the
drug in some other way," the second question asked respondents to specify where this friend or
relative got the prescription drug (e.g., question PRY42C). During the cognitive interviewing
phase, the specifications called for the questionnaire to skip questions PRY42BSP and PRY42C.
(This allowed analysts to avoid learning of others' illegal behavior.) Because this logic was not

removed from the QFT specifications, 17 respondents aged 12 to 17 were skipped out of two
follow-up questions regarding the source of prescription drugs in each of the four prescription
drug main modules (questions PRY42BSP, PRY42C, TRY21BSP, TRY21C , STY26BSP,
STY26C, SVY19BSP, and SV19C). Table 2.5 presents the question text for each of these QFT
items affected by the incorrect logic and the number of QFT respondents who incorrectly
skipped. As Table 2.5 indicates, the number of respondents affected by the inclusion of this
incorrect logic was small, so the impact of this error on the QFT analysis was minimal.
The data structure was changed for question TX10 after QFT data collection. TX10 lists
12 drugs and asks respondents to indicate which for one or more of these drugs the respondent
needed treatment. During the QFT, there were 12 possible responses, but the CAI program only
accepted 10 responses. After a review of 2012 data, it was found that no respondent had entered
more than six responses to question TX10. It is believed that there was no loss of data as a result
of this error in the QFT results. TX10 was updated to accept 12 possible responses for the 2013
DR and the 2015 redesign.
Table 2.5 Questionnaire Field Test Items with Programming Logic Errors

Variable
PRY42BSP

PRY42C
TRY21BSP

TRY21C
STY26BSP

STY26C
SVY19BSP

SV19C

Question Text
Please type in the other way you got the [pain reliever]. You do not
need to give a detailed description — just a few words will be
okay. When you have finished typing your answer, press [ENTER] to
go to the next question.
You reported that you got the [pain reliever] from a friend or relative
for free. How did your friend or relative get the [pain reliever]?
Please type in the other way you got the [tranquilizer]. You do not
need to give a detailed description — just a few words will be
okay. When you have finished typing your answer, press [ENTER] to
go to the next question.
You reported that you got the [tranquilizer] from a friend or relative
for free. How did your friend or relative get the [tranquilizer]?
Please type in the other way you got the [stimulant]. You do not need
to give a detailed description — just a few words will be okay. When
you have finished typing your answer, press [ENTER] to go to the
next question.
You reported that you got the [stimulant] from a friend or relative for
free. How did your friend or relative get the [stimulant]?
Please type in the other way you got the [sedative]. You do not need to
give a detailed description — just a few words will be okay. When
you have finished typing your answer, press [ENTER] to go to the
next question.
You reported that you got the [sedative] from a friend or relative for
free. How did your friend or relative get the [sedative]?

Number of QFT
Respondents
Who Incorrectly
Skipped Item
2

9
1

4
0

4
1

2.4.8.2

Data Collection Issues

Data on callbacks indicate that the distribution of visits to SDUs to complete QFT
screenings and interviews was similar to the 2011 and 2012 quarters 3 and 4 comparison samples
(see Section 4.2 in Chapter 4). Despite these similar callback patterns, overall response rates
were lower for the QFT sample than for the two main study comparison samples. One reason for
this discrepancy was that fewer QFT FIs were available to work the widely dispersed QFT
sample. Field management staff had less flexibility to assign and transfer cases between FIs,
which made the on-the-spot interview and callback attempts less successful than during the main
study data collection. For example, fewer experienced refusal converters were available to be
assigned to refusal conversion efforts. For those QFT segments that were remote, fewer callback
attempts were feasible without having FIs travel long distances for only a few pending cases.
QFT sample partitions 2 and 3 were released on September 28, 2012, when it was
determined that additional sample was needed to ensure the target of 2,000 completed QFT
interviews was met. This additional sample was released in all QFT States, except for
Connecticut, New Mexico, Oregon, South Carolina, Utah, and Virginia. Because data collection
ended on November 3, 2012, FIs did not have as much time to contact these cases in the second
release as in the original release, but all of these cases were contacted at least five times. Overall,
response rates were higher for the original sample release, but the number of SDUs, screenings,
and interviews associated with the additional release were quite small and, therefore, did not
have much of an impact on the overall response rate.

3. Processing and Analysis of the 2012
Questionnaire Field Test Data and 2011 and
2012 Comparison Data
3.1

Overview of Data Processing and Analysis Approach

This chapter describes the procedures followed to process the 2012 Questionnaire Field
Test (QFT), the 2011 National Survey on Drug Use and Health (NSDUH) main study
comparison data, and the 2012 quarters 3 and 4 NSDUH main study comparison data. All data
processing procedures were developed and implemented to provide the greatest possible degree
of comparability among these three datasets to facilitate valid comparisons. Section 3.2 describes
the usable case rules followed, and Section 3.3 details the editing and coding procedures. Section
3.4 presents the imputation procedures, and Section 3.5 describes the weighting steps followed
and the creation of variance estimation strata and replicates. Section 3.6 describes the data file
preparation, and Section 3.7 discusses the data analysis issues.

3.2

Defining Usable Cases

3.2.1

Overview of Defining Usable Cases

A key step in the preliminary data processing procedures established the minimum item
response requirements in order for cases to be used in weighting and further analysis (i.e.,
"usable" cases). These procedures were designed to disregard data from cases with unacceptable
levels of missing data, thereby using data from cases with lower levels of missing data and
reducing the amount of statistical imputation that would be needed for any given record.
3.2.2

Usable Case Definitions

The usable case criteria that were in place for the main survey were used for the 2011
main study and the 2012 quarters 3 and 4 NSDUH main study comparison data, as defined
below:
1. The lifetime cigarette gate question CG01 must be answered as "yes" or "no."
2. At least nine (9) of the following additional gates must have answers of "yes" or "no":
(a) chewing tobacco, (b) snuff, (c) cigars, (d) alcohol, (e) marijuana, (f) cocaine
(in any form), (g) heroin, (h) hallucinogens, (i) inhalants, (j) misuse of pain relievers,
(k) misuse of tranquilizers, (l) misuse of stimulants, and (m) misuse of sedatives.
(For the "multiple gate" modules for hallucinogens through misuse of sedatives, at
least one gate question in the series for that module must have an answer of "yes" or
"no.")

In the 2011 main study, 0.08 percent of all completed interviews (including interviews
from Alaska and Hawaii) did not meet the usable case criteria.6 In the 2012 quarters 3 and 4
NSDUH main study comparison data (which excluded interviews from Alaska and Hawaii),
0.04 percent of the completed interviews did not meet the usable case criteria.
For the QFT, fully defined data for lifetime use or nonuse of cigarettes continued to be a
requirement. Because of changes to the QFT instrument, the following was the second criterion
for usable cases in the QFT:
•

"Usability" must be determined for at least nine (9) of the following additional
modules: (a) smokeless tobacco, (b) cigars, (c) alcohol, (d) marijuana, (e) cocaine
(in any form), (f) heroin, (g) hallucinogens, (h) inhalants, (i) methamphetamine,
(j) pain relievers, (k) tranquilizers, (l) prescription stimulants (i.e., independent of
methamphetamine), and (m) sedatives.

As in the main survey, the usability criterion for smokeless tobacco through heroin was that
lifetime use or nonuse must be determined. For the "multiple gate" modules for hallucinogens
and inhalants, at least one gate question in the series for that module was required to have an
answer of "yes" or "no."
The usability criterion for the prescription drugs in the QFT required that any past year or
lifetime use or nonuse can be determined from the data. Specifically, any of the following met
the usability criteria for prescription drugs:
•

past year use of at least one specific prescription drug in a category (e.g., pain
relievers) is reported in the screener questions; or

•

lifetime use or nonuse of any prescription drugs in the category is reported; or

•

past year nonuse of all specific prescription drugs in the screener is reported,
regardless of whether lifetime use or nonuse can be determined.

One QFT respondent (0.05 percent of the 2,044 completed interviews) did not meet the
usable case criteria and was not included for further analysis. This case failed to meet the
usability criteria for smokeless tobacco, cigars, inhalants, methamphetamine, tranquilizers,
stimulants, and sedatives. This respondent refused most of the questions in the screeners for
tranquilizers, stimulants, and sedatives and refused to report whether he or she had ever used
these prescription drugs.

3.3

Editing and Coding Procedures

3.3.1

Overview of Editing and Coding Procedures

Data that field interviewers (FIs) transmit to RTI are processed to create a raw data file in
which no logical editing of the data has been done. The raw data file consists of one record for

The 2011 comparison dataset (excluding interviews in Alaska and Hawaii) was created from the cases in
the full survey that already been identified as meeting the usable case criteria.

each transmitted interview. Cases were eligible to be treated as final respondents if they met the
usable case criteria described in Section 3.2.
Logical editing was the first step in processing the raw QFT data and the raw comparison
data from 2011 and quarters 3 and 4 of 2012. Logical editing involved using data from within a
respondent's record to (a) reduce the amount of item nonresponse (i.e., missing data) in interview
records, including identification of items that were legitimately skipped; (b) make related data
elements consistent with each other; and (c) identify ambiguities or inconsistencies to be
resolved through statistical imputation procedures (see Section 3.4).
In addition, a limited set of written answers that interviewers or respondents typed for
responses that did not fit any of the listed categories or examples were assigned numeric codes to
facilitate further use of these data in creating final variables or in analysis. These are
subsequently referred to as "OTHER, Specify" data.
3.3.2

Coding of "OTHER, Specify" Data

Written answers that respondents or interviewers typed were assigned numeric codes for
the following: other Hispanic origin, other racial groups, other Asian origin, and other drugs that
respondents used.7 Typed "OTHER, Specify" responses first were compared against databases
for the relevant "OTHER, Specify" variables that contained typed entries and the associated
numeric codes. If an exact match was found between the typed response and an entry in the
system, the response was assigned the appropriate numeric code. Typed responses that did not
match an existing entry were output for manual analyst review and coding.
Coding of data for Hispanic origin, Asian origin, and race made these data available for
creating final demographic variables. Coding of "OTHER, Specify" data for drugs made these
data available for examining the quality of responses to the drug use questions.
Although "OTHER, Specify" data were not coded for other variables, weighted QFT
percentages were generated for affirmative reports to selected lead questions governing
"OTHER, Specify" data, such as reports of obtaining misused prescription drugs "some other
way." Findings for these additional "OTHER, Specify" data are discussed in Section 4.6 in
Chapter 4.
3.3.3

General Editing Principles

To reduce the potential for differences to be attributable to the effects of editing, data for
the main study comparison samples from 2011 and quarters 3 and 4 of 2012 (referred to in the
remainder of Section 3.3 as "comparison" data) and for the QFT were edited in the same manner
wherever possible. If questionnaire changes for the QFT did not permit total comparability
between the editing procedures for the QFT and the comparison data, the aim was to make the
procedures as comparable as possible.

Additional "OTHER, Specify" variables had previously been coded for the 2011 survey. These variables
were not included for the 2011 comparison data analysis because corresponding variables were not coded in the
QFT or the comparison data from quarters 3 and 4 of 2012.

One of the initial steps in the editing involved development and implementation of
procedures for identifying potential patterned responses in the data (subsequently referred to as
data "diagnostics"). Specifically, respondents may enter patterned responses in the core drug use
modules that raise questions about the validity of their answers in a particular module or in the
interview as a whole. The types of patterned responses that were reviewed in the core modules
for the comparison data are documented in the editing and coding section (Section 10) of the
2010 methodological resource book (Kroutil, Handley, & Bradshaw, 2012a). Checks were made
for these same patterns in core QFT modules that did not change (or underwent minimal change)
relative to the main survey. Because the content of the new methamphetamine module in the
QFT was similar to the content in the core modules for marijuana, cocaine, and heroin, the same
types of data checks in these latter modules were implemented for the methamphetamine
module. Particular attention was given to developing specifications and reviewing data for the
QFT prescription drug questions because of changes to these questions for the QFT. Depending
on the results, cases that otherwise met the usable case criteria could be treated as
nonrespondents because their answer patterns raised questions about the overall validity of their
interview data. Alternatively, cases could be kept as final respondents but with all variables in
one or more of their modules being assigned codes for "bad data," provided that these cases still
met the usable case criteria after the assignment of "bad data" codes (see Section 3.2); codes for
"bad data" were treated as missing values in subsequent data processing or analysis. Findings
based on these data diagnostics reviews are discussed in Section 4.6 in Chapter 4.
A key component of the editing procedures for the QFT and comparison data involved
assignment of codes to indicate when it could be determined unambiguously that respondents
legitimately skipped out of questions because of their answers to previous questions. For
example, if respondents answered the lifetime alcohol use question AL01 as "no," all remaining
questions in the alcohol module were skipped. In this situation, the editing procedures assigned
codes to the remaining alcohol variables to indicate that the questions were not applicable
because the respondents never used alcohol. However, if respondents did not know or refused to
report whether they had ever used alcohol, the remaining questions for alcohol use also were
skipped. In this situation, the edited alcohol use variables that had been skipped continued to
have missing values. Determination of whether these respondents were lifetime alcohol users or
nonusers was handed through the imputation procedures described in Section 3.4.
Because the QFT and comparison interviews consisted of "core" sections (i.e., certain
demographic characteristics and use of cigarettes through misuse of sedatives) and noncore
sections starting with the special drugs section, a second key principle of the editing procedures
was that data from supplemental sections typically were not used to edit core data. An exception
discussed in Section 3.3.4 is that comparison data on methamphetamine use from the
supplemental special drugs module along with core data were taken into account in a special set
of edited variables for methamphetamine and stimulants.
However, core drug data could be used to edit supplemental data when respondents were
not asked supplemental questions about a given drug based on their report of most recent use of
that drug in the corresponding core module. For example, respondents in the QFT or
comparisons were not asked questions about cocaine dependence or abuse in the supplemental
substance dependence and abuse module if they last used cocaine or crack cocaine more than
12 months ago. In this situation, the edited variables for cocaine dependence or abuse were
32

assigned codes to indicate that respondents were not asked these questions because the questions
did not apply.
In all core drug modules for the comparison data and in the cigarette through
methamphetamine core QFT modules, respondents were asked "gate" questions to determine
lifetime use or nonuse; because of changes to the questioning strategy and routing logic in the
QFT for prescription drugs, principles for editing the QFT prescription drug variables are
discussed in Section 3.3.4.8 The modules for hallucinogens and inhalants in all datasets and the
prescription drug modules in the comparison data included multiple gate questions about lifetime
use (or misuse) of specific drugs in the category. Respondents who reported lifetime use of the
particular drug (e.g., marijuana) or any drug in the category (e.g., hallucinogens) were asked
when they last used the drug (or any drug in the category). Respondents who did not know or
refused to report when they last used were asked follow-up questions in an attempt to obtain data
on the specific period when they last used (e.g., within the past 30 days, more than 30 days ago
but within the past 12 months, or more than 12 months ago). If these respondents indicated the
specific period when they last used, the data from these follow-up questions were incorporated
into the edited variables for most recent use. If these respondents on follow-up still did not know
or refused to report when they last used, the edited variable for most recent use was assigned a
code to indicate that these respondents logically could be inferred to be users at some point in
their lifetime based on the computer-assisted interviewing (CAI) routing. A definite period of
most recent use was statistically imputed (see Section 3.4).
The CAI program included checks that alerted respondents or interviewers when an
entered answer was inconsistent with a previous answer. In this way, the inconsistency could be
resolved while the interview was in progress. In situations where a "consistency check" was
triggered during the interview, final values from these checks were incorporated into the edited
variables for drugs and selected additional measures in the QFT and comparison data.
Not every inconsistency was resolved during the interviews, and the CAI program did not
include checks for every possible inconsistency that might have occurred in the data. In NSDUH
editing for the main survey, inconsistencies between related variables in core substance use
modules are flagged and the inconsistencies are resolved through statistical imputation
(Kroutil et al., 2012a). To facilitate timely data processing, however, only a limited set of
additional inconsistencies were resolved in the editing procedures. Consequently, inconsistencies
could exist between related variables in the QFT or comparison data that would otherwise have
been handled in the editing procedures for the main study. However, special "flag" variables
were created to alert analysts to the occurrence of these inconsistencies. Findings based on these
flag variables are discussed in Section 4.6 in Chapter 4.
3.3.4

Special Editing Situations

Most editing of the QFT and comparison data followed the principles discussed in
Section 3.3.2. In the alcohol module, the question in the comparison data that was used to define
binge alcohol use asked both males and females about the number of days that they consumed
five or more drinks on the same occasion in the past 30 days. In the QFT, males were asked
8

The text typically mentions "use" when referring both to prescription drugs and other substances.
For prescription drugs, however, this term means "misuse," unless otherwise indicated.

about consumption of five or more drinks on the same occasion, and females were asked about
consumption of four or more drinks on the same occasion. These binge alcohol use variables
were edited in the same manner in both the QFT and comparison data. However, the edited QFT
variable was given a name that was different from the name for the corresponding variable in the
comparison data to indicate the differences in content.
In addition, the following special situations were relevant to editing of the QFT or
comparison data:
•

In the comparison data, respondents were asked separate questions about their use of
snuff or their use of chewing tobacco. In the QFT, respondents were asked about their
use of any smokeless tobacco product (i.e., snuff or chewing tobacco).

•

In all three datasets, respondents could report more recent use of crack cocaine than
they reported for use of any cocaine. Respondents also could report more recent use
of specific hallucinogens (lysergic acid diethylamide [LSD], phencyclidine [PCP], or
Ecstasy in the comparison data; LSD, PCP, Ecstasy, ketamine, dimethyltryptamine
[DMT], alpha-methyltryptamine [AMT], N, N-diisopropyl-5-methoxytryptamine [5MeO-DIPT], or Salvia divinorum in the comparison data) than they reported for use
of any hallucinogen. In addition, respondents in the comparison data could report
more recent misuse or use of OxyContin® or methamphetamine than they reported for
any pain reliever or any stimulant, respectively.

•

In all three datasets, respondents were asked whether they used hallucinogens,
inhalants, pain relievers, tranquilizers, stimulants, or sedatives other than those they
were asked about. Respondents were asked to specify the names of up to five
additional drugs (subsequently referred to as "OTHER, Specify" data). However,
respondents could fail to report use of specific drugs in direct questions about these
drugs and then mention these drugs in the "OTHER, Specify" data.

•

Respondents could indicate that the only prescription drugs they misused in the
lifetime period (for the comparison data) or the past year (for the QFT) were over-thecounter (OTC) medications, despite being instructed not to include use of OTCs in
answering the questions.

•

A new methamphetamine module was added for the QFT. In the comparison data,
methamphetamine questions were included in the core stimulants module, and
methamphetamine was considered to be part of the general category of stimulants.
The comparison data also included methamphetamine questions in the noncore
special drugs module that were used in determining methamphetamine use, stimulant
misuse, and most recent use (or misuse).

•

The focus of the questions for specific prescription drugs in the QFT was on the past
12 months and on the lifetime period in the comparison data. In addition, QFT
respondents first were asked a series of screening questions about any use of specific
prescription drugs in the past 12 months (i.e., use or misuse) or any lifetime use if
they did not report past year use. QFT respondents were asked about misuse in the
past year of any of the specific prescription drugs they reported using in that period.
In contrast, respondents in the comparison data were asked about misuse of specific
prescription drugs in the lifetime period, and questions about more recent misuse
34

applied to the general categories (e.g., past year or past month misuse of any
tranquilizers).
•

Questions in the QFT about use of stimulants with a needle were moved from the
noncore special drugs module to the core stimulants module. These QFT questions
applied only to use of stimulants with a needle in the past 12 months or past 30 days.

•

New questions about methamphetamine dependence or abuse were added to the
substance dependence and abuse module.

•

Sections of the interview in the comparison data that were interviewer-administered
were self-administered in the QFT (e.g., health insurance, income).

For the special editing procedures described in this section that were relevant to the comparison
data, additional details are provided in the editing and coding section of the 2010 methodological
resource book (Kroutil et al., 2012a).
3.3.4.1

Smokeless Tobacco

Editing of the QFT variables for smokeless tobacco use followed the general principles
discussed previously. In the comparison data, variables for any smokeless tobacco use were
created based on the data for use of snuff and use of chewing tobacco. The following principles
were applied in creating the smokeless tobacco variables in the comparison data:
•

Respondents who answered "no" to both questions about lifetime use of snuff and
chewing tobacco were classified as nonusers of smokeless tobacco.

•

Respondents who answered "no" to one of the questions about lifetime use of snuff or
chewing tobacco but who did not know or refused to report whether they ever used
the other type of smokeless tobacco were assigned a missing value for lifetime use or
nonuse of smokeless tobacco. Lifetime use or nonuse was statistically imputed
(see Section 3.4).

•

Respondents who reported use of either snuff or chewing tobacco at a minimum were
classified as lifetime users of smokeless tobacco. The period of most recent use was
determined from respondents' answers to the questions about most recent use of the
smokeless tobacco products.

•

In general, the report of most recent use of either snuff or chewing tobacco was
chosen for the variable pertaining to most recent smokeless tobacco use. If relevant
variables for one of the smokeless tobacco products had missing data, special codes
were assigned for use in statistically imputing a final period of most recent use. For
example, if a respondent reported last using snuff more than 30 days ago but within
the past 12 months but did not know when he or she last used chewing tobacco, the
variable for most recent use of smokeless tobacco was assigned a code to indicate that
the respondent logically last used at some point in the past 12 months. This
respondent could have been a past month user of any smokeless tobacco if he or she
used chewing tobacco in the past month. A specific period of most recent use was
statistically imputed.

3.3.4.2

More Recent Use for General Drug Categories and Specific Drugs

For hallucinogens in the QFT and comparison data and for pain relievers and stimulants
in the comparison data, consistency checks were triggered if respondents reported more recent
use of a specific type of drug in the category (e.g., Ecstasy) than they reported for their last use
of any drug in the category (e.g., any hallucinogen). As noted in the general principles (Kroutil et
al., 2012a), the editing procedures took into account data from these consistency checks. For
example, suppose a respondent reported last using any hallucinogen more than 30 days ago but
within the past 12 months and last using Ecstasy within the past 30 days. If this respondent
reported in the consistency checks that his or her last use of any hallucinogen also was in the past
30 days, the edited variable for most recent hallucinogen use reflected this change, and the data
were no longer inconsistent.
However, if the data continued to indicate more recent use of a specific drug than for use
of any drug in the category despite the respondent being given the opportunity to resolve the
inconsistency, then the editing procedures logically inferred more recent use of any drug in the
category. For example, if a respondent's answers continued to indicate last use of Ecstasy in the
past 30 days and last use of any hallucinogen more than 30 days ago but within the past
12 months, the respondent was logically inferred to have last used any hallucinogen in the past
30 days; a special code was assigned to the variable for most recent hallucinogen use to indicate
that this edit had been performed.
In the comparison data, these principles applied to editing of the variable for most recent
use of any hallucinogen relative to reports of most recent use of LSD, PCP, or Ecstasy.
Questions in the comparison data about most recent use of the hallucinogens ketamine, DMT,
AMT, or 5-MeO-DIPT ("Foxy"), and Salvia divinorum were in the supplemental special drugs
module and therefore were not used in editing the data for most recent use of any hallucinogen.
For the QFT, questions about these three additional hallucinogens were moved from the special
drugs module to the core hallucinogens module. The hallucinogens module for the QFT also
included consistency checks that were triggered if respondents reported more recent use of any
of these three hallucinogens than was reported for most recent use of any hallucinogen.
Consequently, data on most recent use of these additional hallucinogens, along with data on most
recent use of LSD, PCP, or Ecstasy, were used in editing the data for most recent use any
hallucinogen in the QFT. The same principles applied to editing the QFT data when respondents
reported more recent use of any of these additional hallucinogens compared with reports of most
recent use of any hallucinogen.
The cocaine and crack cocaine modules in the QFT and comparison data did not include
consistency checks if respondents reported more recent use of crack cocaine than for cocaine in
general. Consequently, data on the most recent use of crack were used to infer more recent use of
cocaine in general, as per the example discussed previously for hallucinogens. Additional issues
related to the editing of the data for most recent use of methamphetamine and misuse of any
stimulant are discussed in the methamphetamine section.

3.3.4.3

" OTHER, Specify" Data for Drugs

For hallucinogens and inhalants in all three datasets and for prescription drugs in the
comparison data, questions about lifetime use (or misuse) were logically inferred to be "yes" if
respondents originally did not report use of these drugs in the direct questions but reported them
in the "OTHER, Specify" data. Additional details about these editing procedures for the
comparison data are provided in the editing and coding section of the 2010 methodological
resource book (Kroutil et al., 2012a).
As noted previously, QFT respondents were asked about use of specific prescription
drugs in the past year and misuse of those drugs that they used in the past year. Consistent with
the structure of questions in the comparison data, QFT respondents who reported that they
misused "any other" drug in the category (e.g., any other prescription pain reliever) in the past
12 months could specify past year misuse of up to five individual drugs. If a respondent reported
past year use of a specific drug (e.g., the generic pain reliever hydrocodone), did not report
misusing the drug in the past year, but then reported it in the "OTHER, Specify" data, the
response in the edited variable for past year misuse was logically inferred to be "yes"; no editing
needed to be done for the variable pertaining to any use in the past year. If the respondent
reported misuse of a particular drug in the "OTHER, Specify" data but did not report using it in
the past year (and therefore was not asked about past year misuse of the drug), both the variable
for any past year use and the variable for past year misuse of that drug were assigned codes to
indicate that the respondent used and misused that drug in the past year.
3.3.4.4

OTC Misuse

One way that persons can misuse prescription drugs is by taking them without having
their own prescription. Because OTC drugs by definition are available without a prescription,
respondents in both the QFT and comparison data were instructed not to include OTCs when
answering the prescription drug questions. For the comparison data, respondents who specified
that they misused OTCs were logically inferred never to have misused any of the prescription
drugs in the overall category (e.g., pain relievers) if they reported never misusing any of the
specific prescription drugs in the gate questions and the only other "prescription" drugs they
reported misusing in their lifetime were OTCs.
A similar principle was applied to the editing of the QFT prescription drug data, except
that these edits focused on misuse of prescription drugs in the past year. Specifically, QFT
respondents were logically inferred not to have misused any of the prescription drugs in that
category in the past year if they did not use or misuse any of the drugs in that category except for
"any other" drug, and the only other drugs they reported misusing in the past year were OTCs.
However, no editing was done to the screening question about any use of other drugs in that
category in the past year (which resulted in respondents being routed to the question about
misuse of any other drug in the category) because respondents could have used other
prescription drugs in the past year that they did not misuse.
3.3.4.5

Methamphetamine Use

Editing of the methamphetamine variables in the comparison data took into account the
placement of the methamphetamine questions in the core stimulants module. Specifically, the
37

CAI program for the comparison data required answers to questions about methamphetamine use
to be consistent with answers to related questions about misuse of stimulants in general. As noted
previously, for example, a consistency check was triggered if respondents reported more recent
use of methamphetamine than they reported for the most recent misuse of any prescription
stimulant. In keeping with the general editing principles for the comparison data, the editing
procedures took answers in these consistency checks into account when creating the edited
methamphetamine and general stimulant variables. Furthermore, the editing procedures for the
comparison data required misuse of any stimulant always to be as recent as or more recent than
the last use of methamphetamine.
Since 2005, questions about methamphetamine use have been included in the
supplemental special drugs module for respondents who did not previously report
methamphetamine use in the core stimulants module. Because methamphetamine in recent years
has typically been manufactured illegally rather than through the legitimate pharmaceutical
industry, methamphetamine users may fail to report their use when questions about the drug are
asked in the context of questions about misuse of stimulants that are (or have been) available by
prescription in the United States. Data from these methamphetamine questions in the special
drugs module were used to create "core-plus-noncore" (CPN) measures of lifetime and most
recent use of methamphetamine in the comparison data. For example, if respondents in the
comparison data did not report methamphetamine use in the core stimulants module because they
did not think of it as a prescription drug but they reported use in the special drugs module, their
reports for their most recent use of methamphetamine in the special drugs module were
incorporated into the CPN variable for most recent use. In addition, if these respondents who did
not think of methamphetamine as a prescription drug reported more recent use of
methamphetamine in the special drugs module than they reported for their most recent misuse of
any stimulant, the edited CPN variable for most recent stimulant misuse reflected the special
drugs data for methamphetamine.
Editing of the QFT data for lifetime and most recent use of methamphetamine followed
the general principles described in Section 3.3.3. Because the methamphetamine use questions in
the QFT were placed in a module separate from questions about misuse of prescription
stimulants, the edited data for use or most recent use of methamphetamine were not required to
be consistent with data from the core stimulants module. For example, QFT respondents could
report lifetime use of methamphetamine without reporting misuse of prescription stimulants in
their lifetime; these responses were not considered to be inconsistent.
3.3.4.6

Prescription Drugs

drug) were asked whether they ever used any prescription drug in that category. Respondents
who endorsed use of one or more specific prescription drugs in the past 12 months in the
screener questions were asked about past year misuse of the prescription drugs that they reported
using in that period. If respondents reported misuse of any prescription drugs in a given category
in the past 12 months, they were asked whether they misused any prescription drugs in that
category in the past 30 days. Thus, unlike the 12-month questions, misuse in the past 30 days
applied only to the broad prescription drug category rather than to specific prescription drugs.
If respondents used prescription drugs in a given category in the past 12 months but they did not
report misuse, they were asked about lifetime misuse of any prescription drugs in that category.
Similarly, respondents who reported lifetime but not past year use of any prescription drugs in
that category were asked about lifetime misuse. Thus, as for misuse in the past 30 days, lifetime
misuse applied only to the broad prescription drug category.
Consistent with the general editing principles described in Section 3.3.3, an important
component of editing the prescription drug variables in the QFT involved assignment of codes to
indicate when respondents were not asked questions that were not applicable. For example, if
respondents did not report use of a particular drug in the past 12 months, then the corresponding
edited variables for misuse of that drug in the past 12 months were assigned codes to indicate
that the questions did not apply.
As an exception to the general principle of retaining missing values when respondents
answered a question governing a skip pattern as "don't know" (DK) or "refused" (REF), QFT
respondents who had responses of DK or REF in their screener data for past year use of specific
prescription drugs and reported no past year use of other drugs in the screener could answer the
question about lifetime use of any prescription drugs in the category as "no." In this situation, the
report of no lifetime use of any prescription drug in the category took precedence over the
responses of DK or REF in editing the QFT prescription drug variables. Similarly, if respondents
answered one or more questions about past year misuse of specific prescription drugs as DK or
REF and answered questions about past year misuse of other prescription drugs as "no" (or were
skipped out of the past year misuse questions because they did not report any past year use of
these drugs), they were asked whether they ever misused any prescription drug in that category
in their lifetime. Again, if these respondents answered this lifetime misuse question as "no," this
report overruled the responses of DK or REF in editing the past year misuse variables.
Because of the structure of the prescription drug questions in the QFT, respondents were
not asked a specific question for their most recent misuse of any prescription drugs in that
category. Rather, variables for most recent misuse of prescription pain relievers, tranquilizers,
stimulants, and sedatives were created from respondents' answers to questions about misuse of
any prescription drug in the category in the past 30 days, misuse of specific prescription drugs in
a given category in the past 12 months, and lifetime misuse of any prescription drug in the
category. The following general principles were applied in creating the variables for most recent
use of any prescription drugs in a given category in the QFT data:

•

Respondents who reported misuse of prescription drugs9 in the past 30 days were
classified as having last misused prescription drugs in the past 30 days.

•

Respondents who reported misuse of one or more specific prescription drugs in the
past 12 months were classified as having last misused prescription drugs more than
30 days ago but within the past 12 months, provided that they answered "no" to the
question about misuse in the past 30 days.

•

Respondents who reported lifetime (but not past year) misuse of prescription drugs
were classified as having last misused prescription drugs more than 12 months ago,
provided that (a) they answered all applicable questions about misuse of specific
prescription drugs in the past 12 months as "no"; or (b) they reported any use of
prescription drugs in their lifetime and they explicitly reported that they did not use
any prescription drugs in that category in the past 12 months.

•

Respondents who reported that they never used or never misused prescription drugs
were classified as never having misused prescription drugs. (The coding of the
variables for most recent use did not distinguish between respondents who never used
prescription drugs and lifetime users who never misused prescription drugs.)

3.3.4.7

Needle Use

Editing of the needle use data in the QFT and comparison samples principally involved
assignment of the appropriate codes to indicate when respondents were not asked questions that
did not apply. For example, respondents were not asked the needle use questions for a given drug
(e.g., cocaine) if they reported in the corresponding core module that they never used the drug.
Respondents also were not asked the follow-up questions in the special drugs module about most
recent use of a drug with a needle if they used the drug in their lifetime but never used a needle
to inject it.
In addition, "OTHER, Specify" data on use of other drugs with a needle were used to edit
needle use data within the special drugs module. For example, if respondents did not report using
cocaine with a needle but they specified it as some "other" drug they used with a needle, the edits
inferred that these respondents used cocaine with a needle at some point in their lifetime.
Consistent with editing in the core modules (and with general principles of editing
described previously), however, data on needle use from the special drugs module were not used
in editing drug use data from the corresponding core module. For example, if respondents
reported more recent use of cocaine with a needle in the special drugs module compared with
their reports of most recent use of cocaine (including any reports of crack cocaine), the editing
procedures for both the QFT and comparison data did not resolve this inconsistency.
As noted previously, the needle use questions for stimulants in the QFT were moved
from the special drugs module to the core stimulants module. In addition, the questions about use
of stimulants with a needle applied to stimulants that respondents misused in the past 12 months.
Even if the editing procedures allowed editing of core data based on data in the special drugs
9

In this text, "prescription drugs" refers to any prescription drugs in a given category (e.g., any prescription
pain reliever).

module, reports of lifetime use of prescription stimulants with a needle in the "OTHER, Specify"
data for special drugs could not be used to infer past year use of stimulants with a needle or to
infer past year misuse of specific stimulants in the core stimulants module.
3.3.4.8

Methamphetamine and Prescription Stimulant Dependence or Abuse

In the comparison data, because methamphetamine was grouped together with other
stimulants, comparison data respondents who reported past year methamphetamine use were
asked questions about dependence or abuse for prescription stimulants. The QFT included
questions about dependence and abuse for methamphetamine that were separate from questions
about dependence and abuse for prescription stimulants that were misused in the past 12 months.
Consequently, QFT respondents who reported methamphetamine use in the past year but who
did not report past year misuse of prescription stimulants were asked dependence and abuse
questions for methamphetamine but were not asked corresponding questions for stimulants.
QFT respondents who reported past year use of methamphetamine and past year misuse
of prescription stimulants were asked both sets of dependence and abuse questions. For these
respondents, no editing was done to the methamphetamine dependence or abuse variables based
on respondents' answers to questions about corresponding symptoms of dependence or abuse for
prescription stimulants. Similarly, no editing was done to the stimulant dependence or abuse
variables based on respondents' answers to questions about corresponding symptoms of
dependence or abuse for methamphetamine.
3.3.4.9

Interviewer-Administered versus Self-Administered Data

The basic content of the QFT variables for marital status, employment status, health
insurance, and income underwent little or no change relative to the variables in the comparison
data, except that they were self-administered instead of being interviewer-administered.
Consequently, little or no change to the editing procedures for these variables in the QFT were
required relative to the procedures for editing these variables in the comparison data. Editing of
these variables in all three datasets principally involved assignment of codes to indicate when it
could be determined unambiguously that respondents were not asked questions that did not
apply.

3.4

Imputation Procedures

3.4.1

Overview of Imputation Procedures

This section describes the imputation procedures that were implemented for the 2012
QFT data and the two comparison datasets―the 2012 quarters 3 and 4 main study data and the
2011 main study data. The advantages of performing imputation include the following:
(1) reducing bias due to differential nonresponse, (2) allowing all cases to be used for analysis,
and (3) improving the quality of data at the subdomain level. The small QFT sample sizes and
the limited amount of time for imputation make it difficult to implement the standard NSDUH
imputation methods due to sparse donor pools. Because the comparison of the QFT data with the
main study data was performed at a fairly aggregate level, a simple mean imputation procedure
satisfies the needs of the QFT and could be implemented within the short time period for the
QFT. The two main study comparison datasets―2012 quarters 3 and 4 and all quarters from
41

2011―were imputed using the same approach. One of the simplest methods of imputing for
missing data is to replace each missing value with the weighted mean of the observed values for
a variable within a class of respondents containing the respondent with the missing value. This
method provides an unbiased estimate of the overall variable mean either if the probability of the
value being missing is the same for every respondent in a class or if values within a class are not
related to their probabilities of being missing. If neither of these conditions holds, the estimated
variable mean after imputation is biased, but the bias is likely less than if no imputation had
taken place, which is equivalent to treating the entire sample as a single imputation class.
3.4.2

Imputation Methodology

Variables that were imputed include demographics, health insurance, income, and
recency of drug use. The noncore variables associated with drug abuse were not imputed.10
Table 3.1 lists the variables that were imputed for each of the three sets of data. As was done in
the main study, imputation indicators were created for each imputed variable. For the drug use
variables, three variables indicating lifetime use, past year use, and past month use were created
from the imputed recency of use variables. In addition to misuse, the QFT instrument asked
about any use of prescription drugs. These variables were not imputed for this analysis.
Questions about lifetime and past month use of OxyContin® were not included in the QFT
instrument; therefore, only the past year indicator variable for OxyContin® misuse was imputed
for the QFT data. The QFT instrument contained separate modules for methamphetamine and
prescription stimulants. Therefore, an additional recency of misuse of stimulants excluding
methamphetamine was imputed for the QFT only. For the 2011 and 2012 comparison data, the
CPN measures for methamphetamine and misuse of stimulants were created to compare with the
combined stimulants and methamphetamine variables in the QFT.
For categorical variables (including both nominal and ordinal), the weighted percentage
for each variable level within an imputation class was used to impute the missing values.
Imputation classes were based, where possible, on categorical age (12 to 17 years, 18 to 25 years,
and 26 years and older), gender, and four-level race (white, black, Hispanic, and other). For the
race variable imputation, only age group and gender were used to create imputation classes. For
the continuous variable WELMOS—number of months on welfare—the weighted mean was
computed within an imputation class, then used to impute the missing values. Weighted means
were computed using PROC DESCRIPT from SUDAAN® (RTI International, 2008), and
weighted percentages were computed using PROC CROSSTAB. As an example, assume that
among white females aged 26 or older the marital status variable has a complete case weighted
distribution as follows: married (65 percent), widowed (10 percent), divorced (15 percent), and
never married (10 percent). If 20 cases within this imputation class have missing values, then
13 cases would be imputed as married, 2 cases as widowed, 3 cases as divorced, and 2 cases as
never been married. Rounding was used when the percentages did not result in exact numbers of
cases and when there were fewer records with missing values than there were levels of the
10

Variables that regularly undergo imputation, but did not for the QFT include the following: roster
variables; roster pair variables; Hispanic group and immigrant status; personal income variables; "old method"
insurance variables; daily cigarette use, cigar, pipe, chewing tobacco, and snuff use variables; core-only stimulants
and methamphetamine use variables; 12-month and 30-day frequency of drug use variables; age at first drug use
variables; and nicotine dependence variables.

Table 3.1 Imputed Variables
Race
Hispanic Indicator
Marital Status

Demographic Variables
Education
Employment Status

Income Variables
Food Stamps
Welfare Payments
Welfare Services
Number of Months on Welfare
Health Insurance Variables
Private Health Insurance
Medicaid/CHIP (Children's Health Insurance
Program)
Other Health Insurance
Medicare
CHAMPUS (Civilian Health and Medical Program
of the Uniformed Services)
Drug Use Variables
Inhalant Use
Cigarette Use
Marijuana Use
Smokeless Tobacco Use
Core plus Noncore Stimulant Misuse
Alcohol Use
Core plus Noncore Stimulant Misuse, Excluding
Binge Alcohol Use (Past Month Only)
Methamphetamine Use (QFT Only)
Cocaine Use
Core plus Noncore Methamphetamine Use
Crack Use
Pain Reliever Misuse
Hallucinogen Use
OxyContin® Misuse (QFT: Past Year Only)
LSD Use (Lysergic Acid Diethylamide)
Sedative Misuse
PCP Use (Phencyclidine)
Tranquilizer Misuse
Ecstasy Use
Family Income
Wages
Social Security
Supplemental Security

Heroin Use
QFT = Questionnaire Field Test.

imputed value. For example, an imputation class for the four-level recency variable may have
had only two records requiring imputation. In these cases, the distribution of imputed cases may
have looked very different from the distribution of complete cases. However, the rounding
algorithm was such that the distribution of imputed values would match the weighted distribution
of complete values in expectation.
Imputation was occasionally restricted to a few categories when partial information about
the nonrespondent was known or in order to maintain consistency with other variables. For
example, when imputing employment status, if the nonrespondent was known to be employed,
but the level of employment (full-time or part-time) was not known, the weighted percentages
were calculated among employed respondents in each imputation class, and imputation was
restricted to full- or part-time employment.
In a few cases, the imputation class contained only nonrespondents. When this happened,
imputation classes were collapsed by race, then by gender, then by age until at least one
43

respondent was in the imputation class. For example, Exhibit 3.1 shows the imputation classes
for the 12- to 17-year-old age category. If the nonrespondent was a 15-year-old, Hispanic, and
female, and no respondents were in the imputation class for 12- to 17-year-old, Hispanic
females, that class would be merged with the class containing 12- to 17-year-old females of other
races. Collapsing would continue up the hierarchy until at least one respondent was in the
imputation class. Continuing the example above, it may have been necessary to collapse all races
or both genders. Note that if collapsing was necessary, care was taken to collapse as few classes
as possible. As shown in Exhibit 3.1, if collapsing of the race categories was only necessary
among females, parallel collapsing was not done among males. Similarly, if collapsing was only
necessary among 12- to 17-year-olds, no collapsing was done within the other age categories
(Exhibit 3.2).
Exhibit 3.1

Collapsing Imputation Classes: Race

Exhibit 3.2

Collapsing Imputation Classes: Race and Gender

3.5

Weighting Procedures

3.5.1

Overview of Weighting Procedures

Estimates and measures of data quality from the 2012 QFT sample were compared with
those from the 2012 main study during the same quarters (2012 quarters 3 and 4) and from the
full year for the 2011 main study. Analysis weights for those three samples needed to be
developed for the QFT analysis. This section discusses the methods used to develop sample
weights for the 2012 QFT analysis.
For some research questions (Question 1a to 1c), QFT respondents were compared with
the 2012 quarters 3 and 4 and the 2011 NSDUH respondents. To increase the efficiency of the
comparisons by removing the impact of differences between the demographic characteristics of
the three samples caused by random sampling and then exacerbated by nonresponse,
nonresponse-adjusted weights were calibrated for the QFT sample and 2012 quarters 3 and 4
main study sample to distributions of demographic variables from the 2011 sample. Instead of
the full process (Chen et al., 2013) used in developing 12-month analysis weights, where five
adjustment steps were implemented, a shortened process was used similar to producing weights
for the 6-month detailed tables. That is, the design weights were computed for both the QFT
sample and the 2012 quarters 3 and 4 main sample in a manner consistent with 2011 NSDUH
weighting procedures. The design weights were then adjusted for nonresponse at the dwelling
unit and person level, followed by a poststratification adjustment where nonresponse-adjusted

weights were further poststratified to the sum of the analysis weights from the 2011 NSDUH
sample for selected demographic domains.
The final analysis weight was used to calculate the weighted distributions for the 2011
comparison data. For the 2012 QFT and the 2012 quarters 3 and 4 main study data, the final
analysis weights were not available; therefore, the preliminary analysis weights were used
instead. This preliminary weight was created from the person-level sample design weights
adjusted to account for nonresponse at the household level.
3.5.2

Weighting Procedures

This section discusses in detail the procedures used to develop the analysis weights for
the three samples and summarizes the distribution of the QFT analysis weights.
3.5.2.1

2011 NSDUH Sample Weights

The analysis weights (ANALWT) for the 2011 NSDUH sample had 15 weight
components, and among them 5 were adjustment factors at both the dwelling and person levels
(Chen et al., 2013). The generalized exponential model (GEM) (Folsom & Singh, 2000) was
used for the nonresponse and poststratification adjustments within nine model groups
corresponding to nine census divisions. ANALWT is the product of all 15 weight components.
After removing respondents from Hawaii and Alaska, as well as interviews completed
using the Spanish-version questionnaire (LANGVER=2), analysis weights for the remaining
respondents in the 2011 NSDUH were used for the 2012 QFT analyses. The domain-level sums
of the ANALWT for these retained respondents were used as control totals in the
poststratification for the 2012 QFT sample and the 2012 quarters 3 and 4 main study sample as
discussed in the following section.
3.5.2.2

2012 Quarters 3 and 4 Main Study Sample Weights

Design-based weights were computed for the 2012 quarters 3 and 4 main study sample in
a manner consistent with standard NSDUH weighting procedures. To facilitate timely
completion of the QFT analyses, quarter 4 screenings and interviews completed after December
2, 2012, were considered nonrespondents. After December 2, 2012, an additional 2,909
screenings and 604 interviews were completed that would have been included in the 2012
quarters 3 and 4 main study comparison data had the December 2, 2012, cutoff date not been
implemented. The nonresponse adjustments at both the dwelling unit level (DUNR) and person
level (PRNR) for the 2012 quarters 3 and 4 main study sample were similar to those used to
develop the regular 6-month analysis weights. However, the person-level poststratification
(PRPS) for the 2012 quarters 3 and 4 main study sample was different from the regular 6-month
analysis weights, where the nonresponse-adjusted weights were adjusted to the census population
estimates. For the QFT analyses, the person-level poststratification adjusted the weights to match
ANALWT sums for eligible respondents from the 2011 NSDUH sample. GEM was used to
implement all three adjustment steps.

The final analysis weights for the 2012 quarters 3 and 4 main study sample were the
product of various design weights and three adjustment factors. The various design weights were
as follows:
•

inverse probability of selecting census tracts;

•

inverse probability of selecting segments;

•

quarter segment weight adjustment;

•

subsegmentation inflation adjustment;

•

inverse probability of selecting dwelling units;

•

added/subsampled dwelling unit adjustment;

•

dwelling unit sample release adjustment;

•

dwelling unit-level nonresponse adjustment;

•

inverse probability of selecting a person from a dwelling unit;

•

person-level nonresponse adjustment; and

•

person-level poststratification adjustment.

The three adjustment factors were as follows:
•

Dwelling Unit-Level Nonresponse Adjustment (DUNR). One model was used to
account for the failure to obtain screening interviews from eligible dwelling units.
The proposed variables in the model are listed below, and they were all kept in the
final model.
– State,
– quarter,
– population density (metropolitan statistical area [MSA], ≥ 1 million; MSA,
< 1 million; non-MSA, urban; non-MSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of Hispanics in a segment (CH: > 50 percent; 10 to 50 percent;
< 10 percent),
– segment combined median rent and housing value (CV: 1st quintile; 2nd quintile;
3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
47

– CV * CB, and
– CV * CH.
•

Person-Level Nonresponse Adjustment (PRNR). One model was used to adjust
person-level nonresponse, and the proposed variables in the model are listed below
(they were all kept in the final model):
– State,
– quarter,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban;
non-MSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of Hispanics in a segment (CH: > 50 percent; 10 to 50 percent;
< 10 percent),
– segment combined median rent and housing value (CV: 1st quintile; 2nd quintile;
3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB,
– CV * CH,
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender,
– Hispanicity * gender,
– age group * Race3 * Hispanicity,
– age group * Race3 * gender,
48

– age group * Hispanicity * gender, and
– Race3 * Hispanicity * gender.
•

Person-Level Poststratification Adjustment (PRPS). The respondents in the 2012
quarters 3 and 4 main sample from Hawaii and Alaska and interviews completed with
the Spanish-version questionnaire were removed before the PRPS. One model was
used to force the weights of the 2012 quarters 3 and 4 main study sample to sum up to
the ANALWT totals for eligible respondents in the 2011 NSDUH by the following
proposed demographic domains (all proposed variables were kept in the final model):
– State,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 to 64; 65 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender,
– Hispanicity * gender,
– age group * Race3 * Hispanicity,
– age group * Race3 * gender,
– age group * Hispanicity * gender, and
– Race3 * Hispanicity * gender.

3.5.2.3

2012 QFT Sample Weights

Design-based weights for the 2012 quarters 3 and 4 QFT sample were computed in a
manner consistent with standard NSDUH weighting procedures. The three adjustment steps,
DUNR, PRNR, and PRPS, were implemented in a similar fashion as for the 2012 quarters 3 and
4 main study sample weights using GEM. The differences were that fewer variables in the GEM
models were used to develop QFT sample weights because of the relatively small 2012 QFT
sample.
The final analysis weights for the 2012 quarters 3 and 4 QFT sample were the product of
various design weights and three adjustment factors. The various design weights were as follows:
•

inverse probability of selecting QFT State sampling (SS) regions;

•

inverse probability of selecting census tracts;

•

inverse probability of selecting segments;
49

•

quarter segment weight adjustment;

•

subsegmentation inflation adjustment;

•

inverse probability of selecting dwelling units;

•

added or subsampled dwelling unit adjustment;

•

dwelling unit sample release adjustment;

•

dwelling unit-level nonresponse adjustment;

•

inverse probability of selecting a person from a dwelling unit;

•

person-level nonresponse adjustment; and

•

person-level poststratification adjustment.

The three adjustment factors were as follows:
•

Dwelling Unit-Level Nonresponse Adjustment (DUNR). One model was used to
account for the failure to obtain screening interviews from eligible dwelling units.
The variables in the model are listed below, and some two-way interactions of
segment-level variables (CO, CH, CB, and CO) were collapsed in order to get a
convergent model:
– State,
– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban;
non-MSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of Hispanics in a segment (CH: > 50 percent; 10 to 50 percent;
< 10 percent),
– segment combined median rent and housing value (CV: 1st quintile; 2nd quintile;
3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB, and
– CV * CH.

•

– State,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban;
non-MSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB: > 50 percent; 10 to
50 percent; < 10 percent),
– percent of Hispanics in a segment (CH: > 50 percent; 10 to 50 percent;
< 10 percent),
– segment combined median rent and housing value (CV: 1st quintile; 2nd quintile;
3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB,
– CV * CH,
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender, and
– Hispanicity * gender.
•

Person-Level Poststratification Adjustment (PRPS). One model was used to force the
weights of the 2012 quarters 3 and 4 QFT sample to sum up to ANALWT totals for
eligible respondents in the 2011 NSDUH by the following proposed demographic
domains (all variables were kept in the final model):
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 to 64; 65 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),

– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender, and
– Hispanicity * gender.
3.5.3

Distribution of QFT Analysis Weights

The distribution of analysis weights for the 2011 NSDUH sample, 2012 quarters 3 and 4
QFT sample, and 2012 quarters 3 and 4 main study sample are summarized in Table 3.2.
Table 3.2 Weight Distribution of QFT Analysis Weights

Statistics
100% Maximum
99%
95%
90%
75% Quarter 3
50% Median
25% Quarter 1
10%
5%
1%
0% Minimum
n
Mean
Sum of Weights
Unequal Weighting Effect (UWE)1
1

2011 NSDUH
Sample Weights
108,117
28,632
14,867
9,707
3,942
1,501
715
320
196
63
1
65,928
3,688
243,124,072
3.5156

2012 Quarters 3
and 4 QFT Sample
Weights
790,075
481,574
323,750
270,961
152,927
83,482
48,820
35,068
30,391
10,123
4,131
2,044
118,945
243,124,072
1.7172

2012 Quarters 3 and 4
Main Study Sample
Weights
125,076
53,068
30,590
21,027
8,486
3,378
1,729
870
540
237
24
31,213
7,789
243,124,073
3.0279

UWE measures the variation in weights.

3.5.4

Creation of Variance Estimation Strata and Replicates

The nature of the stratified, clustered sampling design of the NSDUH main study and
QFT samples requires that the design structure be taken into consideration when computing
variances of survey estimates. Key nesting variables were created for the QFT and main study
comparison samples to capture explicit stratification and to identify clustering.
To allow for comparisons between the QFT and main study samples, a common set of
stratification and clustering variables were defined. Because State sampling (SS) regions serve as
strata for the main study samples and as primary sampling units (PSUs) for the QFT sample,

there was no direct way of capturing the covariance between the samples and using the entire
main study sample. Instead, the approach used for the 1999 paper-and-pencil interviewing
(PAPI) and CAI mode analysis was followed in developing a design structure that could be used
to simultaneously analyze all three samples (Gfroerer, Eyerman, & Chromy, 2002). Steps in the
process were as follows:
•

Within the QFT sampling strata (census regions), variance strata were generally
formed by assigning two sequential QFT selected SS regions to the same variance
strata on the sorted sampling frame. Each sampled SS region was then assigned to a
replicate (1 or 2). However, there were three QFT SS regions per variance strata for
three randomly selected strata. This was necessary because an odd number of QFT SS
regions were selected in three of the census regions. Within these three strata, the
third SS region was randomly assigned to either replicate 1 or replicate 2. This led to
a total of 105 QFT variance strata, with two replicates per strata.

•

Using the sorted QFT sampling frame of SS regions, the main study SS regions not
selected for the QFT were assigned to QFT sampling strata sequentially, in
accordance with the assignments of selected QFT SS regions. These assignments kept
the number of SS regions per strata as equal as possible given the distribution of QFT
sampled SS regions within the sorted SS region frame. For SS regions not selected for
the QFT sample, the original replicate assignments of either replicate 1 or replicate 2
were maintained. A further discussion of the assignment of main study replicates can
be found in the 2011 sample design report (Morton et al., 2012).

Although this approach to design structure variables does not fit the main study perfectly,
it does capture the total variance and allows for taking advantage of any covariance induced by
the overlapping SS regions between the samples.

3.6

Data File Preparation

Three data files were prepared for the QFT analysis. In order to evaluate the QFT, two
comparison data files for 2011 and 2012 were created based on main study cases.
3.6.1

QFT Data File

The QFT data file was comprised of interviews conducted from September 1, 2012,
through November 3, 2012. No Spanish interviews or interviews in Alaska and Hawaii were
conducted, and these data underwent the normal data quality checks and telephone verification.
The final analysis data file resulted in 2,044 respondents.
3.6.2

2011 Comparison Data File

The 2011 comparison data file was created from the 2011 main study analysis file.
The full set of respondents was subset down to 65,928 by excluding Spanish cases as well as
interviews conducted in Alaska and Hawaii.

3.6.3

2012 Comparison Data File

The 2012 comparison data file was created using most of the 2012 main study cases
worked in quarters 3 and 4. As was done for the 2011 comparison file, Spanish interviews,
Alaska interviews, and Hawaii interviews were also excluded. In order to allow time for analysis
under the QFT schedule, the 2012 comparison file only contains cases with a completed
interview as of December 2, 2012. Because this time frame was prior to completing verification
on the full 2012 main study sample, some decisions were made to exclude cases undergoing field
verifications at the time, based on the following criteria:
•

Cases completed by quarter 3 or 4 field interviewers (FIs) found to have been
falsified as of December 2, 2012. In addition to cases that were determined to have
some form of falsification, cases completed by these same FIs were dropped
whenever it could not be determined whether the interview was actually completed or
whether informed consent was completed. This second set of cases usually resulted
from being unable to contact the respondent.

•

Quarter 4 cases that were worked by FIs whose work was still being field verified as
of December 2, 2012.

•

Quarter 3 interviews for FIs whose work was still being field verified as of December
2, 2012. If falsification of quarter 4 work was found, previous 2012 work completed
by these FIs needed to be field verified.

Interviews scheduled for telephone verification that were not finalized by close of
business on December 2, 2012, and did meet any of the exclusion criteria above were included in
the 2012 quarters 3 and 4 comparison data file. The resulting 2012 quarters 3 and 4 comparison
data file contained 31,213 interviews (see Table 3.3).
Table 3.3 Data Files Created for the 2012 Questionnaire Field Test Analyses
Data File
QFT
2011 Comparison
2012 Comparison

3.7

Data Analysis Issues

3.7.1

Primary Analytic Goals

Data Collection Period
9/1/2012 – 11/3/2012
1/1/2011 – 12/31/2011
7/1/2012 – 12/2/2012

Number of Respondents
2,044
65,928
31,213

The primary goal of the QFT was to measure patterns of effects on NSDUH estimates
due to changes in the protocol planned for the 2015 redesign. Decisions about changes in the
questionnaire and protocol have, for the most part, already been made. As a result, the focus of
the statistical analysis is the measurement of how the collective set of protocol changes could
affect key NSDUH estimates—overall and by the three major age groups—when the new
protocol is implemented in 2015. The QFT sample size was not large enough to permit
quantitative assessments of the impact of individual changes in the protocol because such
analyses would require dedicated samples for assessing each change, unless it were assumed that
54

the effects of changes are uncorrelated with each other—that the effect of each change on
outcomes of interest is independent of the effects of all other changes. To carry out such a design
to estimate the effects of each protocol change would be prohibitively costly and infeasible.
Also, the resources needed to carry out such extensive testing would have risked having an
impact on the main 2012 survey estimates by affecting the availability of interviewers to work on
the main study.
3.7.2

Comparison with Current NSDUH Data

Most of the analyses in this report compare estimates from the 2012 QFT with estimates
from the 2011 NSDUH and quarters 3 and 4 from the 2012 NSDUH. Comparisons between the
2012 QFT and quarters 3 and 4 from 2012 allow for estimating the effects of the overall protocol
change over approximately the same time period, with the QFT being conducted during the last
month of quarter 3 and the first month of quarter 4 of the main study.
An additional point of comparison is provided by estimates from the 2011 NSDUH. Use
of the 2011 NSDUH provides additional sample with which to compare against the QFT sample.
Rather than relying solely on comparisons with the 2012 quarters 3 and 4 sample, survey
designers felt it would be informative to compare estimates from the QFT with the 2011 NSDUH
sample as well. In a manner of speaking, the 2011 NSDUH provides another data point with a
larger sample size with which to compare the QFT. This provides assurance that differences in
estimates between the QFT and the 2012 quarters 3 and 4 sample are not unique to that
comparison. Also, comparisons between the 2011 NSDUH and the QFT sample can be viewed
as an early indicator of what differences in estimates might emerge between the 2014 NSDUH
and the 2015 NSDUH, the first year of the fully implemented redesign. Use of the 2011 NSDUH
as a comparison point assumes that differences in NSDUH estimates between 2011 and 2012 are
generally small.
In addition to comparisons of estimates between the QFT and 2012 quarters 3 and 4 and
2011 NSDUH samples, two other analyses were carried out to rule out potential confounders of
comparisons between the QFT and 2012 quarters 3 and 4 samples.
3.7.2.1

Comparison of QFT Data and 2012 Quarters 3 and 4 Data to Assess "Seasonality"
Effects on Estimates

In principle, the 2012 QFT and comparison cases from quarters 3 and 4 of the 2012
NSDUH generally cover the same time period, late summer and early fall. Estimates from
quarter 3 in the 2012 NSDUH were compared with estimates from quarter 4 in the 2012 NSDUH
as a check for differences in estimates between the two quarters. Because the QFT was
conducted in only 2 months out of the 6 months of quarters 3 and 4, there was concern that the
particular months chosen for the QFT sample (September and October 2012) may not be
representative of all 6 months in the last half of 2012, particularly if there were differences in
estimates between quarters 3 and 4. If there were underlying changes in behavior taking place
throughout the 6 months of quarters 3 and 4, the ideal design would involve collecting data using
the redesigned instrument throughout the same time period. However, due to resource
constraints, the QFT sample could not be fielded in all of the 6 months of quarters 3 and 4 in
2012. If estimates in quarter 3 were similar to those in quarter 4 and there was no underlying

change in the behaviors estimated by NSDUH, the time point at which the QFT was fielded
would be of less concern.
In other words, given that the QFT was conducted during a 2-month period, an
assumption needed to be made that the net impact of the protocol changes will not be different
for the 2 months of the field test than for the other 10 months of the year. This does not imply an
assumption that drug and mental health reporting cannot be affected by the month of data
collection, only that the net impact of the changes in the redesign protocol will not be affected by
the particular month or season chosen.
For the estimates shown in Tables I-1 to I-12 in Appendix I, Tables J-1 to J-12 in
Appendix J, and Tables K-1 to K-4 in Appendix K, significance tests were carried out for
differences between quarters 3 and 4. Overall, very few significant differences emerged,
suggesting that comparisons between estimates from the quarters 3 and 4 2012 NSDUH sample
and the QFT sample are not affected by detectable seasonal differences.
3.7.2.2

Comparison of QFT Outcomes with 2012 Quarters 3 and 4 Main Study Outcomes
to Assess Level of Effort Effects on Estimates

Another concern with comparing estimates from the QFT sample with those from the
2012 quarters 3 and 4 main study sample is that that field efforts for NSDUH are not distributed
equally across the 3 months of each quarter. Typically, many interviews are conducted in the
first month of each quarter, fewer are conducted in the second month, and fewer still in the third
month. First-month responses may be systematically different from third-month responses, given
differences in the level of effort required to screen households and interview selected
respondents in the first month versus the third month. Analyses of the relationship between
indicators related to length of time in the field, such as interview visits, have shown that
respondents requiring more calls to complete the interview may have higher self-reported rates
of illicit drug use (Biemer & Wang, 2006). Given that the QFT data were collected in a
compressed, 2-month time, reduced calling effort may lead to differences between estimates
from the QFT sample and the 2012 quarters 3 and 4 sample.
To investigate this possibility, estimates for a limited number of measures were examined
by the number of visits required to complete the interview for both the QFT and 2012 quarters 3
and 4 samples. Indicators examined were lifetime use measures of hallucinogens, inhalants, any
prescription drug misuse, pain reliever misuse, tranquilizer misuse, and past year and past month
serious psychological distress (SPD). Overall, there was little evidence of strong differences in
estimates by the number of visits and little indication that any such patterns differed by sample.
3.7.3

Comparisons with Other Survey Data

Estimates from the QFT sample were also compared with estimates from other
appropriate sources, such as those shown in Appendix C from the 2010 NSDUH national
findings report (Center for Behavioral Health Statistics and Quality [CBHSQ], 2011). Such
comparisons provide relevant evidence on the effects of changes in the NSDUH data collection
protocol. As noted in the 2010 national findings report, the results of such comparisons may be
difficult to interpret given differences between NSDUH and other data collection systems in a

number of areas, including the population of interest, sample design, data collection periods,
screening and interviewing protocols, and estimation procedures.
The following data sources were used in these comparisons:
•

National Ambulatory Medical Care Survey (NAMCS) and the hospital outpatient
clinic component of the National Hospital Ambulatory Medical Care Survey
(NHAMCS), which mention specific prescription psychotherapeutic drugs;

•

National Health Interview Survey (NHIS), which includes the numbers of doctor
visits, income, education, and cellular telephone coverage; and

•

National Health and Nutrition Examination Survey (NHANES), which includes direct
measures of height and weight.

Results for these comparisons are discussed in Chapter 9.

4. Data Collection Outcomes and Data
Quality Assessment
4.1

Overview of Data Collection and Data Quality Outcomes

This chapter presents a variety of indicators used to assess the quality of the 2012
Questionnaire Field Test (QFT) data. Where feasible and appropriate, data quality outcomes for
the 2012 QFT data are compared with the 2011 main study comparison data and the 2012
quarters 3 and 4 main study comparison data. Examining these indicators identifies the potential
impact of the questionnaire and protocol revisions implemented for the QFT on data quality
when the partial redesign is implemented in 2015.
Section 4.2 presents unit response rates for all three datasets, including both screening
and interviewing response rates. Section 4.3 details imputation rates for variables that were
common to the 2011 comparison data, the 2012 quarters 3 and 4 comparison data, and the QFT
data, while Section 4.4 details missing data rates for new or revised items in the QFT
questionnaire. Section 4.5 presents interview timing results, including comparisons among the
three datasets where appropriate. Section 4.6 describes other data quality indicators for the new
prescription drug modules included in the 2012 QFT questionnaire.

4.2

Unit Response Rates

4.2.1

Screening Response Rates (SRRs) and Number of Visits for Completed and
Noncompleted Screenings

The screening response rate (SRR) is the total number of completed screenings divided
by the total eligible dwelling units. The eligible dwelling units are computed by subtracting the
number of sample dwelling units (SDUs) not eligible to be included in the National Survey on
Drug Use and Health (NSDUH) from the total number of SDUs. Ineligibles include vacant units,
those that are not a primary residence, units that are not dwelling units, group quarters units
(GQUs) listed as housing units (HUs), HUs listed as GQUs, only military units, listing errors,
other ineligibles, and those SDUs where the residents will live there less than half of the quarter.
SRRs were calculated for the 2011 main study comparison sample, the 2012 quarters 3
and 4 main study comparison sample, and the 2012 QFT sample. Response rates for 2011 were
calculated using final 2011 main study data. Data for Alaska and Hawaii were removed to make
rates more comparable with the 2012 QFT. SRRs for the 2012 comparison sample were
calculated based on the preliminary results for quarters 3 and 4 of 2012, with Alaska and Hawaii
removed.11 Screeners associated with field interviewers (FIs) that were subject to field
verification at the time the preliminary data were obtained were considered nonrespondents to
minimize the risk of introducing falsified cases onto the comparison file. Because the 2012
11

Main study screenings completed in Spanish were retained and treated as completions on both the 2011
comparison file and the 2012 comparison file because it was difficult to determine which screenings were completed
in English and which screenings were completed in Spanish.

comparison data were based on the data collected through December 2, 2012, quarter 4
screenings completed after that date were considered nonrespondents for the purposes of the
QFT analysis. Similarly, any screener completions that were later recoded as screener
incompletes (e.g., resulting from falsification detected after December 2, 2012) were treated as
screener completions for the purposes of the QFT analysis.
Table 4.1 lists the sample totals and the national screening and interviewing response
rates for the 2011 main study comparison file, the 2012 quarters 3 and 4 main study comparison
file, and the 2012 QFT. This table provides both the weighted and unweighted screening and
interviewing response rates for each sample. The weighted screening response rates for the 2011
main study comparison file, the 2012 quarters 3 and 4 main study comparison file, and the 2012
QFT were 87.00, 81.77, and 83.58 percent, respectively.
Table 4.1 Screenings, Interviews, and Response Rates for the 2011 Main Study, 2012 Quarters 3
and 4 Main Study, and 2012 Questionnaire Field Test

Selected Dwelling Units
Eligible Dwelling Units
Eligibility Rate
Complete Screenings
Screening Response
Rate
Selected Persons
Completed Interviews
Interviewing Response
Rate
Overall Response Rate

2011 Main Study
Comparison Sample
211,227
174,912
Unweighted
Weighted
82.81%
83.14%
152,333
Unweighted
Weighted

2012 Quarters 3 and 4
Main Study Comparison
Sample
104,618
86,755
Unweighted Weighted
82.93%
83.22%
71,540
Unweighted
Weighted

87.09%
87.00%
86,155
65,928
Unweighted Weighted
76.52%
70.46%
Unweighted
Weighted
66.64%
61.30%

82.46%
81.77%
39,354
31,213
Unweighted
Weighted
79.31%
74.58%
Unweighted
Weighted
65.40%
60.98%

2012 Questionnaire Field
Test
5,358
4,623
Unweighted
Weighted
86.28%
86.24%
3,837
Unweighted
Weighted
83.00%

83.58%
2,823
2,044
Unweighted
Weighted
72.41%
69.04%
Unweighted
Weighted
60.09%
57.71%

One difference between the QFT sample and the two main study samples that could not
be accounted for is the language used to complete the screenings. For the main study, the
screenings could be completed in English or Spanish, and the FI had the ability to switch
languages as needed. As a result, the language used for each screening could not be determined.
For the QFT, no Spanish version of the screening interview was available, so households that
could not complete the screening in English were treated as nonrespondents. This factor reduced
the QFT's SRR relative to the other two samples. An additional factor that could have affected
SRRs was improvements to the QFT lead letter, which were expected to improve SRRs.
Whenever feasible, FIs were required to make at least four callback visits to dwelling
units when attempting to complete the screening and interviewing. In general, callbacks
continued to be made as long as the field supervisor (FS) felt there was a chance that the
screening or the interview could be completed in a cost-effective manner. In some cases, more
60

than 10 visits were made to complete a screening or interview. Table 4.2 presents data on the
number of visits made for successfully completed screenings in each of the three samples. The
overall pattern of visits for completed screenings in the QFT sample looked quite similar to the
2011 and 2012 quarters 3 and 4 comparison samples, with only slight differences for a few
categories. These distributions indicate there were no significant differences in the number of
screenings required to complete household screenings in the QFT data collection compared with
the 2011 and 2012 quarters 3 and 4 comparison samples.
For comparison, Table 4.3 presents data on the number of visits made to dwelling units
that were not successfully screened for each of the three samples. This further comparison allows
for an assessment of how the QFT screening results might have differed from the 2011 and 2012
quarters 3 and 4 comparison samples. For each category of the number of visits made, the
noncompleted screenings in the 2011 and 2012 quarters 3 and 4 comparison samples looked
quite similar. The overall pattern of visits for noncompleted screenings in the QFT sample
looked similar to the 2011 and 2012 quarters 3 and 4 comparison samples. The proportion of
noncompleted screeners appeared to differ for two categories of visits made:
•

A lower proportion of noncompleted QFT screenings were in the single visit category
compared with the 2011 and 2012 quarters 3 and 4 comparison samples.

•

A greater proportion of noncompleted QFT screenings were in the 10 or more
category.

Overall, these results do not suggest systematic differences in the distribution of noncompleted
screeners in each category of visits made for the QFT sample relative to the 2011 and 2012
quarters 3 and 4 comparison samples.
4.2.2

Interview Response Rates (IRRs) and Number of Visits for Completed and
Noncompleted Screenings

The interviewing response rate (IRR) is the number of completed interviews divided by
the total number of eligible respondents chosen through screening. If there are any ineligible
respondents (younger than 12 or actually in the military), these are subtracted from the total. For
the 2012 main study comparison sample, interview status was determined based on the
December 3, 2012, preliminary results. Cases that were undergoing field verification at that time
were treated as nonrespondents. Cases that resulted in interview completions after this date were
treated as nonrespondents, and cases that were classified as interviews on this date that were later
recoded as noncompletes were treated as completed interviews for the purposes of the QFT
analysis. To make the 2011 main study and the 2012 quarters 3 and 4 main study more
comparable with the QFT, interviews completed in Spanish were treated as eligible
nonrespondents and interviews completed in Alaska and Hawaii were excluded.
Table 4.4 presents the unweighted and weighted IRRs by age group for all three samples.
The weighted IRRs for the 2011 main study, the 2012 quarters 3 and 4 main study, and the 2012
QFT were 70.46, 74.58, and 69.04 percent, respectively.

Table 4.2 Number of Visits Made for Completed Screenings for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012
Questionnaire Field Test

Visits
1
2
3
4
5-9
10+
Unknown
Total

2011 Main Study Comparison Sample
Cumulative
Screenings
Percent
Percent
54,976
36.09
36.09
31,785
20.87
56.96
19,143
12.57
69.53
12,090
7.94
77.47
24,707
16.22
93.69
9,632
6.32
100.00
0
0.00
100.00
152,333
100.00
100.00

2012 Quarters 3 and 4 Main Study
Comparison Sample
Cumulative
Screenings
Percent
Percent
26,634
37.23
37.23
14,842
20.75
57.98
8,768
12.26
70.24
5,691
7.95
78.19
11,321
15.82
94.01
4,283
5.99
100.00
1
0.00
100.00
71,540
100.00
100.00

2012 Questionnaire Field Test Sample
Cumulative
Screenings
Percent
Percent
1,442
37.58
37.58
853
22.23
59.81
471
12.28
72.09
299
7.79
79.88
577
15.04
94.92
195
5.08
100.00
0
0.00
100.00
3,837
100.00
100.00

62
Table 4.3 Number of Visits Made for Noncompleted Screenings for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012
Questionnaire Field Test

Visits
1
2
3
4
5-9
10+
Unknown
Total

2011 Main Study Comparison Sample
Noncompleted
Cumulative
Screenings
Percent
Percent
11,500
19.51
19.51
10,847
18.40
37.91
6,698
11.36
49.27
4,890
8.30
57.57
12,922
21.92
79.49
12,089
20.51
100.00
0
0.00
100.00
58,946
100.00
100.00

2012 Quarters 3 and 4 Main Study
Comparison Sample
Noncompleted
Cumulative
Screenings
Percent
Percent
6,249
18.88
18.88
6,253
18.89
37.77
3,643
11.01
48.78
2,721
8.22
57.00
7,337
22.17
79.17
6,849
20.69
100.00
0
0.00
100.00
33,097
100.00
100.00

2012 Questionnaire Field
Test Sample
Noncompleted
Screenings
Percent
220
14.46
259
17.03
187
12.29
141
9.27
359
23.60
355
23.40
0
0.00
1,521
100.00

Cumulative
Percent
14.46
31.49
43.78
53.05
76.65
100.00
100.00
100.00

Table 4.4 Interview Response Rates, by Age, for the 2011 Main Study, 2012 Quarters 3 and 4 Main
Study, and 2012 Questionnaire Field Test (QFT)

Age
Category
12-17
18-25
26-34
35-49
50-64
65+

2011
82.80
78.46
71.46
70.21
68.71
64.09

Unweighted Percent
2012
Quarters 3
and 4
84.50
80.84
76.65
73.31
72.89
68.07

QFT
82.05
75.71
68.07
66.25
67.25
63.68

2011
82.70
77.69
69.86
68.68
68.30
62.96

Weighted Percent
2012
Quarters 3
and 4
84.59
80.76
76.27
72.97
72.46
67.35

QFT
82.25
75.26
68.91
66.32
66.78
63.48

NOTE: Cases where respondents provided only the age category 50+ were counted in the 65+ category.

Table 4.5 presents data on the number of visits made for completed interviews for the
QFT sample and the 2011 and 2012 quarters 3 and 4 comparison samples. Similar to the results
on the number of visits for completed screenings, the proportion of completed interviews in each
category of the number of visits followed a similar pattern across the three samples. The
proportion of completed interviews appeared to differ across the three samples for two
categories:
•

A lower proportion of completed QFT interviews was in the single visit category.
This difference indicates that QFT interviews were less likely to be completed "on the
spot," that is, at the same time the household was screened and one or more
respondents were selected.

•

The proportion of interviews in the 10 or more visits category was greatest for the
2011 comparison sample, somewhat less for the 2012 quarters 3 and 4 comparison
sample, and lower still for the 2012 QFT sample.

Beyond these two differences, the distribution of completed interviews by the number of visits
made for the QFT sample was similar to the 2011 and 2012 quarters 3 and 4 comparison
samples.
Table 4.6 presents results for the number of visits made for selected respondents who
were not successfully interviewed for each of the three samples. This further comparison allows
for an assessment of how the QFT interviewing results might have differed from the 2011 and
2012 quarters 3 and 4 comparison samples. In general, the proportion of noninterviews for the
QFT sample across the categories of visits followed a similar pattern as the 2011 and 2012
quarters 3 and 4 comparison samples. A few categories appeared to differ meaningfully between
the QFT sample and the 2011 and 2012 quarters 3 and 4 comparison samples:
•

About 4 percent more QFT noninterviews were in the three-visit category.

•

About 5 percent more QFT noninterviews were in the five- to nine-visit category.

•

The proportion of QFT noninterviews in the 10-visit or more category was about
4 percent lower than the 2012 quarters 3 and 4 sample and about 8 percent lower than
the 2011 comparison sample.

Table 4.5 Number of Visits Made for Completed Interviews for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012
Questionnaire Field Test

Visits
1
2
3
4
5-9
10+
Unknown
Total

2011 Main Study Comparison Sample
Completed
Cumulative
Interviews
Percent
Percent
36.23
23,884
36.23
22,784
34.56
70.79
7,506
11.39
82.18
3,478
5.28
87.46
5,992
9.09
96.55
2,174
3.30
99.85
110
0.17
100.00
65,928
100.00
100.00

2012 Quarters 3 and 4 Main Study
Comparison Sample
Completed
Cumulative
Interviews
Percent
Percent
11,583
37.11
37.11
10,767
34.50
71.61
3,516
11.26
82.87
1,636
5.24
88.11
2,731
8.75
96.86
910
2.92
99.78
70
0.22
100.00
31,213
100.00
100.00

2012 Questionnaire Field
Test Sample
Completed
Interviews
Percent
700
34.25
726
35.52
243
11.89
126
6.16
192
9.39
55
2.69
2
0.10
2,044
100.00

Cumulative
Percent
34.25
69.77
81.66
87.82
97.21
99.90
100.00
100.00

Table 4.6 Number of Visits Made for Noncompleted Interviews for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012
Questionnaire Field Test

Visits
1
2
3
4
5-9
10+
Unknown
Total

2011 Main Study Comparison Sample
Noncompleted
Cumulative
Interviews
Percent
Percent
1,163
6.30
6.30
2,219
12.02
18.32
1,916
10.38
28.70
1,704
9.23
37.93
6,079
32.93
70.86
5,350
28.98
100.00
0
0.00
100.00
18,485
100.00
100.00

2012 Quarters 3 and 4 Main Study
Comparison Sample
Noncompleted
Cumulative
Interviews
Percent
Percent
525
7.90
7.90
899
13.54
21.44
720
10.84
32.28
645
9.71
41.99
2,181
32.84
74.83
1,636
24.63
100.00
0
0.00
100.00
6,642
100.00
100.00

2012 Questionnaire Field Test Sample
Noncompleted
Cumulative
Interviews
Percent
Percent
41
5.10
5.10
108
13.43
18.53
115
14.30
32.83
77
9.58
42.41
300
37.31
79.72
162
20.15
100.00
0
0.00
100.00
804
100.00
100.00

Overall, these results indicate some differences in the distribution of noninterview cases by the
number of visits made for the QFT sample relative to the 2011 and 2012 quarters 3 and 4
comparison samples. The greatest difference was that a greater proportion of QFT noninterviews
fell within categories for three to nine visits, while a greater proportion of 2011 and 2012
quarters 3 and 4 cases fell within both the single visit category and the 10 or more visit
categories.

4.3

Imputation Rates for Common 2011 Comparison Data, 2012 Quarters 3
and 4 Comparison Data, and QFT Variables

Another indicator of the quality of the QFT data is the proportion of cases for which
imputation was required prior to using specific variables for analysis. For the QFT data, 2011
comparison data, and 2012 quarters 3 and 4 comparison data, records with missing data were
subject to the same imputation procedures. However, when the values of other nonmissing
variables could be used to determine the value of the missing variable, the value was "logically
assigned" instead of imputed.
Tables 4.7a through 4.7d provide rates of imputation and logical assignment that selected
variables underwent in processing the 2011 comparison data, the 2012 quarters 3 and 4
comparison data, and the QFT data. (Section 3.4 in Chapter 3 describes these imputation
procedures.) These tables include the following columns for the variables of interest:
•

respondents in domain (unweighted),

•

unweighted frequency of records imputed or logically assigned, and

•

weighted percentage (relative to their domain size) of records imputed or logically
assigned.

A "domain" in this context is the set of respondents who received a value other than a skip code
for the imputation-revised variable of interest. In other words, a domain is the subset of
respondents for whom the variable of interest is relevant or applicable. In Table 4.7b, for
example, only among respondents aged 15 or older (the domain) is it relevant to ask about
employment status (the variable of interest). Unless otherwise specified, the domain for each
variable includes all respondents. For comparing imputation rates, Tables 4.7a through 4.7d also
include an indicator for whether observed differences in imputation rates between either the 2011
or 2012 quarters 3 and 4 comparison data and the imputation rates for the QFT data are
statistically significant at the 0.05 level.
As Table 4.7a shows, the weighted percentages of cases that were either imputed or
logically assigned in all three datasets were generally low for substance use variables, with
nearly all of the percentages at or below 0.5 percent. Weighted percentages of imputed or
logically assigned cases for the following substance use variables appeared to be slightly higher
for the QFT dataset than for the 2011 and 2012 quarters 3 and 4 comparison datasets:
•

lysergic acid diethylamide (LSD) recency,

•

Ecstasy recency,

Table 4.7a Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test:
Substance Use Variables

Variable (Domain)
Marijuana Recency
Cocaine Recency
Crack Recency
Heroin Recency
Hallucinogen Recency
LSD Recency
PCP Recency
Ecstasy Recency
Inhalant Recency
Cigarette Recency
(Lifetime Cigarette
Users)
Smokeless Tobacco
Recency
Alcohol Recency
Binge Alcohol Use
(Past Month
Alcohol Users)
Pain Reliever Recency
OxyContin® Recency4
OxyContin® Past
Year Use4
Tranquilizer Recency
Sedative Recency
Core Plus Noncore
Stimulant Recency
Core plus Noncore
Methamphetamine
Recency
Stimulants Excluding
Methamphetamine
Recency4

2011 Comparison Data1
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage
65,928
91
0.1
65,928
65
0.1a
65,928
35
0.1a
65,928
37
0.0
65,928
357
0.4
65,928
98
0.2
65,928
74
0.1
65,928
96
0.1
65,928
219
0.2

2012 Comparison Data1,2
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage
31,213
43
0.1
31,213
24
0.1a
31,213
8
0.0
31,213
18
0.0
31,213
151
0.3a
31,213
35
0.1
31,213
38
0.1
31,213
50
0.1
31,213
93
0.1

Respondents
in Domain
2,044
2,044
2,044
2,044
2,044
2,044
2,044
2,044
2,044

QFT1,3
Unweighted
Frequency
2
0
0
1
24
8
2
12
11

Weighted
Percentage
0.1
0.0
0.0
0.0
1.0
0.5
0.2
0.6
0.5

33,754

0.1

15,474

0.0

1,091

0.1

65,928
65,928

70
77

0.1
0.1

31,213
31,213

19
30

0.1
0.1

2,044
2,044

2
1

0.0
0.0

29,249
65,928
65,928

739
473
291

2.2
0.5a
0.3

13,988
31,213
31,213

346
242
147

2.4
0.5a
0.2

925
2,044
N/A

20
34
N/A

1.6
1.4
N/A

N/A
65,928
65,928

N/A
159
191

N/A
0.1
0.2

N/A
31,213
31,213

N/A
70
90

N/A
0.2
0.1

2,044
2,044
2,044

11
11
12

0.7
0.5
0.3

65,928

216

0.2

31,213

0.2

2,044

0.5

65,928

0.1

31,213

0.1

2,044

0.1

N/A

2,044

0.4

LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine; QFT = Questionnaire Field Test.
a

Difference between estimate and QFT estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
QFT data collected from September 1 through November 3, 2012.
OxyContin® recency was only available for the 2011 and 2012 comparison files; the QFT only asked about past year use. Stimulant misuse excluding methamphetamine was only available on the QFT.

2
3
4

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table 4.7b Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test:
Selected Demographic and Socioeconomic Variables

Variable (Domain)
Detailed Race: 15 Levels
Hispanic or Latino Origin
Education Level
Marital Status (Age 15+)
Employment Status
(Age 15+)
Employment Status
(Age 18+)

2011 Comparison Data1
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage
65,928
2,406
3.2
65,928
93
0.1
65,928
3
0.0
54,955
12
0.0a

2012 Comparison Data1,2
Respondents
in Domain
31,213
31,213
31,213
26,036

Unweighted
Frequency
1,218
78
3
1

Respondents
in Domain
2,044
2,044
2,044
1,779

Unweighted
Frequency
96
2
0
8

Weighted
Percentage
3.3
0.0
0.0
0.4

0.1a

26,036

0.1a

1,779

0.4

43,509

0.1a

20,748

0.1a

1,503

0.4

Difference between estimate and QFT estimate is statistically significant at the 0.05 level.

Weighted
Percentage
3.7
0.1
0.0
0.0a

54,955

QFT = Questionnaire Field Test.

QFT1,3

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table 4.7c Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test:
Health Insurance Variables

Variable (Domain)
Respondent Has Health
Insurance
Type of Insurance
Private
Medicare
Military Health Care:
CHAMPUS, TRICARE,
CHAMPVA, VA
Medicaid/CHIP
Other (Respondents
without Private Health
Insurance, Medicare,
Medicaid/CHIP, or
Military Health Care)

2011 Comparison Data1
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

2012 Comparison Data1,2
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

Respondents
in Domain

QFT1,3
Unweighted
Frequency

Weighted
Percentage

65,928

494

0.4a

31,213

315

0.5a

2,044

1.2

65,928
65,928

411
222

0.3a
0.2

31,213
31,213

263
132

0.4a
0.3

2,044
2,044

32
19

0.8
0.7

65,928
65,928

223
511

0.2a
0.4

31,213
31,213

144
328

0.2a
0.5

2,044
2,044

17
29

0.7
1.0

11,149

244

1.2

5,197

149

1.6

431

4.3

CHIP = Children's Health Insurance Program; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Department of
Veteran's Affairs; QFT = Questionnaire Field Test; VA = Department of Veteran's Affairs.
a

Difference between estimate and QFT estimate is statistically significant at the 0.05 level.

2
3

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table 4.7d Cases Imputed or Logically Assigned for the 2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test:
Income Variables

Variable (Domain)
Total Family Income
> or < $20,000
Total Family Income –
Finer Categories
Source of Family Income
Social Security or
Railroad Retirement
Payments
Wages
Public Assistance
Supplemental Security
Income

Food Stamps
Welfare/Job Placement/
Child Care
Number of Months on
Welfare (Family
Receives Public
Assistance or
Welfare/Job
Placement/Child
Care)

2011 Comparison Data1
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage

2012 Comparison Data1,2
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage

2,768

3.8

31,213

1,375

3.9

2,044

4.1

65,928

7,614

14.4

31,213

3,696

14.5

2,044

265

14.1

65,928
65,928
65,928

646
192
521

0.7
0.2a
0.5a

31,213
31,213
31,213

343
105
254

0.6
0.3a
0.4a

2,044
2,044
2,044

33
38
37

1.1
1.2
1.1

65,928

913

0.9a

31,213

461

0.8a

2,044

1.6

2,044

0.6

65,928

267

0.3

31,213

167

0.3

65,928

380

0.4

31,213

193

0.3a

2,044

0.7

4,807

204

3.5a

2,155

118

5.5

160

9.3

Difference between estimate and QFT estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.
2

Weighted
Percentage

65,928

QFT = Questionnaire Field Test.
a

Respondents
in Domain

QFT1,3
Unweighted
Frequency

•

inhalants recency,

•

pain reliever recency,

•

tranquilizer recency, and

•

core-plus-noncore (CPN) stimulant recency (see Table 3.1 in Section 3.4.2).

These differences in rates of imputation or logical assignment for substance use variables
between the QFT dataset and the 2011 and 2012 quarters 3 and 4 comparison datasets were
generally small, from 0.3 percent for multiple variables to 0.9 percent for pain relievers recency.
For one substance use variable, percent binge alcohol use among past month alcohol users, the
imputation or logical assignment rate for the QFT dataset (1.6 percent) appeared to be slightly
lower than the 2011 comparison dataset (2.2 percent) and the 2012 quarters 3 and 4 comparison
dataset (2.4 percent).
The weighted percentages of cases that were either imputed or logically assigned in all
three datasets were relatively low for most of the demographic variables presented in Table 4.7b.
These rates were similar across all three datasets for the first three variables—detailed race,
Hispanic or Latino origin, and education level. Although the imputation rates for the other three
demographic variables—marital status for those aged 15 or older, employment status for those
aged 15 or older, and employment status for those aged 18 or older—were all below 0.5 percent,
the imputation rates for these three variables were significantly higher in the QFT data than in
the 2011 and 2012 comparison data. The QFT imputation rates were 0.4 percent for each of these
three variables. For the 2011 and 2012 comparison data, the imputation rates were 0.1 percent or
lower.
In Table 4.7c, the weighted percentages of cases that were either imputed or logically
assigned in all three datasets were somewhat higher on average compared with the substance use
and demographic variables. These percentages ranged from 0.2 percent for military health care in
the 2011 and 2012 quarters 3 and 4 comparison data to 4.3 percent for other health care in the
QFT data. The weighted percentages of imputed or logically assigned cases were highest for the
other health care variable, and this rate appeared to be higher for the QFT dataset compared with
the 2011 comparison data (1.2 percent) and the 2012 quarters 3 and 4 data (1.6 percent). In
addition, the weighted percentages for whether the respondent has health insurance appeared to
be higher for the QFT dataset (1.2 percent) compared with the 2011 comparison data
(0.4 percent) and the 2012 quarters 3 and 4 data (0.5 percent). The health insurance question was
among the set of items moved from computer-assisted personal interviewing (CAPI) to audio
computer-assisted self-interviewing (ACASI) in the QFT instrument, so the higher imputation
rates observed could have resulted from QFT respondents being more likely to not answer this
question. This outcome could also provide an explanation for other questionnaire items moved
from CAPI to ACASI in the QFT instrument. (See Section 4.4 for the complete results and a
discussion of item missingness rates in the QFT data and the 2011 and 2012 quarters 3 and 4
comparison data.)
Weighted percentages for cases that were either imputed or logically assigned in all three
datasets for income variables are shown in Table 4.7d. Not surprisingly, the weighted
percentages for some of the income variables were relatively high, such as the total family
income's finer categories. For all three datasets, the rates for total family income's finer
70

categories were similar, and all were greater than 14 percent. With the two exceptions of (1) total
family income greater or less than $20,000 and (2) total family income's finer categories, the
rates of imputation or logical assignment appeared to be slightly higher for the QFT dataset than
for the 2011 and 2012 quarters 3 and 4 comparison datasets. The variables presented in
Table 4.7d were all based on questionnaire items moved from CAPI to ACASI administration
for the QFT. Section 4.4 presents and discusses the higher item missingness rates observed for
most of these items when administered in ACASI in the QFT versus CAPI in the 2011 and 2012
comparison data.

4.4

Missing Data Rates for New or Revised QFT Items and Comparisons of
Missing Data Rates for Moved QFT Items with 2011 and 2012
Quarters 3 and 4 Comparison Data

4.4.1

Missing Data Rates for New, Revised, or Moved Items in the QFT Questionnaire

To examine data quality among survey items in the QFT questionnaire that are new
questions or have been revised in some way, this section discuses item missingness rates. The
QFT items met one of the following criteria:
•

the question is new to the instrument,

•

the question or response options have been significantly revised, or

•

the question has been moved from one part of the questionnaire to another, including
either being moved to a different module or moved from CAPI to ACASI
administration.

Table C-1 in Appendix C provides missing data rates for these new, revised, or moved items for
the QFT sample. Missing data rates were relatively low for most of these QFT items, but some
items did produce relatively high missingness rates. For example, health insurance items QHI08,
QHI09, and QHI10—which ask about private health insurance plans covering treatment
for alcohol abuse or alcoholism, drug abuse, or mental or emotional problems—had the highest
missing data rates, from 20 to 25 percent of respondents. However, these high missingness rates
for these items administered via ACASI in the QFT were actually significantly lower than the
missingness rates for these same items administered via CATI in the 2011 and 2012 quarters 3
and 4 comparison data.12 Two questions asking about family income level also had missingness
rates of nearly 10 percent, such as items QI22 and QI23a, which ask about total combined family
income. A few core substance use items showed relatively high missingness rates, but the
number of respondents answering each of these questions was very low, producing an unreliable
estimate for extrapolating missingness rates to the larger NSDUH target population.
4.4.2

Missing Data Rates for Items Moved in the QFT Questionnaire for the QFT Data,
2011 Comparison Data, and 2012 Quarters 3 and 4 Comparison Data

Although valid comparisons of missing data rates for new or revised QFT items between
the QFT data and the two comparison datasets were not possible, items that were moved from
12

For a detailed summary of data quality issues related to moving specific sets of questionnaire items from
CAPI to ACASI, see Appendix R.

CAPI to ACASI administration and were not otherwise changed can be compared. These
comparisons allow assessment of whether item nonresponse rates appear likely to change once
these items are administered via ACASI in the main study beginning in 2015. As Table 4.8
indicates,13 missingness rates for many of these moved items were similar when administered in
ACASI for the QFT as when these were administered by CAPI in the 2011 and 2012 quarters 3
and 4 comparison files. However, some moved items had lower missingness rates in the QFT
data, and several other items had higher missingness rates in the QFT data. This section provides
details on selected moved items that produced statistically different missingness rates than either
the 2011 or 2012 quarters 3 and 4 comparison data.
Two sets of items administered in ACASI for the QFT had significantly lower
missingness rates than in the 2011 and 2012 quarters 3 and 4 comparison files, including the
following:
•

Items QD43, QD44, QD46, QD47, and QD48 on workplace alcohol and drug use
policies had lower item missingness rates in the QFT data compared with the 2011 or
2012 quarters 3 and 4 comparison data. Missingness rates for all of these items were
quite similar in the 2011 and 2012 quarters 3 and 4 comparison data, but
proportionately lower in the QFT data.

•

Items asking about health insurance coverage for treatment of alcohol abuse (QHI08),
drug abuse (QHI09), and mental health issues (QHI10) had lower item missingness
rates in the QFT data than in the 2011 or 2012 quarters 3 and 4 comparison data.
Missingness rates for QHI08 and QHI09 were about 44 or 45 percent in the 2011 and
2012 quarters 3 and 4 comparison data, but only about 27 or 28 percent in the QFT
data. Similarly, the missingness rate for QHI10 was about 27 percent in the 2011 and
2012 quarters 3 and 4 comparison data, but only about 18 percent in the QFT data.

Several types of items that were moved to ACASI for the QFT had significantly higher
missingness rates than the CAPI items from the 2011 and 2012 quarters 3 and 4 comparison
samples, including the following:
•

Item QD07 on marital status, item QD13 on moving home in the past year, and item
QD13a on State of residence 1 year ago all had significantly higher item missingness
rates in the QFT data than in the 2011 or 2012 quarters 3 and 4 comparison data.
Missingness rates for these three items were close to 0.0 percent in the 2011 or 2012
quarters 3 and 4 comparison data, but ranged from 0.4 to 0.8 percent in the QFT data.

•

Item QD19 on full-time or part-time student status, item QD20 on missing school due
to illness or injury, and item QD21 skipping school days all had significantly higher
item missingness rates in the QFT data than in the 2011 or 2012 quarters 3 and 4
comparison data. Missingness rates for these three items were close to 0.0 percent in
the 2011 or 2012 quarters 3 and 4 comparison data, but ranged from 1.0 to 1.5 percent
in the QFT data.

To aid in its readability, the multipage Table 4.8 appears in its entirety at the end of this discussion in
Section 4.4.2.

•

The item asking about work at a job or business at any time in the past week, QD26,
had a significantly higher item missingness rate in the QFT data than in the 2011 or
2012 quarters 3 and 4 comparison data. Missingness rates for this item were close to
0.0 percent in the 2011 or 2012 quarters 3 and 4 comparison data, but 0.2 percent in
the QFT data.

•

Several items that ask about recent employment history, missing workdays, size of
employing organization, and related issues—QD33, QD36, QD38, QD39a, QD40,
QD41, and QD42—had significantly higher item missingness rates in the QFT data
than in the 2011 or 2012 quarters 3 and 4 comparison data. Missingness rates for all
of these items were quite similar in the 2011 and 2012 quarters 3 and 4 comparison
data, but proportionately higher in the QFT data.

•

The item asking about private health insurance coverage, QHI06, had a significantly
higher item missingness rate in the QFT data than in the 2011 comparison data.
Missingness rates for this item were 0.3 percent in the 2011 comparison data and
0.4 percent in the 2012 quarters 3 and 4 comparison data, but 0.7 percent in the QFT
data. Although the missingness rate was about twice as high in the QFT data as in the
2012 quarters 3 and 4 comparison data, this difference was not statistically
significant.

•

Most of the items asking about receipt of various sources of income or participation
in government assistance programs—QI03N, QI05N, QI07N, QI08N, and QI10N—
had significantly higher item missingness rates in the QFT data than in the 2011 or
2012 quarters 3 and 4 comparison data. Missingness rates for all of these items were
quite similar in the 2011 and 2012 quarters 3 and 4 comparison data, but
proportionately higher in the QFT data.

•

Two items on personal income levels—QI20N and QI21A—had significantly higher
item missingness rates in the QFT data than in the 2011 or 2012 quarters 3 and 4
comparison data. The missingness rates for both items were close to 2 percent in the
2011 and 2012 quarters 3 and 4 comparison data, but were 3.7 percent for QI20N and
4.6 percent for QI21A in the QFT data.

The higher missingness rates observed for these sets of items that were moved from CAPI to
ACASI administration in the QFT instrument were not anticipated. All else being equal, higher
item missingness rates could potentially reduce or limit the quality of the data collected in
ACASI mode. For this reason, missingness rates for these sets of items will be closely monitored
in the 2013 Dress Rehearsal (DR) data to see whether similar patterns continue. A detailed report
on the impact of the higher item missingness rates observed for several items moved from CAPI
to ACASI administration in the QFT instrument is included as Appendix R in this report.
In addition, Section 9.4 in Chapter 9 provides the results of further analyses of several of these
items, including benchmarking against other Federal surveys with similar target populations.
These additional analyses provide further evidence on the potential impact on data quality for
selected items moved to ACASI when the redesigned protocol is implemented in 2015.

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data

Instrument Item
Ever used ketamine? (LS01i5)
Ever used DMT, AMT, or Foxy?
(LS01j5)
Ever used Salvia divinorum (LS01k5)
How long has it been since you last used
ketamine? (LS335)
How long has it been since you last used
DMT, AMT, or Foxy? (LS345)
How long has it been since you last used
Salvia divinorum? (LS355)
Ever used a needle to inject any drug
that was not prescribed for you?
(SD156)
Are you now married, widowed,
divorced, or separated, or have you
never married? (QD07)
How many times have you been
married? (QD08)
How many times in the past 12 months
have you moved? (QD13)
In what State did you live in one year
ago today? (QD13a)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)
65,926
105
0.1

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)
31,213
51
0.1

Number of
Cases Asked
the Question
(unweighted)
2,044

QFT1,2
Number of
Cases with
Missing Data4 Missing Data4
(unweighted) (weighted)
2
0.2

65,926
65,926

114
127

0.2
0.1

31,212
31,212

58
70

0.2
0.2

2,044
2,044

3
3

0.2
0.3

656

0.6

321

0.2

0.0*

478

0.1

309

0.2

4.1*

2,583

0.2

1,065

0.1

0.0*

65,926

0.0a

31,213

0.0a

2,044

0.0*

54,954

0.0a

26,036

0.0a*

1,778

0.4

20,247

0.0

9,659

0.0

859

0.2

65,914

0.1a

31,212

0.0a

2,043

0.8

20,017

0.0a

9,585

0.0a

618

0.7
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Were you born in the United States?
(QD14)
Have you lived in the United States for
at least one year? (QD16a)
How many years have you lived in the
United States? (QD16b)
How many months have you lived in the
United States? (QD16c)
Are you now attending or are you
currently enrolled in school? (QD17)
What grade or year of school are you
now attending? (QD18)
Are you a full-time student or a parttime student? (QD19)
During the past 30 days, how many
whole days of school did you miss
because you were sick or injured?
(QD20)
During the past 30 days, how many
whole days of school did you miss
because you skipped or "cut" or just
didn't want to be there? (QD21)
Did you work at a job or business at any
time last week? (QD26)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

65,914

0.0

31,212

0.0*

2,043

0.0

5,101

0.0*

2,437

0.0*

239

0.3

4,872

0.1a

2,337

0.1

227

0.0*

228

0.0*

100

0.0*

19.7*

65,914

0.0

31,212

0.0*

2,043

0.1

34,297

0.0

15,915

0.2

804

0.5

34,297

0.0a

15,915

0.0a

804

1.0

31,249

0.3a

14,472

0.2a

690

1.4

26,816

0.1a

10,528

0.1a

597

1.5

54,944

0.0a

26,035

0.0a*

1,778

0.2
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Even though you did not work at any
time last week, did you have a job or
business? (QD27)
How many hours did you work last
week at all jobs or businesses?
(QD28)
Do you usually work 35 hours or more
per week at all jobs or businesses?
(QD29)
Which one of these reasons best
describes why you did not work last
week? (QD30)
Which one of these reasons best
describes why you did not have a job
or business last week? (QD31)
During the past 30 days, did you make
specific efforts to find work? (QD32)
Did you work at a job or business at any
time during the past 12 months?
(QD33)
How many different employers have
you had in the past 12 months?
(QD36)
During the past 12 months, was there
ever a time when you did not have at
least one job or business? (QD37)
In how many weeks during the past 12
months did you not have at least one
job or business? (QD38)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

25,795

0.0

11,746

0.0

747

0.5

29,144

0.1

14,288

0.1

1,025

0.3

32,036

0.0

15,921

0.1

1,129

0.2

2,892

0.0

1,633

0.1

104

0.0*

22,903

0.1

10,113

0.0a

643

0.8

5,851

0.1

2,607

0.0*

156

0.0*

22,908

0.1a

10,114

0.0a

649

0.6

32,855

0.0a

15,906

0.1a

1,066

0.8

32,036

0.0

15,921

0.0

1,129

0.3

7,023

0.7a

3,615

0.9a

249

4.3
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
In what year did you last work at a job
or business? (QD39a)
In what month in did you last work at a
job or business? (QD39b)
During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)
During the past 30 days, how many
whole days of work did you miss
because you just didn't want to be
there? (QD41)
How many people work for your
employer out of this office, store,
etc.? (QD42)
At your workplace, is there a written
policy about employee use of alcohol
or drugs? (QD43)
Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)
At your workplace, have you ever been
given any educational information
regarding the use of alcohol or drugs?
(QD45)
Through your workplace, is there access
to any type of employee assistance
program or other type of counseling
program for employees who have
alcohol or drug-related problems?
(QD46)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

22,903

0.8a

10,106

0.7a

643

5.2

7,413

0.4

3,335

0.5

175

0.7*

32,036

0.0a

15,921

0.1a

1,129

0.6

32,036

0.0a

15,921

0.0a

1,129

0.5

32,036

0.3a

15,921

0.5a

1,129

1.1

32,036

1,656

4.4a

15,921

872

4.7a

1,129

3.0

23,221

404

2.0a

11,463

198

1.8a

858

0.4

32,036

190

0.7

15,921

107

0.7

1,129

0.4

32,036

4,428

11.8a

15,921

2,231

11.9a

1,129

7.7
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Does your workplace ever test its
employees for alcohol use? (QD47)
Does your workplace ever test its
employees for drug use? (QD48)
Does your workplace test its employees
for drug or alcohol use as part of the
hiring process? (QD49)
Does your workplace test its employees
for drug or alcohol use on a random
basis? (QD50)
According to the policy at your
workplace, what happens to an
employee the first time he or she tests
positive for illicit drugs? (QD51)
Would you be more or less likely to
want to work for an employer that
tests its employees for drug use as
part of the hiring process? (QD52)
Would you be more or less likely to
want to work for an employer that
tests its employees for drug or alcohol
use on a random basis? (QD53)
[SAMPLE MEMBER A] covered by
Medicare? (QHI01)
You have indicated that [SAMPLE
MEMBER B] covered by Medicare.
Is this correct? (QHI01v)
[SAMPLE MEMBER A] covered by
Medicaid? (QHI02)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

32,036

1,805

5.4a

15,921

907

5.3a

1,129

3.2

32,036

1,441

4.3

15,921

741

4.4a

1,129

3.0

14,351

230

2.0

7,214

112

1.8

530

1.2

14,351

806

5.5

7,214

418

5.3

530

3.7

14,351

1,865

14.0

7,214

937

13.0

530

11.3

32,036

0.2

15,921

0.2

1,129

0.5

32,036

0.2

15,921

0.2

1,129

0.3

65,914

193

0.2

31,211

130

0.3

2,042

0.6

1,208

0.0

620

0.1

1.1*

65,914

360

0.3

31,211

235

0.4

2,042

0.8
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
You have indicated that [SAMPLE
MEMBER B] covered by Medicaid.
Is this correct? (QHI02v)
[SAMPLE MEMBER A] currently
covered by [CHIPFILL]? (QHI02A)
[SAMPLE MEMBER A] currently
covered by TRICARE, or
CHAMPUS, CHAMPVA, the VA, or
military health care? (QHI03)
[SAMPLE MEMBER A] currently
covered by private health insurance?
(QHI06)
Was [SAMPLE MEMBER] private
health insurance obtained through
work? (QHI07)
Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)
Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for drug abuse? (QHI09)
Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for mental or emotional
problems? (QHI10)
[SAMPLE MEMBER A] currently
covered by any kind of health
insurance including Indian Health
Insurance? (QHI11)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

0.0*

2.5

663

3.8

142

0.2

2,042

0.6

31,211

261

0.4

2,042

0.7

0.2

19,247

0.2

1,148

0.1

18,327

43.8a

19,247

8,785

44.5a

1,148

322

26.4

40,366

18,195

43.8a

19,247

8,748

44.8a

1,148

330

27.6

40,366

10,900

26.9a

19,247

5,187

26.4a

1,148

209

18.2

10,940

0.2a

5,061

412

102

28,126

567

1.9

13,131

312

65,914

194

0.2

31,211

65,914

382

0.3a

40,366

149

40,366

0.0*

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)
7

220

0.4*

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

0.3

0.0*
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
During the past 12 months, was there
any time when [SAMPLE MEMBER]
did not have any kind of health
insurance or coverage? (QHI13)
During the past 12 months, about how
many months without any kind of
health insurance or coverage?
(QHI14)
About how long has it been since
[SAMPLE MEMBER] last had any
kind of health care coverage?
(QHI15)
Which of these reasons is the main
reason why [SAMPLE MEMBER]
stopped being covered by health
insurance? (QHI17)
Which of these reasons describe
why [SAMPLE MEMBER] never had
health insurance coverage? (QHI187)
In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)
In [YEAR], did you receive
Supplemental Security Income or
SSI? (QI03N)
In [YEAR], did you receive income
from wages or pay earned while
working at a job or business? (QI05N)
In [YEAR], did you receive food
stamps? (QI07N)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

55,956

143

0.2

26,605

0.1

1,685

0.2

4,873

0.6

2,046

0.4

155

1.1

9,498

0.5

4,297

0.2

325

0.8

8,524

0.4

3,857

0.4

258

1.6

974

0.6

440

0.7

0.6*

65,913

616

0.6

31,211

341

0.6

2,042

1.0

65,913

883

0.8a

31,211

459

0.8a

2,042

1.5

65,913

162

0.2a

31,211

103

0.3a

2,042

1.1

65,912

236

0.3

31,211

165

0.3

2,042

0.5
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
At any time during [YEAR], even for
one month, did you receive any cash
assistance from a State or county
welfare program such as
[TANFFILL]? (QI08N)
In [YEAR ], because of low income,
did you receive any other kind of nonmonetary welfare or public
assistance? (QI10N)
For how many months in [YEAR]did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance? (QI12AN)
At any time during [YEAR], even for
one month, did you receive any cash
assistance from a State or county
welfare program such as
[TANFFILL]? (QI08N)
For how many months in [YEAR]did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance, not including food stamps?
(QI12BN)
Before taxes and other deductions, was
your total personal income from all
sources during [YEAR] more or less
than 20,000 dollars? (QI20N)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

65,912

462

0.4a

31,211

239

0.4a

2,042

1.0

65,912

349

0.3a

31,211

191

0.3a

2,042

0.6

1,181

3.0

492

5.3

3.6*

65,912

462

0.4a

31,211

239

0.4a

2,042

1.0

3,583

123

3.0

1,645

5.0

114

5.1*

65,912

785

1.9a

31,211

393

1.9a

2,042

3.7
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Of these income groups, which category
best represents [SAMPLE MEMBER]
total personal income during
[YEAR]?(QI21A)
Of these income groups, which category
best represents [SAMPLE MEMBER]
total personal income during
[YEAR]?(QI21B)
Before taxes and other deductions, was
the total combined family income
during [YEAR] more or less than
20,000 dollars? (QI22)
Of these income groups, which category
best represents your total combined
family income during [YEAR]?
(QI23A)
See notes at end of table.

2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

47,732

581

2.2a

22,448

258

2.2a

17,395

352

2.7

8,370

193

43,440

2,582

7.8

20,458

9,445

605

6.1

4,572

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

1,196

4.6

3.3

769

3.6

1,293

8.1

1,131

9.5

298

6.9

365

9.7
(continued)

Table 4.8 Item Missingness Rates for Moved Items in the 2012 Questionnaire Field Test, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)
2011 Comparison Data1
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

2012 Comparison Data1,3
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

QFT1,2
Number of Number of
Cases Asked Cases with
the Question Missing Data4 Missing Data4
(unweighted) (unweighted) (weighted)

Instrument Item
Of these income groups, which category
best represents your total combined
family income during [YEAR]?
(QI23B)
44,537
2,810
6.4
20,887
1,314
6.3
1,328
87
6.1
* Low precision.
AMT = alpha-methyltryptamine; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the
Veterans Administration; DMT = dimethyltryptamine; QFT = Questionnaire Field Test, VA = Department of Veterans Affairs.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to self- administered.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Missing data include selection of responses of either "don't know" or "refused" for the question. "Missing Data (weighted)" denotes the weighted percentage of missing data.
Denominators for these percentages were based on the total number of cases (i.e., respondents) who were asked the question.
5
For 2011 and 2012 comparison data, these items correspond to items in the special drugs module but were moved to the hallucinogens module in the QFT.
6
For 2011 and 2012 comparison data, this item correspond to special drug item SD05.
7
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not asked the question if all response options were
coded as "blank" (e.g., 98 for 2-digit variables).
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

4.5

Interview Timing Results

4.5.1

Central Tendency Statistics for Overall and Module Timing Results for the 2011
and 2012 Quarters 3 and 4 Comparison Data and the 2012 QFT Data
4.5.1.1 Overall and Module Timing Results for All Respondents in the 2011 and 2012
Quarters 3 and 4 Comparison Data and the 2012 QFT Data

To assess interview timing for the partially redesigned QFT instrument, Tables 4.9a
through 4.9f provide mean and median timing results by module for the 2011 main study
comparison data, the 2012 quarters 3 and 4 comparison data, and the QFT data. These
comparisons include timing results for all respondents in each of the three sets of interviews, as
well as separate timing results for five age categories—aged 12 to 17, 18 to 25, 26 to 49, 50 to
64, and 65 or older. Timing results categorized by age groups provide data on how age is related
to interview duration for the partially redesigned QFT questionnaire and how this compares with
the current main study timing. Respondents with an overall administration time of less than
30 minutes or greater than 240 minutes were classified as outliers and excluded from the timing
results.
Administration times for all three datasets were calculated according to the standard
NSDUH timing data calculation procedures. One necessary variation to the timing calculations
was creating an "administrative residual" category to capture small amounts of additional
interviewing time that did not clearly fall within a defined interview section. Because the
administrative residual timings differed in the revised QFT protocol compared with the 2011
main study and 2012 quarters 3 and 4 protocol, accounting for this time in the three datasets
allowed for more direct and accurate comparisons of overall and section timings across the
datasets. In addition, the administrative residual category provides the ability to add mean
section timings and the administrative residual timing to produce the mean overall timing for the
interviews from each dataset. For each of the three sets of respondents, the mean overall
interview time can be calculated by adding the following mean section times, which are bolded
in Tables 4.9a through 4.9f:14
•

introduction,

•

core demographics,

•

calendar,

•

beginning ACASI,

•

tutorial,

•

total core substances,

•

special drugs to consumption of alcohol,

•

back-end demographics,

•

household roster,

To aid in their readability, Tables 4.9a through 4.9f appear together at the end of this discussion in
Section 4.5.1.1.

•

proxy information/decision,

•

proxy tutorial,

•

health insurance,

•

income, and

•

verification.

Table 4.9a shows that overall interview times were somewhat lower for all QFT
respondents aged 12 or older (mean 59.53, median 55.99) compared with all 2011 respondents
(mean 61.37, median 58.62) and all 2012 quarters 3 and 4 respondents (mean 60.97, median
58.30). Among other factors, the higher item missingness rates observed for multiple
questionnaire items moved from CAPI to ACASI in the QFT instrument (see Section 4.4.2)
could have contributed to the shorter overall administration times for the QFT interviews.
Overall interview times were lower or similar for QFT respondents compared with 2011 and
2012 quarters 3 and 4 respondents for most age groups, as shown in Tables 4.9b through 4.9f.
One exception to this pattern was that the overall timing for QFT respondents aged 65 or older
was actually higher than those 65 or older in the 2011 and 2012 quarters 3 and 4 interviews.
Patterns of overall interview timing across the five age groups were generally similar for the
three sets of respondents, where respondents aged 12 to 17 and those aged 50 or older had higher
overall timings than those aged 18 to 49. For all of the respondent sets, the highest mean and
median overall interview times were greatest for respondents aged 65 or older.
The first five sections in the partially redesigned QFT questionnaire—introduction, core
demographics, calendar, beginning ACASI, and tutorial—took less time to administer for most
respondents compared with the 2011 and 2012 questionnaire. The lower average administration
times among QFT respondents on these early modules were generally small, but also consistent
across age groups. Timings for these sections varied, so a few exceptions to this general pattern
were observed. For example, among respondents aged 50 to 64 and those aged 65 or older,
timings for the tutorial section were actually higher among QFT respondents compared with
2011 and 2012 quarters 3 and 4 respondents.
As expected, the average timing for the total core substance use sections for all
respondents aged 12 or older was higher for the QFT respondents (mean 13.60, median 11.75)
than the 2011 respondents (mean 12.34, median 11.18) and the 2012 quarters 3 and 4
respondents (mean 12.19, median 11.08). Additions and revisions to the hallucinogens, inhalants,
and prescription drug sections in the partially redesigned QFT questionnaire contributed the most
to higher administration times among QFT respondents for the core substance use modules.
Combining the smokeless tobacco items appeared to contribute to lower average timings for the
tobacco section for QFT respondents compared with 2011 and 2012 quarters 3 and 4
respondents, across all age groups. Timing differences between QFT respondents versus 2011
and 2012 quarters 3 and 4 respondents for the remaining core substance use modules—alcohol,
marijuana, cocaine and crack, and heroin—were generally small and inconsequential.
Timings for the redesigned prescription drug modules are of particular interest, given the
considerable changes made to these modules in the QFT questionnaire. The average timing for
the four prescription drug modules for QFT respondents aged 12 or older (mean 5.95,

median 4.92) was clearly higher than the 2011 respondents (mean 5.35, median 4.77) and 2012
quarters 3 and 4 respondents (mean 5.34, median 4.77). Among the redesigned prescription drug
modules, the pain relievers module accounted for the higher administration times for QFT
respondents compared with 2011 and 2012 quarters 3 and 4 respondents. Average timings for the
other three prescription drug modules—tranquilizers, stimulants, and sedatives—were similar or
lower among the three sets of respondents. Administration times did vary across age groups
among the QFT, 2011, and 2012 quarters 3 and 4 respondents. For example, Table 4.9b shows
that QFT respondents aged 12 to 17 actually took less time to complete the four prescription drug
modules than adolescent respondents in the 2011 and 2012 comparison samples. The overall
average timing for the prescription drug modules was increased among QFT respondents by
higher administration times for adult respondents aged 18 or older. In addition, the timing
differences between QFT respondents and the 2011 and 2012 quarters 3 and 4 respondents
increased steadily across the four adult age groups, so that differences among the three sets of
respondents were most pronounced among those aged 65 or older (Table 4.9f). One potential
factor contributing to the increased administration times for the prescription drug modules
among respondents aged 65 or older was the shift in focus from lifetime use to past year use of
prescription medications. Having to report on use of all prescription drugs in the past 12 months
could have increased the time required for older respondents to complete the redesigned
modules.
For sections from special drugs to consumption of alcohol, administration times for all
QFT respondents aged 12 or older varied in relation to the section timings for the 2011 and 2012
quarters 3 and 4 respondents. Sections with lower QFT timings compared with the 2011 and
2012 quarters 3 and 4 interviews included special drugs, prior substance use, youth experiences,
youth mental health service utilization, adolescent depression, and consumption of alcohol. The
lower administration times for special drugs, prior substance use, and youth experiences
appeared likely to result from the deletion of one or more items from these sections in the QFT
questionnaire. QFT administration times were higher than the 2011 and 2012 quarters 3 and 4
interviews for substance dependence and abuse and mental health, despite few changes to these
sections in the QFT questionnaire. For the remaining sections from special drugs to consumption
of alcohol, administration times for QFT respondents were generally similar to the section
timings for the 2011 and 2012 quarters 3 and 4 respondents.
Section timings for the remaining back-end modules also varied for all respondents aged
12 or older when comparing QFT with 2011 and 2012 quarters 3 and 4 respondents, based
mostly on changes made to the QFT questionnaire. For example, under back-end demographics,
the average times for QFT respondents compared with 2011 and 2012 quarters 3 and 4
respondents were higher for education, but lower for employment. These findings are consistent
with the changes to the QFT questionnaire, such as adding new items on disability to the
education section and deleting questions on industry and occupation from the employment
section.
For the health insurance section, a higher average administration time was observed for
QFT respondents compared with the 2011 and 2012 quarters 3 and 4 respondents. The only
change to this section in the QFT questionnaire was moving these questions from CAPI to
ACASI administration. One possible explanation for the increased timing among QFT
respondents was that a higher number of proxy reporters answered these questions in the QFT
86

and the health insurance module is the first section after the proxy tutorial. One consequence of
this sequence is that QFT proxy reporters might have used additional time getting accustomed to
the interview protocol, including the relationship fills.
The income section was also moved from CAPI to ACASI administration in the QFT
questionnaire, and a new question on household telephone service was added to this section.
These changes corresponded with lower timings for QFT respondents compared with 2011 and
2012 quarters 3 and 4 respondents for those aged 12 to 49; similar timings for QFT, 2011, and
2012 quarters 3 and 4 respondents for those aged 50 or older; and higher timings for QFT
respondents compared with 2011 and 2012 quarters 3 and 4 respondents for those aged 65 or
older. The explanation for this unique pattern across age groups is not immediately clear.
Table 4.9a Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
or Older)

Module
Introduction
Core Demographics
Calendar4
Beginning ACASI
Tutorial
Total Core Substances
Tobacco
Alcohol
Marijuana
Cocaine and Crack
Heroin
Hallucinogens
Inhalants
Methamphetamine
Total Prescription Drugs
Pain Relievers (Screener)
Tranquilizers (Screener)
Stimulants (Screener)
Sedatives (Screener)
Pain Relievers (Screener Plus
Main Module)5
Tranquilizers (Screener Plus
Main Module)5
Stimulants (Screener Plus Main
Module)5
Sedatives (Screener Plus Main
Module)5

2011
Main Study1
Mean
Median
1.94
1.73
2.22
1.85
1.67
1.48
2.41
2.20
3.44
3.27
12.34
11.18
2.02
1.70
2.15
1.98
0.49
0.37
0.21
0.13
0.10
0.08
0.83
0.63
1.18
0.92

Q3-Q4 2012
Main Study1,2
Mean
Median
1.78
1.62
2.18
1.82
1.66
1.50
2.38
2.17
3.45
3.27
12.19
11.08
1.96
1.67
2.13
1.98
0.49
0.37
0.21
0.13
0.10
0.08
0.81
0.63
1.15
0.90

2012
Questionnaire
Field Test1,3
Mean
Median
1.52
1.40
2.10
1.73
1.15
1.17
2.22
2.03
3.34
3.15
13.60
11.75
1.83
1.43
2.25
2.07
0.52
0.40
0.22
0.13
0.10
0.08
1.18
0.92
1.35
1.07
0.20
0.15
5.95
4.92
2.42
2.03
0.88
0.70
0.92
0.75
0.81
0.63

5.35

4.77

5.34

4.77

2.09

1.90

2.08

1.88

3.02

2.45

1.15

0.98

1.15

0.98

1.04

0.75

1.16

0.97

1.16

0.97

1.02

0.78

0.95

0.75

0.94

0.75

0.87

0.67
(continued)

See notes at end of table.

Table 4.9a Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
or Older) (continued)

Module
Special Drugs to Consumption of
Alcohol6
Special Drugs
Risk/Availability
Blunts
Substance Dependence and Abuse
Market Information for Marijuana
Prior Substance Use
Special Topics, Drug Treatment
Health Care
Adult Mental Health Service
Utilization
Social Environment
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health Service
Utilization
Adolescent Depression
Consumption of Alcohol
Back-End Demographics (Moves,
Born in United States, Disability,
Education and Employment)7
Education8
Employment

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

21.93
1.60
2.96
0.27
2.19
0.27
1.24
1.63
1.29

20.23
1.47
2.68
0.20
1.58
0.00
0.95
1.35
1.10

21.68
1.59
2.94
0.27
2.13
0.27
1.20
1.61
1.30

20.02
1.45
2.67
0.20
1.56
0.00
0.92
1.33
1.08

20.50
0.57
2.92
0.29
2.29

18.78
0.52
2.62
0.20
1.72

1.09
1.68
2.79

0.92
1.37
2.48

0.80
0.96
0.14
2.79
2.10
1.10

0.63
1.02
0.00
0.00
1.77
0.30

0.79
0.95
0.14
2.78
2.09
1.10

0.63
1.00
0.00
0.00
1.77
0.30

0.85
0.94
0.20
2.10
2.27
1.15

0.70
0.95
0.00
0.00
1.97
0.37

0.64
0.55
0.55

0.00
0.00
0.45

0.64
0.55
0.54

0.00
0.00
0.45

0.48
0.43
0.46

0.00
0.00
0.40

4.45
0.58
3.52

4.42
0.48
3.67

4.51
0.57
3.58

4.53
0.45
3.82

4.00
0.85
1.78

3.65
0.68
1.70
(continued)

See notes at end of table.

Table 4.9a Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
or Older) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
1.64
1.40
0.57
0.32

Q3-Q4 2012
Main Study1,2
Mean
Median
1.69
1.45
0.57
0.33

1.40
3.71
3.01
0.65
61.37

1.40
3.64
3.14
0.70
60.97

1.28
3.23
2.57
NA
58.62

1.28
3.23
2.70
NA
58.30

2012
Questionnaire
Field Test1,3
Mean
Median
1.50
1.28
0.58
0.45
0.74
0.00
1.59
1.37
3.23
2.73
3.31
2.85
0.13
NA
59.53
55.99

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and the Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

Table 4.9b Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
to 17)

2011
Main Study1
Mean
Median
2.00
1.87
2.13
1.75
1.66
1.50
2.44
2.27
3.64
3.55
11.93
11.00
1.77
1.48
1.62
1.40
0.46
0.32
0.18
0.13
0.10
0.08
0.88
0.73
1.37
1.13

Q3-Q4 2012
Main Study1,2
Mean
Median
1.83
1.72
2.09
1.73
1.66
1.52
2.40
2.22
3.70
3.58
11.93
10.98
1.70
1.47
1.60
1.38
0.46
0.33
0.17
0.13
0.10
0.08
0.88
0.73
1.36
1.12

2012
Questionnaire
Field Test1,3
Mean
Median
1.55
1.45
2.01
1.65
1.22
1.23
2.21
2.10
3.41
3.37
11.97
10.83
1.41
1.13
1.62
1.32
0.51
0.42
0.17
0.13
0.10
0.08
1.24
1.03
1.52
1.25
0.22
0.20
5.20
4.52
2.35
2.03
0.81
0.67
0.83
0.72
0.73
0.60

5.56

5.07

5.66

5.15

2.17

2.02

2.18

2.03

2.68

2.32

1.19

1.05

1.21

1.08

0.87

0.68

1.20

1.03

1.23

1.05

0.90

0.73

1.00

0.82

1.03

0.85

0.76

0.62
(continued)

See notes at end of table.

Table 4.9b Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
to 17) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

22.27
1.68
2.97
0.25
0.97
0.20
0.60
1.38
1.33

20.90
1.58
2.77
0.20
0.00
0.00
0.00
1.18
1.17

22.19
1.68
3.03
0.25
0.87
0.20
0.55
1.35
1.34

20.80
1.60
2.80
0.20
0.00
0.00
0.00
1.15
1.18

20.52
0.54
2.85
0.29
0.87

19.15
0.52
2.62
0.20
0.00

0.47
1.31
2.74

0.00
1.12
2.50

8.21

7.83

8.28

7.85

7.83

7.32

1.88
1.62
0.30

1.60
0.63
0.00

1.90
1.65
0.28

1.60
0.63
0.00

1.78
1.61
0.23

1.50
0.60
0.00

2.53
0.88
1.34

1.73
0.82
0.35

2.59
0.85
1.42

1.73
0.80
0.32

3.34
1.27
0.74

3.03
1.17
0.48
(continued)

See notes at end of table.

Table 4.9b Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 12
to 17) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
2.13
1.85
1.00
0.75

Q3-Q4 2012
Main Study1,2
Mean
Median
2.17
1.90
1.00
0.77

1.42
3.97
3.13
0.49
60.74

1.40
3.84
3.20
0.52
60.51

1.28
3.45
2.67
NA
58.70

1.28
3.45
2.75
NA
58.55

2012
Questionnaire
Field Test1,3
Mean
Median
1.94
1.75
0.88
0.75
2.00
1.98
1.75
1.57
3.47
3.00
3.16
2.85
0.12
NA
59.56
57.17

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

Table 4.9c Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 18
to 25)

2011
Main Study1
Mean
Median
1.81
1.67
2.15
1.82
1.64
1.47
2.30
2.12
3.01
2.85
11.77
10.65
2.06
1.77
2.27
2.10
0.55
0.40
0.21
0.12
0.09
0.07
0.76
0.53
0.94
0.73

Q3-Q4 2012
Main Study1,2
Mean
Median
1.67
1.57
2.11
1.80
1.63
1.48
2.28
2.10
2.99
2.83
11.41
10.37
1.96
1.67
2.25
2.08
0.54
0.38
0.20
0.12
0.09
0.07
0.71
0.52
0.90
0.72

2012
Questionnaire
Field Test1,3
Mean
Median
1.32
1.38
1.96
1.70
0.98
0.95
2.19
2.05
2.82
2.67
12.35
10.87
1.85
1.53
2.21
2.10
0.56
0.40
0.21
0.12
0.09
0.07
1.00
0.70
1.04
0.85
0.16
0.12
5.25
4.33
1.98
1.78
0.70
0.58
0.72
0.63
0.61
0.53

4.88

4.35

4.77

4.30

2.00

1.78

1.95

1.73

2.72

2.18

1.04

0.87

1.02

0.87

0.93

0.62

1.04

0.85

1.02

0.85

0.95

0.67

0.80

0.65

0.78

0.65

0.55
(continued)

See notes at end of table.

Table 4.9c Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 18
to 25) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

20.46
1.46
2.61
0.32
3.08
0.45
1.49
1.64
1.03

18.75
1.32
2.37
0.22
2.47
0.00
1.20
1.33
0.90

20.02
1.42
2.54
0.31
2.98
0.46
1.41
1.60
1.02

18.48
1.28
2.33
0.22
2.35
0.00
1.13
1.30
0.90

18.29
0.51
2.48
0.35
3.12

16.72
0.45
2.22
0.23
2.32

1.15
1.64
2.28

0.98
1.30
2.07

1.05
1.31
0.01

0.82
1.18
0.00

1.04
1.29
0.01

0.80
1.17
0.00

0.97
1.07
0.02

0.75
1.00
0.00

2.94
1.52

2.73
0.47

2.89
1.54

2.68
0.47

2.70
1.47

2.50
0.47

0.72

0.60

0.70

0.58

0.54

0.45

5.71
0.66
4.62

5.63
0.57
4.65

5.79
0.65
4.72

5.67
0.53
4.70

4.06
0.77
1.98

3.68
0.65
1.82
(continued)

See notes at end of table.

Table 4.9c Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 18
to 25) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
1.54
1.30
0.39
0.23

Q3-Q4 2012
Main Study1,2
Mean
Median
1.60
1.35
0.40
0.25

1.42
3.61
2.88
0.57
59.27

1.42
3.60
3.03
0.64
58.59

1.33
3.18
2.52
NA
56.58

1.33
3.18
2.67
NA
56.05

2012
Questionnaire
Field Test1,3
Mean
Median
1.48
1.27
0.55
0.42
0.40
0.00
1.46
1.28
2.92
2.45
3.35
2.92
0.13
NA
54.26
50.80

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

Table 4.9d Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 26
to 49)

2011
Main Study1
Mean
Median
1.94
1.70
2.29
1.90
1.65
1.45
2.35
2.13
3.28
3.12
12.18
11.03
2.06
1.78
2.38
2.18
0.46
0.35
0.24
0.13
0.10
0.08
0.77
0.60
1.07
0.85

Q3-Q4 2012
Main Study1,2
Mean
Median
1.77
1.58
2.26
1.88
1.64
1.45
2.31
2.10
3.27
3.07
12.01
10.95
2.02
1.76
2.37
2.18
0.45
0.33
0.23
0.13
0.10
0.08
0.75
0.58
1.02
0.82

2012
Questionnaire
Field Test1,3
Mean
Median
1.56
1.37
2.11
1.72
1.09
1.07
2.07
1.92
3.01
2.88
13.36
11.46
1.89
1.62
2.40
2.18
0.49
0.35
0.23
0.13
0.10
0.08
1.08
0.85
1.21
0.97
0.19
0.13
5.76
4.89
2.28
1.98
0.85
0.70
0.89
0.75
0.77
0.65

5.11

4.53

5.06

4.53

1.99

1.78

1.99

1.78

2.95

2.44

1.11

0.93

1.10

0.93

1.01

0.78

1.12

0.93

1.10

0.93

0.96

0.77

0.89

0.72

0.87

0.72

0.84

0.68
(continued)

See notes at end of table.

Table 4.9d Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 26
to 49) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

21.96
1.55
2.95
0.25
2.74
0.20
1.63
1.81
1.25

19.97
1.40
2.67
0.18
2.17
0.00
1.32
1.47
1.08

21.55
1.54
2.88
0.25
2.73
0.21
1.61
1.77
1.23

19.67
1.40
2.63
0.18
2.18
0.00
1.32
1.43
1.07

20.43
0.57
2.85
0.26
2.80

18.67
0.52
2.53
0.20
2.18

1.33
1.81
2.62

1.18
1.46
2.33

1.25
1.42
0.53

0.95
1.28
0.00

1.21
1.40
0.51

0.93
1.25
0.00

1.16
1.24
0.51

0.88
1.08
0.00

3.16
1.79

2.95
0.53

3.09
1.74

2.87
0.50

3.07
1.71

2.75
0.49

0.63

0.57

0.62

0.57

0.49

0.47

5.62
0.22
5.05

5.52
0.13
5.00

5.60
0.23
5.02

5.45
0.13
4.93

4.13
0.61
2.23

3.72
0.48
2.03
(continued)

See notes at end of table.

Table 4.9d Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 26
to 49) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
1.40
1.22
0.30
0.22

Q3-Q4 2012
Main Study1,2
Mean
Median
1.44
1.25
0.31
0.22

1.32
3.48
2.87
0.90
61.54

1.33
3.43
3.01
0.94
60.87

1.23
3.00
2.42
NA
58.55

1.23
3.03
2.57
NA
57.88

2012
Questionnaire
Field Test1,3
Mean
Median
1.38
1.23
0.41
0.35
0.22
0.00
1.41
1.23
2.96
2.51
3.13
2.73
0.12
NA
57.39
53.90

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

Table 4.9e Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 50
to 64)

2011
Main Study1
Mean
Median
2.09
1.73
2.51
2.00
1.74
1.50
2.60
2.33
4.05
3.95
14.37
12.88
2.41
2.02
2.74
2.52
0.52
0.42
0.30
0.18
0.13
0.10
0.95
0.72
1.31
1.03

Q3-Q4 2012
Main Study1,2
Mean
Median
1.99
1.70
2.42
1.90
1.73
1.52
2.55
2.28
4.13
4.10
14.41
13.08
2.39
2.00
2.78
2.55
0.53
0.43
0.30
0.18
0.13
0.12
0.94
0.72
1.31
1.05

2012
Questionnaire
Field Test1,3
Mean
Median
1.70
1.48
2.24
1.85
1.39
1.48
2.40
2.08
4.26
4.15
16.55
14.40
2.24
1.67
2.86
2.47
0.52
0.47
0.29
0.20
0.13
0.10
1.40
1.10
1.55
1.25
0.23
0.18
7.33
6.22
2.91
2.42
1.14
0.92
1.20
0.93
1.10
0.83

6.02

5.35

6.03

5.43

2.19

1.97

2.23

2.03

3.57

3.03

1.30

1.12

1.31

1.13

1.30

0.98

1.37

1.15

1.36

1.15

1.26

0.97

1.15

0.93

1.13

0.93

1.19

0.90
(continued)

See notes at end of table.

Table 4.9e Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 50
to 64) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

24.19
1.78
3.51
0.22
2.46
0.13
1.84
1.97
1.74

21.80
1.57
3.15
0.18
2.03
0.00
1.50
1.65
1.47

24.21
1.78
3.52
0.22
2.51
0.13
1.80
2.01
1.76

22.00
1.58
3.20
0.18
2.07
0.00
1.50
1.65
1.52

22.14
0.67
3.45
0.22
2.63

20.23
0.62
3.20
0.20
2.12

1.46
1.90
3.52

1.28
1.63
3.23

1.50
1.66
0.27

1.10
1.50
0.00

1.43
1.67
0.29

1.08
1.50
0.00

1.25
1.50
0.24

1.02
1.40
0.00

3.62
1.99

3.27
0.60

3.67
1.99

3.37
0.58

3.17
1.50

2.87
0.52

0.67

0.62

0.67

0.62

0.63

0.53

5.24
0.18
4.79

5.18
0.12
4.82

5.18
0.20
4.70

5.13
0.12
4.75

4.60
0.67
2.50

4.17
0.55
2.32
(continued)

See notes at end of table.

100

Table 4.9e Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 50
to 64) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
1.03
0.85
0.30
0.22

Q3-Q4 2012
Main Study1,2
Mean
Median
1.15
0.92
0.33
0.23

1.38
3.48
3.12
0.87
66.96

1.39
3.48
3.35
0.99
67.30

1.23
3.02
2.60
NA
63.13

1.25
3.03
2.72
NA
63.97

2012
Questionnaire
Field Test1,3
Mean
Median
1.13
0.98
0.50
0.38
0.18
0.00
1.71
1.50
3.45
3.00
3.83
2.95
0.17
NA
66.24
62.25

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

101

Table 4.9f Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 65+)

2011
Main Study1
Mean
Median
2.15
1.80
2.74
2.25
1.89
1.62
3.01
2.68
4.86
4.73
17.26
15.97
2.82
2.33
3.16
2.87
0.47
0.42
0.26
0.23
0.17
0.15
1.19
0.93
1.88
1.48

Q3-Q4 2012
Main Study1,2
Mean
Median
1.92
1.70
2.64
2.17
1.83
1.62
3.05
2.67
4.92
4.75
17.40
16.10
2.85
2.38
3.13
2.89
0.48
0.43
0.27
0.22
0.17
0.15
1.19
0.95
1.89
1.47

2012
Questionnaire
Field Test1,3
Mean
Median
1.69
1.50
2.66
2.30
1.52
1.57
2.89
2.32
5.32
5.13
22.04
19.45
2.57
2.20
3.43
3.25
0.60
0.52
0.31
0.23
0.16
0.15
1.79
1.45
2.29
1.72
0.29
0.23
10.60
8.28
4.28
3.05
1.69
1.27
1.71
1.27
1.62
1.25

7.30

6.68

7.41

6.75

2.49

2.33

2.48

2.33

5.10

3.73

1.63

1.47

1.67

1.52

1.93

1.43

1.66

1.43

1.71

1.47

1.77

1.27

1.52

1.28

1.56

1.32

1.80

1.30
(continued)

See notes at end of table.

102

Table 4.9f Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 65+)
(continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

26.51
2.06
4.59
0.23
1.74
0.02
1.52
2.14
2.47

24.20
1.87
4.05
0.20
0.00
0.00
1.30
1.88
2.15

26.80
2.08
4.53
0.23
1.81
0.02
1.57
2.22
2.56

24.84
1.90
3.98
0.20
1.35
0.00
1.35
1.90
2.18

26.64
0.75
4.36
0.25
2.03

23.87
0.67
3.85
0.20
1.80

1.67
2.36
4.75

1.35
1.95
4.35

1.77
2.29
0.04

1.33
2.02
0.00

1.80
2.24
0.05

1.33
1.98
0.00

1.74
1.96
0.07

1.33
1.77
0.00

4.47
1.62

4.00
0.67

4.60
1.66

4.13
0.65

4.65
1.38

4.25
0.68

0.70

0.65

0.69

0.65

0.67

0.62

2.93
0.16
2.52

1.82
0.12
1.38

3.09
0.16
2.63

1.88
0.12
1.43

5.00
0.90
2.08

4.40
0.68
1.75
(continued)

See notes at end of table.

103

Table 4.9f Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (All Respondents Aged 65+)
(continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Administrative Residual
Overall Questionnaire

2011
Main Study1
Mean
Median
0.82
0.62
0.32
0.20

Q3-Q4 2012
Main Study1,2
Mean
Median
0.87
0.67
0.32
0.20

1.46
3.89
3.62
0.86
72.32

1.49
3.73
3.76
0.88
72.70

1.30
3.28
2.92
NA
68.43

1.32
3.32
3.10
NA
69.39

2012
Questionnaire
Field Test1,3
Mean
Median
0.96
0.73
0.48
0.42
0.32
0.00
2.13
1.93
4.43
3.98
3.98
3.15
0.17
NA
80.24
74.45

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Timings for the module rows in bold are mutually exclusive. However, these timings may not sum exactly to the overall
questionnaire timing because of rounding.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

104

4.5.1.2

Overall and Module Timing Results for Affirmative Gate Respondents in the 2011
and 2012 Quarters 3 and 4 Comparison Data and the 2012 QFT Data

The section and overall timing statistics presented in Section 4.5.1.1 provided results for
all QFT, 2011, and 2012 quarters 3 and 4 respondents. Tables 4.9g through 4.9l show mean and
median timings by module only for "affirmative gate" respondents.15 These comparisons include
timing results only for affirmative gate respondents in each of the three sets of interviews,
including separate timing results for five age categories—aged 12 to 17, 18 to 25, 26 to 49, 50 to
64, and 65 or older. Timing results categorized by age groups provide data on how age is related
to interview duration for affirmative gate respondents using the partially redesigned QFT
questionnaire compared with the current main study questionnaire.
For these tables, affirmative gate respondents were defined as the following subsets of
QFT, 2011, and 2012 quarters 3 and 4 respondents:
1. those who answered affirmatively to at least one gate question within the core
substance questions, or
2. those whose prior responses directed them to complete a specific questionnaire
module.
For example, only respondents who reported smoking part or all of a cigarette in their lifetime
were included in the timing reports for the tobacco use module. Similarly, only respondents who
were administered the parenting experiences module contributed to the mean timing for that
module.
Presenting data only for affirmative gate respondents, Tables 4.9g through 4.9l highlight
timing statistics for respondents whose administration times for a module were beyond the
minimal time taken by those respondents who had no data to report for a given module. These
timing data focus on respondents who actually reported behavior that led to specific sets of
additional questions. As a result, these results provide a sense of the impact of questionnaire
changes for the set of respondents who have behavior to report for each module.
Given that the purpose of these tables is to show timing results for respondents who have
behavior to report for each module, this section focuses primarily on sections where changes
were made in the QFT questionnaire, such as the prescription drug modules and back-end
demographic questions. Overall, among all affirmative gate respondents aged 12 or older, timing
results followed similar patterns for the core substances sections as seen for all respondents in
Section 4.5.1.1. As Table 4.9g shows, the average timing for the total core substances section for
all affirmative gate respondents aged 12 or older was higher for the QFT respondents (mean
13.93, median 12.05) than the 2011 respondents (mean 12.61, median 11.38) and the 2012
quarters 3 and 4 respondents (mean 12.39, median 11.23). Higher administration times were
observed for the hallucinogens, inhalants, and prescription drug sections for QFT respondents,
and lower administration times were observed for the tobacco section for QFT respondents,
compared with the 2011 and the 2012 quarters 3 and 4 respondents. Timing differences between
15

To aid in their readability, Tables 4.9g through 4.9l appear together at the end of this discussion in
Section 4.5.1.2.

105

affirmative gate respondents in the QFT versus 2011 and 2012 quarters 3 and 4 for the remaining
core substance use modules—alcohol, marijuana, cocaine and crack, and heroin—were generally
small and inconsequential.
The impact of changes to the prescription drug modules on timing results was a special
focus for affirmative gate respondents because use of multiple types of prescription drugs could
significantly increase respondent burden in these modules. Among respondents who reported use
and misuse of prescription drugs, average QFT timings for the four prescription drug modules
exceeded the average timings for the 2011 and 2012 quarters 3 and 4 comparison interviews. The
greatest difference was observed among affirmative gate respondents aged 26 or older, for whom
the difference between QFT versus 2011 and 2012 quarters 3 and 4 respondents was over
1 minute. As noted in Section 4.5.1.1, the additional time required to complete the pain reliever
module in the partially redesigned QFT instrument was mitigated by time savings in other
prescription drug modules, resulting in lower overall administration times for the prescription
drug modules for all respondents. For affirmative gate respondents, Table 4.9g shows that the
overall timing for total prescription drugs for QFT respondents (mean 6.46, median 5.42) was
quite similar to the 2011 respondents (mean 6.42, median 5.78) and the 2012 quarters 3 and 4
respondents (mean 6.34, median 5.77).
For back-end demographics, the average times for QFT affirmative gate respondents
compared with 2011 and 2012 quarters 3 and 4 respondents followed patterns to those shown for
all respondents in Section 4.5.1.1. Average administration times for QFT affirmative gate
respondents were higher for education, but lower for employment. The difference between QFT
affirmative gate respondents and 2011 and 2012 quarters 3 and 4 affirmative gate respondents
shown for employment was similarly more pronounced than the difference for education.

106

Table 4.9g Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 12 or Older)

2011
Main Study1
Mean
Median
1.94
1.73
2.22
1.85
1.67
1.48
2.41
2.20
3.44
3.27
12.61
11.38
2.66
2.33
2.58
2.32
0.81
0.67
0.72
0.55
0.51
0.33
1.45
1.22
1.70
1.40

Q3-Q4 2012
Main Study1,2
Mean
Median
1.78
1.62
2.18
1.82
1.66
1.50
2.38
2.17
3.45
3.27
12.39
11.23
2.60
2.28
2.56
2.32
0.80
0.65
0.70
0.55
0.49
0.32
1.40
1.18
1.65
1.37

2012
Questionnaire
Field Test1,3
Mean
Median
1.52
1.40
2.10
1.73
1.15
1.17
2.22
2.03
3.34
3.15
13.93
12.05
2.49
2.15
2.67
2.40
0.82
0.68
0.69
0.57
0.53
0.32
1.71
1.46
1.75
1.45
0.43
0.35
6.46
5.42
2.42
2.03
0.88
0.70
0.92
0.75
0.81
0.63

6.42

5.78

6.34

5.77

3.08

2.78

3.03

2.75

3.02

2.45

1.85

1.65

1.84

1.63

1.04

0.75

1.98

1.72

1.96

1.75

1.02

0.78

1.88

1.63

1.85

1.57

0.87

0.67
(continued)

See notes at end of table.

107

Table 4.9g Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 12 or Older) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

21.93
1.60
2.96
0.54
3.83
1.49
1.65
1.63
1.29

20.23
1.47
2.68
0.47
3.05
1.38
1.32
1.35
1.10

21.68
1.59
2.94
0.53
3.72
1.47
1.61
1.61
1.30

20.02
1.45
2.67
0.45
2.98
1.37
1.30
1.33
1.08

20.50
0.57
2.92
0.61
3.76

18.78
0.52
2.62
0.52
2.98

1.40
1.68
2.79

1.20
1.37
2.48

2.29
1.45
2.52
8.21
3.62
3.21

1.90
1.28
2.20
7.83
3.23
1.30

2.23
1.43
2.43
8.28
3.59
3.22

1.87
1.27
2.13
7.85
3.18
1.33

2.18
1.28
2.46
7.83
3.62
3.18

1.88
1.13
2.03
7.32
3.17
1.39

3.08
2.58
0.79

2.73
1.02
0.63

3.18
2.65
0.77

2.75
1.03
0.63

2.98
2.60
0.63

2.62
1.00
0.53

4.46
0.58
4.22

4.42
0.48
4.33

4.51
0.57
4.30

4.53
0.45
4.40

4.00
0.85
2.05

3.65
0.68
1.88
(continued)

See notes at end of table.

108

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
1.64
1.40
0.57
0.32

Q3-Q4 2012
Main Study1,2
Mean
Median
1.69
1.45
0.57
0.33

1.40
3.71
3.01
61.37

1.40
3.64
3.14
60.97

1.28
3.23
2.57
58.62

1.28
3.23
2.70
58.30

2012
Questionnaire
Field Test1,3
Mean
Median
1.50
1.28
0.58
0.45
0.73
0.00
1.59
1.37
3.23
2.73
3.31
2.85
59.53
55.99

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Some module rows are shown in bold for consistency with Tables 4.9a to 4.9f for all respondents. However, mean
affirmative gate timings in this table for modules in bold are not necessarily mutually exclusive and are not intended to
sum to the overall mean questionnaire timing.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the Tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

109

Table 4.9h Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 12 to 17)

2011
Main Study1
Mean
Median
2.00
1.87
2.13
1.75
1.66
1.50
2.44
2.27
3.64
3.55
12.30
11.27
2.97
2.62
2.47
2.23
1.20
1.07
1.18
1.05
0.73
0.70
1.92
1.68
2.30
1.97

Q3-Q4 2012
Main Study1,2
Mean
Median
1.83
1.72
2.09
1.73
1.66
1.52
2.40
2.22
3.70
3.58
11.98
10.93
2.85
2.55
2.43
2.22
1.17
1.07
1.05
0.94
0.55
0.45
1.90
1.68
2.28
1.98

2012
Questionnaire
Field Test1,3
Mean
Median
1.55
1.45
2.01
1.65
1.22
1.23
2.21
2.10
3.41
3.37
12.04
11.15
2.47
2.07
2.48
2.33
1.19
1.09
0.77
0.77
0.62
0.62
2.05
1.73
2.06
1.81
0.41
0.42
5.69
5.03
2.35
2.03
0.81
0.67
0.83
0.72
0.73
0.60

6.74

6.15

6.74

5.97

3.45

3.16

3.44

3.08

2.68

2.32

2.12

1.95

2.08

1.88

0.87

0.68

2.15

1.87

2.15

1.83

0.90

0.73

2.18

1.88

2.24

1.92

0.76

0.62
(continued)

See notes at end of table.

110

Table 4.9h Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 12 to 17) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

22.27
1.68
2.97
0.69
3.89
1.47
1.37
1.38
1.33

20.90
1.58
2.77
0.60
3.03
1.38
1.07
1.18
1.17

22.19
1.68
3.03
0.68
3.75
1.47
1.34
1.35
1.34

20.80
1.60
2.80
0.60
3.02
1.35
1.03
1.15
1.18

20.52
0.54
2.85
0.79
3.73

19.15
0.52
2.62
0.72
3.08

1.12
1.31
2.74

0.97
1.12
2.50

8.21

7.83

8.28

7.85

7.83

7.32

3.08
2.58
0.85

2.73
1.02
0.57

3.18
2.65
0.84

2.75
1.03
0.55

2.98
2.60
0.68

2.62
1.00
0.43

2.53
0.88
2.62

1.73
0.82
1.42

2.59
0.85
2.80

1.73
0.80
1.50

3.34
1.28
1.44

3.03
1.17
1.13
(continued)

See notes at end of table.

111

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
2.13
1.85
1.00
0.75

Q3-Q4 2012
Main Study1,2
Mean
Median
2.17
1.90
1.00
0.77

1.42
3.97
3.13
60.74

1.40
3.84
3.20
60.51

1.28
3.45
2.67
58.70

1.28
3.45
2.75
58.55

2012
Questionnaire
Field Test1,3
Mean
Median
1.94
1.75
0.88
0.75
2.00
1.98
1.75
1.57
3.47
3.00
3.16
2.85
59.56
57.17

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Some module rows are shown in bold for consistency with Tables 4.9a to 4.9f for all respondents. However, mean
affirmative gate timings in this table for modules in bold are not necessarily mutually exclusive and are not intended to
sum to the overall mean questionnaire timing.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the Health Insurance and Income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

112

Table 4.9i Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 18 to 25)

2011
Main Study1
Mean
Median
1.81
1.67
2.15
1.82
1.64
1.47
2.30
2.12
3.01
2.85
11.99
10.85
2.61
2.33
2.49
2.25
0.83
0.70
0.76
0.58
0.58
0.37
1.47
1.27
1.42
1.22

Q3-Q4 2012
Main Study1,2
Mean
Median
1.67
1.57
2.11
1.80
1.63
1.48
2.28
2.10
2.99
2.83
11.67
10.63
2.52
2.25
2.47
2.25
0.82
0.70
0.74
0.58
0.53
0.36
1.40
1.17
1.46
1.25

2012
Questionnaire
Field Test1,3
Mean
Median
1.32
1.38
1.96
1.70
0.98
0.95
2.19
2.05
2.82
2.67
12.59
11.08
2.43
2.15
2.48
2.28
0.84
0.73
0.79
0.65
0.50
0.31
1.78
1.56
1.69
1.40
0.48
0.40
6.01
5.08
1.98
1.78
0.70
0.58
0.72
0.63
0.61
0.53

6.14

5.53

5.99

5.53

2.97

2.70

2.90

2.67

2.72

2.18

1.75

1.55

1.72

1.53

0.93

0.62

1.90

1.65

1.87

1.70

0.95

0.67

1.79

1.57

1.81

1.68

0.65

0.55
(continued)

See notes at end of table.

113

Table 4.9i Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 18 to 25) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

20.46
1.46
2.61
0.52
4.06
1.45
1.66
1.64
1.03

18.76
1.32
2.37
0.45
3.37
1.35
1.33
1.33
0.90

20.02
1.42
2.54
0.51
3.91
1.44
1.57
1.60
1.02

18.48
1.28
2.33
0.43
3.20
1.35
1.27
1.30
0.90

18.29
0.51
2.48
0.60
3.94

16.72
0.45
2.22
0.53
3.19

1.31
1.64
2.28

1.12
1.30
2.07

2.05
1.31
2.90

1.75
1.18
2.38

2.03
1.29
2.30

1.75
1.17
2.13

1.92
1.07
2.38

1.58
1.00
1.84

3.23
2.84

2.95
1.08

3.18
2.87

2.90
1.17

3.01
2.83

2.73
1.18

0.83

0.68

0.82

0.68

0.65

0.53

5.71
0.67
4.62

5.63
0.57
4.65

5.79
0.65
4.72

5.67
0.53
4.70

4.06
0.77
1.98

3.68
0.65
1.82
(continued)

See notes at end of table.

114

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
1.54
1.30
0.39
0.23

Q3-Q4 2012
Main Study1,2
Mean
Median
1.60
1.35
0.40
0.25

1.42
3.61
2.88
59.27

1.42
3.60
3.03
58.59

1.33
3.18
2.52
56.58

1.33
3.18
2.67
56.05

2012
Questionnaire
Field Test1,3
Mean
Median
1.48
1.27
0.55
0.42
0.40
0.00
1.46
1.28
2.92
2.45
3.35
2.92
54.26
50.80

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Some module rows are shown in bold for consistency with Tables 4.9a to 4.9f for all respondents. However, mean
affirmative gate timings in this table for modules in bold are not necessarily mutually exclusive and are not intended to
sum to the overall mean questionnaire timing.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

115

Table 4.9j Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 26 to 49)

2011
Main Study1
Mean
Median
1.94
1.70
2.29
1.90
1.65
1.45
2.35
2.13
3.28
3.12
12.26
11.12
2.42
2.12
2.50
2.25
0.63
0.48
0.63
0.50
0.40
0.30
1.26
1.08
1.41
1.20

Q3-Q4 2012
Main Study1,2
Mean
Median
1.77
1.58
2.26
1.88
1.64
1.45
2.31
2.10
3.27
3.07
12.08
11.03
2.38
2.08
2.48
2.25
0.62
0.47
0.62
0.50
0.45
0.28
1.25
1.10
1.35
1.17

2012
Questionnaire
Field Test1,3
Mean
Median
1.56
1.37
2.11
1.72
1.09
1.07
2.07
1.92
3.01
2.88
13.45
11.55
2.35
2.00
2.55
2.28
0.69
0.52
0.63
0.52
0.61
0.26
1.56
1.30
1.56
1.34
0.39
0.30
6.00
5.08
2.28
1.98
0.85
0.70
0.89
0.75
0.77
0.65

6.31

5.70

6.26

5.70

2.93

2.65

2.90

2.62

2.95

2.44

1.83

1.62

1.84

1.62

1.01

0.78

1.90

1.67

1.90

1.68

0.96

0.77

1.76

1.55

1.68

1.48

0.84

0.68
(continued)

See notes at end of table.

116

Table 4.9j Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 26 to 49) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

21.96
1.55
2.95
0.48
3.58
1.58
1.73
1.81
1.25

19.97
1.40
2.67
0.42
2.77
1.45
1.40
1.47
1.08

21.55
1.54
2.88
0.46
3.50
1.53
1.71
1.77
1.23

19.67
1.40
2.63
0.40
2.70
1.42
1.40
1.43
1.07

20.43
0.57
2.85
0.51
3.63

18.67
0.52
2.53
0.45
2.92

1.44
1.81
2.62

1.25
1.46
2.33

2.30
1.42
2.44

1.93
1.28
2.15

2.22
1.40
2.37

1.87
1.25
2.08

2.20
1.24
2.39

1.88
1.08
1.93

3.66
3.46

3.33
1.57

3.59
3.42

3.25
1.62

3.59
3.44

3.15
1.77

0.68

0.60

0.67

0.60

0.53

0.50

5.62
0.22
5.06

5.52
0.13
5.00

5.60
0.23
5.02

5.45
0.13
4.93

4.14
0.61
2.23

3.72
0.48
2.03
(continued)

See notes at end of table.

117

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
1.40
1.22
0.30
0.22

Q3-Q4 2012
Main Study1,2
Mean
Median
1.44
1.25
0.31
0.22

1.32
3.48
2.87
61.54

1.33
3.43
3.01
60.87

1.23
3.00
2.42
58.55

1.23
3.03
2.57
57.88

2012
Questionnaire
Field Test1,3
Mean
Median
1.39
1.23
0.41
0.35
0.16
0.00
1.41
1.23
2.96
2.52
3.14
2.73
57.39
53.90

118

Table 4.9k Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 50 to 64)

2011
Main Study1
Mean
Median
2.09
1.73
2.51
2.00
1.74
1.50
2.60
2.33
4.05
3.95
14.43
12.97
2.77
2.33
2.88
2.65
0.71
0.55
0.73
0.57
0.46
0.33
1.53
1.28
1.63
1.40

Q3-Q4 2012
Main Study1,2
Mean
Median
1.99
1.70
2.42
1.90
1.73
1.52
2.55
2.28
4.13
4.10
14.44
13.09
2.79
2.38
2.93
2.68
0.70
0.52
0.71
0.58
0.40
0.32
1.46
1.28
1.65
1.33

2012
Questionnaire
Field Test1,3
Mean
Median
1.70
1.48
2.24
1.85
1.39
1.48
2.40
2.08
4.26
4.15
16.52
14.85
2.84
2.33
3.08
2.78
0.71
0.62
0.63
0.58
0.33
0.34
1.79
1.64
2.03
1.73
0.57
0.38
7.35
6.30
2.91
2.42
1.14
0.92
1.20
0.93
1.10
0.83

7.42

6.68

7.36

6.86

3.39

3.03

3.35

2.98

3.57

3.03

2.11

1.95

2.04

1.89

1.30

0.98

2.33

1.97

2.19

2.03

1.26

0.97

1.83

1.55

1.69

1.43

1.19

0.90
(continued)

See notes at end of table.

119

Table 4.9k Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 50 to 64) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

24.19
1.78
3.51
0.65
3.56
1.77
1.94
1.97
1.74

21.80
1.57
3.15
0.52
2.75
1.60
1.58
1.65
1.47

24.21
1.78
3.52
0.62
3.59
1.85
1.92
2.01
1.76

22.00
1.58
3.20
0.52
2.90
1.71
1.58
1.65
1.52

22.14
0.67
3.45
0.78
3.88

20.23
0.62
3.20
0.63
3.07

1.57
1.90
3.52

1.38
1.63
3.23

2.94
1.66
2.91

2.38
1.50
2.52

2.68
1.67
2.76

2.26
1.50
2.47

2.48
1.50
2.79

2.11
1.40
2.67

4.46
4.03

4.00
1.82

4.52
4.07

4.10
1.73

4.14
3.59

3.63
1.66

0.74

0.65

0.74

0.67

0.73

0.57

5.24
0.18
4.79

5.18
0.12
4.82

5.18
0.20
4.70

5.13
0.12
4.75

4.60
0.67
2.50

4.17
0.55
2.32
(continued)

See notes at end of table.

120

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
1.03
0.85
0.30
0.22

Q3-Q4 2012
Main Study1,2
Mean
Median
1.15
0.92
0.33
0.23

1.38
3.48
3.12
66.96

1.39
3.48
3.35
67.30

1.23
3.02
2.60
63.13

1.25
3.03
2.72
63.97

2012
Questionnaire
Field Test1,3
Mean
Median
1.13
0.98
0.50
0.38
0.18
0.00
1.71
1.50
3.45
3.00
3.83
2.95
66.24
62.25

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Some module rows are shown in bold for consistency with Tables 4.9a to 4.9f for all respondents. However, mean
affirmative gate timings in this table for modules in bold are not necessarily mutually exclusive and are not intended to
sum to the overall mean questionnaire timing.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

121

Table 4.9l Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 65+)

2011
Main Study1
Mean
Median
2.15
1.80
2.74
2.25
1.89
1.62
3.01
2.68
4.86
4.73
17.28
15.85
3.33
2.90
3.49
3.20
0.84
0.66
0.86
0.68
0.46
0.47
1.83
1.28
2.44
2.07

Q3-Q4 2012
Main Study1,2
Mean
Median
1.92
1.70
2.64
2.17
1.83
1.62
3.05
2.67
4.92
4.75
17.35
16.02
3.31
2.93
3.41
3.15
0.81
0.67
0.78
0.68
0.91
0.42
2.42
1.53
2.37
2.03

2012
Questionnaire
Field Test1,3
Mean
Median
1.69
1.50
2.66
2.30
1.52
1.57
2.89
2.32
5.32
5.13
22.36
19.56
3.00
2.45
3.77
3.62
1.09
0.80
1.09
0.88
0.39
0.39
2.02
2.25
1.66
1.66
0.53
0.42
10.67
8.82
4.28
3.05
1.69
1.27
1.71
1.27
1.62
1.25

9.36

8.39

9.05

7.77

4.30

3.98

3.94

3.72

5.10

3.73

3.11

2.53

3.01

2.57

1.93

1.43

2.85

2.48

2.91

2.33

1.77

1.27

3.45

2.12

3.34

1.90

1.80

1.30
(continued)

See notes at end of table.

122

Table 4.9l Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 65+) (continued)

2012
Questionnaire
Field Test1,3
Mean
Median

2011
Main Study1
Mean
Median

Q3-Q4 2012
Main Study1,2
Mean
Median

26.51
2.06
4.59
0.85
3.49
2.14
1.72
2.14
2.47

24.20
1.87
4.05
0.57
2.95
1.84
1.45
1.88
2.15

26.80
2.08
4.53
0.63
3.51
1.71
1.74
2.22
2.56

24.84
1.90
3.98
0.53
2.93
1.18
1.47
1.90
2.18

26.64
0.75
4.36
0.84
3.64

23.87
0.67
3.85
0.64
2.89

1.82
2.36
4.75

1.45
1.95
4.35

3.37
2.29
4.80

2.85
2.02
4.42

3.38
2.24
3.49

2.80
1.98
3.33

3.47
1.96
4.80

3.19
1.77
4.80

5.76
3.80

5.17
1.33

5.90
3.89

5.32
1.33

5.66
2.58

4.93
1.07

0.86

0.73

0.83

0.72

0.80

0.68

2.93
0.16
2.52

1.82
0.12
1.38

3.09
0.16
2.63

1.88
0.12
1.43

5.00
0.90
2.08

4.40
0.68
1.75
(continued)

See notes at end of table.

123

Table 4.9l Overall and Module Mean/Median Timing Data for the 2011 Main Study, Q3-Q4 2012
Main Study, and 2012 Questionnaire Field Test in Minutes (Affirmative Gate
Respondents Aged 65+) (continued)

Module
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance9
Income9
Verification
Overall Questionnaire

2011
Main Study1
Mean
Median
0.82
0.62
0.32
0.20
1.46
3.89
3.62
72.32

1.30
3.28
2.92
68.43

Q3-Q4 2012
Main Study1,2
Mean
Median
0.87
0.67
0.32
0.20
1.49
3.73
3.76
72.70

1.32
3.32
3.10
69.39

2012
Questionnaire
Field Test1,3
Mean
Median
0.96
0.73
0.48
0.42
0.32
0.00
2.13
1.93
4.43
3.98
3.98
3.15
80.24
74.45

ACASI = audio computer-assisted self-interviewing; NA = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Some module rows are shown in bold for consistency with Tables 4.9a to 4.9f for all respondents. However, mean
affirmative gate timings in this table for modules in bold are not necessarily mutually exclusive and are not intended to
sum to the overall mean questionnaire timing.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The calendar appears before the beginning ACASI and tutorial in the 2011 main study and Q3-Q4 2012 main study and is
interviewer-administered. The calendar follows the tutorial in the QFT and is self-administered.
5
Prescription drug modules for the 2011 main study and Q3-Q4 2012 main study include only a main module. For the QFT,
timings for the screener sections are included in the overall screener plus main module timings.
6
These modules comprise the total noncore ACASI for the 2011 main study and Q3-Q4 2012 main study, and this measure
includes timing for the ENDAUDIO question that the interviewer completes to close out the ACASI section. The mean total
noncore ACASI timing for the QFT sections completed only by the respondent is the sum of the mean timings for special
drugs to consumption of alcohol and back-end demographics.
7
The back-end demographics module is interviewer-administered in the 2011 main study and Q3-Q4 2012 main study.
The timing data for the QFT include timing for the ENDAUDIO question.
8
Timings for the education module in the QFT include non-education questions in this section of the questionnaire (marital
status, number of times married, military families).
9
In all datasets, the respondent or an adult proxy who is a family member may complete the health insurance and income
modules. In the 2011 main study and Q3-Q4 2012 main study, the health insurance and income modules are intervieweradministered. In the QFT, these modules are self-administered for the respondent or a proxy.

124

4.5.2

Selected Detailed Interview Timing Data for the 2012 Questionnaire Field Test and
the 2011 and 2012 Quarter 3 and 4 Comparison Data

Administration times for the 2011 and 2012 quarters 3 and 4 comparison samples and the
QFT instrument were calculated according to standard timing data calculation procedures for a
number of specific questionnaire sections. Tables 4.10a through 4.10v present unweighted
overall QFT timing results and results for selected modules for all respondents and for five
separate age groups.16 Timing results by age group for each section are presented in separate
tables for the QFT interviews, the 2011 comparison interviews, and the 2012 quarters 3 and 4
comparison interviews. For each age category, these tables provide the number of interviews, the
number of extreme or missing records, summary statistics, quartiles, percentiles, and the highest
and lowest extreme cases. Respondents with an overall interview administration time of less than
30 minutes or greater than 240 minutes were classified as outliers and were excluded from these
timing results.
As noted in Section 4.5.1.1, the partially redesigned QFT instrument took less than
60 minutes on average to administer among all respondents aged 12 or older, as shown in
Table 4.10a. Examining timing data within age groups reveals that respondents aged 65 or older
experienced the longest average administration times among all age groups, with an overall mean
of more than 80 minutes. Respondents aged 50 to 64 also had a mean administration time that
was considerably higher than the mean for all QFT respondents. Mean interview timings for
respondents aged 12 to 17 were similar to the overall mean for QFT respondents, while the
average times for respondents aged 18 to 25 and those aged 26 to 49 were lower than the overall
mean for QFT respondents. The overall timing patterns across age groups for QFT respondents
were rather consistent with the patterns for the 2011 comparison data interviews and the 2012
quarters 3 and 4 comparison interviews, as shown in Tables 4.10b and 4.10c.
Tables 4.10d through 4.10f provide timing results for the tobacco module for respondents
who answered the question LEADCIG in the QFT interviews, the 2011 comparison interviews,
and the 2012 quarters 3 and 4 comparison interviews. One difference between the QFT
questionnaire and the 2011 and 2012 quarters 3 and 4 questionnaire was that questions about
chewing tobacco and snuff were combined in the tobacco module for the QFT questionnaire.
This change was intended to increase efficiency in collecting age of first use, recency, and
frequency of smokeless tobacco use. In addition, this section in the QFT questionnaire no longer
collected data on the brand of smokeless tobacco that the respondent has used. As expected, the
efficiencies produced by these changes to the QFT questionnaire resulted in a slightly lower
mean timing for this module among QFT respondents (1.83) compared with the 2011
comparison respondents (2.02) and the 2012 quarters 3 and 4 comparison respondents (1.96).
As Tables 4.10g through 4.10j indicate, older respondents generally took more time than
younger respondents to complete the four prescription drug module screeners—pain relievers,
tranquilizers, stimulants, and sedatives. The new screeners included in the QFT questionnaire
asked respondents to report any past year use of prescription pain relievers, tranquilizers,
stimulants, and sedatives. These screener questions then asked respondents to report all use of
16

To aid in their readability, Tables 4.10a through 4.10v appear together at the end of this discussion in
Section 4.5.2.

125

drugs in each category, both those that were prescribed and those that were misused. The mean
pain relievers screener administration time was nearly 2½ minutes, which was the longest of the
four screeners. Because the prescription drug screeners were new in the QFT instrument, timing
data for these sections cannot be compared with the 2011 and 2012 quarters 3 and 4 comparison
interviews.
In the QFT instrument, the four prescription drug main modules followed the screeners
and asked, for each drug used in the past year, whether respondents misused any of them.
Respondents who reported never using a particular class of drug in the past year skip the main
module and are excluded from the timing data for the four prescription drug main modules
presented in presented in Tables 4.10k through 4.10v. These tables provide timing results for the
prescription drug main modules for the QFT interviews, 2011 comparison interviews, and 2012
quarters 3 and 4 comparison interviews. Among QFT respondents who answered questions in the
pain reliever, tranquilizer, and stimulant main modules, those aged 18 to 25 had the longest mean
administration times (Table 4.10k). This finding did not hold in the 2011 and 2012 quarters 3
and 4 comparison samples, where respondents aged 65 or older generally had the longest mean
administration times for these prescription drug modules among all age groups (Tables 4.10l and
4.10m). For the sedatives main module, respondents aged 65 or older had the longest mean
administration times among all age groups for the QFT interviews, 2011 comparison interviews,
and 2012 quarters 3 and 4 comparison interviews (Tables 4.10t, 4.10u, and 4.10v).
Overall, excluding the new prescription drug screeners, the mean timings for each of the
four prescription drug main modules were lower for QFT respondents than for the 2011 and
2012 quarters 3 and 4 comparison respondents. As noted in Section 4.5.1.1, the redesign of the
prescription drug modules was a major factor in increasing the overall burden on respondents
aged 65 or older in completing this questionnaire. Based on the QFT timing data, the additional
amount of time that respondents aged 65 or older took to complete the partially redesigned
questionnaire was significantly longer—about 8 minutes longer—than in the 2011 and 2012
comparison data interviews.

126

Table 4.10a Unweighted Overall Interview Timing Data for the Questionnaire Field Test Protocol
in Minutes, in Total and by Age Groups: All QFT Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1,2
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

59.53
390.36
19.76

59.56
246.82
15.71

54.26
269.00
16.40

57.39
375.11
19.37

66.24
424.96
20.61

80.24
862.65
29.37

228.47
68.27
55.99
46.08
30.13
47.20
198.33

170.48
67.95
57.17
48.53
31.52
64.30
138.97

140.88
62.92
50.80
42.73
30.13
49.72
110.75

191.52
65.90
53.90
44.01
30.13
39.22
161.38

149.88
76.67
62.25
51.97
34.70
.
115.18

228.47
88.07
74.45
62.22
39.97
64.95
188.50

122.97
95.23
82.98
39.07
35.97
31.45

106.88
85.78
79.33
42.40
38.88
34.65

113.00
82.88
74.25
36.73
33.40
30.48

121.88
94.83
80.87
37.63
34.78
31.32

126.15
106.90
94.50
42.72
40.27
35.93

174.25
148.20
112.32
53.98
48.32
41.77

228.47
191.52
174.25
173.52
171.93
30.43
30.30
30.13
30.13
30.13

170.48
135.07
115.90
115.13
107.18
34.52
34.05
33.28
33.20
31.52

140.88
125.35
120.50
116.13
113.00
30.48
30.45
30.45
30.43
30.13

191.52
171.93
148.27
129.47
125.18
31.05
30.85
30.30
30.13
30.13

149.88
126.15
122.97
119.97
119.63
38.07
37.65
36.72
35.93
34.70

228.47
174.25
173.52
168.10
160.88
47.02
46.17
42.87
41.77
39.97

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)

Q = quarter; QFT = Questionnaire Field Test.
1
2

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.
Because the QFT interviews included a higher number of cases with extreme values, which were excluded from this Table 4.10
series of tables (as indicated in footnote 1), the overall mean and median timings for the QFT, 2011 comparison data, and 2012
comparison data interviews were also calculated with the extreme values included. Including the extreme cases had minimal
impact on the overall mean and median interview times for the 2011 and 2012 comparison data. The impact on the overall mean
and median interview times for the QFT was somewhat greater, resulting in decreases of about 0.5 minutes for both the overall
mean and median timing. Given that including the extreme cases resulted in slightly decreased overall mean and median
interview times for the QFT, including the extreme cases would lead to similar conclusions as those drawn from comparing the
QFT timing data with the 2011 and 2012 comparison data interviews with the extreme cases excluded.

127

Table 4.10b Unweighted Overall Interview Timing Data for the 2011 Comparison Protocol in
Minutes, in Total and by Age Groups: All 2011 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,747

22,376

21,582

14,279

4,673

2,837

181

61.37
288.36
16.98

60.74
230.79
15.19

59.27
263.23
16.22

61.54
307.81
17.54

66.96
399.00
19.97

72.32
445.89
21.12

236.17
69.70
58.62
49.67
30.02
55.73
206.15

236.17
68.67
58.70
50.22
30.23
49.92
205.93

234.93
67.28
56.58
48.05
30.02
52.95
204.92

222.57
69.83
58.55
49.43
30.12
54.15
192.45

218.43
76.68
63.13
52.87
30.65
58.20
187.78

194.58
83.07
68.43
57.25
32.05
54.38
162.53

115.32
92.32
82.73
43.03
39.80
34.52

105.90
87.68
80.00
43.73
40.50
35.02

110.25
88.77
79.78
41.68
38.58
33.70

117.50
94.33
83.83
42.92
39.62
34.08

131.70
105.55
93.25
45.57
42.03
36.63

137.32
110.52
99.53
48.88
44.72
38.05

236.17
234.93
228.00
222.63
222.57
30.08
30.07
30.05
30.05
30.02

236.17
228.00
220.82
209.50
207.32
30.45
30.35
30.28
30.28
30.23

234.93
222.63
215.25
209.02
208.87
30.08
30.07
30.05
30.05
30.02

222.57
212.67
211.48
205.88
194.20
30.35
30.25
30.23
30.13
30.12

218.43
215.88
169.97
165.40
163.03
32.33
32.12
31.88
31.45
30.65

194.58
191.63
183.68
177.35
173.73
33.63
32.75
32.40
32.35
32.05

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

128

Table 4.10c Unweighted Overall Interview Timing Data for the 2012 Comparison Protocol in
Minutes, in Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,084

10,438

10,267

6,826

2,189

1,364

129

60.97
291.15
17.06

60.51
242.33
15.57

58.59
253.74
15.93

60.87
306.08
17.50

67.30
385.92
19.64

72.70
474.65
21.79

237.43
69.42
58.30
49.12
30.02
52.28
207.42

237.43
68.53
58.55
49.78
30.55
47.22
206.88

229.95
66.67
56.05
47.63
30.02
50.53
199.93

227.67
69.60
57.88
48.73
30.03
52.13
197.63

202.00
76.93
63.97
53.72
30.80
45.90
171.20

218.40
85.14
69.39
57.28
31.97
43.58
186.43

115.67
91.90
82.23
42.52
39.02
33.97

107.68
87.53
79.63
43.33
39.88
34.68

108.98
87.58
78.57
41.30
37.88
33.55

116.32
93.32
82.43
42.13
38.53
33.77

130.68
102.50
92.83
45.77
42.02
35.77

140.08
111.08
100.07
48.62
43.58
35.55

237.43
229.95
228.20
227.67
225.62

237.43
228.20
225.62
221.42
215.20

229.95
187.40
186.87
178.53
174.98

227.67
204.18
195.47
170.45
168.27

202.00
196.90
179.37
167.33
165.27

218.40
217.73
170.68
167.10
159.80

30.12
30.07
30.05
30.03
30.02

30.70
30.70
30.63
30.55
30.55

30.13
30.12
30.12
30.07
30.02

30.57
30.55
30.38
30.05
30.03

32.47
32.42
32.18
32.05
30.80

33.32
33.18
33.07
32.43
31.97

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

129

Table 4.10d Unweighted Overall Interview Timing Data for the QFT Tobacco Module in Minutes,
in Total and by Age Groups: All QFT Respondents Answering LEADCIG
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

1.83
2.09
1.45

1.41
0.86
0.93

1.85
2.10
1.45

1.89
2.35
1.53

2.24
3.07
1.75

2.57
2.89
1.70

21.68
2.40
1.43
0.88
0.20
0.73
21.48

8.05
1.58
1.13
0.85
0.35
1.10
7.70

13.97
2.63
1.53
0.75
0.22
0.28
13.75

21.68
2.50
1.62
0.89
0.20
0.38
21.48

13.47
3.00
1.67
1.07
0.28
0.83
13.18

11.15
3.27
2.20
1.57
0.32
1.87
10.83

6.65
4.25
3.62
0.53
0.40
0.28

4.95
3.50
2.58
0.65
0.57
0.43

5.97
4.10
3.77
0.43
0.33
0.25

6.82
4.37
3.52
0.43
0.37
0.25

8.68
5.22
4.23
0.70
0.57
0.32

8.97
5.70
4.68
0.73
0.58
0.40

21.68
13.97
13.47
11.98
11.15

8.05
5.52
5.43
5.42
5.37

13.97
11.98
6.00
5.98
5.97

21.68
10.53
8.27
7.07
7.07

13.47
8.68
8.20
8.15
6.80

11.15
8.97
7.32
6.83
6.58

0.23
0.22
0.22
0.22
0.20

0.42
0.42
0.40
0.38
0.35

0.25
0.25
0.25
0.23
0.22

0.23
0.23
0.22
0.22
0.20

0.43
0.35
0.33
0.32
0.28

0.53
0.50
0.42
0.40
0.32

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

130

Table 4.10e Unweighted Overall Interview Timing Data for the 2011 Tobacco Module in Minutes,
in Total and by Age Groups: All 2011 Comparison Respondents Answering LEADCIG
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,737

22,369

21,580

14,278

4,673

2,837

181

2.02
2.12
1.46

1.77
1.40
1.18

2.06
2.28
1.51

2.06
2.05
1.43

2.41
3.31
1.82

2.82
3.53
1.88

28.68
2.60
1.70
1.02
0.07
0.83
28.62

22.08
2.13
1.48
0.97
0.18
0.83
21.90

24.88
2.83
1.77
0.90
0.08
0.42
24.80

24.10
2.68
1.78
1.07
0.07
1.40
24.03

28.68
3.00
2.02
1.30
0.13
1.48
28.55

25.37
3.52
2.33
1.67
0.20
2.10
25.17

6.93
4.70
3.80
0.63
0.48
0.32

6.30
4.08
3.08
0.70
0.60
0.47

6.85
4.82
3.98
0.50
0.38
0.25

6.98
4.58
3.75
0.60
0.47
0.32

8.25
5.45
4.40
0.75
0.57
0.38

9.25
6.13
5.00
1.05
0.78
0.52

28.68
27.12
25.37
24.93
24.88

22.08
19.32
15.23
13.78
12.62

24.88
24.17
21.58
21.27
15.80

24.10
23.98
23.52
16.47
13.70

28.68
27.12
24.93
22.45
22.25

25.37
23.93
20.32
17.77
15.12

0.12
0.10
0.10
0.08
0.07

0.27
0.27
0.27
0.23
0.18

0.12
0.12
0.12
0.10
0.08

0.17
0.15
0.12
0.10
0.07

0.18
0.17
0.17
0.13
0.13

0.35
0.32
0.30
0.27
0.20

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

131

Table 4.10f Unweighted Overall Interview Timing Data for the 2012 Tobacco Module in Minutes,
in Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,055

10,424

10,261

6,820

2,188

1,362

129

1.96
1.87
1.37

1.70
1.19
1.09

1.96
1.99
1.41

2.02
1.84
1.36

2.39
2.91
1.71

2.85
3.10
1.76

22.43
2.52
1.67
0.98
0.12
0.82
22.32

17.28
2.08
1.45
0.97
0.27
0.82
17.02

16.20
2.70
1.67
0.83
0.12
0.43
16.08

20.60
2.63
1.75
1.07
0.13
0.50
20.47

22.43
3.02
2.00
1.30
0.12
1.75
22.32

16.95
3.63
2.38
1.72
0.13
2.07
16.82

6.63
4.57
3.68
0.62
0.47
0.30

5.63
3.75
2.88
0.72
0.60
0.47

6.45
4.60
3.83
0.47
0.37
0.25

6.50
4.57
3.70
0.60
0.45
0.30

8.53
5.52
4.38
0.77
0.57
0.38

8.85
6.07
5.10
1.07
0.80
0.52

22.43
20.60
17.28
16.95
16.27

17.28
14.93
13.65
11.53
11.25

16.20
13.18
12.28
10.77
10.25

20.60
11.78
10.98
10.83
10.70

22.43
13.42
13.27
13.12
12.77

16.95
16.27
12.13
10.52
10.45

0.13
0.13
0.13
0.12
0.12

0.30
0.30
0.28
0.27
0.27

0.17
0.17
0.15
0.13
0.12

0.18
0.18
0.18
0.17
0.13

0.28
0.27
0.27
0.23
0.12

0.43
0.35
0.35
0.28
0.13

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

132

Table 4.10g Unweighted Overall Interview Timing Data for the QFT Pain Relievers Screener in
Minutes, in Total and by Age Groups: All Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

2.42
3.09
1.76

2.35
1.53
1.24

1.98
0.90
0.95

2.28
1.64
1.28

2.91
3.53
1.88

4.28
20.12
4.49

43.75
2.72
2.03
1.57
0.43
1.83
43.32

9.47
2.68
2.03
1.60
0.78
1.40
8.68

10.13
2.28
1.78
1.43
0.43
1.50
9.70

12.58
2.61
1.98
1.53
0.60
1.83
11.98

12.27
3.17
2.42
1.85
0.90
1.90
11.37

43.75
4.28
3.05
2.38
1.20
3.05
42.55

9.18
4.72
3.70
1.27
1.10
0.85

7.95
4.70
3.70
1.30
1.10
0.88

5.45
3.50
2.95
1.13
1.00
0.72

8.77
4.33
3.50
1.25
1.12
0.90

12.22
6.80
4.58
1.50
1.38
1.07

19.43
10.45
8.03
1.97
1.83
1.45

43.75
19.43
16.03
12.58
12.27
0.68
0.62
0.60
0.50
0.43

9.47
9.30
8.78
8.48
8.27
0.87
0.85
0.82
0.82
0.78

10.13
8.27
7.28
5.60
5.45
0.72
0.68
0.62
0.50
0.43

12.58
11.82
10.53
9.43
9.38
0.88
0.80
0.75
0.75
0.60

12.27
12.22
11.02
9.18
9.03
1.30
1.18
1.13
1.07
0.90

43.75
19.43
16.03
12.25
11.83
1.70
1.68
1.52
1.45
1.20

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

133

Table 4.10h Unweighted Overall Interview Timing Data for the QFT Tranquilizer Screener in
Minutes, in Total and by Age Groups: All Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

0.88
0.57
0.75

0.81
0.34
0.59

0.70
0.34
0.58

0.85
0.42
0.65

1.14
0.83
0.91

1.69
1.85
1.36

9.85
1.00
0.70
0.52
0.15
0.57
9.70

6.28
0.92
0.67
0.50
0.20
0.57
6.08

9.85
0.78
0.58
0.47
0.15
0.47
9.70

8.02
0.95
0.70
0.53
0.15
0.58
7.87

5.23
1.23
0.92
0.70
0.30
0.97
4.93

8.33
1.85
1.27
0.88
0.48
1.23
7.85

4.97
1.87
1.42
0.40
0.35
0.27

3.27
1.68
1.30
0.38
0.33
0.27

2.25
1.35
1.12
0.37
0.30
0.22

3.30
1.75
1.35
0.42
0.35
0.28

5.20
2.68
1.75
0.48
0.42
0.33

7.90
4.97
3.60
0.75
0.65
0.52

9.85
8.33
8.02
7.90
6.95

6.28
5.98
4.70
3.85
3.67

9.85
5.10
2.95
2.50
2.25

8.02
6.95
6.12
5.10
4.67

5.23
5.20
5.18
5.18
5.10

8.33
7.90
5.15
5.13
5.07

0.20
0.18
0.17
0.15
0.15

0.27
0.23
0.22
0.22
0.20

0.22
0.22
0.18
0.17
0.15

0.27
0.27
0.27
0.23
0.15

0.38
0.37
0.35
0.33
0.30

0.63
0.62
0.53
0.52
0.48

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

134

Table 4.10i Unweighted Overall Interview Timing Data for the QFT Stimulant Screener in
Minutes, in Total and by Age Groups: All Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

0.92
0.62
0.79

0.83
0.30
0.55

0.72
0.16
0.39

0.89
0.70
0.84

1.20
1.04
1.02

1.71
1.82
1.35

16.55
1.03
0.75
0.55
0.15
0.57
16.40

5.52
0.98
0.72
0.52
0.17
0.58
5.35

3.55
0.83
0.63
0.47
0.17
0.53
3.38

16.55
1.02
0.75
0.57
0.15
0.63
16.40

5.83
1.25
0.93
0.70
0.30
0.93
5.53

6.53
1.73
1.27
0.95
0.47
1.55
6.07

5.23
1.85
1.47
0.40
0.35
0.25

3.08
1.72
1.35
0.40
0.33
0.25

2.22
1.47
1.15
0.35
0.30
0.18

4.38
1.75
1.35
0.42
0.35
0.25

5.58
4.18
1.72
0.57
0.50
0.32

6.22
5.42
3.25
0.82
0.72
0.47

16.55
6.53
6.22
6.05
5.90
0.17
0.17
0.17
0.17
0.15

5.52
5.13
3.80
3.58
3.42
0.23
0.23
0.22
0.22
0.17

3.55
2.90
2.68
2.55
2.22
0.18
0.17
0.17
0.17
0.17

16.55
5.85
5.53
5.42
4.98
0.22
0.22
0.22
0.22
0.15

5.83
5.58
5.53
5.25
5.25
0.40
0.38
0.35
0.32
0.30

6.53
6.22
6.05
5.90
5.50
0.62
0.58
0.58
0.47
0.47

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

135

Table 4.10j Unweighted Overall Interview Timing Data for the QFT Sedative Screener in Minutes,
in Total and by Age Groups: All Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

2,006

539

481

668

189

129

0.81
0.47
0.69

0.73
0.25
0.50

0.61
0.14
0.37

0.77
0.43
0.65

1.10
0.81
0.90

1.62
1.43
1.20

11.77
0.93
0.63
0.47
0.07
0.57
11.70

5.35
0.87
0.60
0.43
0.13
0.40
5.22

4.47
0.72
0.53
0.40
0.07
0.57
4.40

11.77
0.87
0.65
0.48
0.12
0.57
11.65

4.92
1.17
0.83
0.67
0.23
0.83
4.68

6.42
1.67
1.25
0.97
0.45
0.87
5.97

4.55
1.72
1.35
0.35
0.28
0.20

2.62
1.63
1.28
0.33
0.28
0.20

2.08
1.17
0.97
0.30
0.25
0.15

2.42
1.58
1.27
0.37
0.30
0.22

4.92
3.65
1.47
0.50
0.43
0.30

6.13
4.80
3.47
0.72
0.60
0.50

11.77
6.42
6.13
5.35
4.92
0.13
0.13
0.13
0.12
0.07

5.35
4.57
3.52
3.38
2.87
0.18
0.17
0.17
0.15
0.13

4.47
2.62
2.13
2.10
2.08
0.15
0.15
0.15
0.13
0.07

11.77
4.87
4.65
4.42
4.10
0.22
0.17
0.17
0.13
0.12

4.92
4.92
4.85
4.85
4.75
0.38
0.37
0.32
0.30
0.23

6.42
6.13
4.92
4.87
4.82
0.55
0.55
0.55
0.50
0.45

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

136

Table 4.10k Unweighted Overall Interview Timing Data for the Pain Reliever Module in Minutes,
in Total and by Age Groups: All QFT Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

1,150

171

274

476

142

894

1.05
1.73
1.31

1.03
1.42
1.19

1.31
2.63
1.62

0.94
1.55
1.25

0.88
0.78
0.88

1.21
1.72
1.31

12.65
1.10
0.65
0.37
0.07
0.23
12.58

7.58
1.10
0.62
0.43
0.08
0.45
7.50

12.65
1.70
0.67
0.35
0.07
0.28
12.58

11.85
0.98
0.58
0.32
0.07
0.23
11.78

8.02
1.00
0.69
0.47
0.13
0.40
7.88

11.22
1.33
0.95
0.58
0.17
0.42
11.05

7.20
3.62
2.48
0.22
0.15
0.10

7.20
3.32
2.50
0.23
0.17
0.10

8.28
4.03
3.53
0.22
0.13
0.10

5.95
3.50
1.85
0.18
0.15
0.10

5.27
1.82
1.38
0.28
0.23
0.13

11.22
2.97
2.10
0.42
0.33
0.17

12.65
11.85
11.22
10.13
8.52
0.10
0.08
0.07
0.07
0.07

7.58
7.20
6.62
4.45
4.38
0.15
0.13
0.13
0.10
0.08

12.65
8.28
8.28
7.57
6.77
0.10
0.10
0.10
0.07
0.07

11.85
10.13
8.52
8.12
5.95
0.10
0.10
0.10
0.10
0.07

8.02
5.27
3.47
3.45
2.22
0.22
0.18
0.17
0.13
0.13

11.22
4.48
3.37
2.98
2.97
0.33
0.28
0.27
0.20
0.17

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

137

Table 4.10l Unweighted Overall Interview Timing Data for the Pain Reliever Module in Minutes,
in Total and by Age Groups: All 2011 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,745

22,375

21,581

14,279

4,673

2,837

183

2.09
1.37
1.17

2.17
1.34
1.16

2.00
1.33
1.15

1.99
1.31
1.14

2.19
1.49
1.22

2.49
1.59
1.26

37.20
2.57
1.90
1.37
0.02
1.67
37.18

37.20
2.65
2.02
1.47
0.02
1.95
37.18

22.28
2.45
1.78
1.27
0.05
1.57
22.23

27.05
2.40
1.78
1.30
0.08
1.58
26.97

22.83
2.60
1.97
1.47
0.08
1.78
22.75

21.92
3.02
2.33
1.77
0.07
1.90
21.85

6.02
3.97
3.30
0.98
0.77
0.40

5.93
3.97
3.33
1.03
0.78
0.38

5.93
3.98
3.28
0.90
0.70
0.33

5.83
3.88
3.18
0.95
0.78
0.47

6.72
4.12
3.33
1.10
0.93
0.53

6.77
4.20
3.57
1.33
1.05
0.63

37.20
36.30
27.05
26.02
22.88
0.05
0.05
0.05
0.05
0.02

37.20
36.30
21.02
19.70
18.47
0.05
0.05
0.05
0.05
0.02

22.28
21.43
19.03
18.05
17.65
0.07
0.07
0.07
0.07
0.05

27.05
26.02
22.88
20.85
17.60
0.13
0.12
0.12
0.10
0.08

22.83
16.05
15.05
14.95
12.23
0.18
0.12
0.12
0.10
0.08

21.92
20.18
16.33
15.55
12.68
0.10
0.10
0.10
0.08
0.07

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

138

Table 4.10m Unweighted Overall Interview Timing Data for the Pain Reliever Module in Minutes,
in Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,084

10,438

10,267

6,826

2,189

1,364

129

2.08
1.23
1.11

2.18
1.24
1.11

1.95
1.19
1.09

1.99
1.20
1.10

2.23
1.22
1.11

2.48
1.14
1.07

21.67
2.55
1.88
1.37
0.03
1.63
21.63

21.23
2.67
2.03
1.50
0.07
1.80
21.17

18.00
2.38
1.73
1.25
0.03
1.30
17.97

21.67
2.42
1.78
1.30
0.10
1.48
21.57

17.90
2.68
2.03
1.55
0.20
1.78
17.70

12.55
3.05
2.33
1.77
0.20
2.02
12.35

5.85
3.90
3.28
1.00
0.78
0.42

5.98
3.90
3.32
1.05
0.78
0.38

5.68
3.88
3.20
0.90
0.72
0.38

5.82
3.78
3.22
0.98
0.80
0.45

6.33
4.03
3.38
1.17
0.97
0.63

5.85
4.13
3.60
1.35
1.15
0.77

21.67
21.23
18.42
18.00
17.90
0.10
0.08
0.08
0.07
0.03

21.23
18.42
14.80
14.73
14.13
0.12
0.10
0.10
0.08
0.07

18.00
17.10
13.52
11.97
11.78
0.12
0.10
0.10
0.08
0.03

21.67
17.82
13.03
12.13
10.60
0.15
0.12
0.12
0.10
0.10

17.90
13.98
8.78
8.08
7.73
0.48
0.47
0.43
0.38
0.20

12.55
11.50
10.15
9.17
7.58
0.48
0.45
0.42
0.32
0.20

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

139

Table 4.10n Unweighted Overall Interview Timing Data for the Tranquilizer Module in Minutes, in
Total and by Age Groups: All QFT Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

417

196

1,627

0.75
0.80
0.89

0.98
0.96
0.98

1.13
1.40
1.18

0.56
0.58
0.76

0.58
0.35
0.59

0.76
0.28
0.53

6.45
0.80
0.40
0.25
0.05
0.17
6.40

3.78
1.48
0.40
0.30
0.07
0.33
3.72

6.38
1.71
0.57
0.31
0.05
0.25
6.33

6.45
0.53
0.33
0.20
0.05
0.17
6.40

2.93
0.64
0.41
0.26
0.07
0.47
2.87

2.93
0.98
0.64
0.42
0.17
0.70
2.77

4.05
2.60
1.95
0.15
0.12
0.07

3.78
3.25
2.30
0.15
0.12
0.07

6.38
3.40
2.60
0.22
0.12
0.05

4.05
2.08
1.38
0.13
0.10
0.07

2.93
2.32
1.03
0.15
0.08
0.07

2.93
1.73
1.39
0.23
0.18
0.17

6.45
6.38
4.73
4.62
4.05
0.07
0.07
0.07
0.05
0.05

3.78
3.25
2.43
2.30
2.15
0.20
0.15
0.13
0.12
0.07

6.38
4.73
4.62
3.57
3.40
0.12
0.12
0.10
0.08
0.05

6.45
4.05
3.97
2.95
2.67
0.08
0.07
0.07
0.07
0.05

2.93
2.62
2.32
1.52
1.10
0.10
0.08
0.08
0.08
0.07

2.93
1.77
1.70
1.57
1.22
0.25
0.22
0.20
0.17
0.17

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

140

Table 4.10o Unweighted Overall Interview Timing Data for the Tranquilizer Module in Minutes, in
Total and by Age Groups: All 2011 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,744

22,374

21,581

14,279

4,673

2,837

184

1.15
0.58
0.76

1.19
0.50
0.71

1.04
0.60
0.78

1.11
0.53
0.73

1.30
0.62
0.79

1.63
0.89
0.94

39.18
1.48
0.98
0.65
0.02
0.73
39.17

10.97
1.55
1.05
0.68
0.02
0.65
10.95

39.18
1.32
0.87
0.57
0.02
0.55
39.17

11.18
1.40
0.93
0.63
0.03
0.82
11.15

11.60
1.65
1.12
0.77
0.07
0.73
11.53

15.40
2.15
1.47
0.98
0.07
1.20
15.33

3.48
2.48
2.07
0.43
0.33
0.18

3.35
2.43
2.10
0.45
0.35
0.17

3.30
2.32
1.90
0.38
0.30
0.17

3.57
2.40
1.97
0.43
0.35
0.22

3.93
2.68
2.35
0.53
0.42
0.25

4.50
2.93
2.68
0.67
0.53
0.27

39.18
22.78
22.18
15.40
14.77
0.03
0.03
0.03
0.02
0.02

10.97
10.27
9.27
9.03
8.63
0.05
0.05
0.03
0.03
0.02

39.18
22.78
22.18
14.77
13.27
0.05
0.03
0.03
0.03
0.02

11.18
10.58
10.13
8.57
8.40
0.07
0.07
0.05
0.05
0.03

11.60
8.87
7.73
7.60
7.53
0.12
0.10
0.10
0.08
0.07

15.40
9.52
9.00
8.42
8.35
0.08
0.08
0.08
0.07
0.07

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

141

Table 4.10p Unweighted Overall Interview Timing Data for the Tranquilizer Module in Minutes, in
Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,084

10,438

10,267

6,826

2,189

1,364

129

1.15
0.56
0.75

1.21
0.51
0.71

1.02
0.42
0.65

1.10
0.66
0.81

1.31
0.55
0.74

1.67
1.06
1.03

27.42
1.48
0.98
0.65
0.03
0.63
27.38

16.67
1.60
1.08
0.72
0.03
0.82
16.63

8.03
1.28
0.87
0.58
0.05
0.63
7.98

27.42
1.37
0.93
0.63
0.03
0.70
27.38

7.45
1.72
1.13
0.77
0.12
1.08
7.33

22.12
2.27
1.52
0.98
0.13
1.72
21.98

3.35
2.48
2.07
0.43
0.33
0.18

3.23
2.48
2.13
0.47
0.35
0.18

3.25
2.23
1.83
0.38
0.30
0.17

3.38
2.37
1.92
0.43
0.35
0.22

3.62
2.65
2.32
0.55
0.43
0.28

4.27
2.90
2.70
0.67
0.53
0.32

27.42
26.75
22.12
16.67
8.82
0.05
0.05
0.03
0.03
0.03

16.67
8.82
7.28
6.60
6.50
0.07
0.05
0.05
0.05
0.03

8.03
7.80
6.42
5.70
5.67
0.07
0.05
0.05
0.05
0.05

27.42
26.75
8.43
7.28
6.72
0.10
0.08
0.07
0.03
0.03

7.45
7.25
6.58
5.75
5.13
0.20
0.20
0.15
0.15
0.12

22.12
7.95
7.95
6.38
6.30
0.18
0.18
0.18
0.15
0.13

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

142

Table 4.10q Unweighted Overall Interview Timing Data for the Stimulants Module in Minutes, in
Total and by Age Groups: All QFT Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

256

1,788

0.82
1.04
1.02

0.97
0.81
0.90

1.15
1.28
1.13

0.47
0.83
0.91

0.66
0.61
0.78

0.64
0.11
0.33

7.97
1.06
0.38
0.20
0.02
0.17
7.95

3.98
1.15
0.65
0.38
0.10
0.22
3.88

4.02
1.95
0.58
0.20
0.08
0.25
3.93

7.97
0.47
0.23
0.17
0.02
0.17
7.95

3.20
0.73
0.45
0.22
0.05
0.45
3.15

1.32
0.77
0.62
0.37
0.18
0.77
1.13

3.98
3.12
2.25
0.12
0.08
0.05

3.98
3.20
2.10
0.22
0.12
0.10

4.02
3.42
2.98
0.13
0.10
0.08

7.97
1.38
0.83
0.08
0.07
0.02

3.20
3.20
1.63
0.07
0.05
0.05

1.32
1.32
0.98
0.25
0.18
0.18

7.97
4.02
3.98
3.98
3.65
0.07
0.05
0.05
0.03
0.02

3.98
3.20
3.03
2.10
1.82
0.23
0.22
0.22
0.12
0.10

4.02
3.98
3.65
3.48
3.42
0.10
0.10
0.10
0.08
0.08

7.97
3.50
1.87
1.43
1.38
0.07
0.07
0.05
0.03
0.02

3.20
1.63
1.18
0.87
0.73
0.22
0.13
0.12
0.07
0.05

1.32
0.98
0.77
0.77
0.68
0.58
0.52
0.37
0.25
0.18

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

143

Table 4.10r Unweighted Overall Interview Timing Data for the Stimulants Module in Minutes, in
Total and by Age Groups: All 2011 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,744

22,375

21,580

14,279

4,673

2,837

184

1.16
0.65
0.81

1.20
0.61
0.78

1.04
0.56
0.75

1.12
0.64
0.80

1.37
0.87
0.93

1.66
0.99
0.99

30.18
1.50
0.97
0.62
0.02
0.73
30.17

16.17
1.58
1.03
0.63
0.02
0.67
16.15

25.07
1.32
0.85
0.57
0.03
0.58
25.03

30.18
1.40
0.93
0.62
0.03
0.73
30.15

17.23
1.73
1.15
0.75
0.05
0.73
17.18

12.02
2.23
1.43
0.95
0.02
1.02
12.00

3.57
2.70
2.18
0.42
0.32
0.17

3.45
2.67
2.23
0.42
0.30
0.15

3.43
2.42
1.95
0.37
0.28
0.15

3.53
2.55
2.03
0.42
0.33
0.20

4.27
3.00
2.58
0.52
0.40
0.23

4.55
3.18
3.03
0.62
0.48
0.23

30.18
25.07
18.47
17.23
16.28
0.03
0.03
0.03
0.02
0.02

16.17
14.42
10.52
10.37
8.33
0.03
0.03
0.03
0.03
0.02

25.07
14.62
10.98
10.97
10.20
0.05
0.05
0.05
0.03
0.03

30.18
18.47
13.80
11.58
11.40
0.08
0.07
0.07
0.05
0.03

17.23
16.28
10.17
7.68
7.03
0.10
0.10
0.08
0.08
0.05

12.02
9.72
7.67
7.65
7.50
0.07
0.07
0.05
0.03
0.02

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

144

Table 4.10s Unweighted Overall Interview Timing Data for the Stimulants Module in Minutes, in
Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,084

10,438

10,267

6,826

2,189

1,364

129

1.16
0.64
0.80

1.23
0.64
0.80

1.02
0.48
0.69

1.10
0.65
0.80

1.36
0.70
0.84

1.71
1.12
1.06

26.47
1.50
0.97
0.63
0.03
0.75
26.43

21.15
1.65
1.05
0.67
0.03
0.85
21.12

11.63
1.30
0.85
0.55
0.05
0.57
11.58

26.47
1.38
0.93
0.62
0.03
0.68
26.43

9.57
1.77
1.15
0.77
0.10
0.92
9.47

9.17
2.32
1.47
0.93
0.15
0.85
9.02

3.53
2.70
2.17
0.42
0.32
0.17

3.48
2.72
2.27
0.42
0.32
0.15

3.37
2.37
1.90
0.37
0.28
0.13

3.57
2.48
1.93
0.43
0.33
0.20

3.68
3.02
2.53
0.52
0.43
0.27

4.52
3.20
3.05
0.63
0.50
0.27

26.47
21.15
15.07
11.63
11.38
0.05
0.05
0.03
0.03
0.03

21.15
11.38
10.63
8.27
7.55
0.05
0.05
0.05
0.03
0.03

11.63
6.72
6.35
6.13
6.08
0.07
0.07
0.07
0.05
0.05

26.47
15.07
10.33
9.42
8.78
0.08
0.08
0.07
0.05
0.03

9.57
8.00
7.88
7.52
5.95
0.17
0.15
0.13
0.12
0.10

9.17
9.05
8.97
8.87
8.67
0.22
0.22
0.20
0.20
0.15

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

145

Table 4.10t Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes, in
Total and by Age Groups: All QFT Respondents
Overall
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

12-17

18-25

26-49

50-64

65+

243

109

1,801

0.49
0.45
0.67

0.62
0.57
0.76

0.46
0.33
0.58

0.39
0.47
0.68

0.47
0.21
0.46

0.89
0.69
0.83

5.52
0.53
0.28
0.15
0.03
0.13
5.48

2.28
0.70
0.23
0.17
0.05
0.13
2.23

2.83
0.52
0.25
0.17
0.03
0.18
2.80

5.52
0.40
0.23
0.13
0.03
0.08
5.48

2.38
0.53
0.38
0.18
0.08
0.18
2.30

4.05
0.97
0.68
0.38
0.15
0.57
3.90

3.83
1.90
0.97
0.10
0.08
0.03

2.28
2.08
2.07
0.10
0.08
0.05

2.83
1.73
1.25
0.10
0.07
0.03

3.83
0.93
0.62
0.08
0.07
0.03

2.38
1.95
0.93
0.13
0.08
0.08

4.05
2.75
1.83
0.30
0.28
0.15

5.52
4.05
3.83
2.83
2.77
0.05
0.05
0.03
0.03
0.03

2.28
2.08
2.07
1.90
0.87
0.13
0.13
0.10
0.08
0.05

2.83
2.17
1.73
1.70
1.25
0.10
0.10
0.07
0.07
0.03

5.52
3.83
2.77
1.40
1.08
0.05
0.05
0.05
0.03
0.03

2.38
1.95
1.05
0.93
0.67
0.13
0.13
0.12
0.08
0.08

4.05
2.75
1.83
1.33
1.18
0.33
0.32
0.30
0.28
0.15

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

146

Table 4.10u Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes, in
Total and by Age Groups: All 2011 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

65,744

22,375

21,580

14,279

4,673

2,837

184

0.95
0.52
0.72

1.00
0.49
0.70

0.81
0.39
0.62

0.89
0.44
0.66

1.15
0.74
0.86

1.52
1.07
1.03

24.85
1.18
0.75
0.48
0.02
0.48
24.83

11.98
1.32
0.82
0.52
0.02
0.58
11.97

23.67
0.98
0.65
0.43
0.03
0.48
23.63

24.85
1.08
0.72
0.48
0.05
0.52
24.80

24.15
1.45
0.93
0.62
0.05
0.48
24.10

20.28
2.02
1.28
0.82
0.03
0.65
20.25

3.10
2.33
1.83
0.33
0.25
0.13

3.08
2.38
1.95
0.33
0.25
0.13

2.93
1.93
1.48
0.30
0.23
0.13

2.97
2.08
1.62
0.33
0.27
0.15

3.33
2.77
2.22
0.42
0.33
0.20

4.20
3.00
2.85
0.53
0.40
0.20

24.85
24.15
23.67
20.70
20.28
0.03
0.03
0.02
0.02
0.02

11.98
10.52
9.87
9.02
8.80
0.03
0.03
0.02
0.02
0.02

23.67
20.70
11.52
10.70
8.38
0.05
0.05
0.03
0.03
0.03

24.85
10.27
10.02
9.82
8.67
0.07
0.07
0.07
0.07
0.05

24.15
11.50
11.37
8.58
7.42
0.10
0.08
0.08
0.07
0.05

20.28
14.82
14.07
9.62
8.23
0.07
0.07
0.05
0.03
0.03

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

147

Table 4.10v Unweighted Overall Interview Timing Data for the Sedatives Module in Minutes, in
Total and by Age Groups: All 2012 Comparison Respondents
Sample Used in
Analysis
Extreme/Missing
Records1
Summary Statistics
(Minutes)1
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range

Overall

12-17

18-25

26-49

50-64

65+

31,083

10,437

10,267

6,826

2,189

1,364

130

0.94
0.48
0.69

1.03
0.48
0.69

0.78
0.34
0.58

0.87
0.35
0.59

1.13
0.61
0.78

1.56
1.25
1.12

22.12
1.18
0.75
0.50
0.03
0.50
22.08

7.30
1.35
0.85
0.53
0.03
0.43
7.27

22.12
0.95
0.65
0.43
0.05
0.48
22.07

10.18
1.07
0.72
0.48
0.03
0.43
10.15

16.92
1.40
0.93
0.63
0.03
0.92
16.88

15.28
2.14
1.32
0.83
0.07
1.37
15.22

3.05
2.32
1.83
0.33
0.25
0.15

3.10
2.42
2.02
0.35
0.27
0.15

2.82
1.83
1.40
0.30
0.23
0.13

2.93
2.02
1.57
0.35
0.27
0.17

3.15
2.68
2.17
0.42
0.33
0.22

4.10
3.00
2.85
0.53
0.42
0.20

22.12
16.92
15.28
13.53
13.22
0.03
0.03
0.03
0.03
0.03

7.30
6.72
6.47
6.22
5.97
0.05
0.05
0.05
0.03
0.03

22.12
9.88
7.67
6.78
5.83
0.07
0.07
0.07
0.05
0.05

10.18
6.80
6.75
6.68
6.62
0.07
0.05
0.05
0.03
0.03

16.92
7.20
5.03
4.72
4.35
0.12
0.10
0.07
0.05
0.03

15.28
13.53
13.22
10.62
8.45
0.15
0.15
0.15
0.12
0.07

Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest
(Highest)

5 Lowest

(Lowest)
1

Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.

148

4.5.3

Timing Data for High and Low Reports of Numbers of Prescription Drugs Used or
Misused in the Past Year in the QFT Sample
4.5.3.1

Procedures for Categorizing High and Low Reports of Prescription Drugs

Different cut points for extreme high numbers of prescription drugs used or misused were
chosen according to the distributions within age groups so that interview timing data would be
generated for the most extreme reports within a given age group. As much as possible, cut points
were chosen for the respondents in the 95th percentile among the past year users or misusers. For
example, a total of 733 QFT respondents reported any past year use of prescription pain
relievers, and 685 of these past year users (93.5 percent) reported use of one to six pain relievers.
The 12 past year users who reported use of exactly seven pain relievers comprised 1.6 percent of
the past year users, which yielded a cumulative percentage of 95.1 percent of past year users of
pain relievers who reported using one to seven pain relievers. Based on this review, a cut point of
past year use of seven or more pain relievers was chosen for the timing data for the pain relievers
screener among persons aged 12 or older.
Because the cut points for numbers of prescription drugs differ by age group, the sample
sizes for individual age groups do not sum to the total sample sizes used in the analyses for
persons aged 12 or older. For example, if a constant cut point of "seven or more" pain relievers
used in the past year had been picked as per the cut point for respondents aged 12 or older, only
five respondents aged 12 to 17, seven respondents aged 35 to 49, and five respondents aged 50 or
older reported past year use of this many pain relievers. In comparison, analyses of timing data
for the pain relievers screener by age group included 9 respondents aged 12 to 17,
11 respondents aged 35 to 49, and 11 respondents aged 50 or older (Table 4.11a). (To improve
readability, note that Tables 4.11a through 4.11p appear after all discussion of timing data in this
section.)
In addition, if the cut point is lower for a particular age group than for all respondents
aged 12 or older, the maximum interview time shown in that age group may be greater than the
maximum interview time shown for respondents aged 12 or older. For example, the maximum
time required to complete the pain relievers screener among respondents who reported past year
use of seven or more pain relievers was 7.28 minutes (Table 4.11a). A respondent aged 50 or
older who reported use of five or more pain relievers had a corresponding time of 8.03 minutes
but was below the "seven or more" threshold set for respondents aged 12 or older.
For timing data among QFT respondents who reported use or misuse of lower numbers of
prescription drugs, a constant criterion of exactly one pain reliever used or misused was applied
to all groups. For lower reports of use or misuse across all four prescription drug categories,
more variation in the cut points was applied to allow for respondents who might report use or
misuse across more than one drug category. However, upper limits of three prescription drugs
used in the past year and two prescription drugs misused would result in respondents reporting
use or misuse of drugs in less than all four of the categories.
The following timing data were run:

149

•

For respondents who reported any past year use of high numbers of pain relievers:
Pain relievers screener times (Table 4.11a) and total interview times (Table 4.11i).

•

For respondents who reported any past year use of only one pain reliever: Pain
relievers screener times (Table 4.11b) and total interview times (Table 4.11j).

•

For respondents who reported past year misuse of high numbers of pain relievers:
Pain relievers screener and main module times (Table 4.11c) and total interview
times (Table 4.11k).

•

For respondents who reported past year misuse of only one pain reliever: Pain
relievers screener and main module times (Table 4.11d and total interview times
(Table 4.11l).

•

For respondents who reported any past year use of high numbers of any prescription
drugs: All prescription drug screener timings for pain relievers through sedatives
(Table 4.11e) and total interview times (Table 4.11m).

•

For respondents who reported any past year use of lower numbers of any
prescription drugs: All prescription drug screener timings for pain relievers through
sedatives (Table 4.11f) and total interview times (Table 4.11n).

•

For respondents who reported past year misuse of high numbers of any prescription
drugs: All prescription drug screener and main module timings for pain relievers
through sedatives (Table 4.11g) and total interview times (Table 4.11o).

•

For respondents who reported past year misuse of lower numbers of any prescription
drugs: All prescription drug screener and main module timings for pain relievers
through sedatives (Table 4.11h) and total interview times (Table 4.11p).

Unlike the standard timing analyses, timing data from respondents who had extreme low (less
than 30 minutes) or extreme high (greater than 240 minutes) total interview times were retained
for these analyses. The tables indicate the numbers of cases that would have been excluded if
these criteria had been applied.
4.5.3.2

Key Findings on High and Low Reports of Prescription Drugs

In general, there was not much difference in the amount of time needed to complete the
screener sections for pain relievers or for all prescription drugs for respondents who reported use
of high numbers of prescription drugs and those who reported use of lower numbers.
•

The average time to complete the pain relievers screener was 2.48 minutes for
respondents aged 12 or older who reported use of seven or more pain relievers in the
past year (Table 4.11a) and 2.24 minutes for respondents who used only one pain
reliever (Table 4.11b).

•

Maximum times to complete the pain relievers screener according to the number of
drugs that were used were 8.03 minutes for a respondent aged 50 or older who
reported use of at least five but fewer than seven pain relievers, 7.28 minutes for a
respondent aged 12 or older who reported use of seven or more pain relievers, and
11.83 minutes for a respondent who used only one pain reliever.

150

•

The average time to complete all of the QFT prescription drug screeners was 5.33
minutes for respondents aged 12 or older who reported use of 11 or more prescription
drugs of any kind in the past year (Table 4.11e) and 4.69 minutes for respondents
who used one to three prescription drugs (Table 4.11f).

•

Maximum times to complete all of the prescription drug screeners according to the
number of drugs that were used were 13.18 minutes for a respondent aged 12 to 17
who reported use of at least 5 but fewer than 11 prescription drugs in the past year,
10.33 minutes for a respondent aged 12 or older who reported use of 11 or more
prescription drugs, and 28.43 minutes for a respondent who used 1 to 3 prescription
drugs.

•

A more notable pattern for times to complete both the screeners and main modules
was observed according to the numbers of prescription drugs that respondents
misused. However, because of the small sample sizes (especially for respondents who
misused extreme high numbers of prescription drugs) and the variability in the timing
data, caution is advised in interpreting these data. To verify the reproducibility of
these findings, this investigation could be repeated with data from the 2013 DR,
including possible use of combined QFT and DR data to increase the sample sizes.

•

The average time to complete the pain relievers screener and main module was
6.95 minutes for respondents aged 12 or older who reported misuse of eight or more
pain relievers in the past year (Table 4.11c) and 2.18 minutes for respondents who
misused only one pain reliever (Table 4.11d).

•

Maximum times to complete the pain relievers screener and main module according
to the number of drugs that were misused used were 12.45 minutes for a respondent
aged 26 to 34 who reported misuse of seven pain relievers, 11.88 minutes for a
respondent aged 12 or older who misused eight or more pain relievers, and
7.28 minutes for a respondent who misused only one pain reliever.

•

The average time to complete the screeners and main modules for all prescription
drugs was 14.23 minutes for respondents aged 12 or older who reported misuse of
14 or more prescription drugs in the past year (Table 4.11g) and 7.99 minutes for
respondents who misused one or two prescription drugs (Table 4.11h).

•

Maximum times to complete the screeners and main modules according to the
number of drugs that were misused were 28.88 minutes for a respondent aged 18 to
25 who reported misuse of 15 or more prescription drugs in the past year and
25.03 minutes for a respondent aged 35 to 49 who misused 1 prescription drug.

Highlights for the time required to complete the entire interview according to the number
of prescription drugs that were used in the past year include the following:
•

Average times to complete the entire interview were 58.73 minutes for respondents
aged 12 or older who used one pain reliever in the past year (Table 4.11j) and
58.73 minutes for respondents who used one to three prescription drugs in any of the
screeners (Table 4.11n).

•

The shortest time to complete the interview for a respondent who used one to three
prescription drugs was 26.93 minutes (Table 4.11n).
151

•

Among respondents who reported past year use of higher numbers of prescription
drugs, average times to complete the entire interview were 68.28 minutes for
respondents aged 12 or older who used 7 or more pain relievers in the past year
(Table 4.11i) and 68.46 minutes for respondents who used 11 or more prescription
drugs in any of the screeners (Table 4.11m).

•

The shortest time to complete the interview for a respondent who used 11 or more
prescription drugs was 39.60 minutes (Table 4.11m).

On average, therefore, the interview times among persons aged 12 or older differed by about
10 minutes between the timings for respondents who reported use of a low number of
prescription pain relievers or prescription psychotherapeutics (but use of at least one drug) and
those reported use of extreme high numbers of prescription drugs.
Highlights for the time required to complete the entire interview according to the number
of prescription drugs that were misused in the past year include the following. However, note
that the groups of respondents who used high numbers of prescription drugs in the past year and
those who misused high numbers of prescription drugs in that period are not mutually exclusive.
•

Average times to complete the entire interview were 65.41 minutes for respondents
aged 12 or older who misused one pain reliever in the past year (Table 4.11l) and
64.47 minutes for respondents who misused one or two prescription drugs in any of
the modules (Table 4.11p).

•

The shortest time to complete the interview for a respondent who misused one or two
prescription drugs in any category was 27.23 minutes (Table 4.11p).

•

Among respondents who reported past year misuse of higher numbers of prescription
drugs, average times to complete the entire interview were 68.15 minutes for
respondents aged 12 or older who misused 8 or more pain relievers in the past year
(Table 4.11k) and 68.50 minutes for respondents who misused 14 or more
prescription drugs in any of the screeners (Table 4.11o).

•

The shortest time to complete the interview for a respondent who misused 14 or
more prescription drugs in any category was 43.22 minutes (Table 4.11o).

Extreme high interview times were observed regardless of the numbers of prescription
drugs that respondents used or misused. For example, one respondent who used one to three
prescription drugs in the past year had a total interview time of 228.47 minutes (Table 4.11n),
and a respondent who used one pain reliever had a total interview time of 191.52 minutes
(Table 4.11j). Nevertheless, the shortest time to complete the interview for respondents who
misused 14 or more prescription drugs was about 16 minutes longer than the shortest time for
respondents who misused only one or two prescription drugs (Tables 4.11o and 4.11p,
respectively).

152

Table 4.11a Overall Interview Timing Data for the QFT Pain Relievers Screener in Minutes, in
Total and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Pain Relievers Used in the Past Year
Overall, Used
7 or More
Pain Relievers
in the Past
Year1
48
1

12-17, Used 5
or More Pain
Relievers in
the Past Year1
9
0

18-25, Used 8
or More Pain
Relievers in
the Past Year1
17
0

26-34, Used 7
or More Pain
Relievers in
the Past Year1
11
0

35-49, Used 6
or More Pain
Relievers in
the Past Year1
11
0

50+, Used 5 or
More Pain
Relievers in
the Past Year2
11
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

2.48
1.20
1.09

2.04
0.45
0.67

2.25
0.65
0.81

2.43
0.28
0.53

2.80
1.01
1.01

3.10
3.27
1.81

Maximum
Median
Minimum
Range

7.28
2.26
0.45
6.83

3.47
1.73
1.37
2.10

3.70
2.05
1.13
2.57

3.93
2.33
1.88
2.05

4.72
2.68
1.60
3.12

8.03
2.80
1.67
6.37

7.28
4.72
4.15
3.93
3.72
1.50
1.47
1.45
1.13
0.45

3.47
2.70
2.32
1.90
1.73
1.73
1.68
1.62
1.60
1.37

3.70
3.60
3.38
3.30
2.68
1.63
1.50
1.47
1.45
1.13

3.93
2.52
2.48
2.43
2.38
2.30
2.23
2.17
2.10
1.88

4.72
4.15
3.52
3.17
2.80
2.37
2.28
1.83
1.73
1.60

8.03
3.83
3.72
3.22
3.07
2.05
2.02
1.90
1.83
1.67

Sample Used in Analysis3
Extreme/Missing Records4

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or above the 95th percentile for users in this age group.
2
Cases whose number of reported drugs was at or above the 94th percentile for users in this age group.
3

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.
Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

153

Table 4.11b Overall Interview Timing Data for the QFT Pain Relievers Screener in Minutes, in
Total and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Pain Relievers Used in the Past Year

Sample Used in Analysis
Extreme/Missing Records1
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Maximum
Median
Minimum
Range

Overall, Used 1
Pain Reliever
in the Past
Year
335
6

12-17, Used 1
Pain Reliever
in the Past
Year
82
0

18-25, Used
1 Pain
Reliever in
the Past
Year
82
3

26-34, Used
1 Pain
Reliever in
the Past
Year
41
0

2.24
1.79
1.34

2.01
0.57
0.75

1.91
0.81
0.90

1.94
0.61
0.78

2.48
2.67
1.63

2.90
3.75
1.94

11.83
1.95
0.43
11.40

4.75
1.87
0.82
3.93

5.45
1.68
0.43
5.02

4.02
1.75
0.75
3.27

11.82
2.06
0.62
11.20

11.83
2.38
0.90
10.93

11.83
11.82
10.45
8.68
8.10
0.72
0.72
0.68
0.62
0.43

4.75
4.42
3.80
3.70
3.47
1.10
1.08
1.07
1.03
0.82

5.45
4.58
4.57
3.98
3.45
0.80
0.72
0.72
0.68
0.43

4.02
3.88
3.50
3.25
3.18
1.22
1.20
1.18
1.13
0.75

11.82
8.10
4.57
4.42
3.88
1.10
1.07
1.07
0.93
0.62

11.83
10.45
8.68
5.45
5.15
1.38
1.30
1.30
1.07
0.90

35-49, Used
1 Pain
Reliever in
the Past
Year
64
2

50+, Used 1
Pain
Reliever in
the Past
Year
66
1

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

154

Table 4.11c Overall Interview Timing Data for the QFT Pain Relievers Screener and Main Module
in Minutes, in Total and by Age Groups for Respondents Reporting Extreme High
Numbers of Prescription Pain Relievers Misused in the Past Year

Sample Used in Analysis4
Extreme/Missing Records5
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Maximum
Median
Minimum
Range

18-25,
Misused 8 or
More Pain
Relievers in
the Past
Year1
5
0

26-34,
Misused 7
or More
Pain
Relievers in
the Past
Year1
2
0

35-49,
Misused 4
or More
Pain
Relievers in
the Past
Year2
2
0

50+,
Misused 2
or More
Pain
Relievers in
the Past
Year3
2
0

Overall,
Misused 8 or
More Pain
Relievers in the
Past Year1
9
0

12-17,
Misused 8 or
More Pain
Relievers in
the Past
Year1
3
0

6.95
8.10
2.85

8.36
7.10
2.67

6.39
10.24
3.20

8.97
24.27
4.93

5.35
0.22
0.47

7.19
1.65
1.28

11.88
5.48
3.63
8.25

9.90
9.90
5.28
4.62

11.88
5.15
3.63
8.25

12.45
8.97
5.48
6.97

5.68
5.35
5.02
0.67

8.10
7.19
6.28
1.82

11.88
9.90
9.90
6.17
5.48

9.90
9.90
5.28
—
—

11.88
6.17
5.15
5.13
3.63

12.45
5.48
—
—
—

5.68
5.02
—
—
—

8.10
6.28
—
—
—

5.48
5.28
5.15
5.13
3.63

—
—
9.90
9.90
5.28

11.88
6.17
5.15
5.13
3.63

—
—
—
12.45
5.48

—
—
—
5.68
5.02

—
—
—
8.10
6.28

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
— Not applicable.
1

Cases whose number of reported drugs was at or above the 95th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 90th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 70th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

155

Table 4.11d Overall Interview Timing Data for the QFT Pain Relievers Screener and Main Module
in Minutes, in Total and by Age Groups for Respondents Reporting Lower Numbers of
Prescription Pain Relievers Misused in the Past Year
18-25,
Misused 1
Pain
Reliever in
the Past
Year
36
1

26-34,
Misused 1
Pain
Reliever in
the Past
Year
13
0

35-49,
Misused 1
Pain
Reliever in
the Past
Year
11
0

50+,
Misused 1
Pain
Reliever in
the Past
Year
5
0

Overall,
Misused 1 Pain
Reliever in the
Past Year
84
1

12-17,
Misused 1
Pain Reliever
in the Past
Year
19
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

2.18
1.22
1.10

1.71
0.25
0.50

2.13
1.55
1.24

2.18
0.55
0.74

2.83
2.30
1.51

2.85
0.43
0.66

Maximum
Median
Minimum
Range

7.28
1.96
0.72
6.57

3.12
1.43
1.25
1.87

7.28
1.75
0.72
6.57

3.88
2.05
1.33
2.55

7.08
2.33
1.50
5.58

3.83
2.73
2.05
1.78

7.28
7.08
4.58
4.32
3.88
1.18
1.17
1.00
0.87
0.72

3.12
2.48
2.15
2.13
2.10
1.37
1.35
1.28
1.27
1.25

7.28
4.58
4.32
3.67
3.42
1.18
1.17
1.00
0.87
0.72

3.88
3.18
2.58
2.38
2.35
1.83
1.67
1.45
1.33
1.33

7.08
3.43
3.00
2.97
2.57
2.22
2.12
2.07
1.88
1.50

3.83
3.05
2.73
2.60
2.05
3.83
3.05
2.73
2.60
2.05

Sample Used in Analysis
Extreme/Missing Records1

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

156

Table 4.11e Overall Interview Timing Data for All QFT Prescription Drug Screeners in Minutes, in
Total and by Age Groups for Respondents Reporting Extreme High Numbers of
Prescription Drugs Used in the Past Year

Sample Used in Analysis2
Extreme/Missing Records3
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Maximum
Median
Minimum
Range

Overall, Used
11 or More
Prescription
Drugs in the
Past Year1
47
0

12-17, Used
6 or More
Prescription
Drugs in the
Past Year1
9
0

18-25, Used
15 or More
Prescription
Drugs in the
Past Year1
13
0

26-34, Used
11 or More
Prescription
Drugs in the
Past Year1
9
0

35-49, Used
8 or More
Prescription
Drugs in the
Past Year1
10
0

50+, Used 9
or More
Prescription
Drugs in the
Past Year1
8
0

5.33
4.34
2.08

5.18
10.18
3.19

4.40
3.14
1.77

5.41
3.67
1.92

6.31
5.54
2.35

6.77
6.42
2.53

10.33
4.65
2.38
7.95

13.18
4.65
2.70
10.48

9.07
3.88
2.38
6.68

8.93
4.53
3.80
5.13

9.55
5.39
3.85
5.70

10.33
7.02
3.40
6.93

10.33
9.42
9.38
9.07
9.00

13.18
5.93
5.18
4.70
4.65

9.07
6.60
5.17
4.68
4.37

8.93
8.27
5.83
5.02
4.53

9.55
9.42
9.38
7.08
5.65

10.33
9.00
8.58
7.42
6.62

2.88
2.87
2.82
2.70
2.38

4.65
4.03
3.38
2.82
2.70

3.83
3.58
2.88
2.87
2.38

4.53
4.22
4.08
4.03
3.80

5.13
4.63
4.38
4.02
3.85

7.42
6.62
5.08
3.72
3.40

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or above the 95th percentile for users in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

157

Table 4.11f Overall Interview Timing Data for All QFT Prescription Drug Screeners in Minutes, in
Total and by Age Groups for Respondents Reporting Lower Numbers of Prescription
Drugs Used in the Past Year

Sample Used in Analysis4
Extreme/Missing Records5
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Maximum
Median
Minimum
Range

Overall, Used
1 to 3
Prescription
Drugs in the
Past Year1
646
10

12-17, Used
1 or 2
Prescription
Drugs in the
Past Year2
121
0

4.69
7.55
2.75

4.40
2.90
1.70

28.43
4.03
1.12
27.32
28.43
27.52
25.82
23.47
18.22
1.55
1.47
1.28
1.22
1.12

18-25, Used
1 to 3
Prescription
Drugs in the
Past Year1
160
5

26-34, Used
1 to 3
Prescription
Drugs in the
Past Year
106
0

35-49 Used 1
to 3
Prescription
Drugs in the
Past Year1
131
4

50+, Used 1
or 2
Prescription
Drugs in the
Past Year3
98
1

3.79
2.66
1.63

4.08
3.07
1.75

4.89
8.49
2.91

6.60
19.69
4.44

9.98
3.98
2.08
7.90

11.80
3.52
1.12
10.68

14.65
3.75
1.55
13.10

28.43
4.12
1.58
26.85

27.52
5.48
1.90
25.62

9.98
9.68
9.47
8.28
7.88
2.52
2.43
2.23
2.22
2.08

11.80
9.82
8.85
7.40
7.33
1.57
1.47
1.28
1.22
1.12

14.65
8.53
8.50
8.18
7.80
2.25
2.22
2.13
1.98
1.55

28.43
14.75
12.18
12.08
9.23
2.20
2.00
1.95
1.82
1.58

27.52
25.82
23.47
18.22
14.52
2.62
2.53
2.40
1.98
1.90

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was below the 75th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Cases whose number of reported drugs was below the 80th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Cases whose number of reported drugs was below the 65th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within some age groups.

158

Table 4.11g Overall Interview Timing Data for All QFT Prescription Drug Screeners and Main
Modules in Minutes, in Total and by Age Groups for Respondents Reporting Extreme
High Numbers of Prescription Drugs Misused in the Past Year
Overall,
Misused 14 or
More
Prescription
Drugs in the
Past Year1
11
0

Sample Used in Analysis3
Extreme/Missing Records4

12-17,
Misused 16
or More
Prescription
Drugs in the
Past Year1
3
0

18-25,
Misused 15
or More
Prescription
Drugs in the
Past Year1
6
0

26-34,
Misused 8 or
More
Prescription
Drugs in the
Past Year2
4
0

35-49,
Misused 5 or
More
Prescription
Drugs in the
Past Year2
3
0

50+, Misused
2 or More
Prescription
Drugs in the
Past Year2
3
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

14.23
39.27
6.27

16.78
3.44
1.86

14.19
67.77
8.23

13.03
36.16
6.01

9.66
0.06
0.24

14.71
23.03
4.80

Maximum
Median
Minimum
Range

28.88
11.02
7.92
20.97

18.22
17.43
14.68
3.53

28.88
10.53
7.92
20.97

21.93
10.73
8.72
13.22

9.85
9.73
9.38
0.47

20.22
12.45
11.45
8.77

28.88
18.93
18.22
17.43
14.68
10.92
10.17
10.05
8.37
7.92

18.22
17.43
14.68
—
—
—
—
18.22
17.43
14.68

28.88
18.93
11.02
10.05
8.37
18.93
11.02
10.05
8.37
7.92

21.93
10.92
10.55
8.72
—
—
21.93
10.92
10.55
8.72

9.85
9.73
9.38
—
—
—
—
9.85
9.73
9.38

20.22
12.45
11.45
—
—
—
—
20.22
12.45
11.45

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
— Not applicable.
1

Cases whose number of reported drugs was at or above the 95th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 90th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

159

Table 4.11h Overall Interview Timing Data for All QFT Prescription Drug Screeners and Main
Modules in Minutes, in Total and by Age Groups for Respondents Reporting Lower
Numbers of Prescription Drugs Misused in the Past Year
Overall,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
139
1

Sample Used in Analysis3
Extreme/Missing Records4

12-17,
Misused 1
or 2
Prescription
Drugs in the
Past Year2
27
0

18-25,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
66
1

26-34,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
18
0

35-49,
Misused 1
Prescription
Drug in the
Past Year1
14
0

50+, Misused
1
Prescription
Drug in the
Past Year1
7
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

7.99
13.92
3.73

6.72
4.01
2.00

7.13
11.35
3.37

7.79
5.03
2.24

10.64
27.50
5.24

11.86
15.76
3.97

Maximum
Median
Minimum
Range

25.03
7.13
2.57
22.47

11.35
6.75
3.95
7.40

20.80
6.70
2.57
18.23

12.98
7.93
4.42
8.57

25.03
9.43
5.68
19.35

16.53
12.57
6.47
10.07

25.03
20.80
20.28
18.22
18.18
3.95
3.53
3.03
3.02
2.57

11.35
9.87
9.60
9.30
8.70
4.65
4.42
4.32
4.12
3.95

20.80
18.22
17.33
16.70
12.20
4.03
3.53
3.03
3.02
2.57

12.98
10.67
9.50
9.50
9.20
5.90
5.63
5.35
4.47
4.42

25.03
18.18
12.15
11.53
11.08
7.55
7.43
6.52
6.32
5.68

16.53
15.50
13.08
12.57
12.32
13.08
12.57
12.32
6.55
6.47

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or below the 70th percentile for misusers in this age group.

Cases whose number of reported drugs was below the 75th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within some age groups.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

160

Table 4.11i Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Extreme High Numbers of Prescription Pain
Relievers Used in the Past Year
Overall, Used 7
or More Pain
Relievers in the
Past Year1
47
1

Sample Used in Analysis3
Extreme/Missing Records4

12-17, Used 5
or More Pain
Relievers in
the Past
Year1
9
0

18-25, Used
8 or More
Pain
Relievers in
the Past
Year1
17
0

26-34, Used
7 or More
Pain
Relievers in
the Past
Year1
11
0

35-49, Used
6 or More
Pain
Relievers in
the Past
Year1
11
0

50+, Used 5
or More
Pain
Relievers in
the Past
Year2
11
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

68.28
489.80
22.13

64.30
373.31
19.32

64.08
366.73
19.15

63.23
509.28
22.57

64.81
783.46
27.99

83.64
1166.54
34.15

Maximum
Median
Minimum
Range

129.47
62.92
39.60
89.87

103.27
63.33
42.37
60.90

111.50
61.17
41.53
69.97

111.97
56.20
39.60
72.37

129.47
56.17
38.92
90.55

174.25
83.17
45.93
128.32

129.47
111.97
111.50
106.88
103.35
43.22
42.37
41.53
41.00
39.60

103.27
77.65
72.73
70.53
63.33
63.33
55.22
47.87
45.72
42.37

111.50
103.35
80.60
70.02
68.20
52.95
51.30
45.53
43.22
41.53

111.97
97.68
71.07
64.90
57.13
51.73
51.68
49.68
43.93
39.60

129.47
101.73
78.70
61.37
56.95
51.55
50.48
46.62
41.00
38.92

174.25
95.18
90.52
86.65
84.90
75.52
72.25
62.90
48.77
45.93

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or above the 95th percentile for users in this age group.

Cases whose number of reported drugs was at or above the 94th percentile for users in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes (24.6 minutes) and therefore excluded from the analysis of overall interview timing. The
respondent was an 18 to 25 year old and reported past year use of seven pain relievers. Consequently, this case was at the cut point for
respondents aged 12 or older, but was below the cut point extreme for 18 to 25 year olds.

161

Table 4.11j Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Lower Numbers of Prescription Pain Relievers
Used in the Past Year
Overall, Used 1
Pain Reliever
in the Past
Year
335
6

Sample Used in Analysis
Extreme/Missing Records1

12-17, Used 1
Pain Reliever
in the Past
Year
82
0

18-25, Used
1 Pain
Reliever in
the Past
Year
82
3

26-34, Used
1 Pain
Reliever in
the Past
Year
41
0

35-49, Used
1 Pain
Reliever in
the Past
Year
64
2

50+, Used 1
Pain
Reliever in
the Past
Year
66
1

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

58.73
363.18
19.06

56.68
187.15
13.68

53.31
265.03
16.28

55.06
282.39
16.80

62.97
585.75
24.20

66.16
436.51
20.89

Maximum
Median
Minimum
Range

191.52
55.77
27.23
164.28

115.13
55.23
34.05
81.08

113.00
50.57
27.23
85.77

98.18
52.35
30.13
68.05

191.52
60.33
28.48
163.03

150.02
60.20
28.37
121.65

191.52
150.02
123.75
115.13
113.23
29.07
28.63
28.48
28.37
27.23

115.13
88.40
83.27
80.62
80.52
37.68
37.02
36.75
35.72
34.05

113.00
102.78
82.80
80.23
78.75
31.73
31.30
29.80
29.07
27.23

98.18
90.55
87.68
83.20
78.80
34.32
33.45
32.90
32.48
30.13

191.52
123.75
105.63
94.83
93.93
38.20
32.92
32.70
28.63
28.48

150.02
113.23
111.85
109.83
100.30
40.73
40.27
38.97
34.70
28.37

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

162

Table 4.11k Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Extreme High Numbers of Prescription Pain
Relievers Misused in the Past Year
Overall,
Misused 8 or
More Pain
Relievers in the
Past Year1
9
0

Sample Used in Analysis4
Extreme/Missing Records5

12-17,
Misused 8 or
More Pain
Relievers in
the Past
Year1
3
0

18-25,
Misused 8 or
More Pain
Relievers in
the Past
Year1
5
0

26-34,
Misused 7
or More
Pain
Relievers in
the Past
Year1
2
0

35-49,
Misused 4
or More
Pain
Relievers in
the Past
Year2
2
0

50+,
Misused 2
or More
Pain
Relievers in
the Past
Year3
2
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

68.15
569.84
23.87

79.78
435.89
20.88

64.87
721.63
26.86

80.83
1939.61
44.04

64.62
133.93
11.57

79.76
195.03
13.97

Maximum
Median
Minimum
Range

111.50
61.17
43.22
68.28

103.27
72.73
63.33
39.93

111.50
55.52
43.22
68.28

111.97
80.83
49.68
62.28

72.80
64.62
56.43
16.37

89.63
79.76
69.88
19.75

111.50
103.27
72.73
63.33
61.17
61.17
55.52
52.95
49.68
43.22

103.27
72.73
63.33
—
—
—
—
103.27
72.73
63.33

111.50
61.17
55.52
52.95
43.22
111.50
61.17
55.52
52.95
43.22

111.97
49.68
—
—
—
—
—
—
111.97
49.68

72.80
56.43
—
—
—
—
—
—
72.80
56.43

89.63
69.88
—
—
—
—
—
—
89.63
69.88

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
— Not applicable.
1

Cases whose number of reported drugs was at or above the 95th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 90th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 70th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

163

Table 4.11l Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Lower Numbers of Prescription Pain Relievers
Misused in the Past Year
Overall,
Misused 1 Pain
Reliever in the
Past Year
84
1

Sample Used in Analysis1
Extreme/Missing Records2

12-17,
Misused 1
Pain Reliever
in the Past
Year
19
0

18-25,
Misused 1
Pain
Reliever in
the Past
Year
36
1

26-34,
Misused 1
Pain
Reliever in
the Past
Year
13
0

35-49,
Misused 1
Pain
Reliever in
the Past
Year
11
0

50+,
Misused 1
Pain
Reliever in
the Past
Year
5
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

65.41
454.68
21.32

63.01
208.55
14.44

62.59
401.71
20.04

59.07
170.69
13.06

75.80
1246.43
35.30

88.47
161.83
12.72

Maximum
Median
Minimum
Range

171.93
62.45
27.23
144.70

83.02
62.58
40.55
42.47

116.13
60.11
27.23
88.90

85.98
56.10
40.98
45.00

171.93
62.32
47.30
124.63

104.30
86.68
69.93
34.37

171.93
116.13
106.88
104.30
102.78
40.55
39.95
35.05
33.93
27.23

83.02
82.98
80.62
80.52
79.33
48.40
45.72
45.62
42.07
40.55

116.13
106.88
102.78
87.02
84.05
41.35
39.95
35.05
33.93
27.23

85.98
76.17
69.00
65.07
64.90
53.55
47.82
46.40
45.52
40.98

171.93
92.55
91.07
78.40
66.13
61.22
55.80
55.73
51.32
47.30

104.30
95.18
86.68
86.27
69.93
104.30
95.18
86.68
86.27
69.93

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

164

Table 4.11m Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and
by Age Groups for Respondents Reporting Extreme High Numbers of Prescription
Drugs Used in the Past Year
Overall, Used
11 or More
Prescription
Drugs in the
Past Year1
47
0

Sample Used in Analysis2
Extreme/Missing Records3

12-17, Used
6 or More
Prescription
Drugs in the
Past Year1
9
0

18-25, Used
15 or More
Prescription
Drugs in the
Past Year1
13
0

26-34, Used
11 or More
Prescription
Drugs in the
Past Year1
9
0

35-49, Used
8 or More
Prescription
Drugs in the
Past Year1
10
0

50+, Used 9
or More
Prescription
Drugs in the
Past Year1
8
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

68.46
460.65
21.46

70.52
280.79
16.76

62.49
348.93
18.68

64.39
520.32
22.81

75.43
687.04
26.21

77.11
319.05
17.86

Maximum
Median
Minimum
Range

129.47
62.92
39.60
89.87

103.27
72.73
42.37
60.90

111.50
59.58
41.53
69.97

111.97
56.20
39.60
72.37

129.47
70.23
50.48
78.98

95.18
84.03
45.93
49.25

129.47
111.97
111.50
103.35
103.27
43.22
42.37
42.28
41.53
39.60

103.27
77.65
76.12
73.43
72.73
72.73
70.53
63.33
55.22
42.37

111.50
80.60
73.92
68.20
62.92
55.52
52.95
45.53
43.22
41.53

111.97
88.50
72.32
57.13
56.20
56.20
52.38
51.73
49.68
39.60

129.47
101.73
93.42
78.70
76.52
63.95
56.95
51.55
51.55
50.48

95.18
90.52
89.63
84.90
83.17
84.90
83.17
72.25
55.27
45.93

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or above the 95th percentile for users in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

165

Table 4.11n Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Lower Numbers of Prescription Drugs Used in
the Past Year
Overall, Used
1 to 3
Prescription
Drugs in the
Past Year1
646
10

Sample Used in Analysis4
Extreme/Missing Records5

12-17, Used
1 or 2
Prescription
Drugs in the
Past Year2
121
0

18-25, Used
1 to 3
Prescription
Drugs in the
Past Year1
160
5

26-34, Used
1 to 3
Prescription
Drugs in the
Past Year
106
0

35-49 Used 1
to 3
Prescription
Drugs in the
Past Year1
131
4

50+, Used 1
or 2
Prescription
Drugs in the
Past Year3
98
1

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

58.73
394.94
19.87

59.35
227.72
15.09

52.95
246.17
15.69

53.94
265.02
16.28

59.61
494.06
22.23

68.49
685.86
26.19

Maximum
Median
Minimum
Range

228.47
55.55
26.93
201.53

115.13
56.00
34.05
81.08

125.35
50.31
26.93
98.42

108.78
50.57
31.45
77.33

191.52
55.80
28.48
163.03

228.47
62.19
28.37
200.10

228.47
191.52
150.02
125.35
125.18
28.63
28.48
28.37
27.23
26.93

115.13
106.88
100.90
95.55
93.28
37.68
37.02
36.75
35.72
34.05

125.35
113.00
102.78
84.05
82.80
29.90
29.80
29.07
27.23
26.93

108.78
98.92
98.18
90.55
87.68
33.33
32.90
32.48
31.85
31.45

191.52
125.18
123.75
119.80
105.63
30.85
29.98
29.52
28.63
28.48

228.47
150.02
119.63
113.23
111.85
40.22
38.97
36.72
34.70
28.37

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was below the 75th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Cases whose number of reported drugs was below the 80th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Cases whose number of reported drugs was below the 65th percentile for users in this age group but allowed for reporting of use of more than
one drug across all four modules.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within some age groups.

166

Table 4.11o Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Extreme High Numbers of Prescription Drugs
Misused in the Past Year
Overall,
Misused 14 or
More
Prescription
Drugs in the
Past Year1
11
0

Sample Used in Analysis3
Extreme/Missing Records4

12-17,
Misused 16
or More
Prescription
Drugs in the
Past Year1
3
0

18-25,
Misused 15
or More
Prescription
Drugs in the
Past Year1
6
0

26-34,
Misused 8 or
More
Prescription
Drugs in the
Past Year2
4
0

35-49,
Misused 5 or
More
Prescription
Drugs in the
Past Year2
3
0

50+, Misused
2 or More
Prescription
Drugs in the
Past Year2
3
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

68.50
478.57
21.88

79.78
435.89
20.88

67.49
618.54
24.87

72.55
832.11
28.85

79.38
720.64
26.84

81.37
105.33
10.26

Maximum
Median
Minimum
Range

111.50
61.17
43.22
68.28

103.27
72.73
63.33
39.93

111.50
58.34
43.22
68.28

111.97
64.28
49.68
62.28

108.90
72.80
56.43
52.47

89.63
84.60
69.88
19.75

111.50
103.27
80.60
72.73
63.33
59.58
55.52
52.95
49.68
43.22

103.27
72.73
63.33
—
—
—
—
103.27
72.73
63.33

111.50
80.60
61.17
55.52
52.95
80.60
61.17
55.52
52.95
43.22

111.97
76.17
52.38
49.68
—
—
111.97
76.17
52.38
49.68

108.90
72.80
56.43
—
—
—
—
108.90
72.80
56.43

89.63
84.60
69.88
—
—
—
—
89.63
84.60
69.88

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
— Not applicable.
1

Cases whose number of reported drugs was at or above the 95th percentile for misusers in this age group.

Cases whose number of reported drugs was at or above the 90th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within each age group.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

167

Table 4.11p Overall Interview Timing Data for the Full QFT Interview in Minutes, in Total and by
Age Groups for Respondents Reporting Lower Numbers of Prescription Drugs
Misused in the Past Year
Overall,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
139
1

Sample Used in Analysis3
Extreme/Missing Records4

12-17,
Misused 1
or 2
Prescription
Drugs in the
Past Year2
27
0

18-25,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
66
1

26-34,
Misused 1
or 2
Prescription
Drugs in the
Past Year1
18
0

35-49,
Misused 1
Prescription
Drug in the
Past Year1
14
0

50+, Misused
1
Prescription
Drug in the
Past Year1
7
0

Summary Statistics (Minutes)
Mean
Variance
Standard Deviation

64.47
416.50
20.41

64.96
229.41
15.15

59.69
324.82
18.02

59.92
147.95
12.16

73.88
1001.08
31.64

81.64
359.68
18.97

Maximum
Median
Minimum
Range

171.93
61.67
27.23
144.70

106.88
64.43
40.55
66.33

116.13
56.33
27.23
88.90

85.98
58.88
40.98
45.00

171.93
64.23
47.30
124.63

104.30
86.27
45.93
58.37

171.93
125.18
116.13
113.00
110.17
38.62
35.05
33.93
30.50
27.23

106.88
83.02
82.98
80.62
80.52
48.40
45.72
45.62
42.07
40.55

116.13
113.00
106.88
102.78
84.05
38.62
35.05
33.93
30.50
27.23

85.98
78.13
72.32
69.52
69.00
50.03
47.82
46.40
45.52
40.98

171.93
92.55
91.07
78.40
76.72
55.80
55.73
51.55
51.32
47.30

104.30
95.18
86.68
86.27
83.17
86.68
86.27
83.17
69.93
45.93

Extremes
5 Highest

(Highest)

5 Lowest

(Lowest)
1

Cases whose number of reported drugs was at or below the 70th percentile for misusers in this age group.

Cases whose number of reported drugs was below the 75th percentile for misusers in this age group.

Sample sizes for individual age groups do not sum to the sample size for respondents aged 12 or older because different cut points were used
overall and within some age groups.

Overall interview time was less than 30 minutes or greater than 240 minutes, but was included in this particular analysis.

4.6

Other Data Quality Indicators

4.6.1

Overview of Other Data Quality Indicators
Examination of other data quality indicators focused on the following:
•

triggering of inconsistency "flags" in the core drug use data;

•

choosing "other" responses for which respondents subsequently were asked to specify
a written response (i.e., "OTHER, Specify" data), such as other sources of
prescription psychotherapeutic drugs;

•

triggering of "hard errors" in the QFT if respondents reported first misusing specific
prescription drugs at an age that was older than their current age;

•

triggering of consistency checks in the QFT for respondents who reported first misuse
of specific prescription drugs in a year and month that differed from the age they
reported for when they first misused; and

168

•

potential patterned responses in answers to the screening questions for past year
prescription drug use or to the questions for past year misuse.

Identification and handling of potential patterned responses in the 2011 and 2012 comparison
data also are discussed in this section.
4.6.2

Triggering of Inconsistency Flags in Core Drug Use Data

Examination of data from variables that flagged inconsistencies in the core drug modules
focused on the following core modules or core variables, each of which underwent notable
changes that could affect patterns of inconsistent data:
•

smokeless tobacco;

•

binge alcohol use (i.e., based on the threshold of four or more drinks on an occasion
for females);

•

most recent use of hallucinogens (i.e., based on moving questions about most recent
use of three hallucinogens from the noncore special drugs module to the core
hallucinogens module);

•

methamphetamine; and

•

prescription drugs.

Data for inconsistency flags first were examined for the QFT. The decision to examine
inconsistency flag data in the two comparison datasets depended on the occurrence of
inconsistencies in the QFT data. No or low occurrences of inconsistent data in the QFT could be
a function of both the sample size and sample design. Regarding the sample design, persons aged
26 or older were sampled at a higher rate in the QFT than in the main survey. However,
inconsistent response patterns in the main survey often involve reports of initiation of use that is
more recent than the reports of last use. Because most initiation occurs among adolescents and
young adults, having fewer QFT respondents in these two age groups could affect the occurrence
of these patterns of inconsistent reports in the QFT data.
Very small numbers and percentages of QFT respondents had triggered flags for
inconsistent data in the modules for smokeless tobacco, methamphetamine, and prescription
drugs (i.e., fewer than five respondents for any given flag that was set). For prescription drugs,
inconsistencies that were flagged pertained to errors in the computer-assisted interviewing (CAI)
programming that were identified during data editing rather than logical inconsistencies.17 These
programming errors will be fixed for the 2013 DR. In addition, fewer than five respondents each
in the pain relievers, tranquilizers, and stimulants modules reported misuse in the past 30 days
and also reported misuse on "0 days" in that period. This logic was programmed correctly
according to the CAI specifications (i.e., 0 was in the allowable range for the 30-day frequency
17

These programming errors for prescription drugs involved (a) asking the 30-day misuse question when
respondents had already reported initiating misuse of some prescription drug in that category (e.g., pain relievers) in
the past 30 days, which gave respondents the opportunity to answer the 30-day misuse question as "no"; and (b) not
skipping respondents out of subsequent 30-day misuse questions after they had answered the lead 30-day misuse
question as "no," which gave respondents the opportunity to report misuse on 1 to 30 days in the past month.

169

questions). For the 2013 DR, however, the decision has been made to change the allowable range
for the 30-day frequency of misuse to 1 to 30 days because respondents will have been asked a
"yes/no" question for whether they misused any prescription drugs in that category in the past
30 days, or else they may have reported initiating misuse of a specific prescription drug in the
past 30 days.
There were no situations in the QFT data in which the variable for most recent use of any
hallucinogen was logically inferred to be more recent than that reported by respondents based on
reports of more recent use of the specific hallucinogens ketamine, dimethyltryptamine (DMT),
alpha-methyltryptamine (AMT), "Foxy", or Salvia divinorum (i.e., the three hallucinogens that
had been moved from special drugs to the core hallucinogens module). There also were no
situations in the QFT data in which more recent use of any hallucinogen was logically inferred
based on reports of most recent use of lysergic acid diethylamide (LSD), phencyclidine (PCP), or
Ecstasy (i.e., the specific hallucinogens that were included in this module for both the main
survey and the QFT). Most recent use of any hallucinogen was set to an "indefinite" periods of
use (i.e., at some point in the past 12 months or some point in the lifetime) because they had
ambiguous data for most recent use of ketamine or of DMT, AMT, or "Foxy." Similar edits were
implemented for a larger number of QFT respondents (but fewer than 20) based on ambiguous
data for most recent use of LSD or Ecstasy. As noted previously, LSD and Ecstasy were not
among the hallucinogens that had been moved from a noncore module to the core hallucinogens
module for the QFT. Thus, these data suggest that hallucinogens that were already in this module
might have more of an effect on editing of most recent use of any hallucinogen than the three
hallucinogens that were moved from a noncore module.
For binge alcohol use, about 1 percent of QFT respondents had some inconsistency
between their frequency of consumption of five or more drinks (for males) or four more drinks
(for females) and other 30-day alcohol use data. Rates of inconsistent data for binge alcohol use
and other 30-day alcohol use data were similar in the comparison data based on consumption of
five or more drinks for both males and females (2011 comparison data: 0.8 percent; 2012
comparison data: 0.7 percent). The numbers of respondents in the comparison data who had
these patterns of inconsistent data for binge alcohol use were about 10 to 20 times the number of
QFT respondents with inconsistent data.
4.6.3

Responding to Lead Questions for "OTHER, Specify" Data

As noted in Section 3.3.2 in Chapter 3, only the "OTHER, Specify" data for Hispanic
origin, race, and drugs were coded for use in further data processing or analysis. However, data
for variables or response choices that govern whether respondents were asked "OTHER,
Specify" questions provide an indication of data quality. For example, if predefined categories
for a given question or predefined examples in preceding questions (e.g., specific prescription
drugs) are understandable and encompass the bulk of expected responses, then the rates should
be low for the residual "other" responses (e.g., misuse of "any other" pain reliever, obtaining pain
relievers "some other way").
Estimates in Table N-1 in Appendix N for new, moved, or revised items in the QFT
include estimates for the following questions that have associated "OTHER, Specify" data:

170

•

race (question QD05), including other race;

•

past year misuse of specific prescription pain relievers (PRY01 to PRY40), including
misuse of any other prescription pain relievers;

•

reasons for misusing the last pain reliever (PRYMOTIV), including some other
reason;

•

source of the last pain reliever that the respondent misused (PRY42B), including
getting the drug some other way;

•

friend's or relative's source of the pain reliever that the respondent obtained from a
friend or relative for free (PRY42C), including getting the drug some other way;

•

past year misuse of specific prescription tranquilizers (TRY01 to TRY19);18

•

reasons for misusing the last tranquilizer (TRYMOTIV);

•

source of the last tranquilizer that the respondent misused (TRY21B);

•

friend's or relative's source of the tranquilizer that the respondent obtained from a
friend or relative for free (TRY21C);

•

past year misuse of specific prescription stimulants (STY01 to STY24);

•

reasons for misusing the last stimulant (STYMOTIV);

•

source of the last stimulant that the respondent misused (STY26B);

•

friend's or relative's source of the stimulant that the respondent obtained from a friend
or relative for free (STY26C);

•

past year misuse of specific prescription sedatives (SVY01 to SVY17);

•

reasons for misusing the last sedative (SVYMOTIV);

•

source of the last sedative that the respondent misused (SVY19B);

•

friend's or relative's source of the sedative that the respondent obtained from a friend
or relative for free (SVY19C);

•

type of cancer (HLTH26), including other cancer; and

•

born in the United States (QD14).19

Not counting question QD14, which does not offer an explicit choice of "other" (i.e.,
other country or territory is implied by a response of "no"), rates for "other" responses to these
items were low in the QFT relative to rates for predefined prescription drugs or predefined
response categories. These low rates support the overall conclusion that predefined categories or
predefined examples of prescription drugs performed adequately in the QFT.
For past year misuse of specific pain relievers, for example, fewer than 10 QFT
respondents aged 12 or older reported past year misuse of any other prescription pain reliever,
18

"Other" responses for tranquilizers, stimulants, and sedatives correspond to those listed for pain relievers.
Respondents who answer question QD14 as "no" are routed to question QD15, which asks them to
specify the country or territory where they were born.
19

171

for an estimate of 0.2 percent. In comparison, more than 50 respondents reported past year
misuse of Vicodin®, for an estimate of 2.4 percent. An estimated 70.2 percent of persons who
misused pain relievers in the past year reported misusing pain relievers the last time in order to
relieve physical pain, 26.1 percent reported doing so to relax or relieve tension, and 22.3 percent
reported doing so to feel good or get high. Fewer than five QFT respondents reported misusing
pain relievers the last time for some other reason; the corresponding estimate of 2.1 percent
would be suppressed.
More than 50 QFT respondents reported having some type of cancer in their lifetime.
Although this number of respondents allowed acceptable precision for estimating the lifetime
prevalence of cancer among persons aged 12 or older based on data from more than 2,000
respondents, prevalence estimates for specific types of cancer would be suppressed if based on
the denominator of respondents who ever had cancer. Also, fewer than 10 QFT respondents
reported having most specific types of cancer listed in question HLTH26, including other cancer.
In the typed answers to the "OTHER, Specify" question for other forms of cancer, one of the
answers corresponded to a type of cancer in the list in HLTH26. The second response did not
correspond exactly to any of the types of cancer in the list.
Table M-1 in Appendix M shows weighted estimates for question QD14 in the QFT and
in the comparison data for 2011 and 2012. The estimated percentage of persons aged 12 or older
who were born in the United States based on QFT data (87.9 percent) was similar to the
estimates in the 2011 and 2012 comparison data (88.8 and 88.9 percent, respectively). These
findings suggest that moving the question about country of birth from CAPI to ACASI did not
affect reporting of being born in or outside of the United States.
4.6.4

Triggering of Hard Errors Involving Ages at First Prescription Drug Misuse

In the main survey, consistency checks were triggered if respondents reported first
misuse of prescription drugs at an age that was older than their current age. In these consistency
checks, respondents had the option of changing their current age to make it consistent with their
reported age at first misuse (AFU)20 or to change their AFU to make it consistent with their
current age.
For each specific prescription drug that QFT respondents misused in the past year, they
were asked to report the age when they first misused the drug. Unlike the comparison data from
the main survey, "hard errors" were triggered if QFT respondents reported an AFU for a specific
prescription drug that was older than their current age. The message for these hard errors
indicated that the AFU that respondents entered was older than their current age. Respondents
could change their AFU for that prescription drug to make it consistent with their current age,
but they could not change their current age.
The prescription drug variables in the CAI data that were associated with answers to the
AFU questions did not directly capture information to indicate when these hard errors had been
triggered. However, this information was available through the audit trail data, which indicated
each keystroke that respondents made during the interview. The audit trail data for respondents
20

The abbreviation "AFU" (typically, standing for "age at first use" for drugs other than prescription drugs)
also is used in this section to refer to first misuse of prescription drugs.

172

who triggered at least one hard error in their interviews and also reported past year misuse of
prescription drugs were checked by multiple reviewers.
No situations were identified in the audit trail data for the QFT in which respondents
triggered a hard error between the AFU answers for individual prescription drugs and their
current age. Numbers and percentages of respondents in the 2011 and 2012 comparison data who
triggered corresponding consistency checks also were minimal. Fewer than 10 respondents for
pain relievers and fewer than 5 respondents per module for tranquilizers, stimulants, and
sedatives triggered consistency checks between their AFU data and current age in the 2011 or
2012 comparison samples.
4.6.5

Triggering of Specific Consistency Checks in the Prescription Drug Modules

If QFT respondents reported that they first misused a specific prescription drug within
1 year of their current age, they were asked to report the year and then the month when they first
misused that drug (YFU and MFU, respectively).21 A consistency check was triggered if the
AFU reported by the respondent for the specific drug differed from the corresponding age that
was calculated from the YFU, MFU, and birth month.
However, the programming specifications for the YFU and MFU questions for individual
prescription drugs in the QFT were designed to limit the opportunities for respondents to enter
answers in the YFU and MFU questions that were inconsistent with their answer to the
corresponding AFU question. Specifically, the CAI logic typically limited the months that
respondents could choose in the MFU questions based on their interview date, date of birth,
reported AFU, and reported YFU. For example, suppose a respondent reported first misuse of a
prescription drug at his or her current age and in the current year. If the respondent already had a
birthday in the current year, then the only allowable months that the respondent could choose in
the MFU question were from his or her birth month to the interview month. If specific criteria
did not apply for restricting the allowable months in the MFU question, however, the default was
for the MFU question to display all calendar months.
Data from the QFT suggest that the logical constraints for the AFU, YFU, and MFU
questions were successful in reducing inconsistent reporting of initiation data for individual
prescription drugs. Only three QFT respondents triggered consistency checks because of this
pattern of inconsistent reporting. Two of these consistency checks were triggered for different
pain relievers, and one consistency check was triggered for a tranquilizer. No consistency checks
were triggered for prescription stimulants or sedatives. In addition, no more than one of these
consistency checks was triggered for any of these respondents. In the final QFT sample, no
respondents had inconsistent initiation data for individual prescription drugs.
In comparison, nearly 400 respondents in the 2011 comparison data (0.6 percent of all
respondents) and nearly 150 respondents in the 2012 comparison data (0.5 percent) triggered
consistency checks because their reported AFU for any pain reliever or OxyContin® was
inconsistent with the calculated age at initiation based on their initial reports for their YFU and
21

The abbreviations "YFU" (typically, standing for "year of first use" for drugs other than prescription
drugs) and "MFU" (typically, standing for "month of first use") also are used in this section to refer to first misuse of
prescription drugs.

173

MFU. For tranquilizers, the prescription drug category in the comparison data with the second
highest number of inconsistencies between the reported AFU and initiation data based on the
YFU and MFU, nearly 150 respondents in the 2011 comparison data (0.2 percent) and nearly
100 in the 2012 comparison data (0.3 percent) had this initial pattern of inconsistent data.
As noted previously, however, QFT respondents were asked the YFU and MFU questions
for a given prescription drug only if they reported relatively recent initiation of misuse of that
drug. Consequently, the low numbers of QFT respondents who triggered consistency checks
based on their answers to the AFU, YFU, and MFU questions probably reflects the specific
criteria for asking the YFU and MFU questions. Larger numbers of respondents triggering these
consistency checks for prescription drugs would be expected in a full survey sample of
approximately 67,000 respondents, and at least some of these respondents would be expected not
to resolve some inconsistencies in these initiation data. Nevertheless, the findings for these types
of inconsistencies in the prescription drug initiation data in the QFT and comparison data suggest
that the changes to the CAI logic in the QFT will help to reduce the occurrence of these
inconsistencies when the redesigned prescription drug questions are fielded in 2015.
4.6.6

Patterned Responses in the Core Drug Questions for the Comparison Data

As noted in Section 3.3.2 in Chapter 3, core modules in the 2011 and 2012 comparison
data were reviewed for potential patterned responses according to the procedures documented in
the editing and coding section (Section 10) of the 2010 methodological resource book (Kroutil et
al., 2012a). These checks were implemented as part of the general editing procedures for editing
the full 2011 survey data and the 2012 survey data from quarters 3 and 4, regardless of whether
interviews were within or outside of the 48 States of the continental United States. However,
fewer than five cases in the entire 2011 data were classified as nonrespondents even though they
met the usable case criteria because of patterned responses in their core drug data. Similarly,
fewer than five cases in the entire 2011 survey were retained as respondents, but with their
original responses in one or more core drug modules being replaced with "bad data" codes. For
the 2012 survey in quarters 3 and 4, there also were fewer than five cases that met the usable
case criteria but were treated as nonrespondents and fewer than five cases that were retained as
respondents but with their original responses in one or more core drug modules being replaced
with "bad data" codes.
4.6.7

Patterned Responses in the Drug Use Questions for the QFT Data

The checks for patterned responses that were used for the comparison data also were
implemented for core QFT modules that did not change (or underwent minimal change) relative
to the comparison data. Because the content of the new methamphetamine module for the QFT
was similar to the content of other modules in the comparison data, the relevant checks for the
comparison data were run for the methamphetamine data in the QFT.
Changes to the prescription drug questions for the QFT had the potential to yield some
results in which the pattern of responses could call into question the overall validity of the data
for prescription drugs. Therefore, particular attention was given to identifying the occurrence of
the following patterns in the prescription drug data and examining the results if these patterns
occurred:

174

•

keying responses of "1" (and only "1") to all screener questions for a given
prescription drug category;

•

keying responses of "2" (and only "2") to all screener questions for a given
prescription drug category; and

•

reports of high numbers of individual prescription drugs that were misused relative to
the overall distribution of the number of drugs that were misused within a given
category, with all AFUs being within 1 year of each other (including those in which
all AFUs were at the same age).

4.6.7.1

Background on Patterned Responses in the QFT Prescription Drug Data

In modules preceding the screening questions for pain relievers, for example, responses
of "2" in "gate" questions (e.g., any lifetime use of specific inhalants, any lifetime use of
methamphetamine) meant "no." In the screeners for prescription drugs, however, responses of
"2" typically meant use in the past year of a specific prescription drug. For example, a response
of "2" in the first screening question for pain relievers meant use in the past year of the pain
reliever Lortab®. Thus, if lifetime nonusers of drugs in modules that preceded the prescription
drug screening questions failed to recognize that "2" no longer meant "no" in these screening
questions, they might continue to key responses of "2," thinking incorrectly that this meant that
they did not use any of the drugs in a given question.
Similarly, responses of "1" in gate questions for modules preceding the prescription drug
screening questions meant "yes." On the one hand, a response of "1" in the screening questions
for past year use of prescription drugs could correctly mean that respondents used that particular
prescription drug in the past year. However, there were 11 questions in the screener for pain
relievers about past year use. The remaining screeners for tranquilizers, stimulants, and sedatives
each included six questions about past year use of prescription drugs in their respective
categories. Consequently, keying responses only of "1" to every single screening question for a
given prescription drug category would be highly unlikely; in questions where respondents could
report use of more than one prescription drug in the past year, responses only of "1" would mean
that the respondent used the first (and only the first) prescription drug shown in each question.
Again, if some respondents failed to recognize that "1" no longer meant "yes" in the prescription
drug screeners, they might think incorrectly that "1" meant "yes" to use of any of the drugs in a
given question. Furthermore, if respondents keyed answers of "1" (and only "1") in screening
questions to mean that they used at least one of the drugs in the list, it could not be determined
which specific drugs they actually used.
As noted previously, QFT respondents were asked to report their ages when they first
misused each of the prescription drugs that they reported misusing in the past 12 months. This
could involve misuse of up to 40 pain relievers, 19 tranquilizers, 24 stimulants, and 17 sedatives.
An underlying assumption for asking the initiation questions for each individual prescription
drug was that most respondents would report past year misuse of relatively few prescription
drugs, if any. Nevertheless, if respondents reported misuse of a relatively high number of
prescription drugs within a category in the past year but provided little or no variation in their
reported ages when they initiated misuse of each drug, concern could be raised about the validity
of the self-reported initiation data. For example, some respondents could report the same

175

initiation data for each drug in order to get through the questions faster. Even if respondents were
attempting to answer each individual initiation question as accurately as possible, concern also
could be raised about respondents' ability to provide accurate self-reports in each set of initiation
questions when they reported misuse of relatively high numbers of prescription drugs.
4.6.7.2

Actions Based on Patterned Responses in the QFT Prescription Drug Data

No cases were dropped from the QFT data (i.e., treated as nonrespondents) because of
patterned responses. However, patterns of responses in the QFT prescription drug data were
reported to SAMHSA for a total of 22 cases. For five of these respondents, edited variables for
one or more categories of prescription drugs were assigned "bad data" codes because of
patterned responses in their prescription drug data. These included three respondents who keyed
only responses of "2" wherever possible in the screening questions and two respondents who
keyed only responses of "1" wherever possible in the screening questions. One of these QFT
respondents who keyed only responses of "1" in the screening questions had additional patterned
responses in the questions about misuse, including endorsing all five ways of misuse in the past
year for all four prescription drug categories (i.e., without a prescription, in greater amounts,
more often, longer than told to take the drug, or in some other way not directed by a doctor) and
endorsing all possible motivations for misuse in the past year for all four prescription drug
categories. These results suggest the potential for patterned responses to occur more frequently in
the redesigned prescription drug questions when the partially redesigned questionnaire is
implemented in 2015. Unlike the lead questions in prior modules, responses of "1" or "2" in the
screener questions do not mean "yes" or "no," respectively. Therefore, patterns of keying only
"1" or only "2" wherever possible suggest that these respondents may not have noticed the
change in meaning of these responses when they reached the prescription drug screener
questions. This potential data quality issue warrants further monitoring in the 2013 DR data and
the 2015 main study data.
4.6.7.3

Initiation Patterns in the QFT Prescription Drug Data

A total of 14 QFT respondents (including some of those who keyed responses of only "1"
in the screening questions) reported past year misuse of four or more individual prescription
drugs within a given prescription drug category, and they also reported no more than 1 year of
variation in the answers to the individual AFU questions. These included respondents who
reported first misuse of all prescription drugs within a category at the same age or often across
multiple categories of prescription drugs.
A cut point of four or more was chosen based on the distributions for the numbers of
individual prescription drugs for which respondents reported past year misuse. Specifically,
percentages of QFT respondents reporting past year misuse of zero to three individual
prescription drugs were 98.7 percent for pain relievers, 99.5 percent for tranquilizers and
stimulants, and almost all respondents for sedatives (i.e., the percentage shown to one decimal
place rounded to 100.0). For QFT respondents who were above this cut point, 26 reported past
year misuse of four or more individual pain relievers, including 9 respondents who reported
misuse of eight or more. For tranquilizers, 10 respondents reported past year misuse of four or
more individual drugs, including 3 respondents who reported misuse of eight or more.

176

For stimulants, 10 respondents reported past year misuse of four or more individual drugs,
including 7 respondents who reported misuse of six or seven stimulants.
One of these 14 respondents keyed responses of only "1" in the screening questions for
all four categories of prescription drugs. Consequently, all edited prescription drug variables for
this respondent (including the variables associated with the AFU questions) were assigned codes
of "bad data," as described previously. No further editing was done to the data on initiation of
misuse for the remaining 13 respondents. However, some of these respondents reported initiation
of misuse of all prescription drugs at the same age more than 10 years prior to the interview date;
AFUs for some of these prescription drugs also would have translated to initiation of misuse
prior to the availability of these drugs by prescription in the United States. Other respondents not
only reported initiation of misuse of all drugs at the same age but also reported initiation of
misuse of all prescription drugs in the same year and month or keying of the response for "don't
know" (DK) for the MFU questions after the first couple of times of being asked questions for
the AFU, MFU, and YFU. This latter pattern could suggest either annoyance or fatigue
associated with the respondent repeatedly asked about first misuse.
An additional five QFT respondents were identified with reports of past year misuse of
relatively high numbers of individual prescription drugs. Unlike the previous 14 respondents,
these respondents provided more variation in their initiation data. One of these five respondents
also had codes of "bad data" assigned to prescription drug variables because the respondent
keyed only responses of "1" wherever possible in the screening questions. No further editing was
done to the data on initiation of misuse for the remaining four respondents.
4.6.7.4

Measurement Issues for Initiation of Prescription Drug Misuse in the QFT

The assumed primary analytic aim of the questions about initiation of misuse of
prescription drugs is to distinguish between respondents who first misused all prescription drugs
within a given category within the past 12 months (i.e., past year initiates) and those who
initiated misuse of some prescription drugs in that category more than 12 months ago. If that is
the case, then respondents' ability to recall accurately the exact ages when they first misused each
individual prescription drug would become a secondary concern. In particular, if respondents can
recall accurately that they first misused some prescription drugs in that category more than
12 months prior to being interviewed, then they by definition would not be past year initiates,
even if there is some inaccuracy in their self-reports of when they first misused every individual
drug.
On the surface, if respondents reported past year initiation of misuse for all individual
prescription drugs in a category that they misused in the past year, then it would appear that these
respondents could be classified as past year initiates of misuse for that category. For example,
suppose a respondent reported misuse of four different prescription pain relievers and reported
first misuse of all four at his or her current age. By definition, initiation of misuse for each of
these pain relievers would have occurred in the past 12 months.
Because QFT respondents were asked questions about their first misuse of the
prescription drugs that they misused in the past 12 months, a limitation of these initiation
questions is that they do not capture information about other prescription drugs in the category

177

that respondents may have last misused more than 12 months ago. In the preceding example, if
the respondent who misused four pain relievers at his or her current age misused a fifth pain
reliever at some point in his or her lifetime but not in the past 12 months, the pain reliever
questions in the QFT would not capture information about this additional prescription pain
reliever. By definition, however, a respondent who misused any prescription drugs within a
category (e.g., pain relievers) more than 12 months ago could not be a past year initiate for the
overall category. A respondent who reported first misusing a prescription drug with a particular
active ingredient (e.g., the pain reliever hydrocodone, such as Vicodin® or the generic equivalent
hydrocodone with acetaminophen) or within a given prescription drug subcategory (e.g.,
benzodiazepine tranquilizers such as Xanax® or the generic equivalent alprazolam) also could
not be classified with certainty as a past year initiate for the more narrowly defined subcategory.
As for the definition of past year initiation for the overall prescription drug category, the
respondent could have misused similar drugs in a subcategory (e.g., other pain relievers
containing hydrocodone) more than 12 months ago but not in the past 12 months and therefore
would not have been asked about these other drugs in the QFT.
4.6.8

Issues to Consider for the Dress Rehearsal

Based on the review of responses to the prescription drug questions in the QFT, two
issues may be particularly relevant to the design of these questions for the 2013 DR:
1. alerting respondents that responses of "1" or "2" in the prescription drug screening
questions do not necessarily mean "yes" or "no," respectively; and
2. capturing information about potential initiation of prescription drug misuse more than
12 months ago for those respondents who reported past year initiation of all
prescription drugs in a category that they misused in the past year.
4.6.8.1

Alerting Respondents to Content Changes for Prescription Drugs

At a minimum, revisions to the prescription drug questions for the 2013 DR in response
to the first issue could involve an introductory screen prior to the start of the screener for pain
relievers to inform respondents of the change in meaning of responses of "1" or "2." Ideally, this
would slow down respondents sufficiently to pay attention to this change.
However, if respondents are hurrying through the core drug questions without paying
close attention to changes in the content—especially if they have become conditioned to expect
that "2" means "no"—they still may fail to pay sufficient attention to a new introductory screen
immediately prior to the prescription drug screeners. Therefore, an additional option for the 2013
DR would be inclusion of new logic relatively early in the screening questions for a given
prescription drug category to alert respondents if they appear to be falling into a pattern of
keying responses of only "1" or only "2" in the screener. For example, if a respondent entered
answers of only "2" in the first two screening questions about past year use of pain relievers, the
respondent might be prompted about what these responses of "2" mean (e.g., past year use of
Lortab® and Percocet®, respectively, based on the content of the QFT questions). The respondent
then would be asked whether these answers are correct. In case respondents have gotten
conditioned to associate responses of "1" with "yes" and responses of "2" with "no," the question
asking respondents to indicate whether these previous answers were correct could involve use of
178

a response other than "1" for "yes" if respondents want to confirm their answer and a response
other than "2" for "no" if they want to indicate that their previous answers were not correct.
Respondents who indicate that their previous answers were not correct would be re-asked the
relevant screener questions to allow them to change their answers to these questions.
The decisions were made not to implement either of these changes for the 2013 DR.
However, continued monitoring of the occurrence of these patterns is planned for the DR.
4.6.8.2

Refining the Initiation Questions for Prescription Drugs

In keeping with the aim of distinguishing between past year initiates of misuse of any
prescription drug within a category and respondents who initiated misuse of some prescription
drugs in that category more than 12 months ago, it would be necessary in the 2013 DR to collect
additional initiation data only from those respondents who reported past year initiation of misuse
for all of the prescription drugs in a category that they misused in the past year. If DR
respondents continue to be asked initiation questions for each prescription drug that they misused
in the past year, then any respondents who first misused any of these drugs more than 12 months
prior to the interview date are not past year initiates. If first misuse in the past 12 months is the
only initiation that respondents report for prescription drugs that they misused in that same
period, they could be asked a follow-up question to determine if they ever misused any
prescription drugs in that category more than 12 months ago.
Follow-up questions have been added to the 2013 DR instrument for respondents who
report only past year initiation of specific prescription drugs in a given category (e.g., pain
relievers).22 These respondents will be asked whether they ever misused any prescription drug in
that category more than 12 months prior to the interview date. Respondents who answer this
follow-up question as "no" can be classified as past year initiates of misuse for any prescription
drug in that category. Those who answer the follow-up question as "yes" can be classified as not
being past year initiates. As noted previously, it will not be necessary to ask this follow-up
question if respondents reported initiating misuse more than 12 months ago for any prescription
drugs that they also misused in the past year. By definition, these respondents are not past year
initiates.

Included in the classification of respondents who reported only past year initiation are those who had
missing data on initiation for some drugs in a given category (i.e., responses of "don't know" or "refused") and
reported past year initiation for the remaining prescription drugs in that category that they misused in the past year.

179

180

5. Assessments of the Redesigned Protocol
5.1

Overview of QFT Protocol Assessment

This chapter presents the results of four efforts to assess the partially redesigned protocol
used for the 2012 Questionnaire Field Test (QFT) data. The overall purpose of these assessments
was to ensure that the revised questionnaire and protocol used for the 2012 QFT will facilitate
continued high quality and efficiency in National Survey on Drug Use and Health (NSDUH) data
collection when the partial redesign is implemented in 2015. Section 5.2 presents complete
results of field observations of QFT field interviewers (FIs). Section 5.3 provides selected data
compiled from FI debriefing items completed for QFT cases. Section 5.4 presents findings from
two surveys on new equipment used by FIs in the QFT. Section 5.5 provides key findings from
three focus groups conducted with QFT FIs about their experiences using the redesigned
NSDUH interview protocol and tablet computer for screening.

5.2

Summary of Results from Field Observations of QFT Field Interviewers

This section summarizes the results of the field observations described previously in
Section 2.4.7.2 of Chapter 2. All field observations were completed between September 4 and
September 17, 2012. During this time period, a total of 20 field observations were completed
with 20 different FIs. These FIs completed 34 screenings and 28 interviews. Substance Abuse
and Mental Health Services Administration (SAMHSA) staff observed 5 of the 20 FIs
completing 10 screenings and 5 interviews. The remaining observations were conducted by RTI
staff, which included observations by one FS, two regional supervisors (RSs), and two other RTI
staff members. This section summarizes the field observation procedures followed and the errors
observed. It also includes comments from observers and FIs about the new materials, procedures,
and equipment used for the QFT data collection.
Several trends emerged among the QFT field observation data. The majority of FIs
displayed positive behaviors when conducting screenings (see Appendix D). Of the 21 items
listed on the QFT field observation screening checklist, only 2 items were observed being
conducted incorrectly more than 5 percent of the time:
•

not asking all roster questions verbatim, and

•

not reading verification instructions verbatim when no household members were
selected for an interview (code 22, 25, 26, or 30).

These errors were not specifically related to the QFT and could have occurred during a main
study observation. Based on observation of these errors, no changes to the equipment or
materials are anticipated. Items were added to the QFT field observation screening checklist to
reflect changes to the screening procedures, project information, and use of specific QFT
materials. There was only one error recorded for these items (see Table 5.1) in which an FI did
not correctly answer a respondent's questions using the QFT-specific information.

181

Table 5.1 Screening Errors Specific to the Questionnaire Field Test
Screening Error
Not including name, RTI International, DHHS, and lead letter in introduction
Not providing respondent with correct QFT materials
Answer questions correctly and thoroughly, referencing correct QFT details (e.g., RTI
International, DHHS, did not mention QFT or field test, sample size, or payment)
TOTAL

Error
Rate, %
0.00
0.00

Errors
Observed
0
0

2.94
0.98

1
1

DHHS = U.S. Department of Health and Human Services; QFT = Questionnaire Field Test.
NOTE: The error rate equals the percentage of observed cases where the error was observed. A total of 34 interviews were
observed.

The majority of FIs also displayed positive behaviors when conducting interviews (see
Appendix D). Of the 14 items listed on the QFT field observation interviewing checklist, only
3 items were observed being conducted incorrectly at least 5 percent of the time:
•

not explaining the purpose of the study thoroughly to an interview respondent who
was not the screening respondent;

•

not handing the QFT study description to the respondent; and

•

not reading all screens verbatim.

As with the observed screening errors, these errors were not related specifically to the QFT and
could have occurred during a main study observation. In instances where an error was recorded
for the FI not handing the QFT study description to the respondent, the FI did not hand any study
description to the respondent. This error was not attributed to the QFT procedures.
Items were added to the QFT field observation interview checklist to reflect changes to
the interview procedures, project information, and use of specific QFT materials. Two errors
were recorded on these items, as noted in Table 5.2. For both of these errors, the FI used
procedures or language from the main study instead of following QFT procedures.
Table 5.2 Interview Errors Specific to the Questionnaire Field Test
Interview Error
Not following the proper QFT quality control form and incentive procedures
Not answering respondent questions correctly and thoroughly, referencing the appropriate
QFT details (e.g., RTI International, DHHS, did not mention QFT or field test, sample
size, or payment)
Not providing respondent with correct QFT materials
TOTAL

Error
Rate, %
3.57

Errors
Observed
1

3.57
0.00
2.38

1
0
2

The field observations show that FIs generally did well at following both new procedures
specific to the QFT and procedures carried over from the main study. Although it is a cause for
concern to see any violations of protocol, errors were relatively infrequent during the QFT field

182

observations. The results do not indicate that the majority of these errors were the result of any
new field procedures specific to the QFT.
Observers were also asked to evaluate the performance of the QFT equipment (i.e., tablet
and laptop) and materials (i.e., QFT lead letter, QFT study description, and "question & answer"
[Q&A] brochure) while in the field. There were no additional comments or concerns from
observers about the performance of the QFT materials during their observations. Three
comments were provided about the performance of the tablet in the field. One FI was concerned
that there was more glare on the tablet screen in direct sunlight than typically observed with the
current iPAQ device. Another FI suggested that a new functionality be added to the tablet
program, removing finalized cases from the "select case" screen when transmitted. This change
does not need to be made for the 2013 Dress Rehearsal (DR) because this functionality is already
available on the tablet. The view/sort function on the tablet already allows FIs to select whether
they want to view pending or final cases on the select case screen. Two FIs had issues
troubleshooting unexpected events with the tablet, such as an alarm going off during a screening.
These troubleshooting issues are to be addressed during the 2013 DR training, and
documentation will be added to the FI handbook on how to resolve these occurrences. The QFT
field observations did not uncover any serious concerns about the QFT equipment or materials.
Observers did witness some respondent confusion during the interview. Respondents
asked FIs for assistance with or were obviously confused by the following questions:
•

GOTDOG:
You answer questions by putting in the number that is shown next to your answer.
The numbers are located in the second row of the keyboard.
To answer a question, you first press the correct number and then press [ENTER].
Do you have a dog?

One respondent pressed F2 instead of 2 to answer this question and needed FI assistance.
•

AL08:
During the past 30 days, that is, since [DATEFILL], on how many days did you have
[IF QD01=5 THEN FILL 5 IF QD01=9 THEN FILL 4] or more drinks on the same
occasion? By "occasion," we mean at the same time or within a couple of hours of
each other.

One respondent asked what the definition of "occasion" was for this question.
•

Pain Relievers Module:

One respondent asked the interviewer to explain the difference between Tylenol® with
Codeine 3 and Tylenol® with Codeine 4.
One respondent asked if he should be reporting pain relievers he was prescribed by a
doctor and read the question out loud to the FI.
•

SP09:
In [STATE FILL FROM FIPE4], has marijuana been legally approved for medical
use?
183

One respondent did not know how to answer this question. She asked the FI, and the FI
instructed her to use the "Don't Know" option.
•

HLTH19:
During the past 12 months, how many times have you visited a doctor, nurse,
physician assistant or nurse practitioner about your own health at a doctor's office, a
clinic, or some other place?

One respondent asked if she should include all trips to the doctor because she is pregnant
and goes to the doctor regularly.
•

QD35:
How many different employers, including yourself, have you had in the past
12 months?

One respondent was confused on how to answer this question if he or she had only one
employer.
•

Household Roster:

One respondent was confused on how to answer the relationship questions in this section,
which asks about the ages and relationships of household members.
These experiences suggest that respondents might express similar confusion on these questions
in the main study data collection. However, the main study field observations do not provide
comparison data on how many times respondents were confused or what comments respondents
made on these same issues.
Several respondents also made comments as they completed the interview. These
comments do not necessarily indicate confusion or issues with the questionnaire, but they do
give some insight into how respondents reacted to the instrument.
•

ACASI (audio computer-assisted self-interviewing)—One respondent commented
that the drug names made him laugh.

•

ACASI—One respondent volunteered that she was a nurse and had not heard of all
the drugs included in the ACASI. She commented that it was "an education."

•

ACASI—One respondent laughed at the marijuana and crack availability questions,
which ask how easily one could obtain these drugs.

•

ACASI—One respondent commented, "I'm sure there are people who take all of
these, but this is insane. I can't imagine."

•

Household Roster—One respondent wondered why they had to repeat this
information about household members from the screening and commented that it was
repetitive.

•

Household Roster—One respondent commented that the relationship questions were
"unusual."

184

Observer comments also suggested changes that could be made to the computer-assisted
interviewing (CAI) instrument. In two cases, it was suggested that a transition statement or
instructions be added to the end of the interview to provide some context for the FI tasks. This
statement would allow the end of the interview to flow more naturally and not leave the
respondent sitting in silence while the FI finishes his or her tasks.
Despite issues with respondent confusion or misunderstanding, FI performance during
field observations met the expected quality standards. Out of a possible 714 screening errors in
the QFT field observations (34 completed screenings multiplied by 21 possible errors on the
QFT field observation screening checklist), field observers noted 8 errors, or 1.12 percent of the
possible screening errors. Out of a possible 392 interviewing errors in the QFT field observations
(28 completed interviews multiplied by 14 possible errors on the QFT field observation
interviewing checklist), field observers noted 17 errors, or 4.34 percent of the possible interview
errors.
Overall, the 20 completed field observations provided an important opportunity to see
firsthand how the QFT instrument, materials, and equipment performed in the field. These items
all performed well, and only minimal changes were suggested. Several items that observers were
instructed to observe went so smoothly that there were no reported issues or comments,
including the flow of the screening presentation, overall issues with the tablet or tablet case, and
issues transitioning between the screening and the interview. The lack of comments on these
items, combined with the few comments and issues reported on other QFT-specific items,
indicates the instruments, equipment, and materials performed well in the field. Although some
small errors were observed, the QFT FIs also performed well while working with the new
instrument, materials, and equipment. Because these observations were conducted with
experienced FIs and from a nonrandom selection, they may not be generalizable to the NSDUH
main study FI population. These field observation data did not produce any suggestions for
significant changes to the 2013 DR or the 2015 redesign.

5.3

QFT Field Interviewer Debriefing Results

Additional insight on the redesigned protocol in 2015 was obtained from FI debriefing
questions that were administered at the end of each interview. Debriefing items (shown in
Appendix E) were included in the QFT protocol. Debriefing items asked FIs to note whether
respondents expressed any difficulties or reactions to certain features of the revised protocol,
such as the electronic version of the reference calendar, the electronic pill images, proxy use of
ACASI, and the new contact materials (Q&A brochure). In addition, FIs also responded to
debriefing items about the screening respondent's recall of the lead letter. Although this reporting
depends on unprompted information being supplied by QFT screening and interview
respondents, these items provide information that can be used to identify potential problems with
the new features of the redesigned protocol in an unobtrusive manner.
Tables 5.3 through 5.8 present information on FI reports of screening respondent recall of
the lead letter. FIs reported that older screening respondents (those 26 or older) were more likely
to recall seeing the lead letter than younger screening respondents (18 to 25 years old). To
examine screening respondent recall of the lead letter more closely, a three-category measure of
interview status at the dwelling unit level was created, as follows:
185

•

Not Selected – Dwelling units in which the screening was completed and no one was
selected for the interview.

•

Selected and Not Interviewed – Dwelling units in which the screening was completed
and at least one person was selected for the interview but no interviews were
completed. Interviews were not completed for several reasons, including refusal,
noncontact, and language barriers.

•

Selected and Interviewed – Dwelling units in which the screening was completed and
at least one interview was completed.

Recall of the lead letter appeared to be associated with willingness to do the interview.
Table 5.4 shows that FI reports that the screening respondent recalled the lead letter were lower
when the dwelling unit was selected for an interview but not interviewed than when an interview
was completed in the dwelling unit. Tables 5.5 through 5.8 show that this pattern did not vary a
great deal by the age of the screening respondent, with the notable exception of cases where the
age of the screening respondent was 65 or older. As shown in Table 5.8, for screening
respondents aged 65 or older, there was little difference in the recall of the lead letter between
those in households where an interview was completed (57.5 percent) and those where a person
was selected but no interviews were completed (55.2 percent).
Table 5.3 Screening Respondent Recall of Lead Letter, by Screening Respondent Age
QFTDBF1 - Did the respondent
remember receiving the lead
letter?
Yes
No

18 to 25
(n = 353)
%
N
131
37.1
222
62.9

Screening Respondent Age
26 to 49
50 to 64
(n = 1,576)
(n = 1,054)
%
%
n
n
809
51.3
589
55.9
767
48.7
465
44.1

65 or Older
(n = 818)
%
n
422
51.6
396
48.4

Overall
(n = 3,801)
%
n
1,951
51.3
1,850
48.7

NOTE: Screening respondent age was missing for 28 completed screenings.

Table 5.4 Screening Respondent Recall of Lead Letter, by Dwelling Unit Interview Status

QFTDBF1 - Did the respondent
remember receiving the lead letter?
Yes
No

Not Selected1
(n = 1,931)
%
n
1,002
51.9
929
48.1

Dwelling Unit Interview Status
Selected & Not
Selected &
Interviewed2
Interviewed3
(n = 459)
(n = 1,443)
%
%
n
n
194
42.3
767
53.2
265
57.7
676
46.9

Overall
(n =3,833)
%
n
1,963
51.2
1,870
48.8

Dwelling units in which the screening was completed and no one was selected for the interview.
Dwelling units in which the screening was completed and at least one person was selected for the interview but no interviews
were completed.
3
Dwelling units in which the screening was completed and at least one interview was completed.
2

186

Table 5.5 Recall of Lead Letter among Screening Respondents Aged 18 to 25, by Dwelling Unit
Interview Status

QFTDBF1 - Did the respondent
remember receiving the lead letter?
Yes
No

Dwelling Unit Interview Status
Selected & Not
Selected &
Interviewed2
Interviewed3
(n = 51)
(n = 237)
%
%
n
n
13
25.5
87
36.7
38
74.5
150
63.3

Not Selected1
(n = 65)
%
n
31
47.7
34
52.3

Overall
(n = 353)
%
n
37.1
131
62.9
222

Table 5.6 Recall of Lead Letter among Screening Respondents Aged 26 to 49, by Dwelling Unit
Interview Status

QFTDBF1 - Did the respondent
remember receiving the lead letter?
Yes
No

Dwelling Unit Interview Status
Selected & Not
Selected &
Interviewed2
Interviewed3
(n = 239)
(n = 768)
%
%
n
n
99
41.4
422
55.0
140
58.6
346
45.1

Not Selected1
(n = 569)
%
n
288
50.6
281
49.4

Overall
(n = 1,576)
%
n
51.3
809
48.7
767

Table 5.7 Recall of Lead Letter among Screening Respondents Aged 50 to 64, by Dwelling Unit
Interview Status

QFTDBF1 - Did the respondent
remember receiving the lead letter?
Yes
No

Dwelling Unit Interview Status
Selected & Not
Selected &
Interviewed2
Interviewed3
(n = 110)
(n = 272)
%
%
n
n
49
44.6
165
60.7
61
55.4
107
39.3

Not Selected1
(n = 672)
%
n
375
55.8
297
44.2

Overall
(n = 1,054)
%
n
589
55.9
465
44.1

187

Table 5.8 Recall of Lead Letter among Screening Respondents Aged 65 or Older, by Dwelling Unit
Interview Status
Dwelling Unit Interview Status
Selected &
Selected & Not
Interviewed3
Interviewed2
(n = 58)
(n = 153)
%
%
n
n
32
55.2
88
57.5
26
44.8
65
42.5

Not Selected1
(n = 607)
%
n
302
49.8
305
50.3

QFTDBF1 - Did the respondent
remember receiving the lead letter?
Yes
No

Overall
(n = 818)
%
n
422
51.6
396
48.4

Additional tabulations of the information presented in Tables 5.3 to 5.8 are shown in
Table 5.9 as the rates at which interviews were completed in households selected for interviews,
conditional on whether or not the lead letter was recalled. Overall, among those who were
selected for the interview, when the screening respondent mentioned recalling the lead letter,
80.3 percent of the dwelling units had at least one completed interview (767 out of 955). In
contrast, when the screening respondent did not mention recalling the lead letter, about
71 percent of dwelling units completed at least one interview (668 out of 933). When this is
examined by screening respondent age groups, the differences range from about 7 percentage
points for the 18 to 25 age group to about 13 percentage points for the 50 to 64 screening
respondent age group. In contrast, there is only a small difference in the percentages of
households interviewed by recall of the lead letter when the screening respondent was 65 or
older.
Table 5.9 Interview Status, by Recall of Lead Letter and Screening Respondent Age

Selected
Interviewed
Percent
Interviewed

18 to 25
Recalled Lead
Letter?
Yes
No
100
188
87
150
87.0%

79.8%

26 to 49
Recalled Lead
Letter?
Yes
No
521
486
422
346
81.0%

71.2%

50 to 64
Recalled Lead
Letter?
Yes
No
214
168
165
107
77.1%

63.7%

65 or Older
Recalled Lead
Letter?
Yes
No
120
91
88
65
73.3%

71.4%

Total
Recalled Lead
Letter?
Yes
No
955
933
767
668
80.3%

71.6%

Tables 5.10 and 5.11 provide more details on the screening respondent comments on the
lead letter as reported by the FIs. Not surprisingly, the selected but not interviewed households
had lower rates of screening respondents looking forward to the visit, expressions of interest in
the study, and willingness to participate in the study than screening respondents in dwelling units
where no one was selected for an interview or in dwelling units where at least one person was
selected for the interview and at least one interview was completed. Screening respondents in
dwelling units that were selected for an interview but did not complete an interview also had
higher rates of not wanting anyone to come to their homes, expressions of confusion, reports of
not having all questions about participation answered, and doubts about the confidentiality of
their information. Additional details on the lead letter comments and on the Q&A brochure, as
well as the length of the interview, are provided in Tables 5.10 through 5.15.
188

Table 5.10 Screening Respondent Comments on Lead Letter, by Screening Respondent Age
QFTDBF2 - What comments, if any, did the
respondent [R] make about the lead letter or
in response to the lead letter?
R did not make any comments about the
lead letter.
R was looking forward to your visit/been
waiting for you.
R was interested in the study.
R would like to participate in the study.
R does not believe the government is paying
$30/waste of tax dollars.
The letter answered the R's
questions/concerns.
R did not want someone coming to home
without permission.
R was confused by the letter.
The letter did not answer all of the R's
questions/concerns.
R does not believe the survey is confidential.
R thought this was a scam.
R does not open anything addressed to
"resident."
Other

18 to 25
(n = 131)

Screening Respondent Age
26 to 49
50 to 64
(n = 809)
(n = 589)

65 or Older
(n = 422)

Overall
(n = 1,951)

74.1

566

70.0

390

66.2

283

67.1

1,336

68.5

18
10
8

13.7
7.6
6.1

93
70
50

11.5
8.7
6.2

95
48
33

16.1
8.2
5.6

56
27
21

13.3
6.4
5.0

262
155
112

13.4
7.9
5.7

0.0

0.1

0.7

1.2

0.5

0.0

0.5

0.2

1.4

0.6

0
4

0.0
3.1

6
12

0.7
1.5

7
10

1.2
1.7

9
6

2.1
1.4

22
32

1.1
1.6

1
0

0.8
0.0

18
5

2.2
0.6

13
7

2.2
1.2

13
7

3.1
1.7

45
19

2.3
1.0

0.0

0.7

1.4

0.8

0
4

0.0
3.1

4
32

0.5
4.0

8
27

1.4
4.6

1
24

0.2
5.7

13
87

0.7
4.5

Table 5.11 Screening Respondent Comments on Lead Letter, by Dwelling Unit Interview Status

QFTDBF2 - What comments, if any, did the
respondent [R] make about the lead letter or in
response to the lead letter?

Not Selected1
(n = 1,002)
n
%

Dwelling Unit Interview Status
Selected & Not
Selected &
Interviewed2
Interviewed3
(n = 194)
(n = 767)
n
%
n
%

R did not make any comments about the lead letter.
R was looking forward to your visit/been waiting
for you.
R was interested in the study.
R would like to participate in the study.

673

67.2

139

71.7

529

69.0

1,341

68.3

146

14.6

9.8

101

13.2

266

13.6

78
54

7.8
5.4

5
5

2.6
2.6

76
56

9.9
7.3

159
115

8.1
5.9

7
8

0.7
0.8

1
1

0.5
0.5

2
2

0.3
0.3

10
11

0.5
0.6

13
16

1.3
1.6

7
4

3.6
2.1

2
12

0.3
1.6

22
32

1.1
1.6

2.1

3.1

2.4

2.3

14
12
8
45

1.4
1.2
0.8
4.5

4
2
1
12

2.1
1.0
0.5
6.2

2
2
4
31

0.3
0.3
0.5
4.0

20
16
13
88

1.0
0.8
0.7
4.5

R does not believe the government is paying
$30/waste of tax dollars.
The letter answered the R's questions/concerns.
R did not want someone coming to home without
permission.
R was confused by the letter.
The letter did not answer all of the R's
questions/concerns.
R does not believe the survey is confidential.
R thought this was a scam.
R does not open anything addressed to "resident."
Other

Overall
(n = 1,963)
n
%

Dwelling units in which the screening was completed and no one was selected for the interview.

Dwelling units in which the screening was completed and at least one person was selected for the interview but no interviews were completed.

Dwelling units in which the screening was completed and at least one interview was completed.

189

Table 5.12 Timing of Providing Q&A Brochure
QFTDBF3 - When did you give the respondent (or parent/guardian of youth
respondent) the Q&A [question and answer] brochure?
Before the interview
During the interview
At the end of the interview
TOTAL

n
517
35
1,488
2,040

25.3
1.7
72.9
99.9

NOTE: Percentages do not sum to 100 percent due to rounding.

Table 5.13 Comments on Q&A Brochure
QFTDBF3a - What comments, if any, did the respondent [R] (or parent/guardian) make
about the Q&A [question and answer] brochure?
There were no comments about the Q&A brochure.
The brochure did not answer all of the R's questions about the study.
The brochure addressed the R's questions.
The R was confused by the brochure.
The brochure encouraged the R to participate.
Other

n
1,911
16
53
2
40
32

93.7
0.8
2.6
0.1
2.0
1.6

NOTE: Percentages are based on 2,040 respondents; more than one response could be selected.

Table 5.14 Comments on Q&A Brochure, by Timing of Providing Brochure

QFTDBF3a - What comments, if any, did the respondent [R]
(or parent/guardian) make about the Q&A [question and answer]
brochure?
There were no comments about the Q&A brochure.
The brochure did not answer all of the R's questions about the
study.
The brochure addressed the R's questions.
The R was confused by the brochure.
The brochure encouraged the R to participate.
Other

When Brochure Was Provided
During
Before Interview
Interview
End of Interview
(n = 517)
(n = 35)
(n = 1,488)
%
%
%
n
n
n
433

83.8

85.7

1,448

97.3

11
39
2
36
9

2.1
7.5
0.4
7.0
1.7

0
3
0
1
1

0.0
8.6
0.0
2.9
2.9

5
11
0
3
22

0.3
0.7
0.0
0.2
1.5

NOTE: Percentages are based on responses to QFTDBF3; more than one response could be selected.

Table 5.15 Respondent Comments on the Interview Being Too Long
QFTDBF9 - Did the respondent make any comments about the interview being too
long?
Yes
No
TOTAL

n
261
1,779
2,040

%
12.8
87.2
100.0

Table 5.16 shows that a larger percentage of persons aged 50 to 64 (18 percent) and those
aged 65 or older (29 percent) made comments about the interview being too long compared with
other age groups (10 to 12 percent). These comments are consistent with the timing data
presented in Table 4.9a in Section 4.5, which shows that respondents in the 65 or older age
group had the highest mean and median interview times among all age groups in the sample.

190

Table 5.16 Respondent Comments on the Interview Being Too Long, by Interview Respondent
Age
QFTDBF9 - Did the respondent make
any comments about the interview
being too long?
Yes
No

12 to 17
(n = 539)
%
n
58
10.8
481
89.2

Interview Respondent Age
18 to 25
26 to 49
50 to 64
(n = 504)
(n = 678)
(n = 190)
%
%
%
n
n
n
50
9.9
81
12.0
35
18.4
454
90.1
597
88.1
155
81.6

65 or Older
(n = 129)
%
n
37
28.7
92
71.3

Table 5.17 shows that more than 2 times as many interview respondents with less than a
high school education reported that the interview was too long compared with respondents with
higher levels of education overall. These comments cannot be directly compared with interview
timing data because the timing data were not calculated by respondent education level.
Table 5.17 Respondent Comments on the Interview Being Too Long, by Interview Respondent
Education

QFTDBF9 - Did the respondent make any comments about
the interview being too long?
Yes
No

< High
School
(n = 187)
%
n
50
137

26.7
73.3

Interview Respondent Education
High School
Some
Graduate
College
(n = 425)
(n = 531)
%
%
n
n
62
363

14.6
85.4

50
481

9.4
90.6

College
Graduate
(n = 538)
%
n
41
317

11.5
88.6

NOTE: Interview Respondent Education is shown only for persons aged 18 or older.

Comments on the prescription drug questions were recorded by FIs, and the 207
responses were coded into the general themes displayed in Table 5.18. The most frequent type of
comment recorded by FIs was the number of prescription drugs asked in these modules. Among
those respondents for whom any comment was recorded, about 40 percent provided a comment
consistent with this theme. In some cases, the comments were expressions that the number of
prescription drug items was burdensome, but in other cases respondents simply expressed
surprise at the numbers of prescription drugs available.
Table 5.18 Classification of Open-Ended Comments on Prescription Drug Questions
Please describe the respondent's [R's] comments about the prescription drug questions.
Comment on numbers of drug questions
Concepts of prescription drug use and misuse
Navigation issues/code 95 for have not used in past 12 months
Drug classification issues (e.g., uncertainty on reporting over-the-counter medications; categories
in which certain drugs might fit)
Personal experiences/circumstances with drug use
Comment on specific drug(s)
Comprehension comments
Comment that R requested help from someone to answer
Unclassified
TOTAL

191

n
80
48
14

%
38.6
23.2
6.8

10
12
12
9
7
15
207

4.8
5.8
5.8
4.3
3.4
7.2
100.0

The next most frequent type of comment was on the concepts of use and misuse of
prescription drugs, accounting for 23 percent of the comments in this category (see Tables 5.19
and 5.20). Many of the comments focused on whether respondents had a prescription at some
point and having questions about what should be recorded, but it was not always clear if these
comments were referring to the drug screening items or to the follow-up items.
Table 5.19 Interview Respondent Questions or Comments on Prescription Drug Questions
QFTDBF10 - Did the respondent have any questions or comments about the
prescription drug questions in the ACASI [audio computer-assisted selfinterviewing] section of the questionnaire?
Yes
No
TOTAL

%
10.1
89.9
100.0

n
207
1,833
2,040

Table 5.20 Interview Respondent Questions or Comments on Prescription Drug Questions, by
Interview Respondent Age
QFTDBF10 - Did the respondent
have any questions or comments
about the prescription drug
questions in the ACASI [audio
computer-assisted selfinterviewing] section of the
questionnaire?
Yes
No

12 to 17
(n = 539)

n
31
508

%
5.8
94.3

18 to 25
(n = 504)

%
7.9
92.1

n
40
464

Respondent Age
26 to 49
(n = 678)

%
11.1
88.9

n
75
603

50 to 64
(n = 190)

n
23
167

%
12.1
87.9

65 or Older
(n = 129)

n
38
91

%
29.5
70.5

Table 5.21 Interview Respondent Questions or Comments on Prescription Drug Questions, by
Interview Respondent Education
QFTDBF10 - Did the respondent have any questions or
comments about the prescription drug questions in the
ACASI [audio computer-assisted self-interviewing] section
of the questionnaire?
Yes
No

< High
School
(n = 187)
%
n
29
15.5
158
84.5

Education
High School
Some
Graduate
College
(n = 425)
(n = 531)
%
%
n
n
43
10.1
52
9.8
382
89.9
479
90.2

College
Graduate
(n = 538)
%
n
52
14.5
306
85.5

NOTE: Interview Respondent Education is shown only for persons aged 18 or older.

Finally, a small number of respondents (14) reported confusion about the use of "95" in
the drug screening questions to indicate that they have not used a particular drug in the past
12 months (data not shown). These respondents felt that "95" was not an intuitive number to
indicate nonuse, preferring either "0" or the next number in the sequence (i.e., if four drugs are
listed as 1, 2, 3, and 4, 5 would be the choice for never having used in the past 12 months). Given
the small number of respondents who expressed confusion about the use of "95" in the drug
screening questions to indicate nonuse, it was decided not to change this response option for the
2013 DR.

192

Tables 5.22 to 5.25 provide details regarding the comments on the on-screen calendars.
Overall, very few comments were made by respondents about the on-screen calendars. The lack
of comments suggested that respondents were able to understand and use the on-screen calendars
with relative ease.
Table 5.22 Any Interview Respondent Questions or Comments on On-Screen Calendars
QFTDBF11 - Did the respondent have any questions or comments about the onscreen calendars in the ACASI [audio computer-assisted self-interviewing]
section of the questionnaire? If the respondent asked how to access the calendar
at any time during the ACASI portion of the interview, select "YES."
Yes
No
TOTAL

21
2,019
2,040

1.0
99.0
100.0

Table 5.23 Any Interview Respondent Questions or Comments on On-Screen Calendars, by
Interview Respondent Age
QFTDBF11 - Did the respondent have any
questions or comments about the on-screen
calendars in the ACASI [audio computerassisted self-interviewing] section of the
questionnaire? If the respondent asked how
to access the calendar at any time during the
ACASI portion of the interview, select
"YES."
Yes
No

Respondent Age
26 to 49
(n = 678)

12 to 17
(n = 539)

18 to 25
(n = 504)

50 to 64
(n = 190)

65 or Older
(n = 129)

6
533

1.1
98.9

5
499

1.0
99.0

5
673

0.7
99.3

1
189

0.5
99.5

4
125

3.1
96.9

Table 5.24 Any Interview Respondent Questions or Comments on On-Screen Calendars, by
Interview Respondent Education
QFTDBF11 - Did the respondent have any questions
or comments about the on-screen calendars in the
ACASI section of the questionnaire? If the
respondent asked how to access the calendar at any
time during the ACASI portion of the interview,
select "YES."
Yes
No

< High School
(n = 187)

Education
High School
Graduate
Some College
(n = 425)
(n = 531)

College
Graduate
(n = 538)

5
182

2.7
97.3

5
420

1.2
98.8

3
528

0.6
99.4

2
356

0.6
99.4

NOTE: Interview Respondent Education is shown only for persons aged 18 or older.

Table 5.25 Types of Interview Respondent Questions or Comments on On-Screen Calendars
QFTDBF11a - What comments did the respondent [R] make about the on-screen calendars?
The R asked how to access the calendar.
The R asked how to close the calendar.
The R did not see the reference dates on the calendar.
The calendar helped the R answer the question.
The calendar covered the questions or the images on the screen.
Other

n
4
1
1
5
1
13

NOTE: Percentages are based on the 21 "Yes" answers to QFTDBF11; more than one response could be chosen.

193

%
19.1
4.8
4.8
23.8
4.8
61.9

Table 5.26 shows that for about 10 percent of the interviews, the FI recorded that the
respondent had trouble understanding questions besides those on prescription drugs. The most
noteworthy problem mentioned in response to QFTDBF12 ("Did the respondent have trouble
understanding any other questions asked during the interview?") was with the new PLAYINFO
item in the ACASI tutorial. The new question asks respondents, "In the past 30 days, on how
many days did you eat any kind of fried potatoes?" and instructs the respondent to use the F2 key
to bring up additional information on what is meant by "fried potatoes." A total of 19
respondents (less than 1 percent) reported a problem in answering the question or using the F2
key. In some cases, respondents were not clear what to do after entering F2. Some respondents
perhaps did not realize that they must enter a response after seeing the pop-up instruction box.
Based on these results, the wording of PLAYINFO will be revised for the 2013 DR to explain
more clearly the steps respondents must take to enter a response for these questions.
Table 5.26 Interview Respondent Troubles with Other Questions
QFTDBF12 - Did the respondent have trouble understanding any other
questions asked during the interview?
Yes
No
TOTAL

n
193
1,847
2,040

%
9.5
90.5
100.0

Information on interviewer reports of the use of proxies for reporting on income and
health insurance items, respondent views on the use of proxies to provide this information, and
reported problems with proxy reporting are shown in Tables 5.27 to 5.34. Table 5.29 shows that
interviewers did not report any respondents with concerns about whether the proxy respondent
could see responses to questions answered by the respondent (which the instrument did not
allow), and very few respondents (2.3 percent) had any questions or comments about the proxy
interview (Table 5.30).
Table 5.27 Proxy Used for Income and Health Insurance Questions
QFTDBF13 - Was a proxy used for the income and health insurance questions?
Yes
No
TOTAL

n
602
1,438
2,040

%
29.5
70.5
100.0

Table 5.28 Proxy Used for Income and Health Insurance Questions, by Interview Respondent Age
QFTDBF13 - Was a proxy used
for the income and health
insurance questions?
Yes
No

12 to 17
(n = 539)
%
n
452
83.9
87
16.1

18 to 25
(n = 504)
%
n
81
16.1
423
83.9

194

Respondent Age
26 to 49
(n = 678)
%
n
45
6.6
633
93.4

50 to 64
(n = 190)
%
n
10
5.3
180
94.7

65 or Older
(n = 129)
%
n
14
10.9
115
89.2

Table 5.29 Interview Respondent Concerns about Revealing Answers to Proxy Respondent
QFTDBF14 - Did the respondent have any questions or concerns about his/her
answers being revealed to the proxy?
Yes
No
TOTAL

n
0
604
604

%
0.0
100.0
100.0

Table 5.30 Interview Respondent Questions or Comments about Proxy Interview
QFTDBF15 - Did the respondent have any other questions or comments about
the proxy interview?
Yes
No
TOTAL

%
2.3
97.7
100.0

n
14
590
604

Similarly, as shown in Tables 5.31 to 5.34, interviewers reported very few problems with
proxy respondents using the proxy ACASI tutorial or with answering questions in ACASI.
Problems in using the proxy ACASI tutorial were reported in only 3.5 percent of interviews in
which a proxy was used (Table 5.31). Problems with answering questions on health insurance
and income by proxy respondents were only mentioned in 5.5 percent of interviews in which a
proxy was used (Table 5.33).
Table 5.31 Problems with Proxy on ACASI Tutorial
QFTDBF16 - Were there any problems with the proxy's understanding of the
ACASI [audio computer-assisted self-interviewing] tutorial?
Yes
No
TOTAL

n
21
583
604

3.5
96.5
100.0

Table 5.32 Types of Problems with Proxy on ACASI Tutorial
QFTDBF16a - Which of the following describes the problems with the proxy's
understanding of the tutorial?
The proxy did not understand how to answer the questions.
The proxy did not know why he/she was asked to answer these questions.
Other

10
4
9

47.6
19.1
42.9

NOTE: Percentages are based on 21 reports of problems with proxy understanding in QFTDBF16; more than one response could
be chosen.

Table 5.33 Problems with Proxy Use of ACASI to Answer Income and Health Insurance
Questions
QFTDBF17 - Were there any problems with the proxy's use of ACASI [audio
computer-assisted self-interviewing] to answer the income and health insurance
questions?
Yes
No
TOTAL

195

n
33
571
604

%
5.5
94.5
100.0

Table 5.34 Types of Problems with Proxy Use of ACASI to Answer Income and Health Insurance
Questions
QFTDBF17a - Which of the following describes the problems with the proxy's use
of ACASI [audio computer-assisted self-interviewing] in answering the income and
health insurance questions? Check all that apply.
The proxy did not know the answers to the questions.
The proxy did not know how to enter his/her answers to the questions.
The proxy refused to answer some questions.
The proxy did not know why he/she was asked to answer these questions.
Other

4
5
0

12.1
15.2
0.0

4
24

12.1
72.7

NOTE: For responses of "OTHER," follow-up information was not collected.

As Table 5.34 shows, over 70 percent of the responses provided regarding problems with
proxy use of ACASI to answer the income and health insurance questions were in the "other"
category. Open-ended "other" responses were not captured and coded for the 2012 QFT, but
these "other" responses will be captured for the 2013 DR.
Tables 5.35 to 5.38 present information on interview locations, interviewer ratings of
privacy, and reports of other persons in the presence of the interview. Overall, the distributions
of responses to these debriefing items from the QFT were similar to those from the comparison
samples.
Table 5.35 Interviews Conducted at Respondent's Home for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test (QFT)
Did you conduct this interview at the respondent's home,
either inside or outside?
Yes
No

2011 Main Study
%
n
64,933
98.5
976
1.5

2012 Quarters 3
and 4 Main Study
%
n
30,687
98.3
522
1.7

2012 QFT
%
n
1,998
97.9
42
2.1

Table 5.36 Interview Location Not at Respondent's Home for the 2011 Main Study, 2012
Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test (QFT)
2011 Main Study
%
n
216
22.1
131
13.4

Where did you conduct this interview?
At the respondent's workplace
At the home of the respondent's relative or friend
In some type of conference room in a residence hall, school or
apartment complex
At a library
In some type of common area, such as a lobby, hallway,
stairwell, or laundry room
Some other place

196

2012 Quarters 3
and 4 Main Study
%
n
99
19.0
51
9.8

2012 QFT
%
n
10
23.8
9
21.4

248
159

25.4
16.3

127
103

24.3
19.7

12
6

28.6
14.3

72
150

7.4
15.4

75
67

14.4
12.8

2
3

4.8
7.1

Table 5.37 Field Interviewer (FI) Evaluation of Interview Privacy in Respondent's Home for the
2011 Main Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field
Test (QFT)
Please indicate how private the interview was. Do not count
yourself or a project observer as another person in the room.
Completely private–no one was in the room or could overhear
any part of the interview
Minor distractions–person(s) in the room or listening less than
1/3 of the time
Person(s) in the room or listening about 1/3 of the time
Serious interruptions of privacy more than half the time
Constant presence of other person(s)

2011 Main Study
%
n

2012 Quarters 3
and 4 Main Study
%
n

54,544

82.8

25,630

82.1

1,617

79.3

8,406
1,080
236
1,643

12.8
1.6
0.4
2.5

4,154
546
129
750

13.3
1.7
0.4
2.4

277
45
13
88

13.6
2.2
0.6
4.3

2012 QFT
%
n

Table 5.38 Field Interviewer (FI) Reports of Others Present during Interview for the 2011 Main
Study, 2012 Quarters 3 and 4 Main Study, and 2012 Questionnaire Field Test (QFT)

Not including yourself or project observers, other people present
or listening to the interview were:
Parent(s)
Spouse
Live-in partner/ boyfriend/ girlfriend
Other adult relative(s)
Other adult(s)
Child(ren) under 15
Other

2011 Main
Study
%
n
5,227
46.0
1,538
13.5
642
5.6
1,404
12.4
1,058
9.3
3,791
33.4
379
3.3

2012 Quarters 3
and 4 Main
Study
%
n
2,522
45.2
744
13.3
335
6.0
677
12.1
531
9.5
1,776
31.8
191
3.4

2012 QFT
%
n
179
42.3
70
16.6
30
7.1
47
11.1
34
8.0
150
35.5
15
3.6

The findings that older respondents (those aged 50 or older) and those with less than a
high school education were both more likely to comment that the interview was too long suggest
that these respondents may face greater cognitive burden than other respondents and that steps
could be taken to either address these concerns or that additional items should be added to the
survey to account for differences in cognitive abilities and familiarity with computers. For
example, in a study of 18 to 40 year olds in the Chicago area, Johnson, Fendrich, and MackesyAmiti (2010) found that computer literacy is related to accuracy of self-reporting of cocaine use
on an ACASI survey. Accuracy of self-report was assessed using urine and saliva testing. The
study found a positive relationship between computer literacy and the accuracy of cocaine use
reports. Another possibility is that older respondents and those with less than a high school
education experienced greater overall burden by receiving more questions. Respondents who
report higher use of substances will receive more questions. No plans are in place to attempt to
address this issue in the 2013 DR protocol, but this issue could be investigated further with the
2013 DR data in combination to the 2012 QFT data.

197

5.4

QFT Equipment Surveys

5.4.1

Purpose and Development of the Equipment Surveys

As part of NSDUH's equipment evaluation for the 2015 NSDUH redesign, a new
device—the Samsung Galaxy Tab 7.0"— was selected for conducting household screenings for
further field-based evaluation in the 2012 QFT. This tablet was chosen for its small size, light
weight, and bright, easily readable screen display, which made it the most portable and easiest to
see and maneuver among a variety of devices, including Android tablets and Windows-based
convertible laptops that were assessed during previous evaluation phases.
A new Android-based screening program was developed for the tablets used for the QFT.
A total of 159 NSDUH FIs used this new program to collect data from 5,358 screened
households throughout the continental United States. The user interface on the new screening
program was designed to match as closely as possible NSDUH's existing screening program in
order to take advantage of the FIs' familiarity with the current program and to minimize the
amount of training and programming effort required.
To gather feedback from FIs about the tablet as a screening device, a brief electronic user
satisfaction questionnaire was administered before and after QFT data collection. The survey
questions included a combination of customized questions used in previous equipment
evaluations, as well as a number of questions adapted from the System Usability Scale,23 an
industry standard scale for measuring usability of hardware and software first developed and
published by engineers at the Digital Equipment Corporation (DEC) in 1986. In the first survey,
FIs were asked about their experience using touch screen devices, such as smart phones or tablets
and not including the NSDUH iPAQ. Several additional questions were included to evaluate FI
satisfaction with the QFT training program and materials. For the second survey, wording
changes were made to several questions about the QFT training session and handbook to reflect
the change in time periods between the first and second surveys. These wording changes were
also facilitated to gauge FI opinion on specific topics of interest, such as the amount of training
provided on the tablet, transmission, and troubleshooting. No revisions were made to questions
about the tablet between surveys. The complete sets of questions asked on the first and second
QFT equipment surveys are provided along with FI responses to each question in Appendix F.
5.4.2

Procedures for Conducting the Equipment Surveys

The first survey was administered at the conclusion of the QFT training sessions on
August 26 and 29, 2012. All results were completed and transmitted to RTI by September 6,
2012. The second survey was released toward the end of QFT data collection on October 8,
2012, and was completed by October 15, 2012. FIs received both surveys on their QFT laptops
via the NSDUH transmission process and were given 1 week to complete the survey and transmit
results to RTI. An introduction screen explained the purpose of the survey and the confidentiality
of individual responses. Results were sent back to RTI via the NSDUH transmission system.
All 160 QFT FIs who attended the QFT training session completed the first survey at the end of
training. The second survey was completed by 153 FIs who worked QFT cases in the field.
Seven FIs did not complete the second survey for the following reasons:
23

See http://hell.meiert.org/core/pdf/sus.pdf.

198

5.4.3

•

One FI did not successfully complete the QFT training and therefore did not work on
the QFT.

•

Five FIs did not complete the second survey because they had dropped out of the
QFT after training or did not work any QFT cases.

•

One FI was on medical leave at the time the second survey was administered and was
therefore unable to complete the survey.

Summary and Discussion of Results from the Equipment Surveys

A summary of FI feedback on the tablet used in the QFT is provided below. The
percentages included in this summary are from the second QFT survey administered near the end
of QFT data collection and indicate FI opinions on the tablet after having used it in a realistic
field setting. Table 5.39 provides the combined counts of FIs who strongly agreed or agreed to
each of the statements in the questionnaire, while Table 5.40 shows how often FIs used the QFT
handbook.
•

Overall, 27 percent of QFT FIs had never previously used a touch screen device, such
as a smart phone or tablet (excluding the NSDUH iPAQ), while 37 percent had used
one "a lot." See Exhibit 5.1 for the distribution of touch screen device experience
among QFT FIs.

•

Overall, FIs were highly satisfied with the tablet as a screening device. The vast
majority indicated they would like to use the tablet on a regular basis for fieldwork
(76 percent), found it intuitive (84 percent) and easy to use (88 percent), and learned
to use it quickly (93 percent).

•

The majority of FIs liked the layout of the screening program (80 percent), reported
they could efficiently complete screenings using the tablet (95 percent), and felt
confident using the tablet (93 percent).

•

FI responses were mixed with regard to navigation features on the tablet. A minority
of FIs preferred to navigate through the screening program using swipe gestures
(22 percent) rather than "Next" and "Previous" buttons (42 percent), while 36 percent
remained neutral.

•

With regard to data input methods, the majority of FIs preferred to use a stylus
(55 percent) rather than their fingers (24 percent) to tap on the screen, while
20 percent reported being neutral. With regard to keyboard input, a majority of FIs
(80 percent) reported they were able to easily type record of call (ROC) notes or
comments using the tablet keyboard.

•

The majority of FIs were satisfied with the design of the carrying case provided for
the tablet (72 percent). Several FIs commented they would like to have a pen holder
added to the carrying case, which would be helpful for writing on appointment cards.

•

FIs were highly satisfied with the QFT training program. The vast majority enjoyed
attending the training program (93 percent) and reported that the training prepared
them to properly complete QFT tasks (98 percent).

199

Table 5.39 Field Interviewer Opinions on Use of the Tablet before Questionnaire Field Test (QFT)
Data Collection and after QFT Data Collection

Comment on the Tablet
I (would) like using the tablet on a regular basis
for my field work.
The tablet is easy to use.
I can use the tablet without needing technical
assistance.
I like the layout of the screening program.
I learned to use the tablet quickly.
I am able to efficiently complete screenings
using the tablet.
I find the tablet intuitive, in that it's clear what I
need to do.
I feel confident using the tablet.
I think veteran interviewers will be able to use
the tablet without much training.
I think the tablet will work well in a variety of
weather conditions such as sunshine, rain, and
snow.
I can easily type ROC notes or comments using
the keyboard on the tablet.
I prefer to move through the screening program
using swipe gestures rather than the Next or
Previous buttons.
I prefer to tap the screen with my finger rather
than use a stylus.
The weight of the tablet is suitable for screening
at the door.
I am satisfied with the design of the carrying
case provided for the tablet.

QFT Equipment Survey 1
(August 2012)
Agree or Strongly Agree
(n = 160)1
%

QFT Equipment Survey 2
(October 2012)
Agree or Strongly Agree
(n = 153)2
%

135
142

84
89

117
134

76
88

125
139
140

78
87
88

134
122
143

88
80
93

146

145

132
142

83
89

129
142

84
93

122

129

137

123

125

114

127

110

FI = field interviewer; ROC = record of call.
1
Of the 160 QFT FIs who attended the QFT FI training sessions, 159 FIs successfully completed the training. One
FI demonstrated significant performance issues during the QFT training session and therefore did not successfully
complete the training.
2
Six FIs did not complete the second survey conducted after data collection because they did not successfully the
QFT training or had dropped out of the QFT after successfully completing training. One FI was on medical leave at
the time of the second survey administration and was unable to complete the survey.

Table 5.40 Field Interviewer (FI) Expectations on Referencing the Questionnaire Field Test (QFT)
Handbook before QFT Data Collection and FI Need to Reference the QFT Handbook
after QFT Data Collection
QFT FI Survey 1: How often do you think you will
reference the QFT FI Handbook?
QFT FI Survey 2: How often did you reference
the QFT FI Handbook?
Each day with QFT work
Two to three times a week
Rarely, when unusual situations arise
Never

QFT FI Survey 1
(August 2012)
(n = 160)
30
65
65
0

200

%
19
41
41
0

QFT FI Survey 2
(October 2012)
(n = 153)
5
18
99
31

%
3
12
65
20

Exhibit 5.1

Field Interviewer (FI) Experience with Touch Screen Devices before Questionnaire
Field Test (QFT) Training

As noted in Section 2.3.2 of Chapter 2, the QFT FIs were not selected randomly from the
set of all NSDUH FIs, but were selected based on their experience on the project, history of
reliable performance, and proximity to the QFT segments. Therefore, results from the equipment
surveys might not represent the full range of opinions among more recently hired FIs. Given the
popularity and increasing prevalence of tablet devices, it seems likely that the tablet would be
similarly well-received among NSDUH FIs who did not work on the QFT data collection. Unlike
more experienced FIs, those who were hired more recently have not been accustomed to using
the iPAQ device for several years on NSDUH.
5.4.4

FI Comments on the Tablet, Screening Program, and Tablet Accessories

The equipment surveys included one open-ended question that allowed FIs to comment
on any aspect of the tablet, screening program, or accessories, such as the carrying case. In the
first survey, 102 FIs made comments, while 91 FIs made comments on the second survey.
Comments were loosely grouped based on their content into the following areas: (a) general
comments about the tablet or screening program, (b) specific features and functions of the
screening, (c) accessories (stylus and carrying case), and (d) training. The comments were
diverse and individualized, and it was not possible to identify any recurrent or pervasive themes
shared by significant numbers of FIs. Issues raised by a small number of FIs for each category
are summarized in this section. The complete set of raw comments from each survey is included
in Appendix F.
FIs provided the following general comments on using the tablet devices:

201

•

Tablet Size and Maneuverability. While 78 percent of FIs agreed the weight of the
tablet was sufficient for screening at the door, some FIs commented that the tablet
was larger and more cumbersome than the iPAQ, which fits easily in the hand. This
made it more difficult to wear around the neck, protect in the rain, or see in bright
sun. On the other hand, a number of FIs emphasized they "loved" the larger display,
buttons, and font size, which made the tablet easier to read and navigate in the field.
Additionally, some FIs mentioned that the larger display size made it easier to show
the screen to respondents, who could easily see what they were doing, and that they
felt more "professional."

•

Touch Screen Sensitivity. In the second equipment survey, some FIs remarked that
the touch screen was highly sensitive, which made it too easy to tap inadvertently and
enter something they did not intend or move to a different screen. Others liked that
the tablet was more "responsive" and "efficient" than the iPAQ.

FIs provided the following comments on specific features or functions of the screening
program:
•

Select Case Screen. A few FIs stated that they wanted to highlight cases, and a
couple of others noted that they preferred the table format used on the iPAQ. For the
2013 DR, cases will remain highlighted for a period after being selected. One FI
noted there was "too much information" on each line, making it "hard to distinguish"
between cases, and another suggested bolding the address rather than the case ID.
Two FIs suggested that finalized cases should be removed from the select case screen.
FIs can remove final cases from the select case screen display by setting the view
function on the tablet to show only "pending cases."

•

Selection Screen and ROC Screen. Two FIs noted they would like to see the full
case ID displayed on the respondent selection and ROC screens as it is on the iPAQ.
For the 2013 DR, the screening program will display the full case ID on the
respondent selection and ROC screens.

•

Call Distribution. Two FIs noted it would be useful to have the call distribution
feature available on the tablet so that they could review the different days and times
they had visited households. Because of time constraints in the development of the
QFT screening program, the call distribution feature that is currently on the iPAQ was
not implemented. The same is true for the appointment calendar function. These
functions will be implemented in the 2013 2013 DR tablet screening program.

•

View Letters. A few FIs mentioned they would like the ability to view when their
field supervisor (FS) sends the unable-to-contact or refusal conversion letters as they
can on the iPAQ screening program. This function was implemented in the QFT
screening program. It only appears as an option once the letter has been sent by the
FS, so some FIs did not recognize that it had been implemented. The view letters
function will be implemented in the 2013 DR version of the screening program, and
the 2013 DR FI handbook and training sessions will clarify how to use it.

•

Transmission Feedback. Some FIs mentioned that they would like to receive
feedback regarding the number of cases added and removed on their tablet when they

202

transmit. This information will be integrated into the 2013 DR screening program and
will be displayed after each transmission.
•

Debriefing Questions. One FI remarked that he or she "loved" completing the
interview debriefing questions on the tablet rather than on the laptop. These questions
will continue to be included on the tablet during the 2013 DR.

FIs provided the following comments on two tablet accessories—the carrying case and
the stylus:

5.4.5

•

Carrying Case. Several FIs indicated that the carrying case could be improved by
adding a pen holder in addition to the stylus holder so that they could have easy
access to a pen for writing on appointment cards. Although a couple of FIs indicated
that the neck strap was too wide on the case and that the snap was hard to use,
a number of FIs commented that they were happy the Velcro® closure had been
removed. Because the carrying case was customized for the tablet used in the QFT,
which will also be used in the 2013 DR, no changes will be made to the carrying case
for the 2013 DR data collection. Adjustments to the design of the carrying case—such
as adding a pen holder and a thinner neck strap—will be considered as part of the new
equipment purchase for the 2015 main survey.

•

Stylus. Two FIs indicated that the stylus was too short and would prefer a longer
pen-sized stylus.

FI Feedback on the QFT Handbook

In addition to the questions about the satisfaction with the tablet, the survey also included
several questions about the QFT handbook that described QFT procedures and protocols and the
QFT training program. Table 5.40 (shown earlier) provides the FIs' responses to questions on
their anticipated use of the QFT handbook before data collection from the August 2012 survey
and their actual use of the QFT handbook during data collection from the October 2012 survey.

5.5

Focus Groups with QFT Field Interviewers

5.5.1

Purpose of the Focus Groups

The purpose of the three QFT focus group discussions was to obtain direct feedback from
FIs on their experiences collecting data using the redesigned NSDUH interview protocol and
tablet computer for screening. The complete set of protocol and equipment changes is presented
in Section 2.4.1. The goal of the focus groups was to gather feedback from FIs on the following
topics:
•

significant questions or concerns raised by members of sampled households about the
redesigned contact materials;

•

challenges encountered using the tablet computer to conduct household screenings;

•

challenges encountered in administering the redesigned questionnaire or protocol; and

203

•

significant questions or concerns that respondents raised about specific aspects of the
redesigned questionnaire or protocol, specifically the prescription drug modules and
the overall length and burden of the interview.

The results of the three focus groups were used to inform potential changes to the preparations,
protocol, and procedures for the 2013 DR.
5.5.2

Sites and Participants

Focus groups were conducted in three regional locations—Washington, DC; Chicago,
Illinois; and Irvine, California. RTI identified up to 15 QFT FIs who would be most able to
attend the group discussion for each of the three locations, based on proximity to each focus
group location. Up to 12 of the QFT FIs identified for each site were invited to attend the group
discussion (see Table 5.41).
Table 5.41 Sites and Number of Participants for QFT Focus Groups
Site
Washington, DC
Chicago, IL
Irvine, CA

Number of Participants
11
8
12

NOTE: Each focus group discussion was video recorded, and a note-taker was present to capture key points from
the group.

5.5.3

Focus Group Protocol and Procedures

Moderators began each focus group with an introduction that lasted about 5 minutes and
was intended to set up the discussion rules and familiarize the participants in each group.
Discussion about the redesigned contact materials was allotted 15 minutes and covered how
respondents reacted to the lead letter and Q&A brochure. The next 15 to 20 minutes of each
session were devoted to discussion about using the tablet to administer household screenings.
Topics included features of the tablet, training on the tablet computer, respondent reactions to the
naming of the "U.S. Department of Health and Human Services (DHHS)" as the study sponsor
(vs. the "U.S. Public Health Service"), and other materials, such as the new tablet carrying case
and portfolio. Over 30 minutes were devoted to topics surrounding questionnaire administration
using the redesigned methods and protocol. These topics included respondent comments about
the electronic reference date calendar, whether respondents asked questions about specific
modules within the instrument, and the experience of proxy respondents. The penultimate section
called for a discussion about the prescription drug modules specifically. The moderator asked
questions about the length of administration time, electronic pill cards, and the questions
designed to capture misuse. The last section asked FIs to share general comments or concerns
about the partially redesigned questionnaire, including interview length and burden. The
concluding section was intended to give both participants and observers a final opportunity to
ask questions or make comments. The moderator's guide for the QFT focus group is included in
Appendix G.

204

5.5.4

Focus Group Results by Topic
5.5.4.1

Reactions to the Redesigned Contact Materials

FIs nearly all responded positively to the changes to the lead letter and the Q&A
brochure. When discussing the lead letter, some mentioned that they appreciated that the letter
was addressed to "[NAME County/Parish/District] Resident at:" and did not just say "Resident."
Others mentioned that they liked the color picture on the letter and that overall the letter looked
more professional. A few FIs felt that the letter gave too much information, such as details about
the study topics, to respondents before the FI had an opportunity to speak to them, while most
FIs felt that the additional information increased the odds that a respondent would choose to
participate. One FI felt that not featuring a date on the letter made it feel generic.
Respondents who indicated they had read the letter responded positively. FIs agreed that
the proportion of respondents recalling the letter was about the same as in the main study.
Respondents did not go so far as to comment on any other aspects of the letter, with one
exception. FIs reported that respondents mentioned the incentive that was explained in the letter.
One FI said that, similar to the main study, respondents had an expectation of receiving an
incentive for completing the screening. FIs felt that the sooner they visited an address after
sending the letter, the more likely the respondent was able to recall the letter.
FIs were also asked about reactions to the Q&A brochure. FIs reported that respondents
did not make comments or have questions about the brochure more often than main study
respondents. One FI thought that respondents, while not commenting, spent more time with the
brochure and reviewed it more thoroughly. All FIs agreed that the brochure looked more
professional, expensive, and official, which lent more legitimacy to the study and possibly
contributed to higher levels of cooperation. During the main study, respondents commented that
the FI could have printed the brochure at home.
One FI reported the wording inside the brochure is more convincing, and she used this
verbiage to convert potential refusals. Other FIs had a positive reaction to the way the project
Web site is listed. FIs thought that more respondents reported visiting the Web site than recent
respondents in the main survey. No respondent questions about the brochure were reported.
When discussing the study sponsor change from the "U.S. Public Health Service" to the
"U.S. Department of Health and Human Services," FIs had a number of reactions. Many thought
this change did not have an impact, while others reported some respondents thought that
"DHHS" was social services. When announcing the visit, respondents would say, "Social
services is here." Or they would refer to it as "child protective services." Some FIs mentioned
that the DHHS title was more official. One FI noted that, in a graphic in the redesigned Q&A
brochure, a respondent is pictured using a paper reference date calendar. Based on this
observation, this picture was removed from the Q&A brochure and replaced with another picture
that does not show the paper reference date calendar. This revised brochure will be used in the
2013 DR.
5.5.4.2

Reactions to Using the Tablet to Administer Household Screenings

FIs confirmed that the QFT training program adequately discussed the goals of the field
test. They agreed that the training agenda provided enough time and instruction to ensure
205

competent use of the tablet in the field. FIs pointed out the pros and cons of the new portfolio
that was provided at training. Some said they disliked the portfolio enough to revert to using the
old one, which they viewed as sturdy and professional. The new portfolio was characterized by
some FI as being slippery and difficult to hold. These FIs also noted that the tablet, when placed
on the portfolio, fell off and the materials fell out of it. FIs also indicated that the closure is
flimsy. These FIs would have preferred a zip closure similar to the main study portfolio. Further
comments indicated that the portfolio was difficult to write on, such as when filling out the
quality control letters. FIs did, however, like the number of slots in the portfolio and the clear
pockets for easier access to materials. The features and costs of other portfolios with multiple
pockets that are sturdier will be investigated for use in the 2013 DR.
FIs also provided feedback on the tablet computer. They reported that the training on how
to use the tablet was effective and that from the beginning of their fieldwork they felt
comfortable using the tablets. Some FIs would have preferred more training on administrative
and troubleshooting issues before entering the field. These FIs reported getting into programs or
onto screens early in their fieldwork that they had not seen in training and did not know how to
return to the screening program. Although they felt comfortable conducting the screening with
the tablet, they would have preferred more hands-on training on how to deal with these
unexpected FI navigational problems. The training agenda developed for the 2013 DR will
address this issue. However, overall, they thought the tablet was easier to use than the iPAQ. It is
faster, easier to tap out the letters, and readable without the use of glasses. Many liked the size
and weight of the tablet. At first it felt big, but the size turned into an advantage once FIs became
accustomed to it. They appreciated the clear visibility and larger text. They also liked that more
information fit on the screen. In the iPAQ, only the first part of the address with the case ID is
shown on the screen. On the tablet, FIs can see all of the information, including case status.
An unexpected benefit of the size of the equipment was also noted. Because of the size of
the tablet, screening respondents were more engaged in the screening. They looked at the screen
and did not remove their gaze throughout the screening. It is easier to show respondents the
screen, and respondents reacted well when looking at it. FIs shared tips to respond to the
challenge of keeping the select case information from their view. Only one FI reported disliking
the tablet and would have preferred a smaller device, such as a smart phone for screenings.
FIs also liked the case that was designed for the tablet. It was easy to flip the cover open
to charge. Many FIs reported disliking the strap for the tablet, felt that it was too bulky and thick,
and indicated that it interfered with badges and necklaces. Some reported they would like a pen
holder on the side of the case opposite the stylus. Several FIs preferred the magnetic snap closure
to the Velcro® closure on the current iPAQ case. As noted in Section 5.4.4, the carrying case was
customized for the tablet used in the QFT, and the same tablet will be used in the 2013 DR.
For this reason, no changes will be made to the carrying case for the 2013 DR data collection.
Adjustments to the design of the carrying case—such as adding a pen holder and a thinner neck
strap—will be considered as part of the new equipment purchase for the 2015 main survey.
FIs noted a few issues with the screening program that were problematic. FIs would like
to be able to edit a status code. They reported that they could delete a code and add a new one,
but did not have the capability to change an existing code. All FIs agreed that they did not like
this feature. This capability would be helpful, for example, to change a screening result code 10
(vacant) to a 13 (not a primary residence). The screening program will be modified for the 2013

206

DR to include the ability to edit existing ROC codes. Also, FIs stated that it was tricky to
navigate back to the verification screen for the vacant dwelling units. Navigating to the
verification screen for a vacant unit is achieved by selecting the case on the select case screen
and selecting "View Verification Information" from the pop-up actions menu. It seems likely that
some FIs did not clearly understand these steps. Therefore, the 2013 DR training program will
provide clearer instructions about how to view verification information for any case. Table 5.42
provides a list of modifications to the screening program/tablet functionality mentioned by FIs in
the focus groups. The screening program will be modified for the 2013 DR to address some of
these issues, such as enabling edits to the screening ROC code and adding the call distribution.
However, other items, such as revising the tablet keypad layout, changing the default tablet
calendar, or continuously highlighting selected cases, are not possible on the Android platform.
Appendix X provides a complete summary of potential changes to tablet functions that were
identified during the QFT and indicates which changes will be implemented for the 2013 DR.
Table 5.42 QFT FIs' "Wish List" for Modifications to Tablet Functions

•

Revisions to symbols available on the primary
keyboard

•

Improve calendar usability

•

Ability to continuously highlight the selected
case on the select case screen

NOTE: The item in boldface will be implemented for the 2013 DR data collection.

FIs also provided feedback on the keypad. FIs noted that they would like to have the
apostrophe and quotation marks available and would like unnecessary symbols removed from the
keypad. They also said that the question mark was hard to find and requested that the period
should be placed on the same keypad as the letters and should also be available if a user inserts
two spaces after a sentence. Given that the layout and design of the default keypad on the tablet
cannot be altered, other keypad options have been investigated for use in the 2013 DR. In
addition to training 2013 DR FIs on using the default tablet keypad, a second keypad (called the
"hacker's" keypad) will be loaded onto tablets as an alternative for the 2013 DR.
The debriefing items were not challenging to complete, and FIs reported preferring to
answer these questions on the tablet rather than on the laptop at the end of the interview. FIs who
work in rural segments had some difficulty finding a place to complete these questions after
leaving a respondent's home. Some FIs suggested adding a field to record comments about the
case. This open-ended field has been added to the 2013 DR debriefing questions.
FIs strongly wished they had access to the call distribution feature and felt that this was
the primary capability that was missing compared with the iPAQ. FIs were happy with the ability
to pull up the refusal letters that have been sent to households, but not all were aware of these
capabilities. More detail will be provided on this feature in the 2013 DR FI training and
handbook. The development schedule leading up to the QFT did not allow for the addition of the
call distribution feature. This will be added to the 2013 DR program.
The stylus received mixed reviews. Some liked it and used it. Others did not use the
stylus, saying it was hard to insert into the holder on the case, was slippery, and caused the
holder on the case to tear.

207

Moderators also asked about instances where FIs called NSDUH technical support staff
for help with equipment problems. One FI in the Washington, DC, focus group reported a glitch
where ROCs were not transferred along with cases. This was corrected during the QFT. Several
FIs mentioned that a car charger would be appreciated because the battery did not last all day.
Because car chargers for the iPAQ are provided for the main study, the addition of a car charger
is being considered for the 2013 DR.
FIs were asked whether they would benefit from a more expansive suite of features with
the tablet, such as predictive typing or alternating between landscape and portrait orientation.
A couple of FIs wanted predictive typing. Others were not as enthusiastic. Several FIs would like
a larger calendar on the tablet to record future appointments. They would also like several of the
iPAQ features to be transferred to the tablet (e.g., the case ID remains at the top of the screen
during a screening, and a selected line remains highlighted on the select case screen). One FI said
that it was better to not make the devices sophisticated. Larger calendars will be implemented for
the 2013 DR. Although it is not possible to have a selected case remain highlighted, the
highlighting will remain for a longer time.
5.5.4.3

Administering the Redesigned Questionnaire and Protocol

A discussion about respondent feedback on the electronic reference date calendar opened
this section. Although FIs reported that respondents did not have any comments or questions on
the reference date calendar, the FIs themselves reported liking it. They said that no one looks at
the paper calendar, so it is an improvement that it is now on screen. An FI did suggest a darker
color to highlight dates because the current colors are difficult to see in sunlight.
Respondents did have questions while completing the computer tutorial. Some did not
understand how to enter the answer after the F2 box closes. Others asked what potatoes have to
do with the study, and if this was related to targeting McDonalds regarding nutrition issues.
Some FIs suggested that the tutorial be clearly labeled as a practice session or that the
introduction be emphasized. They reported that respondents struggled with providing accurate
answers to questions and were confused by the lack of concordance with the question topics and
the NSDUH study description. In response, each question in the ACASI tutorial has been labeled
as a practice question in the 2013 DR questionnaire.
In general, respondents asked about the same number of questions and had a similar
volume of comments compared with the main study. Some FIs expected fewer comments from
respondents, while others expected more.
FIs reported that the interview felt longer because it was not broken up by the computerassisted personal interviewing (CAPI) as in the main study. They recognized this could be their
perception, as opposed to the respondent's.
There were no comments about the proxy introduction. Some proxy respondents reported
the sound had been turned off, or the FI was able to pick up on cues that the volume was not
playing. In these cases, FIs turned the volume on using F7. Others used Fn+Page Up or the
sound dial on the headphones.

208

FIs reported on issues that respondents had with questions, not all of which were
redesigned or new questions. A bulleted list of comments or issues follows:
•

One respondent said, "I didn't drink in the past year; why is it asking me about the
past 30 days?"

•

Seniors did not know what "enrolled in school" means.

•

Some had questions about what the word "kicks" means.

•

One child asked questions about what "h-e-r-o-i-n" means.

•

Minor respondents (i.e., adolescents) often asked about the meaning of "seldom."

•

In response to the question "How many times have you moved?" a 12-year-old
respondent said, "I move all of the time."

•

Many respondents had questions and problems with the self-help group questions.

•

Three respondents wanted to know about the 95 response option in the prescription
drug modules.

5.5.4.4

Reactions to the Redesigned Prescription Drug Modules

In general, very few comments about the prescription drug modules were reported.
FIs mentioned that some respondents said there were missing pills or asked about pills not
referenced in the interview. Others thought there were more comments about the length of the
interview as compared with the main study, but it was not apparent that these comments related
specifically to the prescription drug questions. Only one FI expected the respondents to react to
the length of time for the prescription drug modules. Others did not share this same expectation.
When asked whether the respondents had comments about the electronic pill images in
the questionnaire, FIs responded they did not. FIs believed the electronic images felt more
private. Others said they received more comments on the main study showcards as compared
with the electronic images.
One focus group participant noted that a 13-year-old respondent asked him if Tylenol®
was a prescription drug. Another asked a lot of questions about what class of drugs particular
pills were. One respondent was angry about the detailed information asked in this section
because he was suspicious that it would be used to help the pharmaceutical industry.
5.5.4.5

Overall Reactions to the Redesigned Questionnaire

FIs who participated in the focus groups had some additional feedback on the QFT
procedures. They recommended adding more language prior to the FI-administered household
roster to inform the respondents that they or another household member would be given the
computer to complete another part of the interview. This may help respondents manage their
expectations about the remainder of the interview.
No respondents reacted strongly to switching back and forth between ACASI and CAPI,
and FIs acknowledged this could be their perception. For households with only one resident, FIs
felt that switching the laptop back was awkward and would like the second ACASI portion to be

209

combined with the first in these cases. Also, FIs do a good job of warning proxies or parents that
they may be called upon later to assist. This discussion led to an additional topic. FIs were
concerned about the availability of the parent who may best serve as a proxy. They mentioned
challenges associated with making sure that the parent does not leave the household or become
unavailable before the child reaches the back end of the instrument. They recommended moving
the proxy section to the beginning of the interview. The proxy section will remain in the same
part of the interview for the 2013 DR as the QFT, but FIs will be reminded in FI training to
confirm that the parent will remain in the house or be available for the entirety of the interview.
In general, FIs had mixed experiences with interview timing. Some thought it seemed
longer than the main study, while others reported it was shorter. One FI noted there appeared to
be timing differences between younger and older youths. Younger youth respondents took the
interview quite seriously and seemed to take longer to complete it, while older teenagers seemed
to move through the interview quickly.
When asked about their expectations about the interview, several FIs mentioned
expecting more comments and questions about the interview than what were received. One FI
expected the interview to be longer than it was.
The moderator asked FIs how they would feel about having an additional tool available to
help with doorstep screenings. This tool would consist of a 20- to 30-second video clip of the
NSDUH press conference, would be available on the tablet, and could help with gaining
cooperation. FIs were enthusiastic about this idea, if the video was optional and not a required
part of the screening. One FI suggested having multiple videos designed to address common
respondent concerns, such as confidentiality, or targeted to specific populations, such as parents
or elderly persons. They said respondents would think that if it is on television, it is true. It
would also help with legitimacy and would be short enough to use at the doorstep. Addition of
this video will be revisited during planning for the 2015 survey.
5.5.4.6

Other General Feedback

FIs had some other general comments about features of the new protocols. FIs liked that
the income questions are now in the ACASI portion of the interview because they thought this
mitigated social desirability concerns and ensured better data quality. FIs would prefer not to be
privy to this information and reported that some parents clearly do not want their child to know.
Overall, FIs shared fewer ideas for improving the questionnaire as opposed to improving the
functionality of the tablet. They indicated that they would like to do away with the showcards
and rearrange the demographic questions to be self-administered. Despite this feedback, these
changes will not be made for the 2013 DR.

210

6. QFT Estimates Compared with NSDUH
Estimates: Substance Use Items Other than
Methamphetamine and Prescription Drugs
6.1

Overview of QFT Estimates Compared with NSDUH Estimates for
Substance Use Items Other than Methamphetamine and Prescription
Drugs

This chapter presents findings for core substance use estimates from the 2011 National
Survey on Drug Use and Health (NSDUH) comparison data, the 2012 NSDUH quarters 3 and 4
comparison data, and the 2012 Questionnaire Field Test (QFT) data for substances other than
methamphetamine and prescription drugs. The tables in Appendix I provide lifetime, past year,
and past month estimates for use of these substances for all persons aged 12 or older and for
three separate age groups of interest. Section 6.2 provides estimates for marijuana, cocaine, and
heroin for all three datasets. Section 6.3 presents results for hallucinogens and inhalants.
Estimates for multiple definitions of use of "any illicit drug" are discussed in Section 6.4.
Section 6.5 presents results for tobacco use, focusing on cigarette use and smokeless tobacco
use. Finally, Section 6.6 provides findings on alcohol use, including binge alcohol use, as
defined in that section.

6.2

Marijuana, Cocaine, and Heroin

This section presents findings on marijuana, cocaine, and heroin use from the 2011
comparison data and 2012 quarters 3 and 4, as well as the QFT data. Tables I-1 through I-4 in
Appendix I provide estimates for lifetime use of these substances for all persons aged 12 or
older, adolescents aged 12 to 17, young adults aged 18 to 25, and adults aged 26 or older.
Likewise, Tables I-5 through I-8 provide estimates for past year use of these substances, and
Tables I-9 through I-12 provide estimates for past month use of these substances. No changes
were made in the QFT instrument for the questions on marijuana, cocaine (including crack), and
heroin use. However, these estimates are examined in this report because changes were made to
other elements of the survey design, including changes to the contact materials and interview
protocol, that have some potential to affect these estimates in ways that are difficult to predict
and cannot easily be addressed by other analyses.
•

There were no statistically significant differences in estimates of marijuana use across
all three reporting periods (lifetime, past year, and past month) and over all age
groups between the QFT data and both the 2011 and 2012 comparison data.

•

For cocaine, there were statistically significant differences for adolescents aged 12 to
17 for lifetime use between the QFT and both sets of comparison data. Statistically
significant differences also were shown in Table I-6 for past year cocaine use and
Table I-10 for past month cocaine use among adolescents. However, both of the QFT
estimates of 0.0 percent would be suppressed. Therefore, these QFT estimates would

211

not be shown in published estimates, nor would any statistically significant
differences be presented.
•

Among young adults aged 18 to 25, the rate of past month cocaine use in the 2011
comparison data was higher than the rate in the QFT (1.3 vs. 0.4 percent)
(Table I-11).

•

For crack, there were statistically significant differences for adolescents aged 12 to 17
for past year use between the QFT and 2011 comparison data (0.0 vs. 0.1 percent, but
with the QFT estimate suppressed) (Table I-6). The difference between the QFT
estimate and the estimate for the 2012 comparison data approached statistical
significance (0.0 vs. 0.1 percent; p = 0.055).

•

Also for crack, there were statistically significant differences for persons aged 12 or
older for past month use between the QFT (0.0 percent) and both the 2011
(0.1 percent) and 2012 (0.1 percent) comparison data (Table I-9), as well as for
persons aged 26 or older for past month use between the QFT (0.0 percent) and both
the 2011 (0.1 percent) and 2012 (0.1 percent) comparison data (QFT estimate
suppressed for adults aged 26 or older) (Table I-12).

•

For heroin, there were statistically significant differences for both past year and past
month use for persons aged 26 or older, although the QFT estimates would be
suppressed (Tables I-8 and I-12).

As noted in this section, some differences between the estimates for cocaine and heroin
use were statistically significant between the QFT and comparison data despite the content of
these modules not changing for the QFT. However, many of the relevant QFT estimates would
be suppressed, such that these apparent differences would not be published in a summary of
findings from the QFT. Nevertheless, further examination of estimates of cocaine and heroin use
in the 2013 Dress Rehearsal (DR) will be important for assessing the likelihood that the trend
data for these drugs will not be disrupted in 2015.

6.3

Hallucinogens and Inhalants

As noted in Section 2.4.1, questions currently in the special drugs module for the
hallucinogens ketamine, tryptamines (dimethyltryptamine [DMT], alpha-methyltryptamine
[AMT], and N, N-diisopropyl-5-methoxytryptamine [5-MeO-DIPT], also known as "Foxy"), and
Salvia divinorum were moved to the core hallucinogens module for the QFT. These included
questions about lifetime and most recent use of these additional hallucinogens. For inhalants,
questions about lifetime use of markers and computer keyboard cleaner (also known as "air
duster") were added to the QFT questionnaire. Questions did not differ between the main study
and the QFT for respondents who reported lifetime use of one or more inhalants (e.g., first use,
most recent use).
6.3.1

Hallucinogens
•

Estimates of lifetime use of any hallucinogen, lysergic acid diethylamide (LSD),
phencyclidine (PCP), and Ecstasy did not differ between the QFT and the 2011 or
2012 comparison data for persons aged 12 or older (Table I-1). For example, the

212

estimates of lifetime use of any hallucinogen among persons aged 12 or older were
16.2 percent for the QFT, 14.8 percent for the 2011 comparison data, and
15.0 percent for the 2012 comparison data.
•

Among adolescents aged 12 to 17, the estimate of lifetime use of hallucinogens was
greater in the QFT (6.5 percent) than in the 2011 and 2012 comparison data (3.7 and
3.2 percent, respectively) (Table I-2). However, lifetime estimates of use of LSD,
PCP, or Ecstasy among adolescents were not significantly different between the QFT
and the comparison data.

•

Lifetime estimates of hallucinogen use—including LSD, PCP, and Ecstasy—did not
differ for adults aged 18 to 25 (Table I-3) or those aged 26 and older (Table I-4)
between the QFT and the comparison data.

•

Most estimates of use of hallucinogens, LSD, PCP, or Ecstasy in the past year or past
month did not differ between the QFT and comparison data for persons aged 12 or
older or within the age groups. For example, the estimates of past year use of any
hallucinogen among persons aged 12 or older were 2.1 percent for the QFT and
1.6 percent in both the 2011 and 2012 comparison data (Table I-5).

•

Among adolescents, the QFT estimate of past year LSD use (0.2 percent) was lower
than the estimates of 0.6 percent for both the 2011 and 2012 comparison data
(Table I-6). Also, the estimate of past month use of Ecstasy among persons aged 12
or older was lower in the QFT than in the 2011 comparison data (0.1 vs. 0.2 percent),
but the estimate for the 2012 comparison data (also 0.2 percent) was not significantly
different from the QFT estimate (Table I-9). In addition, the estimate of past month
use of Ecstasy among adults aged 26 or older was lower in the QFT (0.0 percent) than
in the comparison data (0.1 percent in each year), but the QFT estimate would be
suppressed (Table I-12).

The estimates for hallucinogen use in the comparison data that were described previously
were based only on reports of use from the core module. These estimates did not include data on
the use of ketamine, tryptamines, and Salvia divinorum that were in the supplemental (i.e.,
noncore) special drugs module. Therefore, core-plus-noncore (CPN) measures of hallucinogen
use that included data from these three additional hallucinogens also were created for the 2011
and 2012 comparison data. These CPN estimates were compared with the QFT estimates based
on core data and are included in Tables I-18 to I-20 in Appendix I.
•

Inclusion of noncore hallucinogens data did not affect most patterns of differences
between the QFT and comparison data for lifetime, past year, or past month estimates
of any hallucinogen use among persons aged 12 or older and within the age groups.
For example, the estimate of lifetime hallucinogen use among persons aged 12 or
older was 16.2 percent for the QFT. Corresponding CPN estimates were 15.4 percent
for the 2011 comparison data and 15.5 percent for the 2012 comparison data. The
QFT and CPN estimates of past year hallucinogen use were 2.1 percent for the QFT,
1.9 percent for the 2011 comparison data, and 1.8 percent for the 2012 comparison
data.

•

Among adolescents aged 12 to 17, the CPN estimate of lifetime use in the 2011
comparison data (4.5 percent) was no longer significantly different from the QFT
213

core estimate of 6.5 percent. However, the CPN estimate of lifetime use in the 2012
comparison data (3.6 percent) continued to be lower than the QFT estimate.
In addition, respondents in the main survey and the QFT were asked about lifetime use of
"any other" hallucinogen besides the ones they had seen in the preceding questions. Respondents
who reported use of other hallucinogens could specify use of up to five other hallucinogens that
they had ever used (subsequently referred to in this section as "OTHER, Specify" data). The
questions about ketamine, tryptamines, and Salvia divinorum had been included in the main
survey since 2006 because of evidence from their "OTHER, Specify" data that these could be
additional important substances for understanding hallucinogen use, especially among
adolescents and young adults aged 18 to 25 (Kroutil, Vorburger, & Aldworth, 2007).
Consequently, moving the questions about these hallucinogens from the special drugs module in
the main survey to the core hallucinogens module in the QFT could reduce the reporting of use
of "other" hallucinogens. Also, moving the questions for these three hallucinogens from the
special drugs module to the core hallucinogens module could affect lifetime reporting because of
their earlier placement in the QFT.
Therefore, estimates of lifetime use of ketamine, tryptamines, Salvia divinorum, and
other hallucinogens were compared for the QFT and the data from 2011 and quarters 3 and 4 of
2012. Estimates are shown in Table I-13 in Appendix I.
•

Estimates of lifetime use of ketamine, tryptamines, and Salvia divinorum were not
significantly different between the QFT and the comparison data for persons aged
12 or older or within the age groups.

•

Estimates of lifetime use of other hallucinogens were lower in the QFT than in the
2011 or 2012 comparison data for persons aged 12 or older, young adults aged 18 to
25, and adults aged 26 or older. For persons aged 12 or older, the estimate of lifetime
use of other hallucinogens was 0.6 percent for the QFT and 1.6 percent for both the
2011 and 2012 comparison data. Among young adults, the estimate of other
hallucinogen use decreased from 3.8 percent in the 2011 comparison data and
3.4 percent in the 2012 comparison data to 1.7 percent in the QFT.

At least for adults, moving the additional hallucinogen questions from the special drugs
module to the core hallucinogens module in the QFT appears to have affected the reporting for
the residual "other" hallucinogen category. Benefits of this change are that analysts have more
information about the specific hallucinogens that persons have used, whereas the category for
other hallucinogens can be a "catchall" for a wide variety of possible substances. Furthermore,
this change could reduce the amount of data review and coding of "OTHER, Specify" data that is
needed for hallucinogens when the redesigned questionnaire is fielded in 2015. An additional
noteworthy finding from these analyses is that moving the questions for these three
hallucinogens from the special drugs module to the core hallucinogens module did not appear to
affect lifetime reporting because of their earlier placement in the QFT. However, the effect of
this change in the placement of these questions could warrant further investigation in the
2013 DR and in preliminary data from the 2015 survey (e.g., from the first two quarters).

214

6.3.2

Inhalants

Questions about lifetime use of felt-tip pens and computer keyboard cleaner (air duster)
were added to the inhalants module for the QFT because review of "OTHER, Specify" data
suggested that these could be other important inhalants that persons used to get high.
Furthermore, prior research has shown that NSDUH respondents are more likely to report use of
a substance if they are asked a direct "yes/no" question about the substance than if they need to
type in its name as part of "OTHER, Specify" questions (Kroutil, Vorburger, Aldworth, &
Colliver, 2010). Therefore, even though the only change to the inhalants module for the QFT was
the addition of the questions about lifetime use of these two inhalants, increased reporting of
lifetime use could translate to increased reporting of use in more recent periods.
•

Estimates of lifetime use of inhalants were greater in the QFT than in the 2011 and
2012 comparison data for persons aged 12 or older, adolescents aged 12 to 17, and
adults aged 26 or older (Tables I-1, I-2, and I-4). For example, 11.1 percent of
persons aged 12 or older in the QFT were lifetime users of inhalants compared with
8.2 percent for the 2011 comparison data and 8.3 percent for the 2012 comparison
data (Table I-1).

•

For adolescents aged 12 to 17, the QFT estimate of lifetime use of inhalants was
11.7 percent (Table I-2). In comparison, 7.5 of adolescents in the 2011 comparison
data and 5.7 percent of those in the 2012 comparison data were estimated to be
lifetime users. For young adults aged 18 to 25, the estimate of lifetime inhalant use in
the QFT also was greater than the estimate in the 2012 comparison data (11.7 vs.
7.9 percent) (Table I-3).

•

Estimates of past year and past month use of inhalants did not differ significantly
between the QFT and comparison data for persons aged 12 or older, adults aged 18 to
25, and those aged 26 or older (Tables I-5, I-7, and I-8, respectively, for the past year
and Tables I-9, I-11, and I-12 for the past month). For example, the estimates of use
of inhalants in the past year among persons aged 12 or older were 0.9 percent for the
QFT, 0.7 percent for the 2011 comparison data, and 0.6 percent for the 2012
comparison data (Table I-5).

•

For adolescents aged 12 to 17, the QFT estimate of past year use of inhalants was
greater than the estimate for the 2012 comparison data (4.1 vs. 2.1 percent)
(Table I-6). However, the estimate for the 2011 comparison data (3.0 percent) was
not significantly different from the QFT estimate. Estimates of use of inhalants in the
past month among adolescents did not differ between the QFT and comparison data
(Table I-10).

As for the hallucinogen data described previously, adding the questions to the QFT about
lifetime use of felt-tip pens or computer keyboard cleaner could affect reporting of the lifetime
use of "other" inhalants. Also, computer keyboard cleaner is an aerosol product. Therefore,
asking about lifetime use of computer keyboard cleaner could affect estimates for lifetime use of
other aerosol sprays (i.e., other than spray paint in the main study and other than spray paint or
computer keyboard cleaner in the QFT).

215

Estimates of lifetime use of felt-tip pens and computer keyboard cleaner were made for
the QFT. Estimates of lifetime use of other aerosol sprays and other inhalants also were
compared for the QFT and the data from 2011 and quarters 3 and 4 of 2012. These estimates are
shown in Table I-14 in Appendix I.
•

The prevalence of lifetime use of felt-tip pens based on the QFT data was 3.3 percent
for persons aged 12 or older, 9.4 percent for adolescents aged 12 to 17, 5.8 percent for
young adults aged 18 to 25, and 2.0 percent for adults aged 26 or older.

•

Relative to the estimate of 11.7 percent for lifetime use of any inhalant among
adolescents (Table I-2), the 9.4 percent who ever inhaled felt-tip pens appeared to
comprise a substantial portion of the adolescent lifetime inhalant users. The
5.8 percent of young adults who ever inhaled felt-tip pens (Table I-14) appeared to
comprise about half of the 11.7 percent of lifetime users of inhalants in this age group
(Table I-3).

•

The prevalence of lifetime use of computer keyboard cleaner based on the QFT data
was 1.2 percent for persons aged 12 or older, 1.1 percent for adolescents, 2.4 percent
for young adults, and 1.0 percent for adults aged 26 or older.

•

Among young adults aged 18 to 25, the QFT estimate for lifetime use of other aerosol
sprays (0.7 percent) was lower than the estimates in the 2011 and 2012 comparison
data (1.8 and 1.5 percent, respectively). The QFT estimate for other inhalants
(0.1 percent) also was lower than the comparison data estimates for 2011
(0.8 percent) and 2012 (0.7 percent) for this age group.

To further understand the estimates in Table I-14 and in anticipation of effects on
estimates of inhalant use in 2015, further analyses of the QFT data were conducted that
categorized users into two groups: (1) lifetime users of felt-tip pens or computer keyboard
cleaner (which could include persons who used other inhalants in addition to these two); and
(2) lifetime users of other inhalants, excluding use of felt-tip pens and computer keyboard
cleaner. Estimates for these two groups of lifetime users were made for persons aged 12 or older
and for each age group. Estimates of persons aged 12 or older who reported past year use also
were made for these two groups of lifetime users; corresponding past year estimates were not
made by age group because of small sample sizes.
Estimates for these further analyses are shown in Table I-17 in Appendix I. Statistical
testing was not conducted to identify any age group differences in the estimates presented in this
table or differences in the past year estimates. Also, the QFT questions did not allow
determination of the specific inhalants that were used in the past year.
•

An estimated 4.1 percent of persons aged 12 or older were lifetime users of felt-tip
pens or computer keyboard cleaner, and 7.0 percent were lifetime users of inhalants
but not these two.

•

Percentages of persons who were lifetime users of felt-tip pens or computer keyboard
cleaners were 10.0 percent for 12 to 17 year olds, 7.4 percent for 18 to 25 year olds,
and 2.8 percent for adults aged 26 or older. Percentages of persons who were lifetime
users of other inhalants (but not these two) were 1.8 percent for 12 to 17 year olds,
4.3 percent for 18 to 25 year olds, and 8.1 percent for adults aged 26 or older.
216

•

Among persons aged 12 or older who were lifetime users of felt-tip pens or computer
keyboard cleaners, 12.8 percent used some inhalant in the past year. For lifetime users
of other inhalants excluding these two, 5.0 percent used inhalants in the past year.

Although age group differences were not tested, lifetime use of felt-tip pens or computer
keyboard cleaner appears to be more common among adolescents and young adults than among
adults aged 26 or older. In addition, the findings for past year use of inhalants among lifetime
users of felt-tip pens or computer keyboard cleaner and among lifetime users of inhalants (but
not these two) may be affected by age-related differences in reporting of lifetime use of specific
inhalants and also age-related differences in the proportions of lifetime users who also used in
the past year. For example, QFT estimates in Tables I-2 and I-6 indicate that 11.7 percent of
12 to 17 year olds were lifetime users of inhalants, and 4.1 percent were past year users.
Corresponding QFT estimates in Tables I-4 and I-8 for persons aged 26 or older were
10.9 percent for lifetime use and 0.4 percent for past year use.
Taken together, these findings suggest that adding the questions about lifetime use of
felt-tip pens and computer keyboard cleaner may affect data trends in lifetime use of inhalants
once the new questionnaire is fielded for the 2015 survey, including trends for adults aged 26 or
older. These findings also suggest that this questionnaire change could affect trends for past year
use of inhalants among adolescents aged 12 to 17. However, estimates for past month use of
inhalants appeared unlikely to be affected by this change. Because NSDUH national reports tend
to focus on estimates of past month use (i.e., current use), inclusion of these two additional
inhalants in the 2015 survey might have a small impact on trends in the past month use of
inhalants. Because long-term trends in lifetime use and past year use of inhalants are typically
included in annual NSDUH detailed tables and reports of findings, it will be important for the
Substance Abuse and Mental Health Services Administration to consider how to handle any
disruption in the trends for lifetime and past year use of inhalants in the 2015 detailed tables.

6.4

Illicit Drug Summary Measures

This section presents comparisons of estimates between the QFT and comparison data for
2011 and 2012 for several summary measures of illicit drug use. The standard definition of any
illicit drug use captures use of any of one of nine categories of illicit drugs: marijuana, cocaine
(including crack), heroin, hallucinogens, inhalants, and misuse of any one of four classes of
psychotherapeutics (i.e., pain relievers, tranquilizers, stimulants, and sedatives). The standard
definition of any illicit drug use also includes use of methamphetamine reported in the noncore
questions added in 2005 and 2006 and the new methamphetamine module in the QFT. In
addition, because marijuana use has historically been the most prevalent form of illicit drug use,
a summary measure of illicit drug use other than marijuana is a standard NSDUH measure that
allows for the detection of trends in any illicit drug use that may be masked by trends in
marijuana use.
Because of extensive changes to questions asking about prescription drug misuse
(including the addition of a new methamphetamine module), the standard definitions of any
illicit drug use (and any illicit drug use other than marijuana) were modified for this analysis to
exclude the use of methamphetamine and the misuse of any prescription drugs. Alternate
Definition 1 of any illicit drug use covers any use of marijuana, cocaine (including crack),
217

heroin, hallucinogens, and inhalants. Comparisons between the QFT sample and the 2011 and
2012 samples for this measure are free of any measurable differences in the use of
methamphetamine and the misuse of psychotherapeutics. Alternate Definition 3 for any illicit
drug use includes use of marijuana, cocaine (including crack), heroin, hallucinogens, inhalants,
and methamphetamine. Similarly, the Alternate Definition of any illicit drug use other than
marijuana covers any use of cocaine (including crack), heroin, hallucinogens, and inhalants.24
In addition, as noted in Sections 6.3.1 and 6.3.2, the modules for hallucinogens and
inhalants were modified by explicitly asking respondents about hallucinogens that had
previously been asked about in the special drugs module and asking direct questions about
specific additional inhalants. Thus, Alternate Definition 2 of any illicit drug use is similar to
Alternate Definition 1 except that the use of hallucinogens and inhalants is ignored. Similarly,
ignoring any reported use of hallucinogens and inhalants leads to a measure of any illicit drug
use other than marijuana that only contains two categories of drugs: cocaine (including crack)
and heroin.
Table 6.1 summarizes these measures, which were all were constructed for the lifetime,
past year, and past month reporting periods. These estimates are shown in Tables I-1 to I-12 in
Appendix I and Tables J-1 to J-12 in Appendix J. Estimates from the tables in Appendix I are
discussed in this section and focus on the effects on summary estimates of illicit drug use that
could be attributed to changes to the hallucinogens and inhalants modules in the QFT (or other
differences), separate from any effects on these estimates that could be attributed to changes to
questions for methamphetamine and prescription drugs. Estimates from the tables in Appendix J
are discussed in Chapter 7 in the context of a discussion of the changes to the QFT questions for
these substances and the effects of these changes on estimates.
Table 6.1 Substances Included in Definitions of Illicit Drugs and Illicit Drugs Other than
Marijuana

Substance
Marijuana
Cocaine (including
Crack)
Heroin
Hallucinogens
Inhalants
Prescription Drug
Misuse
Methamphetamine
Estimates Shown in:

Standard
Definition


Illicit Drugs
Alternate
Alternate
Definition
Definition
1
2



Illicit Drugs Other than Marijuana
Alternate
Definition
3


Standard
Definition

Alternate
Definition

Cocaine or
Heroin















Appendix I



Appendix J


Appendix I

Appendix I


Appendix J

 = Use of this substance is included in the summary measure.

Note that a respondent who is considered a user of illicit drugs other than marijuana may have used
marijuana, but he or she would have used one of the other substances to be considered a user of illicit drugs other
than marijuana. Similarly, information on the use of methamphetamine and the misuse of psychotherapeutics is
ignored in creating these measures.

218

6.4.1

6.4.2

Any Illicit Drug
•

Summary estimates of lifetime use of illicit drugs based on Alternate Definition 1
(i.e., including hallucinogens and inhalants but not methamphetamine or prescription
drugs) and Alternate Definition 2 (i.e., excluding hallucinogens and inhalants in
addition to methamphetamine and prescription drugs) did not differ between the QFT
and comparison data for persons aged 12 or older (Table I-1), adults aged 18 to 25
(Table I-3), or adults aged 26 or older (Table I-4). Among persons aged 18 to 25, for
example, lifetime estimates based on Alternate Definition 1 were 56.0 percent in the
QFT, 54.5 percent in the 2011 comparison data, and 54.2 percent in the 2012
comparison data (Table I-3). Corresponding estimates based on Alternate Definition
2 were 52.2 percent in the QFT, 53.1 percent in the 2011 comparison data, and
53.0 percent in the 2012 comparison data.

•

Among adolescents aged 12 to 17, the summary estimate of lifetime use of illicit
drugs based on Alternate Definition 1 was higher in the QFT (26.7 percent) than in
the 2011 and 2012 comparison data (22.3 and 20.0 percent, respectively) (Table I-2).
When hallucinogens and inhalants were removed for Alternate Definition 2, however,
the estimates of lifetime use of illicit drugs among adolescents no longer differed
between the QFT and comparison data.

•

Consistent with the pattern observed for lifetime use, the prevalence of past year and
past month use of illicit drugs based on Alternate Definition 1 and Alternate
Definition 2 did not differ between the QFT and comparison data for persons aged
12 or older (Tables I-5 and I-9), adults aged 18 to 25 (Tables I-7 and I-11), or adults
aged 26 or older (Tables I-8 and I-12). Among persons aged 12 or older, estimates of
past year illicit drug use based on Alternate Definition 1 ranged from 12.8 to
13.5 percent (Table I-5). Past year estimates for persons aged 12 or older based on
Alternate Definition 2 ranged from 12.3 to 12.7 percent.

•

Among adolescents aged 12 to 17, the estimate of past year use of illicit drugs based
on Alternate Definition 1 in the QFT (18.2 percent) was greater than the estimate in
the 2012 comparison data (14.2 percent), but it was not significantly different from
the estimate in the 2011 comparison data (15.8 percent) (Table I-6). Estimates of past
year use of illicit drugs for adolescents based on Alternate Definition 2 did not differ
between the QFT and comparison data.

•

Estimates of past month use of illicit drugs among adolescents aged 12 to 17 did not
differ between the QFT and comparison data for Alternate Definition 1 or Alternate
Definition 2 (Table I-10). For example, estimates of past month use among
adolescents based on Alternate Definition 1 ranged from 7.2 to 8.5 percent in these
three datasets.

Illicit Drugs Other than Marijuana

As noted previously, marijuana historically has been the most commonly used illicit drug.
Consequently, similar estimates of any illicit drug use in the QFT and comparison data for
Alternate Definitions 1 and 2 could be explained by a corresponding lack of significant
differences for marijuana use. Changes to the QFT questions for hallucinogens and inhalants

219

could have more of an effect on estimates of use of illicit drugs other than marijuana (or even
more of an effect on these estimates for adolescents aged 12 to 17). Higher rates of use of
cocaine, crack, and heroin in the QFT that were reported in Section 6.2 also affect estimates for
use of illicit drugs other than marijuana, independent of the changes to the modules for
hallucinogens and inhalants.
•

Rates of lifetime use of illicit drugs other than marijuana based on the Alternate
Definition that included hallucinogens and inhalants but not methamphetamine or
prescription drugs were not significantly different between the QFT and comparison
data (Table I-1). However, the differences approached statistical significance for the
QFT (25.0 percent) and the 2011 comparison data (22.4 percent; p = 0.077) and for
the QFT and 2012 comparison data (22.3 percent; p = 0.066). Estimates of lifetime
use of illicit drugs other than marijuana that were limited to cocaine (including crack)
and heroin among persons aged 12 or older ranged from 14.3 to 14.9 percent and did
not differ between the QFT and comparison data.

•

Among adolescents aged 12 to 17, the rate of lifetime use of illicit drugs other than
marijuana based on the Alternate Definition that included hallucinogens and inhalants
was greater in the QFT (16.3 percent) than in the 2011 or 2012 comparison data
(10.3 and 8.2 percent, respectively) (Table I-2). In contrast, the QFT estimate of
lifetime use of cocaine or heroin among adolescents (0.5 percent) was lower than the
corresponding estimates in the comparison data for 2011 (1.4 percent) and 2012
(1.3 percent).

•

For young adults aged 18 to 25, the lifetime estimate for the Alternate Definition of
any illicit drugs other than marijuana in the QFT (28.8 percent) was higher than that
in the 2012 comparison data (23.6 percent) (Table I-3). The difference in estimates
between the QFT and 2011 comparison data (24.0 percent) approached statistical
significance (p = 0.060).

•

Lifetime estimates of use of cocaine or heroin among 18 to 25 year olds did not differ
between the QFT and comparison data and ranged from 10.5 to 12.7 percent
(Table I-3).

•

Estimates of lifetime use of illicit drugs other than marijuana based on the Alternate
Definition or for cocaine or heroin did not differ between the QFT and comparison
data for adults aged 26 or older (Table I-4). For example, the Alternate Definition
estimates ranged from 23.7 to 25.5 percent.

•

Estimates of past year use of illicit drugs other than marijuana based on the Alternate
Definition or for cocaine or heroin did not differ between the QFT and comparison
data for persons aged 12 or older (Table I-5), adults aged 18 to 25 (Table I-7), or
adults aged 26 or older (Table I-8). Among persons aged 12 or older, the Alternate
Definition estimates ranged from 3.2 to 3.5 percent.

•

Among adolescents aged 12 to 17, the QFT estimate of past year use based on the
Alternate Definition was greater than the estimate for the 2012 comparison data
(7.0 vs. 4.2 percent), but it did not differ from the estimate of 5.3 percent for the 2011
comparison data (Table I-6). In contrast, the QFT estimate of past year use of cocaine
or heroin among adolescents (0.2 percent) was lower than the estimate from the 2011
220

comparison data (1.0 percent), and the difference between the QFT and 2012
comparison data (0.8 percent) approached statistical significance (p = 0.072).
•

Estimates of past month use of illicit drugs other than marijuana based on the
Alternate Definition did not differ between the QFT and comparison data for persons
aged 12 or older (Table I-9) or among any age groups (Tables I-10 to I-12).
Estimates of past month use of cocaine or heroin also did not differ between the QFT
and comparison data for persons aged 12 or older and adults aged 26 or older.

•

There were some significant differences in estimates of past month use of cocaine or
heroin between the QFT and comparison data for adolescents aged 12 to 17
(Table I-10) and young adults aged 18 to 25 (Table I-11). Although the QFT estimate
for adolescents (0.0 percent) was lower than the estimates in the comparison data for
2011 (0.3 percent) and 2012 (0.1 percent), the QFT estimate would be suppressed.
For young adults, the QFT estimate was lower than the estimate in the 2011
comparison data (0.7 vs. 1.5 percent), but it was not significantly different from the
estimate in the 2012 comparison data (1.2 percent).

Taken together, these findings suggest that changes to the modules for hallucinogens and
inhalants could affect trend data for the use of illicit drugs and illicit drugs other than marijuana
in 2015, especially for adolescents. Effects on these illicit drug use estimates because of the
changes for hallucinogens and inhalants will warrant further investigation in the 2013 DR and in
preliminary data for 2015. Although the cocaine and heroin modules did not change for the QFT,
some significant differences also were observed for aggregate estimates of use of cocaine or
heroin. As noted previously, further examination of estimates of cocaine and heroin use in the
2013 DR will be useful for assessing the likelihood that data for these two substances also will
not disrupt the trends in 2015.

6.5

Tobacco

This section presents findings on tobacco use from the 2011 comparison data and 2012
quarters 3 and 4 comparison data, as well as the QFT data. Estimates for use of cigarettes and
smokeless tobacco are presented in Appendix I for each of the three datasets. Tables I-1 through
I-4 provide estimates for lifetime use of these tobacco products for all persons aged 12 or older,
adolescents aged 12 to 17, young adults aged 18 to 25, and adults aged 26 or older, respectively.
Likewise, Tables I-5 through I-8 provide estimates for past year use, and Tables I-9 through
I-12 provide estimates for past month use.
Questions on cigarette use were not changed for the QFT instrument, so the expectation
was that the QFT estimate would be very similar to the estimates for the 2011 comparison data
and 2012 quarters 3 and 4 comparison data. In the main survey, however, respondents are asked
separate sets of questions about their use of snuff and about their use of chewing tobacco. In the
QFT, respondents were asked a single set of questions about use of any smokeless tobacco
product. Smokeless tobacco for the QFT also was defined somewhat differently than in the main

221

survey and included use of snuff, dip, chewing tobacco, or "snus."25 These changes could affect
estimates of smokeless tobacco use.
6.5.1

Cigarettes

Consistent with expectations, the QFT estimates for cigarette use were similar to the 2011
comparison estimates and 2012 quarters 3 and 4 comparison data estimates. None of the small
differences in cigarette use across the three samples was statistically significant. This pattern
held for lifetime, past year, and past month cigarette use estimates and held for estimates across
all age groups.
•

For all persons aged 12 or older, the prevalence of lifetime cigarette use was
62.5 percent for the QFT sample, 63.9 percent for the 2011 comparison data, and
63.2 percent for the 2012 quarters 3 and 4 comparison data (Table I-1). Estimates for
lifetime cigarette use ranged from less than 20 percent for adolescents aged 12 to 17
in all three samples (Table I-2) to about 70 percent for adults aged 26 or older for all
three samples (Table I-4).

•

The estimate of past year cigarette use for all persons aged 12 or older was
28.0 percent for the QFT sample, 26.5 percent in the 2011 comparison data, and
26.1 percent for the 2012 comparison data (Table I-5). Estimates for past year
cigarette use ranged from less than 13 percent for adolescents aged 12 to 17 in all
three samples (Table I-6) to more than 40 percent for young adults aged 18 to 25 in
all three samples (Table I-7).

•

The rate of past month cigarette use for all persons aged 12 or older was 24.2 percent
for the QFT sample, 22.5 percent for the 2011 comparison data, and 22.2 percent for
the 2012 comparison data (Table I-9). Estimates for past month cigarette use among
adolescents aged 12 to 17 (Table I-10) appeared to be higher in the 2011 comparison
data (7.8 percent) than in the QFT data (6.1 percent), but as previously noted, this
difference was not statistically significant. Estimates of past month cigarette use
among young adults aged 18 to 25 ranged from 31.8 to 34.0 percent in all three
samples (Table I-11).

Given the lack of changes to questions on cigarette use and the similarity of estimates
across all three datasets, these results do not suggest any changes to these questions are
warranted for the 2013 DR. Based on these findings, it seems likely that the trend for estimates
of cigarette use will continue when the partially redesigned instrument and protocol are
implemented in 2015.
6.5.2

Smokeless Tobacco

Lifetime estimates of smokeless tobacco use did not differ between the QFT and
comparison data for persons aged 12 or older or within any of the three age groups. However,
estimates of past year and past month use were greater in the QFT than in the comparison data
for persons aged 12 or older and adults aged 26 or older. For adolescents aged 12 to 17 and
25

"Snus" is a type of Swedish snuff. The question in the QFT is as follows: "The next questions are about
your use of 'smokeless' tobacco such as snuff, dip, chewing tobacco, or 'snus.'"

222

young adults aged 18 to 25, the estimates of past year and past month smokeless tobacco use did
not differ between the QFT and comparison data. Thus, the higher estimates among adults aged
26 or older appear to be driving the higher past year and past month estimates for persons aged
12 or older in the QFT.
•

Estimates of lifetime smokeless tobacco use among persons aged 12 or older were
17.4 percent in the QFT, 18.8 percent in the 2011 comparison data, and 18.4 percent
in the 2012 comparison data (Table I-1). Lifetime estimates ranged from 6.4 to
8.3 percent among adolescents aged 12 to 17 (Table I-2). Among adults aged 26 or
older, estimates ranged from 18.0 to 20.0 percent (Table I-4).

•

The estimate of past year use of smokeless tobacco for persons aged 12 or older in the
QFT was 6.8 percent compared with estimates of 4.7 percent in each of the
comparison samples (Table I-5). Among adults aged 26 or older, the rate of past year
use was 6.6 percent in the QFT compared with 3.9 percent in the 2011 comparison
data and 4.0 percent in the 2012 comparison data (Table I-8).

•

The estimate of past month use of smokeless tobacco for persons aged 12 or older in
the QFT was 5.2 percent compared with estimates of 3.4 to 3.5 percent in the
comparison samples (Table I-9). Among adults aged 26 or older, the rate of past
month use was 5.5 percent in the QFT compared with rates of 3.1 to 3.3 percent in the
comparison data (Table I-12).

These findings suggest that trends could be disrupted for past year and past month use of
smokeless tobacco for all persons aged 12 or older and among adults aged 26 or older in 2015.
Given that respondents had two opportunities to report past year or past month use of smokeless
tobacco in the comparison data, it also is noteworthy that the QFT estimates of past year and past
month use (which were based only on one set of questions) were higher than the comparison
estimates for persons aged 12 or older and adults aged 26 or older. All other things being equal,
providing respondents with multiple opportunities to report use would be expected to yield
higher estimates than questions that allow respondents only a single opportunity to report use in
a given period.26
One possible explanation for these findings is that it may be less of a challenge for some
respondents to determine that they used some type of "smokeless tobacco" in the past year or
past month than to determine whether the product specifically was "snuff" or "chewing tobacco."
This explanation is consistent with main survey data for the brand of snuff or chewing tobacco
that respondents reported using most often in the past 30 days. Specifically, respondents could
specify a brand of snuff as some "other" brand of "chewing tobacco" they used most often, or
vice versa (Kroutil et al., 2012a). Although respondent difficulties in distinguishing between
snuff and chewing tobacco in the main survey can be identified only for the past 30 days, they
also are likely to be occurring for reports of these types of smokeless tobacco use that occurred
less recently than the past 30 days but within 12 months of the interview.

Although estimates of past year use also include reports of use in the past month, QFT respondents had
only a single opportunity to report that they used smokeless tobacco in the past 30 days or more than 30 days ago
but within the past 12 months.

223

6.6

Alcohol

Tables I-1 through I-4 provide estimates for lifetime alcohol use for all persons aged 12
or older, adolescents aged 12 to 17, young adults aged 18 to 25, and adults aged 26 or older,
respectively. Likewise, Tables I-5 through I-8 provide estimates for past year alcohol use, and
Tables I-9 through I-12 provide estimates for past month alcohol use. In addition, Table I-15
provides estimates for past month alcohol use by age and gender, and Table I-16 presents
estimates for binge alcohol use in the past month by age and gender. All of these tables provide
estimates for the 2011 comparison data and 2012 quarters 3 and 4 comparison data, as well as the
QFT data.
Because the primary questions for lifetime, past year, and past month alcohol use were
not changed for the QFT instrument, QFT estimates for these items were expected to be very
similar to the 2011 comparison data and 2012 quarters 3 and 4 comparison data. One notable
change in the QFT instrument involved the definition of binge alcohol use. In the 2011 and 2012
quarters 3 and 4 instruments, binge alcohol use is defined as drinking five or more drinks on one
occasion for both male and female respondents. In the QFT instrument, the definition of binge
alcohol use was changed to drinking four or more drinks on one occasion for female respondents.
This change had the potential to increase reports of binge alcohol use by lowering the threshold
for the minimum number of drinks for females.
6.6.1

Any Alcohol Use

Consistent with expectations, the QFT estimates for alcohol use were very similar to the
2011 comparison estimates and 2012 quarters 3 and 4 comparison estimates across all age groups
within the lifetime, past year, and past month periods. Similarly, no significant differences were
observed for any alcohol use in the past month among males and females (Table I-15).
•

For all persons aged 12 or older, the rate of lifetime alcohol use was 81.8 percent for
the QFT sample, 83.2 percent for the 2011 comparison data, and 83.4 percent for the
2012 quarters 3 and 4 comparison data (Table I-1). Estimates for lifetime alcohol use
ranged from about 33 percent for adolescents aged 12 to 17 in all three samples
(Table I-2) to nearly 90 percent for adults aged 26 or older in all three samples
(Table I-4).

•

The estimate of past year alcohol use for all persons aged 12 or older was 66.8
percent for the QFT sample, 67.1 percent in the 2011 comparison data, and 67.6
percent for the 2012 comparison data (Table I-5). Estimates for past year alcohol use
ranged from about one fourth of adolescents aged 12 to 17 in all three samples
(Table I-6) to about three fourths of young adults aged 18 to 25 in all three samples
(Table I-7).

•

Rates of past month alcohol use for all persons aged 12 or older were 51.6 percent for
the QFT sample, 53.0 percent for the 2011 comparison data, and 53.4 percent for the
2012 comparison data (Table I-9). The estimate for past month alcohol use among
adolescents aged 12 to 17 was higher in the 2011 comparison data (13.4 percent) than
in the QFT data (10.3 percent) (Table I-10).

224

The lack of significant differences in most rates of any alcohol use between the QFT and
comparison data suggests that trends in any alcohol use generally will be maintained in 2015.
However, examination of estimates of past month alcohol use among adolescents aged 12 to 17
will warrant further attention in the 2013 DR to assess whether the significant difference between
the QFT and 2011 comparison data is repeated for other comparisons in the 2013 DR, or if this
difference was an anomaly.
6.6.2

Past Month Binge Alcohol Use

There were no significant differences in estimates of binge alcohol use in the past month
regardless of gender for persons aged 12 or older or in any of the three age groups (Tables I-9 to
I-12). However, differences approached statistical significance for adults aged 26 or older
(Table I-12).
•

Rates of binge alcohol use in the past month among all persons aged 12 or older were
23.9 percent for the QFT sample, 22.3 percent for the 2011 comparison data, and
22.9 percent for the 2012 comparison data (Table I-9).

•

Among adults aged 26 or older, the differences in estimates of binge alcohol use
approached statistical significance for the QFT and both comparison samples (QFT
and 2011 comparison: 23.2 and 21.4 percent; p = 0.074; QFT and 2011 comparison:
23.2 and 22.1 percent; p = 0.084) (Table I-12).

Table I-16 contains two sets of estimates of binge alcohol use by age group and gender.
The first set of estimates is based only on core data. As noted previously, binge alcohol use in the
comparison data was defined for males and females as drinking five or more drinks on the same
occasion on at least 1 day in the past 30 days based on their reports in the core alcohol module.
For the QFT, binge alcohol use was defined for males in the same manner as in the comparison
data. For females, binge alcohol use in the QFT was defined as drinking four or more drinks on
the same occasion based on their reports in the core alcohol module.
Table I-16 also contains core-plus-noncore (CPN) estimates for the 2011 and 2012
comparison data. In addition to reports of consumption of five or more drinks on a single
occasion on at least 1 day in the past 30 days, these CPN measures took into account females'
reports of usual consumption of four or more drinks on the days that they drank alcohol in the
past 30 days (from the core alcohol module) or their consumption of four or more drinks on the
same occasion on at least 1 day in the past 30 days (from the noncore consumption of alcohol
module). These CPN measures were created to further gauge the potential effects on estimates of
binge alcohol use because of the change to the threshold for females. For males in the
comparison data, the CPN measure was the same as the measure based only on core data. QFT
estimates based on core alcohol use data (i.e., including the "four or more" criterion for females)
are repeated for comparison with the CPN estimates.
•

Among all persons aged 12 or older in the QFT, the rates of binge alcohol use in the
past month were 30.1 percent for males and 18.2 percent for females.

225

•

Estimates of binge alcohol use among males aged 12 or older were similar between
the QFT (30.1 percent) and the comparison data for 2011 (29.3 percent) and 2012
(30.4 percent).

•

The estimate of binge alcohol use in the past month for females aged 12 or older in
the QFT (18.2 percent) was in the direction of being higher than the core-only
estimates for the 2011 comparison sample and the 2012 quarters 3 and 4 comparison
sample (15.8 percent in each sample). However, differences between the QFT and
comparison data were not statistically significant.

•

Although the measure of binge alcohol use in the past month for males was the same
in the QFT and comparison data, the difference between the estimates for males aged
12 to 17 in the QFT and the 2011 comparison data approached statistical significance
(23.9 and 22.3 percent, respectively; p = 0.097).

•

Among females aged 26 or older, the difference between the QFT estimate of binge
alcohol use (16.8 percent) and the core estimate in the 2011 comparison data
(14.0 percent) approached statistical significance (p = 0.085).

•

The CPN estimates of binge alcohol use for females aged 12 or older in the 2011 and
2012 comparison data (20.7 and 20.8 percent, respectively) that took into account
reports of consumption of four or more drinks on an occasion were not significantly
different from the QFT estimate from the core alcohol module (18.2 percent).
However, these differences between the QFT and comparison data approached
statistical significance (QFT vs. 2011 comparison: p = 0.067; QFT vs. 2012
comparison: p = 0.060).

•

The difference between the QFT and CPN estimate of binge alcohol use in the 2011
comparison data also approached statistical significance for all adolescents aged 12 to
17, regardless of gender (5.6 and 7.5 percent, respectively; p = 0.061).

These findings suggest that lowering the threshold for binge alcohol use among females
to consumption of four or more drinks on an occasion may not affect the trends in binge alcohol
use among all persons aged 12 or older or among all persons within most age groups (i.e.,
regardless of gender). Although statistical testing was not conducted to identify whether rates of
binge alcohol use in the QFT were higher among males than among females even with the lower
threshold for females, the relatively higher (but not necessarily significant) rate of binge alcohol
use among males aged 12 or older in the QFT than among females suggests that binge alcohol
use among males will continue to drive the overall rates of binge alcohol use in 2015.
Adults aged 26 or older may provide an exception to this general conclusion. If the QFT
sample size of adults in this age group had been similar to the sample sizes in the comparison
data, the apparently higher rate in the QFT may have been statistically significant. The finding
that the differences in core-only estimates of binge alcohol use among females aged 26 or older
approached statistical significance between the QFT and both comparison samples also suggests
that the planned change in the definition of binge alcohol use among females in 2015 may affect
trends for females in this age group. The lower threshold for binge alcohol use among females
may be more important for estimating binge alcohol use among adults aged 26 or older (both
overall and for females) than it is for other age groups.

226

7. QFT Estimates Compared with Current
NSDUH Estimates: Methamphetamine and
Prescription Drug Items
7.1

Overview of QFT Estimates Compared with NSDUH Estimates for
Methamphetamine and Prescription Drug Items

As noted in Sections 2.4.1 and 3.3.3, the following changes to the questions for
methamphetamine and prescription drugs were made for the Questionnaire Field Test (QFT):
•

A new methamphetamine module was added instead of questions about
methamphetamine use being included as part of the stimulants module.

•

The definition, approach, and terminology for measuring misuse of prescription drugs
were revised.

•

Modules were added that asked respondents about any use of pain relievers,
tranquilizers, stimulants, and sedatives, as opposed to just misuse.

•

The focus of the prescription drug modules was on a 12-month reference period
rather than the lifetime reference period used in the current questionnaire.

•

Electronic images of prescription drugs replaced the current hard-copy pill card
versions, and the images included more than just pills.

•

Questions about discontinued prescription drugs were deleted, and questions were
added for other prescription drugs not previously included in the questionnaire.

•

Questions about prescription drugs that were included in supplemental sections of the
current questionnaire were moved to the appropriate prescription drug module.

These changes are planned for implementation in the redesigned National Survey on Drug Use
and Health (NSDUH) questionnaire in 2015 and are likely to affect estimates of
methamphetamine use and misuse of prescription drugs starting in 2015.
This chapter presents findings on methamphetamine use and prescription drug misuse
from the comparison data for 2011 and quarters 3 and 4 of 2012 and from the QFT. Detailed
tables containing these estimates are included in Appendix J. For each relevant measure, data are
presented in the detailed tables for use or misuse in the lifetime, past year, and past month
periods, as well as for the following age groups: 12 or older, 12 to 17, 18 to 25, and 26 or older.
Variables for all drug use estimates presented in this chapter were edited according to the
procedures described in Section 3.3 and were imputed according to the procedures described in
Section 3.4. Consequently, these drug use measures had no missing data.
Findings also note whether differences in estimates between the QFT and the comparison
data were statistically significant at the 0.05 level of significance (see Section 3.7). In addition,
some differences are presented that approached but did not attain statistical significance

227

(i.e., 0.05 < p < 0.1). Because of the smaller sample sizes for the QFT, differences that
approached statistical significance in these comparisons could become significant with a sample
size of approximately 67,000 respondents in 2015. Otherwise, statements in this chapter such as
"estimates did not differ significantly between the QFT and comparison data" indicate
differences in which p > 0.1.

7.2

Estimates for Methamphetamine Items

A consequence of the placement of questions about methamphetamine use within the
current NSDUH module for misuse of prescription stimulants is that misuse of any stimulant
always will be as recent as or more recent than the last use of methamphetamine in the edited and
imputed data. Furthermore, as noted in Section 3.3.4.5, a consistency check is triggered in the
core stimulants module in the main survey if respondents report more recent use of
methamphetamine than they reported for most recent use of any prescription stimulant. Some
respondents in these consistency checks may change their answer for methamphetamine to
indicate less recent use than they had originally reported. Because the methamphetamine
questions in the QFT were in a module separate from the questions about misuse of prescription
stimulants, respondents could report lifetime use or more recent use of methamphetamine
without needing to report lifetime misuse of stimulants or misuse of stimulants as recently or
more recently than when they last used methamphetamine.
Also, respondents who receive the current NSDUH questionnaire may fail to report
methamphetamine use when questions about this drug are asked in the context of questions about
misuse of prescription stimulants. Therefore, the methamphetamine use measures for the
comparison data (i.e., 2011 and quarters 3 and 4 of 2012) were based on reports of
methamphetamine use in the core stimulants module plus reports of use from the supplemental
(or noncore) special drugs module (i.e., core plus noncore, or CPN). However, additional
respondents who reported lifetime use of methamphetamine in the special drugs module were
included in the CPN measures only if their reason for not previously reporting methamphetamine
use was that they did not think of methamphetamine as a prescription drug; respondents who
reported use in the special drugs module were not counted as users if they reported that they did
not previously report methamphetamine use because they "made a mistake" when answering the
methamphetamine questions in the stimulants module or for reasons other than not thinking of
this as a prescription drug (Kroutil, Handley, Bradshaw, Chien, & Felts, 2012b). Consequently,
these CPN measures of methamphetamine use in the comparison data still might underestimate
the prevalence of use.
For the QFT, the methamphetamine use measures were based only on data from the new
methamphetamine module in the core section of the QFT questionnaire. Although QFT
respondents did not have the same multiple opportunities to report methamphetamine use as in
the comparison data, there also was no question (and no need) to check for the reason that some
respondents did not previously report methamphetamine use.
•

The estimate of lifetime methamphetamine use among persons aged 12 or older was
greater in the QFT than in the 2012 comparison data (6.5 vs. 4.8 percent) (Table J-1).
The estimate for 2011 (also 4.8 percent) was not significantly different from the QFT
estimate but approached statistical significance (p = 0.062).
228

7.3

•

Among persons in the three age groups, estimates of lifetime methamphetamine use
did not differ significantly between the QFT and comparison data. Estimates for
adults aged 26 or older were 5.6 percent in 2011 and in the 2012 comparison data and
7.7 percent in the QFT (Table J-4). Again, these differences approached statistical
significance (p = 0.069 for QFT vs. 2011; p = 0.052 for QFT vs. 2012).

•

Estimates of methamphetamine use in the past year among persons aged 12 or older
and in each of the three age groups did not differ significantly between the QFT and
comparison data. Estimates for persons aged 12 or older were 0.4 percent in 2011 and
in the 2012 comparison data and 0.5 percent in the QFT (Table J-5). However, the
difference between the estimates of past year use for adolescents aged 12 to 17 in the
QFT (0.2 percent) and the 2011 comparison data (0.4 percent) approached statistical
significance (p = 0.095) (Table J-6).

•

Estimates of methamphetamine use in the past month among persons aged 12 or older
and in each of the three age groups did not differ significantly between the QFT and
comparison data. Among persons aged 12 or older, the difference between the QFT
estimate (0.4 percent) and the estimate for the 2012 comparison data (0.1 percent)
approached statistical significance (p = 0.077) (Table J-9).

Estimates for Prescription Drug Items

principally through their response to the question, "How long has it been since you last used any
prescription pain reliever that was not prescribed for you or that you took only for the experience
or feeling it caused?" Only those respondents in the comparison data who reported lifetime
misuse of the pain reliever OxyContin® have an additional opportunity to report past year misuse
through a corresponding question about the last time they used OxyContin® that was not
prescribed for them or that they took only for the experience or feeling the drug caused.
As noted previously, the definition of misuse also was changed for the QFT. The
definition of misuse in the main survey combines a behavior (use of a prescription drug that was
not prescribed for the respondent) and a motivation for misuse (use of a prescription drug only
for the experience or feeling that it caused). In the QFT, the definition of misuse "in any way a
doctor did not direct you to use it" focuses on behaviors. The following examples are given to
QFT respondents for behaviors that constitute misuse:
•

(use) without a prescription of your own;

•

(use) in greater amounts, more often, or longer than you were told to take it; or

•

(use) in any other way a doctor did not direct you to use it.

Especially for misuse of prescription pain relievers, alerting QFT respondents that overuse of
prescribed medication (e.g., use in greater amounts or more often than prescribed) constitutes
misuse also could increase reporting of misuse in the QFT.
7.3.1 Any Prescription Psychotherapeutic Drug
•

The estimate of lifetime misuse of any prescription psychotherapeutic drug (i.e., pain
relievers, tranquilizers, stimulants, or sedatives) among persons aged 12 or older was
lower in the QFT than in the 2012 comparison data (17.9 vs. 21.0 percent)
(Table J-1). The estimate for 2011 (20.5 percent) was not significantly different from
the QFT estimate but approached statistical significance (p = 0.062).

•

Adults aged 26 or older had a lower estimate of lifetime misuse of any prescription
drug in the QFT than in the 2012 comparison data (17.7 vs. 21.2 percent) (Table J-4).
Estimates approached statistical significance for adolescents aged 12 to 17 in both the
2011 and 2012 comparison data (p = 0.057 for QFT vs. 2011; p = 0.077 for QFT vs.
2012) (Table J-2) and for adults aged 26 or older in the 2011 comparison data
(p = 0.090) (Table J-4).

•

Estimates of misuse of any prescription drug in the past year were greater in the QFT
than in the 2011 and 2012 comparison data for persons aged 12 or older (8.1, 5.7, and
5.9 percent, respectively) (Table J-5) and young adults aged 18 to 25 (22.8, 13.0, and
13.2 percent, respectively) (Table J-7), but not for adolescents aged 12 to 17 or adults
aged 26 or older (Tables J-6 and J-8).

•

Among persons aged 12 or older, estimates of misuse of any prescription drug in the
past month approached statistical significance between the QFT (3.2 percent) and
both sets of comparison data (2.4 percent in each comparison dataset; p = 0.088 for
QFT vs. 2011; p = 0.096 for QFT vs. 2012) (Table J-9). Estimates also approached
statistical significance for adults aged 18 to 25 in the QFT (7.4 percent) and both
230

comparison datasets (2011: 5.0 percent, p = 0.064; 2012: 4.9 percent, p = 0.063)
(Table J-11).
•

The estimate of past month misuse of any prescription drug for adolescents in the
QFT (1.3 percent) was lower than the estimates in the comparison data for 2011
(2.7 percent) and 2012 (2.5 percent) (Table J-10).

Given that estimates of past month misuse of any prescription drug were in the direction
of being greater in the QFT than in the comparison data (but did not attain statistical
significance) for persons aged 12 or older and those aged 18 to 25, the finding of lower estimates
in the QFT than in the comparison data for adolescents aged 12 to 17 is counterintuitive. As
noted in Chapter 6, however, additional illicit drug use estimates in the QFT were lower among
adolescents. Therefore, further examination of estimates of prescription drug misuse using data
from the 2013 Dress Rehearsal (DR) will be important for adolescents.
7.3.2 Pain Relievers
Estimates for misuse of prescription pain relievers followed the same general pattern as
misuse of any prescription drug, with some lower estimates of lifetime misuse in the QFT than in
the 2012 comparison data, higher estimates of past year misuse in the QFT than in both
comparison datasets for persons aged 12 or older and young adults aged 18 to 25, and lower
estimates of past month misuse among adolescents aged 12 to 17 in the QFT than in the two
comparison datasets. Highlights are presented in the remainder of this section for past year
misuse.
•

An estimated 6.0 percent of persons aged 12 or older were past year misusers of pain
relievers according to the QFT compared with 4.3 percent for the 2011 comparison
data and 4.4 percent for the 2012 comparison data (Table J-5). Among young adults
aged 18 to 25, 15.2 percent were past year misusers of pain relievers according to the
QFT compared with 10.0 percent for 2011 and 9.3 percent for the 2012 comparison
data (Table J-7).

•

The estimate of past year misuse of pain relievers among adults aged 26 or older
approached statistical significance for persons aged 12 or older data between the QFT
and 2011 comparison data (p = 0.089) (Table J-8).

•

Estimates of past year misuse of OxyContin® among persons aged 12 or older were
1.1 percent for the QFT, 0.6 percent for the 2011 comparison data, and 0.5 percent for
the 2012 comparison data (Table J-5).27

•

Estimates of past year misuse of OxyContin® among young adults aged 18 to 25 were
2.9 percent for the QFT, 1.9 percent for the 2011 comparison data, and 1.4 percent for
the 2012 comparison data (Table J-7). The difference between the QFT and 2012
comparison data estimates approached statistical significance (p = 0.092).

Because of the changes to the prescription drug questions, it was possible to estimate only the past year
prevalence of OxyContin® misuse for the QFT.

231

7.3.3 Tranquilizers
•

Estimates of lifetime misuse of tranquilizers in the QFT were lower than the
corresponding estimates from the 2011 and 2012 comparison data for persons aged 12
or older and all age groups except adolescents aged 12 to 17. Among persons aged 12
or older, the estimate for lifetime tranquilizer misuse was 5.6 percent compared with
estimates of 8.8 and 9.3 percent in the 2011 and 2012 comparison data, respectively
(Table J-1).

•

Young adults aged 18 to 25 were more likely to be past year misusers of tranquilizers
based on the QFT (7.8 percent) than in the 2011 and 2012 comparison data (4.6 and
4.9 percent, respectively) (Table J-7). Rates of past year misuse of tranquilizers did
not differ significantly between the QFT and the comparison data for persons aged 12
or older and the other age groups (Tables J-5 to J-8).

•

The prevalence of misuse of tranquilizers in the past month was similar between the
QFT and the comparison data for persons aged 12 or older and all age groups
(Tables J-9 to J-12).

7.3.4 Sedatives
•

Unlike the general pattern for other prescription drugs, the estimate of lifetime misuse
of sedatives among young adults aged 18 to 25 in the QFT was greater that the
estimate in the 2012 comparison data (2.6 vs. 1.1 percent) (Table J-3). Otherwise,
estimates of lifetime misuse of sedatives were similar between the QFT and the two
comparison datasets.

•

Estimates of past year sedative misuse in the QFT were greater than corresponding
estimates in the 2011 and 2012 comparison data for all groups except adolescents
aged 12 to 17 (Tables J-5 to J-8).

•

The prevalence of misuse of sedatives in the past month was similar between the QFT
and the comparison data for persons aged 12 or older and all age groups (Tables J-9
to J-12).

232

•

Inclusion of data for Ambien® raised the CPN estimates of lifetime misuse of
sedatives in the comparison data to the point that these estimates were now greater
than the QFT estimates for all groups except young adults aged 18 to 25.
Furthermore, this pattern of differences between the CPN and QFT estimates is
consistent with the general pattern elsewhere for prescription drugs, with estimates of
lifetime misuse in the QFT tending to be lower than corresponding estimates in the
comparison data.

•

Among young adults aged 18 to 25, CPN estimates of lifetime misuse of sedatives
were 4.1 percent in the 2011 comparison data and 3.7 percent in the 2012 comparison
data (Table J-16). As noted previously, the corresponding QFT estimate of lifetime
misuse in this age group was 2.6 percent.

•

Ambien® data in the CPN estimates of past year misuse appeared to erase the
differences in prevalence between the QFT and comparison data that were observed
for comparison data estimates based only on core sedatives module data (or, in some
instances, to reverse the direction of the differences). Among persons aged 12 or
older, for example, the CPN estimates of past year misuse of sedatives in the 2011
and 2012 comparison data (0.9 and 0.7 percent, respectively) were similar to the QFT
estimate (0.8 percent) (Table J-17). Without the Ambien® data, the estimate of past
year misuse of sedatives was 0.2 percent in each comparison dataset. In addition, the
CPN estimate of past year sedative misuse among 12 to 17 year olds was greater than
the QFT estimate (0.8 vs. 0.3 percent).

•

Inclusion of Ambien® data in the CPN estimates had little apparent effect on
estimates of past month sedative misuse or differences between the QFT and
comparison data for past month misuse (Table J-18).

Although the estimate of lifetime misuse of sedatives was greater in the QFT than in the
comparison data for young adults aged 18 to 25, including the noncore Ambien® data in the CPN
estimates for sedatives in the comparison data erased this difference. Findings that including
reports of Ambien® misuse in the CPN estimates of past year misuse appeared to remove the
differences in prevalence between the QFT and comparison data also underscore the likely
importance of including questions about Ambien® for estimating sedative misuse. Given the
potential for changes in the prescription drug market and the prescription drug market share, a
further implication of these findings for sedatives is the need for regular monitoring of changes
in prescription drug availability beyond the redesign of the prescription drug questions in 2015.
The Substance Abuse and Mental Health Services Administration (SAMHSA) plans to
implement procedures for monitoring prescription drug changes in connection with the redesign.
7.3.5 Stimulants
Because the estimates of methamphetamine use in the 2011 and 2012 comparison data
were based on CPN measures of methamphetamine use (see Section 7.2), the corresponding
estimates of any stimulant misuse in the comparison data included these CPN methamphetamine
use data. These CPN measures are referred to as the "Standard Definition" of stimulant misuse in
the Appendix J tables. To produce estimates of stimulant misuse for the QFT that were as
analogous as possible to these estimates in the comparison data, the "standard definition"
estimates of stimulant misuse were based on data from the core methamphetamine and
233

prescription stimulants modules. A "QFT definition" of stimulant misuse also was created for the
QFT based on data in the core stimulants module but not including data on methamphetamine
use. Because it is not possible to disentangle methamphetamine use from misuse of other
stimulants in the comparison data, however, this QFT definition measure was not created for the
comparison data.
•

Estimates of lifetime stimulant misuse based on the standard definition including
methamphetamine were similar between the QFT and comparison data. For young
adults aged 18 to 25, however, the differences between the QFT estimate
(13.1 percent) and the comparison data estimates (9.5 percent in each dataset)
approached statistical significance (p = 0.064 for QFT vs. 2011; p = 0.058 for QFT
vs. 2012) (Table J-3).

•

The standard definition estimates of past year stimulant misuse in the QFT were
greater than the corresponding estimates in the comparison data for persons aged 12
or older and young adults aged 18 to 25 (Tables J-5 and J-7). Among young adults in
particular, the standard definition estimates for past year misuse were 9.1 percent for
the QFT, 3.2 percent for the 2011 comparison data, and 3.8 percent for the 2012
comparison data.

•

Estimates of stimulant misuse in the past month based on the standard definition were
greater in the QFT than in the 2011 comparison data for persons aged 12 or older
(0.8 vs. 0.4 percent) (Table J-9). The prevalence of stimulant misuse in the past
month based on the standard definition also was greater for persons aged 18 to 25 in
the QFT (2.7 percent) than in the 2011 or 2012 comparison data (1.0 percent in each
year) (Table J-11). The difference in the past month prevalence for persons aged 12
or older between the QFT and the 2012 comparison data (0.4 percent) approached
statistical significance (p = 0.053) (Table J-9).

For the QFT, statistical tests were not conducted between estimates of stimulant misuse
based on the standard definition that included methamphetamine and the QFT definition that did
not include methamphetamine. Nevertheless, these data provide some indication of the potential
effect if methamphetamine use is no longer included in estimates of stimulant misuse in 2015
and beyond.
•

Estimates of lifetime stimulant misuse in the QFT for persons aged 12 or older were
9.0 percent for the standard definition that included methamphetamine and
3.9 percent for the QFT definition that did not include methamphetamine (Table J-1).

•

An estimated 9.1 percent of persons aged 26 or older were lifetime misusers of
stimulants based on the standard definition, and 2.9 percent were lifetime misusers
based on the QFT definition (Table J-4). Among young adults aged 18 to 25,
estimates of lifetime stimulant misuse based on the standard and QFT definitions
were 13.1 and 11.0 percent, respectively (Table J-3). Among adolescents aged 12 to
17, the estimates were 2.2 percent for the standard definition and 1.9 percent for the
QFT definition (Table J-2).

•

Among persons aged 12 or older, the standard definition estimate of past year
stimulant misuse for the QFT was 2.1 percent, and the QFT definition estimate was

234

1.8 percent (Table J-5). Data for other age groups followed a similar pattern. Among
young adults aged 18 to 25, for example, the standard definition estimate for the QFT
was 9.1 percent, and the QFT estimate was 8.9 percent (Table J-7).
•

The standard definition estimate in the QFT for past month stimulant misuse among
persons aged 12 or older was 0.8 percent, and the QFT definition estimate was
0.5 percent (Table J-9).

Inclusion of data for Adderall® had relatively little effect on whether differences in
lifetime stimulant misuse between the QFT and comparison data were statistically
significant (Table J-13). Among adolescents aged 12 to 17, the estimates of lifetime
stimulant misuse based on the standard definition were not significantly different
between the QFT and comparison data. However, the standard definition plus
noncore Adderall® estimate for this age group in the 2011 comparison data was
greater than the QFT standard definition estimate (3.6 vs. 2.2 percent). The difference
between the QFT and 2012 estimate that included Adderall® (3.5 percent) also
approached statistical significance (p = 0.061).

•

Among young adults aged 18 to 25, differences between the QFT and both the 2011
and 2012 comparison estimates for the standard definition of lifetime stimulant
misuse approached statistical significance (p = 0.064 and p = 0.058, respectively).
In contrast, the standard definition estimate of lifetime misuse among young adults in
the QFT (13.1 percent) was not significantly different from either estimate in the
comparison data that included Adderall® (2011: 15.4 percent; 2012: 16.0 percent),
nor did these differences approach statistical significance (Table J-13).

•

For persons aged 12 or older and young adults aged 18 to 25, inclusion of data for
Adderall® appeared to erase the differences in the prevalence of past year misuse that
were observed between the QFT and comparison data for the standard definition
estimates (Table J-14). Among persons aged 18 to 25, for example, the estimates of
past year stimulant misuse in the 2011 and 2012 comparison data that included
noncore Adderall® data (6.3 and 7.0 percent, respectively) were not significantly
different from the QFT estimate based on the standard definition (9.1 percent);
however, the difference between the QFT and 2011 comparison data approached
statistical significance (p = 0.097). Without the Adderall® data, the estimates of past
year misuse of stimulants in this age group were 3.2 percent in the 2011 comparison
data and 3.8 percent in the 2012 comparison data.

235

•

Among persons aged 12 or older, the standard definition estimate of past month
stimulant misuse was greater in the QFT (0.8 percent) than in the 2011 comparison
data (0.4 percent) and approached statistical significance relative to the estimate of
0.4 percent for the 2012 comparison data (p = 0.053) (Table J-15). In contrast, the
comparison data estimates for 2011 and 2012 that included noncore Adderall® data
(0.6 percent in each dataset) were similar to the standard definition estimate in the
QFT.

•

Among young adults aged 18 to 25, the estimates of past month stimulant misuse that
included Adderall® were 1.9 percent in the 2011 comparison data and 2.0 percent in
the 2012 comparison data (Table J-15). These estimates were not significantly
different from the past month estimate for young adults in the QFT based on the
standard definition (2.7 percent). In contrast, the estimates of past month misuse in
this age group based on the standard definition were 1.0 percent in each year of the
comparison data and were lower than the corresponding QFT estimate.

Although the estimates of past year misuse of stimulants based on the standard definition
(i.e., including methamphetamine) were greater in the QFT than in the comparison data for
persons aged 12 or older and for young adults aged 18 to 25, these differences no longer
remained when noncore Adderall® data were included in the CPN estimates for the comparison
data. These findings underscore the likely importance of including questions about Adderall® for
estimating misuse of prescription stimulants.

7.4

Effects of Methamphetamine and Prescription Drugs on Illicit Drug Use
Estimates

As noted in Section 6.4 in Chapter 6, the measures of use of any illicit drug and illicit
drugs other than marijuana in current published NSDUH estimates include use of
methamphetamine and misuse of prescription drugs. The changes to the methamphetamine and
prescription drug questions that were summarized in Section 7.1 for the QFT (and, by extension,
for the redesigned questionnaire in 2015) also could affect estimates for these other summary
measures of illicit drug use.
Therefore, alternate measures of use of any illicit drug and illicit drugs other than
marijuana were created that did not include data for methamphetamine or prescription drugs (see
Appendix H). Estimates based on these alternate measures are presented in Chapter 6 and in the
detailed tables in Appendix I.
A third alternate definition for any illicit drug use was developed that included
methamphetamine but did not include prescription drugs (subsequently referred to as Alternate
Definition 3). In addition, measures of use of illicit drugs and illicit drugs other than marijuana
were created based on the standard NSDUH definitions that included both methamphetamine and
prescription drugs. Estimates based on Alternate Definition 3 for illicit drug use and the standard
definitions are presented in this section and in the detailed tables in Appendix J.
•

Estimates of lifetime use were not significantly different between the QFT and the
comparison data for persons aged 12 or older, adults aged 18 to 25, and adults aged

236

26 or older for the illicit drug Alternate Definition 3 or for the standard definitions of
use of illicit drugs or illicit drugs other than marijuana (Tables J-1, J-3, and J-4).

7.5

•

Among adolescents aged 12 to 17, the Alternate Definition 3 estimate of lifetime use
of illicit drugs was greater in the QFT (26.7 percent) than in the 2011 or 2012
comparison data (22.4 and 20.1 percent, respectively) (Table J-2). The standard
definition estimates in the QFT for lifetime use of illicit drugs (28.5 percent) and
illicit drugs other than marijuana (19.1 percent) also were greater than the
corresponding estimates in the 2012 comparison data (23.4 and 14.1 percent,
respectively).

•

As for the lifetime period, estimates of past year use of illicit drugs based on the
standard definition or Alternate Definition 3 were not significantly different between
the QFT and comparison data for persons aged 12 or older, but did differ between the
QFT and 2012 comparison data for adolescents aged 12 to 17 (Tables J-5 and J-6).
For adolescents, the standard definition estimate of past year illicit drug use was
20.6 percent, and the Alternate Definition 3 estimate was 18.2 percent. Corresponding
estimates in the 2012 comparison data were 16.6 and 14.2 percent, respectively.

•

The estimates of use of illicit drugs other than marijuana in the past year based on the
standard definition were greater in the QFT than in the 2011 or 2012 comparison data
for persons aged 12 or older and young adults aged 18 to 25 (Tables J-5 and J-7).
Among young adults, the estimates were 25.3 percent for the QFT, 17.7 percent for
the 2011 comparison data, and 17.9 percent for the 2012 comparison data. The
difference between the estimates for illicit drugs other than marijuana among 12 to 17
year olds in the QFT (11.6 percent) and the 2012 comparison data (8.3 percent) also
approached statistical significance (p = 0.064) (Table J-6).

•

Most estimates of past month use of illicit drugs or illicit drugs other than marijuana
did not differ significantly between the QFT and comparison data, regardless of the
definitions. Among adolescents aged 12 to 17, however, the estimate of use of illicit
drugs other than marijuana based on the standard definition was lower in the QFT
than in the 2011 comparison data (2.5 vs. 4.0 percent) (Table J-10). The difference in
standard definition estimates for past month use of illicit drugs other than marijuana
among young adults aged 18 to 25 in the QFT (9.0 percent) and 2012 comparison
data (6.6 percent) also approached statistical significance (p = 0.072) (Table J-11).

Methamphetamine, Prescription Drug, and Illicit Drug Estimation
Issues to Consider for the 2013 Dress Rehearsal and 2015 Redesign

This section highlights findings from Sections 7.2 to 7.4. Particular attention is given to
findings that have implications for the 2013 DR in 2013 and estimates from the redesigned
questionnaire for the 2015 survey, including implications for reporting trends in drug use or
misuse.
7.5.1 Methamphetamine
Although past year and past month estimates of methamphetamine use did not differ
significantly between the QFT and comparison data, the estimate of lifetime use for persons aged
237

12 or older was greater in the QFT than in the comparison data. Estimates by age group suggest
that this difference was largely being driven by patterns of lifetime use among adults aged 26 or
older.
In contrast, published NSDUH trend data indicate that the prevalence of lifetime
methamphetamine use among persons aged 12 or older decreased from 6.5 percent in 2002 to
4.6 percent in 2011 (Center for Behavioral Statistics and Quality [CBHSQ], 2012e). The estimate
of lifetime use from the 2012 QFT for persons aged 12 or older was the same as the point
estimate in 2002. As was noted in Section 7.2, inclusion of additional questions about
methamphetamine in a supplemental section of the main survey since 2005 may not fully capture
reports of methamphetamine use from respondents who do not think of this drug in the context of
questions about prescription stimulants.
If the prevalence of lifetime methamphetamine use is higher than in recent years for
persons aged 12 or older or within different age groups because of changes to the questionnaire
in 2015, SAMHSA will need to decide how to handle the reporting of trends in lifetime use.
One option would be not to report trend data for lifetime methamphetamine use between 2015
and earlier years or to discontinue the reporting of lifetime trend data for methamphetamine
altogether from 2015 onward. Alternatively, SAMHSA could start a new baseline for lifetime
methamphetamine use beginning in 2015. Other, more sophisticated options could involve
statistical procedures to adjust the trend data for 2002 to 2014.
Although data on trends in lifetime prevalence may be of interest for examining historical
changes in the popularity of different drugs, data on trends in the prevalence of
methamphetamine use in the past year and past month are likely to be of more importance to
policymakers, the public health sector, the criminal justice sector, and others because of the
demands that methamphetamine users may place on the criminal justice system, the health care
delivery system (including substance abuse treatment), and systems for providing social services
(including services to dependents of adult substance users). The prevalence of methamphetamine
use in the past year among persons aged 12 or older has remained fairly stable since 2008, at
0.3 to 0.5 percent. The prevalence of past month methamphetamine use among persons aged 12
or older also has remained fairly stable since 2007, at 0.1 to 0.2 percent. Similar trends for past
year and past month use are observed for most age groups (CBHSQ, 2012e).
If trends in past year and past month use of methamphetamine continue to remain fairly
stable based on NSDUH data for 2012 to 2014, then moving the methamphetamine questions to
a separate module in 2015 may not disrupt the trend data for past year and past month use.
Because of the relatively small number of QFT respondents, however, it cannot be established
conclusively that these findings from the QFT will translate to similar relationships between
estimates in 2014 and 2015. Advance monitoring of estimates of methamphetamine use from the
2015 survey (e.g., based on the first two quarters of data) will be important for anticipating
potential disruptions in the trend data because of the changes to the methamphetamine questions
in 2015.

238

7.5.2 Prescription Drugs
The general findings of lower estimates of lifetime misuse of prescription drugs but
higher past year estimates in the QFT relative to the comparison data are expected, given the
changes to the prescription drug questions for the QFT. The structure of the current questionnaire
provides respondents with multiple opportunities to report lifetime misuse of specific
prescription drugs but less opportunity to report past year misuse. This situation was reversed for
the QFT, with respondents having more opportunity to report past year misuse of specific
prescription drugs and limited opportunity to report misuse of any prescription drugs that
occurred more than 12 months prior to the interview—including misuse of prescription drugs
that are no longer available by prescription in the United States.
A notable finding for the lifetime estimates was that most estimates of lifetime misuse of
tranquilizers were lower in the QFT than in both sets of comparison data. Some lifetime
estimates of misuse in the QFT were lower than in the comparison data for other prescription
drug categories, but not to the extent of the differences that were observed for tranquilizers.
As noted in Section 7.3, however, estimates of lifetime misuse for other prescription drug
categories were in the direction of being lower in the QFT than in the comparison data but did
not meet the criteria for statistical significance. The QFT sample of only 2,044 respondents may
not have allowed sufficient statistical power to detect additional differences in lifetime misuse. If
the prescription drug modules for the 2013 DR undergo minimal or no change relative to the
modules in the QFT, then the prescription drug data from the 2012 QFT and 2013 DR could be
combined to increase the sample size for further analysis.
Nevertheless, these findings support the conclusion to start a new baseline in 2015 for
trends in prescription drug misuse. It also may be useful for SAMHSA to consider whether to
discontinue reporting trend data for lifetime misuse of prescription drugs after 2014 because of
questions about the accuracy of respondent self-reports of misuse of prescription drugs more
than 12 months prior to the interview.
Principally because of scheduling issues for analyzing and reporting of QFT data to
inform SAMHSA's decision making for the 2013 DR, QFT data on initiation of misuse in the
past year were not analyzed. As noted in Section 4.6.5.4, however, changes to the questions in
the QFT for initiation of misuse of prescription drugs have important implications for measuring
and estimating initiation for prescription drugs in 2015 and beyond. These changes also may
have implications for measuring and estimating initiation of illicit drug use in general. In the
QFT, the following numbers of respondents provided valid data for their age at first misuse of at
least one prescription drug in the overall category: 144 for pain relievers, 71 for tranquilizers,
56 for stimulants, and 18 for sedatives. Therefore, the QFT sample size would be adequate for
conducting further analysis of the initiation data for pain relievers. SAMHSA could investigate
the initiation data in the 2012 QFT and 2013 DR for pain relievers, tranquilizers, stimulants, and
any prescription drug to examine this issue further. If similar numbers of 2012 QFT and 2013
DR respondents provide initiation data for the misuse of sedatives, the number of respondents in
the combined 2012 QFT and 2013 DR data still would not be adequate for analyzing the
initiation data for sedatives.

239

7.5.3 Illicit Drugs
Many estimates of the use of illicit drugs or the use of illicit drugs other than marijuana
were not significantly different between the QFT and comparison data when data for
methamphetamine or prescription drugs (or both) were included in the QFT estimates.
Nevertheless, some estimates were affected, especially for adolescents aged 12 to 17 and young
adults aged 18 to 25. However, changes to the methamphetamine and prescription drug use
questions were not the only changes made to the questionnaire for the QFT. In particular,
changes also were made to the hallucinogens and inhalants modules in the QFT that could affect
estimates of the use of illicit drugs and illicit drugs other than marijuana (see Section 2.4.1 and
Chapter 6). Therefore, additional analysis of 2012 QFT and 2013 DR data (including combined
2012 QFT and 2013 DR data, where applicable) will be important for assisting SAMHSA in
deciding how to create these summary illicit drug use measures in 2015 and how to report trends
for these measures.

240

8. QFT Estimates Compared with NSDUH
Estimates: Noncore Items
8.1

Overview of QFT Estimates Compared with NSDUH Estimates for
Noncore Items

This chapter summarizes Questionnaire Field Test (QFT) estimates compared with the
2011 comparison estimates and the 2012 quarters 3 and 4 comparison estimates for selected
noncore items. Section 8.2 describes the estimates for substance dependence and abuse.
Section 8.3 presents estimates for the needle use items. Section 8.4 examines comparisons of
medical marijuana reports by State in reference to the current laws in each State. Section 8.5
describes selected estimates for the noncore demographic and household items. Section 8.6
presents estimates for selected items subject to context effects due to the questionnaire redesign.
Section 8.7 discusses estimates for new, revised, and moved items in the QFT instrument,
including how QFT estimates for moved items align with the 2011 and 2012 quarters 3 and 4
comparison estimates. The chapter concludes with Section 8.8, which provides a comparison of
the distribution of relationships for proxy respondents and estimates for selected items based on
the proxy report status.

8.2

Estimates for Substance Dependence and Abuse

Estimates of substance dependence and abuse were examined for the QFT and
comparison data for 2011 and 2012 based on the following changes to the QFT questionnaire
that had the potential to affect estimation:
•

The focus of the prescription drug modules shifted to use and misuse of specific
prescription drugs in the past 12 months rather than the lifetime period.

•

The introductions to questions for prescription drugs in the substance dependence and
abuse module were changed to reflect the revised definition of misuse in the QFT.

•

Additional questions that captured information about specific past year use or misuse
of hallucinogens (e.g., Ecstasy), prescription stimulants (e.g., Adderall®), and
prescription sedatives (e.g., Ambien®) that were in a supplemental section of the
interview in the main survey were moved to the respective core modules.

•

A new methamphetamine module was added to the core drug modules, and separate
questions about methamphetamine dependence or abuse were included in the
substance dependence and abuse module. The redesigned stimulants module no
longer includes questions related to the use of methamphetamine.

•

Respondents who reported past year use of methamphetamine but not past year
misuse of prescription stimulants were not asked questions about stimulant
dependence or abuse.

•

Although the question for most recent use of inhalants was not changed for the QFT,
new questions were included about lifetime use of two additional inhalants.
241

In particular, as noted in Section 7.3 in Chapter 7, the shift in emphasis in the QFT from
a lifetime to a past year period for capturing data on misuse of specific prescription drugs
resulted in many estimates of prescription drug misuse in the past year being higher in the QFT
than in the comparison data for 2011 and 2012. In turn, the increased reporting of past year
misuse of prescription drugs in the QFT could yield higher estimates of dependence or abuse for
prescription drugs. Estimates of dependence or abuse for prescription stimulants could be
affected because QFT respondents who reported past year use of methamphetamine but not past
year misuse of prescription stimulants were not asked these questions for stimulants.
This section presents findings on substance dependence and abuse from the comparison
data for 2011 and quarters 3 and 4 of 2012 and from the QFT. Detailed tables containing these
estimates are included in Tables K-1 to K-4 in Appendix K.
The computer-assisted interviewing (CAI) instrumentation for both the main survey and
the QFT for the National Survey on Drug Use and Health (NSDUH) included questions that
were designed to measure alcohol and illicit drug dependence and abuse. Dependence and abuse
questions were based on the criteria in the Diagnostic and Statistical Manual of Mental
Disorders, 4th edition (DSM-IV) (American Psychiatric Association [APA], 1994). Additional
details about measurement of substance dependence and abuse in NSDUH are provided in the
public use file codebook for the 2011 NSDUH and in the 2011 report on national findings
(Center for Behavioral Health Statistics and Quality [CBHSQ], 2012c, 2012e). Information on
measures of dependence and abuse used in this report also is included in Appendix H of this
report, particularly for the development of measures for methamphetamine dependence and
abuse.
In both the main study and the QFT, persons are defined as having abuse if they met one
or more of the four criteria for abuse included in the DSM-IV, and the definition of dependence
was not met for that substance. For measurement of abuse that encompasses multiple drug
categories (e.g., prescription drugs), respondents who were defined as having abuse met the
criteria for abuse for at least one drug (or drug category) but did not meet the criteria for
dependence for any of the drugs or categories that were included. For example, a respondent who
met one or more criteria for prescription pain reliever abuse but did not meet the criteria for pain
reliever dependence would be defined as having pain reliever abuse. However, if this respondent
with pain reliever abuse but not dependence met the criteria for dependence for another
prescription drug category (e.g., tranquilizers), then he or she would be defined as having
dependence for any prescription drug and by definition would not be defined as having abuse for
any prescription psychotherapeutic drug. Consequently, this respondent would be defined as
having abuse for pain relievers but also as having dependence for prescription drugs as a whole.
Therefore, estimates of abuse for some drugs (or groups of drugs) within a broader category
(e.g., pain relievers within the broader category of prescription drugs as a whole) could be larger
than the estimate for abuse for the more aggregated category (e.g., prescription drugs).
•

For persons aged 12 or older in Table K-1 and for each of the age groups in
Tables K-2 to K-4, there were no significant differences in estimates of illicit drug
dependence, illicit drug abuse, or illicit drug dependence or abuse between the QFT
and corresponding estimates from the 2011 or 2012 comparison data. There also were
no significant differences in estimates of dependence, abuse, or dependence or abuse
242

between the QFT and comparison samples for marijuana, prescription drugs,
prescription pain relievers, illicit drugs other than marijuana, or illicit drugs excluding
marijuana28 among persons aged 12 or older. Although differences between the QFT
and the 2011 or 2012 comparison data for these estimates were not statistically
significant by age group, some differences approached conventional significance
levels.
•

The estimate of hallucinogen dependence among persons aged 12 or older was less
than 0.05 percent based on the QFT data and was significantly different from the
corresponding estimate of 0.1 percent in the 2011 comparison data (Table K-1).
However, the estimate of hallucinogen dependence in the 2012 comparison data also
was less than 0.05 percent and was not significantly different from the QFT estimate.

•

Estimates for adolescents aged 12 to 17 were lower in the QFT than in the 2011
comparison data for prescription drug dependence or abuse (0.2 vs. 1.2 percent), pain
reliever dependence or abuse (0.2 vs. 1.0 percent), and dependence or abuse for illicit
drugs other than marijuana (0.8 vs. 1.7 percent) (Table K-2). In addition, the
difference between the estimates for prescription drug dependence or abuse among
adolescents in the QFT (0.2 percent) and the 2012 comparison data (0.5 percent)
approached statistical significance (p = 0.086). No adolescents in the QFT were
defined as having dependence for pain relievers or abuse for prescription drugs.

•

Among adults aged 26 or older, estimates were lower in the QFT than in the 2012
comparison data for prescription drug dependence (0.2 vs. 0.6 percent), dependence
for illicit drugs other than marijuana (0.4 vs. 0.9 percent), and dependence or abuse
for illicit drugs other than marijuana (0.6 vs. 1.2 percent) (Table K-4).

•

For the QFT and 2011 comparison data, the difference between the estimate of
prescription drug dependence among adults aged 26 or older approached statistical
significance (0.2 and 0.5 percent, respectively; p = 0.078). The following differences
between the QFT and 2012 comparison data for adults aged 26 or older also
approached statistical significance: illicit drug dependence (0.9 and 1.1 percent;
p = 0.087); pain reliever dependence (0.2 and 0.5 percent; p = 0.077); dependence for
illicit drugs excluding marijuana (0.4 and 0.8 percent; p = 0.055); and dependence or
abuse for illicit drugs excluding marijuana (0.6 and 1.0 percent; p = 0.088).

•

Additional estimates for dependence, abuse, or dependence or abuse in the QFT
would have been suppressed but were lower than in one or both comparison datasets
for persons aged 12 to 17 (Table K-2), those aged 18 to 25 (Table K-3), or those aged
26 or older (Table K-4). For example, suppressed QFT estimates for adolescents aged
12 to 17 were significantly different from estimates in the 2011 or 2012 comparison
data for pain reliever dependence, hallucinogen abuse, and prescription drug abuse.
However, statistically significant differences typically are not reported if one or both
estimates is suppressed.

Estimates for illicit drugs excluding marijuana included dependence or abuse for cocaine, heroin,
hallucinogens, inhalants, or prescription psychotherapeutic drugs but also required persons not to have dependence
or abuse for marijuana.

243

•

Only 12 QFT respondents were asked questions about methamphetamine dependence
or abuse because they reported past year use in the core methamphetamine module.
Consequently, no QFT respondents were defined as having methamphetamine
dependence.

Lower QFT dependence and abuse estimates discussed in this section for any prescription
drug and pain relievers for some age groups relative to estimates in the comparison data are
counterintuitive, given the higher estimates of past year misuse in the QFT (see Chapter 7 and
Appendix J). That is, respondents who reported past year misuse of any prescription drug within
a given category (e.g., past year misuse of any pain reliever) were routed into the corresponding
questions for dependence or abuse in both the QFT and main survey. Therefore, higher estimates
of past year misuse in the QFT could correspond to more respondents reporting misuse in the
QFT than in the comparison data. If that is the case, more respondents in the QFT than in the
comparison data would have had the opportunity to report symptoms of dependence or abuse
attributable to their past year misuse of prescription drugs within a given category. Furthermore,
the dependence and abuse estimates for prescription drugs and pain relievers were not
significantly different between the QFT and comparison data. These findings suggest that the
smaller QFT sample size and its effect on the numbers of respondents who reported sufficient
numbers of problems to be classified with dependence or abuse for prescription drugs could have
contributed to the observed differences within age groups.
However, an alternative explanation for these dependence or abuse findings for
prescription drugs is that the respondent burden involved in answering the questions about past
year misuse of prescription drugs in the QFT could have suppressed reporting of dependence or
abuse symptoms for prescription drugs. As noted in Section 4.5.1 in Chapter 4, when
respondents reported use and misuse of prescription drugs, the QFT timings exceeded those for
the 2011 and 2012 comparison samples, with the greatest difference occurring among adults
aged 26 or older. Consequently, some QFT respondents who reported past year misuse of one or
more prescription drugs could have been prone to answer the dependence and abuse questions as
"no" to reduce the number of additional questions they were asked. These findings for
prescription drug dependence or abuse will be examined further in the analysis of data from the
Dress Rehearsal (DR), including analysis of combined data from the QFT and the DR, where
applicable.
Findings of no significant differences between the estimates in the QFT and comparison
data for any illicit drug dependence, illicit drug abuse, and illicit drug dependence or abuse may
be driven by the contributions of marijuana dependence or abuse to these estimates. The
marijuana module for the QFT did not change relative to the module in the main study, and no
changes to this module are planned as part of the redesign of the questionnaire in 2015. If similar
findings for illicit drug dependence or abuse estimates are observed once the DR data are
available, then these findings could suggest that questionnaire changes in 2015 will not
appreciably affect substance use disorder (i.e., dependence or abuse) trends for any illicit drug.
However, if substance use disorders for prescription drugs—especially prescription pain
relievers—contribute more substantially to estimates of substance use disorders for illicit drugs
other than marijuana, then changes to the prescription drug modules in 2015 could affect
dependence or abuse trends for illicit drugs other than marijuana. The relatively small QFT
sample size and the corresponding lack of statistical significance for most comparisons do not
244

ensure that no differences will be observed for dependence and abuse estimates in 2015. Again,
analysis of DR data will provide further opportunity to explore potential effects of the redesign
on these estimates for illicit drugs other than marijuana. Analysis of data from the first two
quarters of 2015 also can assist the Substance Abuse and Mental Health Services Administration
(SAMHSA) in anticipating any effects on dependence or abuse trends for illicit drugs other than
marijuana and for prescription drugs.

8.3

Estimates for Needle Use Items

Specific questions about use of a needle to inject heroin and to inject cocaine in the QFT
were unchanged relative to the main survey. However, the addition of the new methamphetamine
module to the core drug modules in the QFT could affect the number of respondents who were
asked questions about use of methamphetamine with a needle. Also, QFT questions about use of
prescription stimulants with a needle were moved from the supplemental special drugs module to
the core stimulants module and focused on use of stimulants with a needle in the past year or past
month, but not lifetime use of stimulants with a needle.
In addition, the order and context for questions about needle use differed between the
QFT and the main survey, although the question wordings were the same for use of heroin or
cocaine with a needle. In the QFT, all respondents first were asked questions in the noncore
special drugs module about use of over-the-counter (OTC) cough and cold medicines to get high.
QFT respondents who reported lifetime use of OTC cough and cold medicines to get high were
asked to report their most recent use, and those who reported use at some point in the past
12 months were asked to specify the names of up to five OTC medicines that they used in the
past 12 months to get high. Following the question(s) about OTC cough and cold medicines,
QFT respondents were asked about their lifetime use of gamma hydroxybutyrate (GHB), and if
applicable, their most recent use of GHB. Depending on whether they reported lifetime use, QFT
respondents then were asked questions about needle use or other drug use behaviors in the
following order: (a) use of cocaine with a needle;29 (b) smoking heroin; (c) sniffing or "snorting"
heroin; (d) use of heroin with a needle;30 (e) use of methamphetamine with a needle; (f) use of
any other drug with a needle (or any drug with a needle if respondents did not report use of
cocaine, heroin, or methamphetamine with a needle); and (g) if applicable, needle use behaviors
the last time that respondents injected drugs (e.g., reuse of a needle they had used before, sharing
of needles).
In the main survey, depending on reports of lifetime use or misuse in the corresponding
core modules, respondents first were asked about their behaviors associated with (a) heroin use
(i.e., smoking, sniffing, or injection); (b) use of methamphetamine with a needle (i.e., if
respondents had previously reported methamphetamine use in the core stimulants module) or
methamphetamine use in general (i.e., if respondents had not reported methamphetamine use in
the core stimulants module); (c) use of (other) stimulants with a needle, and (d) use of cocaine
with a needle. All main survey respondents then were asked whether they ever used a needle to
inject any drug (or any other drug), and needle users were asked about their needle use the last
29

Respondents also were asked questions about the most recent time they engaged in a particular behavior
(e.g., use of cocaine with a needle) if they reported engaging in that behavior in their lifetime.
30
Respondents in both the QFT and main survey who reported lifetime use of heroin but did not report
smoking, sniffing, or injecting it were asked follow-up questions to determine how they used heroin.

245

time they injected drugs. Questions about use of GHB and use of cough and cold medicines to
get high were asked later in the special drugs module (i.e., after the questions about needle use).
Because of these differences, this section presents findings on injection drug use (i.e., use
of a needle to inject drugs) from the comparison data for 2011 and quarters 3 and 4 of 2012 and
from the QFT. Estimates for persons aged 12 or older are shown in Table K-5 in Appendix K.
Estimates of needle use by age group are not presented because of the low prevalence of needle
use in the general population. In 2011, for example, 0.7 percent of persons aged 12 or older had
ever used a needle to inject heroin, 0.8 percent had ever used a needle to inject cocaine, and
0.5 percent had ever used a needle to inject methamphetamine; among adolescents aged 12 to 17,
the lifetime needle use estimates for these three drugs were 0.1 percent or less (CBHSQ, 2012e).
Therefore, the QFT sample could not support estimates of needle use by age group, especially
for the past year and past month periods. Because of the changes to the questions for use of
stimulants with a needle that were described previously, estimates for use of prescription
stimulants with a needle and use of heroin, cocaine, methamphetamine, or prescription
stimulants with a needle are presented in Table K-5 only for the past year and past month.
•

Lifetime estimates of needle use among persons aged 12 or older were similar
between the QFT and the 2011 and 2012 comparison data. Lifetime estimates for use
of heroin with a needle were 0.7 percent for the QFT and 0.8 percent in the 2011 and
2012 comparison data. Estimates for use of cocaine with a needle were 1.0 percent for
the QFT and 0.8 percent in each comparison dataset. Lifetime estimates of
methamphetamine use with a needle ranged from 0.6 to 0.8 percent in the QFT and
comparison data.

•

Percentages of persons in the 2011 and 2012 comparison data who used a needle to
inject heroin, cocaine, methamphetamine, prescription stimulants, or any of these
drugs in the past year or past month were 0.1 percent or less. No QFT respondents
reported past year or past month use of cocaine or prescription stimulants with a
needle.

•

Estimates of use of a needle to inject any of these four drugs (i.e., heroin, cocaine,
methamphetamine, or prescription stimulants) with a needle were similar between the
QFT and the 2011 and 2012 comparison data. Past year estimates for use of any of
these drugs with a needle were 0.2 percent in the QFT and both comparison datasets,
and past month estimates were 0.1 percent in each of these three datasets.

Two-year trends (e.g., 2010 and 2011) in the lifetime prevalence of needle use are
presented in the NSDUH detailed tables (CBHSQ, 2012d). On the one hand, findings from
Table K-5 suggest that planned changes to the questionnaire in 2015 will not affect the 2-year
trends for heroin, cocaine, or methamphetamine between 2014 and 2015. However, continued
investigation of needle use estimates with data from the DR will be useful using the combined
QFT and DR data. Also, changes to the questions for injection of stimulants could require
creation of new trend data for 2002 to 2015 for lifetime use of a needle to inject cocaine, heroin,
or methamphetamine (i.e., without data on use of stimulants with a needle). Because of the
decision to ask about use of stimulants with a needle only for the past year or past month periods
in the redesigned questionnaire, estimates for injection of stimulants that are presented in
NSDUH detailed tables would require establishment of a new baseline in 2015.
246

8.4

Comparisons of Medical Marijuana Reports by State in Reference to
Current State Laws

To examine how reports of using marijuana for medical purposes aligned with the current
State laws where respondents reported use, responses to question MJMM on the medical use of
marijuana, which was added to the blunts module of the QFT questionnaire, were examined by
State. Overall, a total of 15 QFT respondents answered question MJMM affirmatively, indicating
that at least some of their marijuana use in the past year was allegedly recommended by a
doctor. Of these 15 respondents, 7 respondents reported living in a State that had a medical
marijuana law in effect in 2012 (not counting Massachusetts).31 The remaining 8 respondents did
not live in States that had a medical marijuana law in effect in 2012.
Because question MJMM asks about use in the past 12 months, some or all of the
8 respondents who reported use of marijuana for medical purposes in States that did not have a
medical marijuana law in effect in 2012 could have been referring to prior use in the past year in
a different State with a medical marijuana law in effect. For this reason, question QD13a in the
back-end demographics about moves in the past year was examined to determine whether any of
these 8 respondents had lived 1 year prior to the interview date in a State with a medical
marijuana law. Adding this check to the analysis did not identify any additional respondents who
were living in a State with a medical marijuana law 1 year prior to their QFT interview.
One further possibility is that the reports of using marijuana for medical purposes from
the 8 respondents who did not live in States that had a medical marijuana law in effect in 2012
reflected access to marijuana in neighboring States that had a medical marijuana law. Each of
these 8 respondents lived in States that border at least one State that had a medical marijuana law
in effect in 2012. Table 8.1 shows the current State of residence for each of these respondents
and the current or former bordering States with a medical marijuana law in effect in 2012.
Table 8.1 Current State of Residence without a Medical Marijuana Law in Effect and Current or
Former Bordering States with Medical Marijuana Laws in Effect for Eight QFT
Respondents Reporting Medical Use of Marijuana
Respondent Reporting
Medical Use of Marijuana
1
2
3
4
5
6
7
8
1

Respondent's Current State of
Residence without Medical
Marijuana Laws
Indiana
Maryland
New York
North Carolina
Ohio
Oklahoma
Pennsylvania
Wisconsin

Bordering States to Respondent's
Current or Prior State of Residence with
Medical Marijuana Laws
Michigan
Delaware, District of Columbia
Connecticut, New Jersey, Vermont
Michigan1
Michigan
New Mexico, Colorado
Delaware, Maryland, New Jersey
Michigan

This respondent reported in question QD13 residing in Indiana 1 year prior to the QFT interview.

A ballot initiative allowing use of marijuana for medical reasons was approved in Massachusetts in
November 2012 but did not take effect until January 2013.

247

Overall estimates for the medical use of marijuana are presented in Table M-1 in
Appendix M. Given that question MMJM was included in the 2013 main study instrument, early
review of the 2013 data (including analysis of data from the first two quarters of 2013) will allow
for an examination of the alignment between reports of using marijuana for medical purposes
with the current State laws where respondents report use for a larger number of respondents and
States.

8.5

Estimates for Noncore Demographic and Household Items

This section examines whether QFT estimates of selected demographic and household
items differed from the 2011 and 2012 quarters 3 and 4 comparison estimates. A notable change
in the QFT instrument was moving questions on health insurance coverage and family income
from interviewer administration using computer-assisted personal interviewing (CAPI) to
self-administration using audio computer-assisted self-interviewing (ACASI). As a result, some
differences could be observed on these demographic items between the QFT estimates and the
2011 and 2012 quarters 3 and 4 comparison estimates if QFT respondents systematically
answered these items differently in ACASI mode.
Estimates for selected demographic and household items for each of the three datasets are
presented in Appendix K. Tables K-6 through K-13 provide estimates for demographic and
household items for all persons aged 12 or older, adolescents aged 12 to 17, young adults aged
18 to 25, and adults aged 26 or older, respectively. Demographic questions that were not asked
for specific age groups are indicated by "N/A" ("not applicable ") in these tables. For example,
in Table K-7, education is indicated to be "N/A." NSDUH national estimates by education are
limited to adults aged 18 or older because most adolescents aged 12 to 17 would not have
finished high school based on their age.
For most demographic and household items, the estimates from the QFT data were
similar to the 2011 and 2012 quarters 3 and 4 comparison estimates. The majority of differences
observed indicated that the QFT sample members were associated with lower socioeconomic
status. For example, the QFT estimates for participating in government programs such as food
stamps were significantly higher than those for the 2011 and 2012 quarters 3 and 4 comparison
data. Differences in missingness rates and estimates for items that were most highly correlated
with socioeconomic status could have been affected by these observed differences in
socioeconomic status between the QFT sample and the 2011 and 2012 quarters 3 and 4
comparison samples. Given that the noncore demographic and household questions were
administered via ACASI for QFT respondents and via CAPI for 2011 and 2012 quarters 3 and 4
respondents, the effects of this mode difference cannot be disentangled from the effects of
differences in socioeconomic status. It is also not clear how much these differences can be
attributed to differences in the samples, such as those produced by the differential response rates,
which were not accounted for by the QFT weighting process.
•

For all persons aged 12 or older (Table K-6), the estimate for participation in
government assistance programs was 32.2 percent for the QFT sample compared with
25.4 percent for the 2011 comparison sample and 26.4 percent for the 2012 quarters 3
and 4 comparison sample. The differences between the QFT estimate and the
estimates for the two comparison samples were statistically significant.
248

This difference between the QFT sample and the 2011 and 2012 quarters 3 and 4
comparison samples was also observed across all three age groups (Tables K-7
through K-9).
•

No differences were observed among the three datasets on receiving income from
social security or welfare payments for all persons aged 12 or older. However, QFT
estimates for receiving supplemental security income (SSI) and participating in food
stamp programs were higher than estimates from the 2011 comparison sample, but
not the 2012 quarters 3 and 4 comparison sample. For all persons aged 12 or older,
the QFT estimate of 68.6 percent for receiving income from wages was significantly
less than the estimate of about 82 percent for both the 2011 and 2012 quarters 3 and 4
comparison samples. This pattern of differences between the QFT sample and the
2011 and 2012 quarters 3 and 4 comparison samples was also repeated for the three
age groups.

•

One further difference for all persons aged 12 or older was that QFT respondents
were more likely than 2011 and 2012 quarters 3 and 4 respondents to use a proxy
reporter for demographic and household items. Among QFT respondents,
15.7 percent reported using a proxy compared with 13.7 percent among 2011
comparison sample respondents and 13.9 percent among 2012 quarters 3 and 4
comparison sample respondents.

•

Among adult respondents aged 18 or older, the QFT estimate for education level
differed significantly from the 2011 and 2012 quarters 3 and 4 samples. Table K-10
provides unweighted and weighted estimates for the QFT sample and the 2011 and
2012 quarters 3 and 4 comparison samples for (1) a four-category education variable,
(2) a four-category employment status variable, (3) four geographic regions, and
(4) three county types. This table was produced to provide a clearer sense of
differences between the QFT sample and the 2011 and 2012 quarters 3 and 4
comparison samples and how such differences could have affected key estimates.
Consistent with the results presented in Table K-6, Table K-10 shows that the QFT
estimates produced higher proportions for the less than high school and some college
categories, a lower proportion for the high school graduate category, and a slightly
lower proportion for the college graduate category. These differences were observed
both among the unweighted and weighted estimates.

•

Estimates for the four-category employment variable showed significantly different
employment patterns for the QFT sample versus the 2011 and 2012 quarters 3 and 4
samples, but only for the unweighted data. The two main differences observed in the
unweighted estimates were that the QFT sample produced a slightly higher proportion
for being employed full time (as opposed to part time) and a slightly lower proportion
for being unemployed. Weighting the estimates for employment status eliminated
statistically significant differences among the three samples.

•

Similar to the estimates for employment status, estimates of unweighted proportions
in one of four geographic regions—Northeast, Midwest, South, and West—differed
between the QFT sample versus the 2011 and 2012 quarters 3 and 4 samples.
Specifically, the QFT sample produced a slightly higher proportion for the South
region and a slightly lower proportion for the West region. Weighting the estimates

249

for geographic region eliminated statistically significant differences among the three
samples.
•

No significant differences among the unweighted or weighted data were observed
between the QFT sample versus the 2011 and 2012 quarters 3 and 4 samples with
respect to the distribution of proportions across large metropolitan, small
metropolitan, and nonmetropolitan counties.

The smaller sample size for the QFT makes it difficult to conclude whether estimates of
participation in government programs and receipt of specific types of income will change
significantly when the partially redesigned instrument and protocol are implemented in 2015.
The results for the demographic and household items discussed in this section suggested that the
following changes be made to some of these questions for the DR:
•

editing the ranges for height in feet and inches for accuracy in the height question;

•

increasing the upper weight limit in the weight question;

•

moving the definition of "immediate family" from the "Help" screen to the question
text in the military family questions, making other minor wording changes to these
questions for clarity, and adding an "Other, Specify" item to this series of questions;

•

removing the "Help" instructions in item QHI06 on private health insurance, and
moving key terms into the question itself;

•

deleting item QI05N on income from wages or pay, and adding this to the list of
potential sources of household income in the introductory item INTRTINN;

•

editing the wording of item QI03N on the receipt of SSI for accuracy;

•

editing the wording of item QI07N on the receipt of food stamps for accuracy; and

•

reordering the list of income sources in INTRTINN.

Regardless of whether any changes are made to the demographic and household questions for the
DR, differences noted between the QFT versus the 2011 and 2012 quarters 3 and 4 samples will
be reexamined for all of these estimates with the DR and 2012 and 2013 comparison data.

8.6

Estimates for Selected Items Potentially Subject to Context Effects Due
to Questionnaire Redesign

The introduction of new items in the questionnaire may lead to changes in estimates that
follow the new items due to context effects. Context effects may be said to take place between
two survey questions when a change introduced to the first (or contextual) item affects the
response process for the subsequent (target) item, which in turn may lead to a different response
than if the change had not been made. The potential presence of such effects cannot be
distinguished from changes in estimates due to the complete set of changes made to the QFT
survey protocol and questionnaire. Nevertheless, estimates for data collected in the QFT were
compared with data from the comparison samples for the following variables (shown in
parentheses).

250

•

The first variables of the risk availability module may be affected by changes to items
in the special drugs module (RK01a, RK01b, RK01c).

•

Change to the stimulant questions in the substance dependence and abuse module
may affect responses to the prior substance use items. The questions administered in
this module are also dependent upon earlier reports of use. This analysis focused on
age of last use reports of all substances reported.

•

Changes to the prior substance use questions may affect responses to the substance
treatment module (TX01, TX02, TX03).

•

Extensive changes to the health module may affect answers in the adult mental health
service utilization module (ADMT01, ADMT02, ADMT04) and the youth mental
health service utilization module (YSU01, YSU02, YSU04, YSU05).

•

Items from the mental health, adult depression, and adolescent depression modules
are crucial outcomes in the survey. Estimates were compared for key measures, such
as Kessler-6 (K6) scores, serious psychological distress (SPD), limitation of activities
because of psychological distress (as measured by World Health Organization
Disability Assessment Schedule [WHODAS] scores), suicide (ideation, plans, and
attempts), and major depressive episode (MDE).

•

Initial items in the special topics module on being arrested and booked in the lifetime
and past 12 months were compared.

Comparisons between the QFT sample and the 2011 and 2012 quarters 3 and 4 samples
are shown in Tables K-14 to K-21. Overall, very few differences were observed between the
QFT and main study samples for the items examined.
One notable difference was the average number of years since last use for hallucinogens
between the 2011 comparison sample (11.3 years) and the QFT sample (9.6 years). One
explanation for this difference is that the 2011 comparison data do not take into account reports
of lifetime use of ketamine, DMT/AMT/"Foxy," or Salvia divinorum from the noncore special
drugs module.32 That is, respondents in the 2011 comparison data who did not report lifetime use
of hallucinogens in the core but who reported lifetime use of one or more of these drugs in the
special drugs module were not asked the prior substance use questions for hallucinogens.
In short, the universes of respondents being asked the prior substance use questions differed
between the two samples. Also, comparison data respondents could report less recent use of
hallucinogens in the core than they reported for the three hallucinogens in the special drugs
module.
In the QFT, the logic for asking the prior substance use questions for hallucinogens
would appear on the surface to be the same as in the main survey. However, the three
hallucinogens mentioned previously were moved from the noncore special drugs module to the
core hallucinogens module in the QFT. Also, years since last use was defined as zero (0) for past
year and past month users. Consequently, users of these hallucinogens that previously were
"noncore" were eligible in the QFT to be administered the prior substance use questions for
32

DMT is an abbreviation for dimethyltryptamine, and AMT is an abbreviation for alphamethyltryptamine.

251

hallucinogens. Reports of past year or past month use of these previous noncore hallucinogens
could further decrease the mean in the QFT.
Another contributing factor to the difference between reports of years since last use of
hallucinogens between the QFT and the 2011 comparison sample is that the largest increase in
lifetime hallucinogen use was for adolescents aged 12 to 17 (2011: 3.7 percent; 2012 quarters 3
and 4: 3.2 percent; 2012 QFT: 6.5 percent). For young adults aged 18 to 25, the difference was
18.1 versus 19.4 percent, and the difference was 15.7 versus 16.9 percent for adults aged 26 or
older. By definition, younger people have a smaller range of answers for years since last use than
older persons. Some of the decline in "years since last use" may be due to a higher relative
proportion of lifetime users within the younger ages than previously observed. Overall, the
reasons for the decrease in average years since last use of hallucinogens appear to be due to
factors other than context effects.
There were also differences in several statistically significant mental health measures
between both the 2011 and 2012 comparison samples and the QFT sample. Past month SPD
among adults 18 years or older was lower in the QFT sample (3.6 percent) than in either the
2011 comparison sample (4.7 percent) or the 2012 comparison sample (5.3 percent). Similar
differences were found for past year SPD. At this point, it is unclear why such differences could
emerge due to context effects. Context effects have been suspected of producing differences in
responses to the K6 mental health items (which are used to measure SPD) in previous years,
most notably in the 2004 survey in which changes in the content of questions prior to the K6
items were thought to have affected respondent interpretation of the K6 items (Aldworth,
Chromy, Foster, Heller, & Novak, 2005). It is not clear how changes in question items preceding
the K6 items in the QFT sample might have led respondents to interpret the K6 items differently
from those in the 2012 and 2011 comparison samples. Demographic differences noted in
Section 8.5 between the QFT sample and the 2012 and 2011 comparison samples may have
contributed to differences in responses to the K6 items, but such an inference may require an
additional analysis. These findings for past year and past month SPD will be examined further in
the analysis of DR data, including analysis of combined QFT and DR data, where applicable.

8.7

Estimates for New, Revised, and Moved Items in the QFT Instrument

As noted in Section 4.4.1 in Chapter 4, the QFT instrument included items that differed
from the 2011 and 2012 quarters 3 and 4 instrument in one of three ways:
•

the question was new to the instrument,

•

the question or response options were significantly revised, or

•

the question was moved from one part of the questionnaire to another, including
either being moved to a different module or moved from CAPI to ACASI
administration.

This section provides estimates for questionnaire items that fall under one of these categories—
new items and moved items. For items moved in the QFT questionnaire, but otherwise
unchanged, this section also provides comparisons of the QFT estimates to the 2011 and 2012
quarters 3 and 4 comparison estimates. As presented in Table 4.8 in Chapter 4, missingness rates

252

for some of the moved items were significantly higher in the QFT data than in the 2011 and 2012
quarters 3 and 4 comparison data. For this reason, in addition to comparisons of QFT estimates
for moved items with the 2011 and 2012 quarters 3 and 4 comparison estimates, further analyses
of selected moved items included examining the role of proxy reports in generating these
estimates (see Section 8.8) and benchmarking the QFT estimates for these items against other
survey data (see Sections 9.3 and 9.4 in Chapter 9).
Table M-1 in Appendix M presents weighted estimates, standard errors, and unweighted
number of respondents for the new questionnaire items in the QFT that were also added to the
2013 main study questionnaire. Because the QFT was the first data collection to field these
items, these results provide an initial look at the estimates for these items and how they might
look in the 2013 data. Given that these items were new additions to the questionnaire, no
comparisons of these QFT estimates could be made to the 2011 and 2012 quarters 3 and 4
comparison data. To determine how well the QFT results match current estimates for other
national surveys collecting the same data, estimates for some of these new items were
benchmarked to other survey estimates including height and weight (see Section 9.3) and receipt
of social security or railroad retirement payments (see Section 9.4).
For items that were moved in the QFT questionnaire, Table N-1 in Appendix N presents
estimates and standard errors for the QFT data, the 2011 comparison data, and the 2012 quarters
3 and 4 comparison data. These results highlight a few more items that were moved from CAPI
to ACASI administration in the QFT questionnaire and produced significantly different QFT
estimates compared with the 2011 and 2012 quarters 3 and 4 comparison data:
•

The QFT estimate (15.6 percent) for persons not having at least one job or business
during the past 12 months (item QD37) was significantly higher than the 2011
comparison estimate (12.4 percent) and the 2012 quarters 3 and 4 comparison
estimate (12.3 percent).

•

The QFT estimate (13.8 percent) for the average number of weeks during the past
12 months persons did not have at least one job or business (item QD38) was
significantly lower than the 2011 comparison estimate (17.1 percent) and the 2012
quarters 3 and 4 comparison estimate (17.9 percent).

•

The QFT estimate (18.6 percent) for persons working for an employer with 25 to
99 employees (item QD42) was significantly lower than the 2011 comparison
estimate (22.3 percent) and the 2012 quarters 3 and 4 comparison estimate (21.4
percent). No differences were observed between the QFT and the 2011 and 2012
quarters 3 and 4 comparison data for the other four categories of number of
employees, indicating that overall differences were small in the distribution of
employer size between the QFT data and the 2011 and 2012 quarters 3 and 4
comparison data.

•

The QFT estimate (2.3 percent) for persons working for an employer that has a
written policy about employee use of alcohol or drugs that only covers drugs (item
QD44) was significantly lower than the 2012 quarters 3 and 4 comparison estimate
(3.5 percent). The QFT estimate was not significantly different from the 2011
comparison estimate (3.0 percent).

253

Without additional corroborating estimates for these questions, it is not possible to determine
whether moving these items from CAPI to ACASI administration in the QFT questionnaire
played any role in these observed differences or whether the differences made the estimates more
accurate or less accurate. Given that many more items used to produce these estimates had higher
missingness rates in the QFT data than in the 2011 or 2012 comparison data, differential
missingness rates could have contributed to observed differences in estimates. Even though some
of these items did not have missingness rates that were significantly higher in the QFT than in
the 2011 or 2012 comparison data, the overall pattern that was observed was that greater
missingness rates occurred in the ACASI mode versus the CAPI mode for these items.
(See Section 4.4 and Appendix R for more details on data quality issues for items moved from
CAPI to ACASI administration for the QFT.) These differences are highlighted to provide some
indication of how estimates for these items moved from CAPI to ACASI administration might
look different than current CAPI estimates when the partially redesigned questionnaire is
implemented in 2015, assuming further changes are not made to these items.
Table O-1 in Appendix O presents estimates and standard errors for all new, revised, or
moved items from the QFT data only among persons aged 12 or older. This complete set of
estimates for all new, revised, or moved items includes the smaller subsets of new items
presented in Table M-1 and moved items presented in Table N-1. These estimates provide a
comprehensive sense of how the data might look for all of these items when the partially
redesigned instrument and protocol are implemented in 2015, assuming further changes are not
made to these items.

8.8

Comparison of the Distribution of Relationships for Proxy Respondents
and Estimates for Selected Items Based on Proxy Report Status

Two sets of questionnaire items that were moved from CAPI to administration in the
QFT questionnaire—health insurance and income—allowed for a proxy respondent to answer
these questions in lieu of the primary respondent. For example, about 75 percent of youth
respondents aged 12 to 17 nominate a parent or other adult in their household to answer these
questions instead of them. As noted in Section 8.5 and presented in Table K-6, QFT respondents
were significantly more likely to use a proxy reporter for these questions than 2011 and 2012
quarters 3 and 4 comparison respondents. Given this difference, reporting patterns among
proxies could be one possible source of observed differences between QFT estimates and 2011
and 2012 quarters 3 and 4 comparison estimates for these items. This section presents and
discusses two types of data on proxy reports in the QFT data compared with the 2011 and 2012
quarters 3 and 4 comparison data:
•

the distribution of proxy relationships to the primary respondent and

•

estimates for proxy reports versus respondent reports for these items.

These analyses will provide some insight on whether the greater use of proxy reporters in the
QFT appeared to have any impact on differences observed QFT estimates and 2011 and 2012
quarters 3 and 4 comparison estimates for these items.

254

Table P-1 in Appendix P shows the distribution of respondents' relationships with their
proxy reporters for youths aged 12 to 17 and adults aged 18 or olde, orwhetr for the QFT sample,
the 2011 comparison sample, and the 2012 quarters 3 and 4 comparison sample. Overall, the
distributions of proxy relationships across 11 types of relationships were very similar across all
three datasets for both youths and adults. For youths aged 12 to 17 in all three samples, a little
over two thirds of proxies were mothers of the primary respondents, and about one quarter were
fathers. For adults aged 18 or older in all three samples, about 60 percent of proxies were
spouses, and about 23 percent were mothers. Proportions for other relationship categories for
both youths and adults were relatively small. Only one difference among all relationship
categories was statistically significant. For adult respondents, the QFT sample proportion
(0.2 percent) for using another adult relative as a proxy was significantly lower than the 2011
comparison sample proportion (1.5 percent). This proportion was 1.0 percent for the 2012
quarters 3 and 4 comparison sample, but the difference between the QFT and the 2012 quarters 3
and 4 proportions was not statistically significant. The lack of significant differences in the
distribution of respondents' relationships with their proxy reporters across the three datasets
indicates that proxy relationships to those respondents who used proxies were not a factor in
explaining differences in estimates between the samples for items where proxy reporting was
allowed.
Although the relationship of proxy reporters to primary respondents was not a factor in
observed differences in relevant estimates among the three datasets, the higher overall use of
proxy reporters could have been a contributor to these observed differences. To explore this
possibility, Tables P-2 through P-4 in Appendix P compare estimates from proxy reports versus
primary respondent reports for three age group categories: all respondents aged 12 or older,
youth respondents aged 12 to 17, and adult respondents aged 18 or older. If the greater use of
proxy reporters in the QFT was at least partly responsible for differences in estimates between
the QFT sample and the 2011 and 2012 quarters 3 and 4 comparison samples, significant
differences in the relevant estimates would be expected among the proxy reports and small or no
differences would be expected among the primary respondent reports. These results revealed two
important patterns among estimates that differed significantly between the QFT sample and the
2011 and 2012 quarters 3 and 4 comparison samples.
One pattern observed for several estimates was differences between the QFT and the
2011 and 2012 quarters 3 and 4 comparison samples being of similar magnitude for both proxy
and nonproxy reports. For example, the QFT estimate among all respondents aged 12 or older
(Table P-2) for having private health insurance that includes coverage for treatment of alcohol
abuse or alcoholism (item QH108) was 73.7 percent for data reported by proxies. The QFT
proportion was significantly lower than the proxy-reported estimates for the 2011 comparison
sample (84.7 percent) and the 2012 quarters 3 and 4 comparison sample (85.1 percent).
Looking at the same estimates for data reported by the primary respondents, the QFT estimate
(76.8 percent) was similarly lower than the 2011 comparison sample (84.0 percent) and the 2012
quarters 3 and 4 comparison sample (84.2 percent). The greater use of proxies among QFT
respondents was clearly not a significant factor in explaining differences between the three
datasets for items where this pattern of results was observed.
A second pattern observed for some items was QFT proxy and nonproxy estimates being
different from each other, but still significantly different from the parallel 2011 comparison and
255

2012 quarters 3 and 4 comparison estimates. For example, Table P-2 shows that the QFT
proportion for receiving income from wages or pay earned from working at a job or business
(item QI05N) was 63.8 percent for data reported by proxies. The QFT proportion was
significantly lower than the proxy-reported estimates for the 2011 comparison sample
(84.9 percent) and 2012 quarters 3 and 4 comparison sample (86.3 percent). For the same
estimates for data reported by the primary respondents, the QFT estimate (71.6 percent) was
significantly higher than the QFT proxy estimates, but still significantly lower than the 2011
comparison sample (87.2 percent) and the 2012 quarters 3 and 4 comparison sample
(87.5 percent). A similar pattern was observed for receipt of food stamps (item QI07N), where
the difference between QFT estimates for proxy reports compared with the 2011 and 2012
quarters 3 and 4 comparison estimates was significantly greater than the difference in estimates
for nonproxy reports, but still significantly different. The greater use of proxies among QFT
respondents appeared to be a factor in explaining differences between the three datasets for items
where this pattern of results was observed. For these items, proxy reports exacerbated differences
between QFT estimates versus 2011 and 2012 quarters 3 and 4 comparison estimates, but did not
fully account for these differences.
Another important conclusion from Tables P-2 through P-4 is that the two patterns
identified above appeared to hold for both youth respondents aged 12 to 17 than among adult
respondents. Estimates for nonproxy reports for several of these items for respondents aged 12 to
17 were of low precision due to low numbers of respondents in this category (Table P-3).
These low precision estimates prohibited conclusions to be reached on the statistical significance
of observed differences for youth respondents, but the proportions for both proxy and nonproxy
reports appeared to fit the two main patterns.

256

9. Selected QFT Estimates Compared with
Other Survey Estimates
9.1

Overview of Selected QFT Estimates Compared with Other Survey
Data

This chapter presents comparisons of estimates from the 2012 Questionnaire Field Test
(QFT) with estimates from other data sources. Comparable statistics from other surveys can be
used as benchmark tools for evaluating the validity of estimates from the QFT. Such
comparisons take into consideration that the external data used in the comparisons have their
own error properties and influences, such as mode of administration (e.g., self-administration vs.
interviewer administration, or paper-and-pencil questionnaires vs. computer-assisted
interviewing). These differences must be considered regardless of how similar or dissimilar the
estimates are from the compared data sources. Section 9.2 presents comparisons between data
from the QFT with estimates from the National Ambulatory Medical Care Survey (NAMCS) and
the National Hospital Ambulatory Medical Care Survey (NHAMCS) on prescription drug use.
This section also presents comparisons of estimates from the QFT with those from Monitoring
the Future (MTF), a school-based survey on drug use. In Section 9.3, selected health and
demographic estimates from the National Health Interview Survey (NHIS) are compared with
estimates from the 2012 QFT. Section 9.4 presents additional comparisons for five sets of QFT
demographic and household estimates with parallel estimates from the 2011 and 2012 quarters 3
and 4 comparison sample and from other national surveys.

9.2

Estimates for Prescription Drug Misuse

Estimates from data sources other than National Survey on Drug Use and Health
(NSDUH) can provide external checks of the validity of the QFT estimates for prescription drug
use and misuse. As noted in Section 3.7.3 in Chapter 3, comparisons with other data sources can
pose challenges when there are methodological or other differences between NSDUH and these
external data sources. A further challenge is whether suitable data on prescription drug use or
misuse are available from other sources for comparison with the QFT estimates. For example,
commercial market data on drug sales or prescriptions dispensed in the United States would
provide market share information for prescription drugs of interest. However, these data may not
be publicly available, or only limited information may be accessible. The National Center for
Health Statistics (NCHS) within the Centers for Disease Control and Prevention (CDC) makes
public use data available for two health care surveys: the NAMCS and the NHAMCS. Although
NAMCS and NHAMCS data are publicly available for analysis, prescription drug data from
these two sources do not allow direct estimates to be made of the prevalence of actual
prescription drug use or estimates of the numbers of prescriptions for different medications that
were dispensed.
Similarly, limited data on prescription drug misuse are available at the national level for
comparison with QFT data (e.g., as opposed to surveys within a single school district, university,
or State). The MTF is principally a school-based survey that collects national data on

257

prescription drug misuse through surveys of 8th, 10th, and 12th graders. It also includes a
longitudinal component in which samples of respondents who completed the survey as 12th
graders are administered follow-up surveys into adulthood. However, the MTF does not survey
dropouts or include students who were absent from school on the day of the survey. NSDUH has
shown dropouts to have higher rates of illicit drug use (Gfroerer, Wright, & Kopstein, 1997).
Therefore, the population of inference for the MTF school-based data collection is adolescents
who were in the 8th, 10th, and 12th grades. Depending on the effects of the exclusion of
dropouts and frequent absentees, data from the MTF may not generalize to the population of
adolescents as a whole, especially for older adolescents. Similarly, because the longitudinal
component of the MTF is drawn from 12th graders who were still in school when the survey was
administered, adolescents who had already dropped out of school are not eligible to be included
for longitudinal follow-up. Even among adolescents at the 12th grade level (i.e., including
dropouts who would be at this grade level if they had remained in school), dropouts are likely to
raise the estimated percentages of substance use only modestly compared with estimates based
on 12th graders who were in school. Excluding data from dropouts may have a more notable
effect on estimates of the numbers of adolescent substance users, especially for less prevalent
substances such as cocaine (Center for Behavioral Health Statistics and Quality [CBHSQ],
2012a).
Although the Drug Abuse Warning Network (DAWN) provides population estimates
through 2010 of visits to hospital emergency departments (EDs) that are attributable to misuse of
prescription drugs, DAWN does not directly measure the prevalence of prescription drug misuse.
Depending on the levels of risk of adverse events associated with misuse, estimated numbers and
rates of ED visits in DAWN for misuse of certain prescription drugs also may be
disproportionately high relative to their actual prevalence of misuse in the general population.
The National Epidemiologic Survey on Alcohol and Related Conditions (NESARC)
conducted by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) also provides
data on the misuse of prescription drugs among adults in the civilian, noninstitutionalized
population of the United States. However, NESARC data were not examined because the data
are less current. Specifically, the first wave was conducted in 2001 and 2002, and the second
wave was conducted in 2004 and 2005 (Grant & Dawson, 2006). Although a 1-year data
collection period for the next wave of the survey (NESARC-III) began in 2012, these data were
not available.
Therefore, despite these limitations and considerations, NAMCS and NHAMCS were
chosen for estimating mentions of prescription drugs for comparison with QFT data on past year
use because of the availability of public use data for these two surveys. The MTF was chosen for
comparison with QFT data because the survey provides national estimates.
9.2.1

NAMCS and NHAMCS

NAMCS and NHAMCS are national probability sample surveys. For NAMCS, a national
sample of office-based and community health center-based physicians provide data on patients'
outpatient visits. In 2010, a total of 31,229 patient record forms (PRFs) were received from the
physicians who participated in NAMCS (NCHS, 2012a). The 2010 NHAMCS included
34,718 PRFs from samples of patient records at hospital outpatient departments (NCHS,

258

2012b).33 These datasets provide information on medications mentioned in outpatient office
visits (for NAMCS) or hospital outpatient records (for NHAMCS). Data are available for
specific medications mentioned and for therapeutic categories of medications (e.g.,
benzodiazepines) based on the Multum Lexicon classifications. As noted previously, NAMCS
and NHAMCS allow weighted estimates to be created for numbers of mentions of specific drugs
or categories of drugs rather than estimates of the prevalence of actual use. These data also may
not directly translate to patients actually being prescribed or filling a prescription for a particular
medication. However, the relative order of mentions of prescription drugs in these datasets can
be compared with the relative order of prevalence estimates of any past year use in the QFT.
9.2.2

Prescription Drug Use and Misuse in the QFT and Prescription Drug Mentions in
NAMCS and NHAMCS

Tables L-1 to L-3 in Appendix L show QFT estimates for any past year use, past year use
without misuse, and past year misuse. These tables also show estimates of the numbers of
mentions of these drugs in the 2010 NAMCS data and NHAMCS outpatient hospital data
(subsequently referred to as NHAMCS).34
Because NAMCS and NHAMCS data are expressed as numbers of mentions, QFT
estimates in these tables represent the estimated numbers of persons aged 12 or older (in
thousands) in the civilian, noninstitutionalized population of the United States who were past
year users or misusers. Data in these tables include estimates for all of the specific prescription
drugs in the QFT questionnaire. Because of the small numbers of QFT respondents (or no
respondents) reporting any past year use for some prescription drugs, estimates were limited to
the overall NSDUH sample of persons aged 12 or older. Estimated numbers in the QFT and
standard errors that are indicated with "0 (0)*" represent situations where no respondents
reported use or misuse of that particular prescription drug; as indicated by the asterisk, these
estimates would be suppressed (i.e., not published) under standard NSDUH suppression rules for
unreliable estimates. Estimated numbers that are shown as zero with a standard error of zero but
would not be suppressed represent situations where a very small number of QFT respondents
reported use or misuse; in these situations, the estimated number and standard error were less
than 500 and rounded to zero when shown to the nearest 1,000 persons.
NAMCS and NHAMCS estimates in these tables are for the universe of annual outpatient
office visits (NAMCS) or hospital outpatient department visits in the United States for persons
aged 12 or older.35 Data for a given drug or drug category in these tables represent the estimated
number of times that a particular drug (or category) was mentioned in all outpatient office visits
or hospital outpatient department visits in the United States in 2010. NCHS considers an estimate
in NAMCS or NHAMCS to be unreliable if it has a relative standard error (RSE, or the standard
33

NHAMCS also collects data on patient visits to hospital EDs, but these ED data were not included in the

analysis.
34

The weighted number of mentions in NAMCS and NHAMCS could include duplicate counts if a drug
(or related drugs, such as pain relievers containing the same active ingredient) was mentioned more than once in an
outpatient visit. However, most drugs or related drugs were mentioned only once in an outpatient visit.
35
The NAMCS and NHAMCS also include data for patients younger than age 12. Outpatient visits were
restricted to those for persons aged 12 or older to match the NSDUH target population.

259

error divided by the estimate) greater than 0.3 or if it was based on fewer than 30 records,
regardless of the magnitude of the RSE. As for the QFT, NAMCS and NHAMCS estimates that
did not meet these standards for reliability are shown but are indicated with an asterisk (*).
Although QFT respondents were asked separate questions about their use or misuse of
tranquilizers and sedatives, Table L-2 in Appendix L includes data for both of these prescription
drug categories. This was done because anxiolytics, sedatives, and hypnotics are classified
together in NAMCS and NHAMCS. The aggregate benzodiazepine category in these two
datasets also does not differentiate between benzodiazepines that are indicated for use as
tranquilizers (e.g., Xanax® or alprazolam) and those that are indicated for use as sedatives
(e.g., Restoril® or temazepam).
In this section, terms such as "highest," "second highest," "greater than," "less than," or
other similar terms are used to indicate the relative magnitude of the estimates. However, testing
was not conducted for these estimates to identify statistically significant differences. Unlike
other sections of this report where weighted prevalence estimates are presented, therefore, these
terms do not indicate statistical significance. Readers are advised not to infer that any differences
or relative order of estimates described in this section are statistically significant.
Given the numbers of estimates presented in these tables (many of which are very small,
particularly for the QFT), the discussion of findings also is not intended to be exhaustive. Rather,
the focus is on overarching themes and highlights from these data, with examples being given as
needed for illustration.
9.2.2.1

Creation of QFT Measures

Estimates in Tables L-1 to L-3 for past year misuse of any prescription drug in a category
for the QFT (e.g., any prescription pain reliever) used the same imputed data for past year misuse
(see Section 3.4 in Chapter 3) that were used for the prescription drug estimates presented in
Chapter 7 and Appendix J. However, data were not imputed for past year use of any prescription
drug in a given category, past year use of specific prescription drugs, or past year misuse of most
specific prescription drugs.36 Rather, the prescription drug estimates for the QFT that are shown
in Tables L-1 to L-3 used data that had been edited but had not been imputed (see Section 3.3 in
Chapter 3).
Measures of "no past year misuse" were created from reports of past year use and past
year misuse. These measures were created because past year use of prescription drugs as directed
by the person for whom the medications were prescribed and past year misuse are not mutually
exclusive, such as if a person usually took the medication as prescribed but sometimes took more
than the prescribed dosage. The measures of past year misuse and no past year misuse among

The exception is that an imputed measure was created in the QFT for past year misuse of the pain
reliever OxyContin® because analogous measures were available for 2011 and the quarter 3 and quarter 4 data in
2012. For consistency with the data for other individual prescription drugs, however, edited (but not imputed) data
were used for the estimate of OxyContin® misuse in Table L-1 in Appendix L. Consequently, the estimate for past
year misuse of OxyContin® in Table L-1 (0.8 percent) is not identical to the corresponding estimate in Table J-5 in
Appendix J that was based on the imputed measure (1.1 percent).

260

past year users were mutually exclusive.37 However, the sum of the estimated numbers for past
year misuse and no past year misuse could differ from the overall estimated number for any past
year use because of rounding.
The edited variables from which these QFT estimates were made could have missing data
because most data had not been imputed (see Sections 3.3 and 3.4 in Chapter 3). If respondents
reported any past year use of a given drug but had missing data for past year misuse, they also
were treated as having missing data for no past year misuse. Respondents with missing data for a
given drug use measure were excluded from the estimate.
9.2.2.2

Creation of NAMCS and NHAMCS Measures

For a given outpatient visit reported on a PRF, the physician could record the names of
up to eight drugs mentioned in the visit; the drugs mentioned could be brand-name drugs (e.g.,
Vicodin®) or the generic equivalent of a brand-name drug (e.g., hydrocodone plus
acetaminophen). These variables were used to identify specific drugs mentioned in the NAMCS
and NHAMCS that corresponded to the specific drugs included in the QFT. These variables also
were used for creating aggregate measures of use of any of the specific "named" drugs (e.g.,
Vicodin®, Lortab®, Lorcet®, or hydrocodone) to correspond to the specific drugs that QFT
respondents were asked about. Other variables in these datasets were used for aggregate
measures of any drug within a broad therapeutic class (e.g., benzodiazepines).
In some situations, however, the QFT questionnaire included more detail than was
available in these other data. For example, QFT respondents were asked about their use and
misuse (if applicable) of the brand-name sedative Ambien®, the generic equivalent zolpidem,
the brand-name extended-release formulation Ambien® CR, and the generic extended-release
zolpidem. The NAMCS and NHAMCS had codes for the first three of these sedatives. When
zolpidem was mentioned, however, the codes did not distinguish between whether drug being
referred to was the standard formulation or the extended-release formulation. For this reason,
Table L-2 in Appendix L shows an entry of "N/A" ("not applicable") for mentions of extendedrelease zolpidem in the NAMCS and NHAMCS.
As noted previously, the NAMCS and NHAMCS also included variables for therapeutic
categories of medications based on the Multum Lexicon classifications. These therapeutic
category variables were used for the following NAMCS and NHAMCS estimates:
•

narcotic analgesics (Table L-1).

•

anxiolytics, sedatives, and hypnotics (Table L-2), including the following:
– benzodiazepines,
– barbiturates, and
– miscellaneous anxiolytics, sedatives, and hypnotics.

•

muscle relaxants (Table L-2), including the following:

For brevity, references are made to "no past year misuse" in the remainder of this section rather than to
"no past year misuse among past year users."

261

– neuromuscular blocking agents,
– skeletal muscle relaxants, and
– skeletal muscle relaxant combinations.
•

central nervous system (CNS) stimulants (Table L-3).

9.2.2.3

Use and Misuse of Specific Prescription Drugs in the QFT

Estimates from the QFT, NAMCS, and NHAMCS for pain relievers (Table L-1),
tranquilizers and sedatives (Table L-2), and stimulants (Table L-3) provide the following
highlights for the use and misuse of prescriptions drugs:
•

For pain relievers, tranquilizers, and sedatives, most past year use was accounted for
by use without any misuse. In Table L-1 in Appendix L, for example, an estimated
30.2 million persons aged 12 or older reported any use of OxyContin®, Percocet®,
Percodan®, Tylox®, or oxycodone in the past year, including 25.2 million who did not
report misuse and 5.0 million who reported misuse. Thus, more than 80 percent of the
past year users of these oxycodone products did not misuse them.

•

Misuse appeared to be fairly common among some past year users of stimulants.
For example, 5.4 million persons reported past year use of Adderall®, including
3.1 million who reported past year misuse and 2.3 million who were not misusers
(Table L-3).

•

Because the QFT estimates are based on respondents' self-reports, respondents may
report use or misuse of a drug they recognize by name rather than the actual drug they
took. For example, 11.5 million persons were estimated to be past year users of
Xanax®, and the estimate for the generic equivalent alprazolam was 3.7 million
(Table L-2). If the market share for the generic drug is greater than that of the brandname drug (e.g., because of lower insurance co-pays for generic drugs), then some of
the reports for Xanax® could reflect use of the generic drug.

•

Including multiple opportunities for respondents to report use or misuse of
prescription drugs containing a common active ingredient is likely to be important,
particularly for estimating the prevalence of misuse. For example, the estimated
numbers of persons from the QFT who misused specific pain relievers in the past year
that contain hydrocodone were 5.8 million for Vicodin®, 2.3 million for Lortab®,
0.6 million for Lorcet®, and 4.7 million for generic hydrocodone. An estimated
9.2 million persons aged 12 or older misused any of these pain relievers in the past
year. Thus, relying on reports of misuse of only a single drug with a given active
ingredient could underestimate the prevalence of past year misuse of any prescription
drug containing that ingredient. For example, the estimate of 5.8 million persons who
reported past year misuse of Vicodin® would fail to account for about one third of the
estimated 9.2 million persons who misused any of the four hydrocodone products
shown in Table L-1.

•

Including as comprehensive of a list of prescription drugs as possible (within reason)
in the QFT and the Dress Rehearsal (DR) can be helpful to the Substance Abuse and
Mental Health Services Administration (SAMHSA) for identifying the most
262

important prescription drugs within a category to include in the 2015 partial redesign
and which drugs might be less important (at least in the short term). For example, an
estimated 14.6 million persons aged 12 or older were past year misusers of any
prescription pain reliever, including 9.2 million who misused hydrocodone products,
5.0 million who misused oxycodone products, 4.1 million who misused codeine
products, and 2.4 million who misused tramadol products. In contrast, only 310,000
persons misused pain relievers containing propoxyphene (which has since been
withdrawn from the market), about 170,000 misused products containing fentanyl,
and only about 60,000 persons misused pain relievers containing pentazocine (i.e.,
Talacen®, Talwin®, or Talwin® NX) (Table L-1).
•

Estimates of the numbers of persons who misused prescription drugs in an overall
category or with specific active ingredients may be important for documenting the
magnitude of problems in a way that percentages might not. For example, the QFT
estimate of 6.0 percent of persons who were past year misusers of prescription pain
relievers (Table J-5) corresponds to nearly 15 million persons. The estimate of
4.1 million persons who misused codeine products in the past year represents less
than 2 percent of the population aged 12 or older but is larger than the population of
the city of Los Angeles (U.S. Census Bureau, 2013).

On the one hand, low estimates for specific prescription drugs in the QFT—particularly
for past year misuse—could be informative to SAMHSA for identifying prescription drugs that
could be dropped for the 2015 partial redesign without seriously sacrificing the validity of
prevalence estimates. Doing so could reduce respondent burden and fatigue while still obtaining
sufficiently complete data for valid estimates.
However, the finding that any of the 2,044 QFT respondents reported use or misuse of
some of these prescription drugs also is an issue for consideration. Small numbers of respondents
reporting use or misuse of some of these individual drugs in the QFT could translate to larger
numbers in 2015. Additional analysis of data from the DR will be useful for assessing whether
additional reports of use or misuse are obtained for some of these less commonly reported
prescription drugs and (to the extent possible) whether there are notable changes in reports for
these drugs. Furthermore, low prevalence estimates for use or misuse could reflect the length of
time that a particular drug has been on the market. For example, the U.S. Food and Drug
Administration approved the pain reliever Opana® in 2006 and the extended-release formulation
Opana® ER in 2011 (U.S. Food and Drug Administration, 2013). Including pain relievers such as
oxymorphone products in NSDUH before they start becoming more commonly misused
prescription drugs could be important to SAMHSA for staying "ahead of the curve" in terms of
the content of the prescription drug questions. Furthermore, prescription drugs with a lower
prevalence of misuse still could contribute cumulatively to overall estimates of misuse.
An additional consideration is that a drug with an apparent low prevalence could pose a
more serious public health threat than a drug with a higher prevalence. For example, of the
approximately 360,000 estimated ED visits in 2010 involving misuse of narcotic pain relievers,
approximately 66,000 involved misuse of methadone, or nearly 20 percent of these ED visits
(CBHSQ, 2012b). In comparison, of the estimated 14.6 million persons who misused
prescription pain relievers in the past year based on the QFT data, only 636,000 misused

263

methadone (Table L-1 in Appendix L), or less than 5 percent of the number who misused any
pain reliever. Furthermore, capturing information on the misuse of extended-release formulations
is important, especially for pain relievers, where tampering with the extended-release mechanism
of drug delivery (e.g., crushing, chewing) to release a higher dosage of the drug more quickly
can result in a life-threatening or fatal overdose. Thus, having as comprehensive a list of
prescription drugs as possible (within reason) can be important for ensuring that reports of
prescription drug misuse in NSDUH are as complete and accurate as possible and for ensuring
that the survey captures information about misuse for the prescription drugs that are especially
important from a public health standpoint.
Although misusers appeared to account for a notable proportion of the past year users of
some stimulants (e.g., Adderall®, Adderall® XR; see Table L-3 in Appendix L), these findings
need to be interpreted with caution. In particular, the QFT definition of misuse includes both use
without a prescription and use of prescribed medications in ways other than directed. Some users
of these stimulants may have used these drugs as prescribed and also may have misused them on
occasion in the past year. Thus, for example, the estimate of approximately 3.1 million persons
who misused Adderall® in the past year ought not to be interpreted to mean that all of these
persons used Adderall® without a prescription.
As noted previously, respondents may report the name of a drug they recognize despite it
not being the actual drug that they took. This issue may be particularly relevant for persons
attempting to recall which prescription drugs they misused. Based on respondent self-reports in
the QFT, for example, about 3.1 million of the 5.4 million past year users of Adderall® misused
it and 2.3 million did not. In comparison, an estimated 1.8 million persons reported using the
generic equivalent of Adderall® (i.e., mixed amphetamine-dextroamphetamine combinations) in
the past year, including about 600,000 who reported misuse and 1.2 million who reported no
misuse (Table L-3). Some of the QFT respondents who reported past year misuse of Adderall®
may have chosen to report misuse of this drug because of name recognition or because its name
is simpler than that of the generic equivalent,38 even if they actually may have misused the
generic. In addition, estimates for use or misuse of related stimulants containing amphetamine or
dextroamphetamine (i.e., Adderall®, Adderall® XR, Dexedrine®, dextroamphetamine, or
amphetamine-dextroamphetamine combinations) rounded to the nearest 0.1 million were
7.9 million persons who used at least one of these stimulants in the past year, 4.0 million who
used but did not misuse any of them, and 3.8 million who misused any of them (Table L-3). This
summary measure may more accurately reflect the relative prevalence of use without misuse and
past year misuse compared with the prevalence estimates for individual drugs in this category
(e.g., Adderall®).
Even if QFT respondents misreported the exact drug they used or misused in the past
year, however, estimates for any drug containing a given active ingredient may still be reliable
for reporting purposes. For the example of misuse of amphetamine or dextroamphetamine
stimulants, the important issue for analysis and reporting is more likely to be whether
respondents can correctly recall if they used or misused some kind of amphetamine or
38

In the screening questions for any past year of prescription stimulants, for example, the generic
equivalent of Adderall® is presented in the response choice as "Mixed amphetamine-dextroamphetamine pills other
than Adderall (generic)."

264

dextroamphetamine stimulant, even if they do not perfectly recall which exact stimulant it was
(e.g., Adderall® or the generic drug).
9.2.2.4

Relative Order of Past Year Use in the QFT and Mentions in the NAMCS and
NHAMCS

Tables 9.1 through 9.3 summarize the data presented in Tables L-1 to L-3 in Appendix L.
These summary tables present data according to common active ingredients (e.g., pain relievers
containing hydrocodone, such as Vicodin®, Lortab®, Lorcet®, or hydrocodone in Table 9.1) or
other chemically related drugs (e.g., benzodiazepines in Table 9.2). These summary tables also
are designed to facilitate comparison of the relative order of any past year use of prescription
drugs in the QFT data with the relative order of mentions of these drugs in outpatient visits in the
NAMCS and NHAMCS data.
Summary data from the QFT, NAMCS, and NHAMCS for pain relievers (Table 9.1),
tranquilizers and sedatives (Table 9.2), and stimulants (Table 9.3) provide the following
highlights on the prevalence of use or misuse (NSDUH) or the number of mentions (NAMCS
and NHAMCS) of each type of prescription drug:
•

Prescription pain relievers were the most commonly used category of
psychotherapeutic drugs in the QFT. Estimated numbers of persons in the QFT who
were past year users of any drugs in the general prescription drug categories were
94.0 million persons aged 12 or older who used pain relievers (Table 9.1);
46.6 million persons who used any tranquilizer or sedative39 (Table 9.2); and
14.5 million persons who used stimulants (Table 9.3).

•

Estimated numbers of mentions of tranquilizers, sedatives, or similar drugs were the
most commonly mentioned category of psychotherapeutic drugs in outpatient visits in
2010 for the NAMCS and NHAMCS. Estimated numbers for the NAMCS were
77.2 million for narcotic analgesics (Table 9.1); 114.2 million for tranquilizers,
sedatives, hypnotics, or muscle relaxants (Table 9.2); and 17.1 million for CNS
stimulants (Table 9.3). Estimated numbers of mentions in outpatient hospital clinic
visits in 2010 for the NHAMCS were 8.7 million for narcotic analgesics; 13.1 million
for tranquilizers, sedatives, hypnotics, or muscle relaxants; and 1.4 million for CNS
stimulants. The numbers of mentions of tranquilizers, sedatives, or similar drugs in
the 2010 NAMCS and NHAMCS were somewhat greater than the numbers of
mentions for narcotic analgesics.

•

The four most commonly used groups of prescription pain relievers in the past year
for the QFT in Table 9.1 were Vicodin®, Lortab®, Lorcet® or hydrocodone
(61.1 million persons); OxyContin®, Percocet®, Percodan®, Tylox®, or oxycodone
(30.2 million persons); Tylenol® with codeine 3 or 4 or codeine pills (27.7 million
persons); and Ultram®, Ultram® ER, Ultracet®, Ryzolt®, or tramadol (15.3 million
persons).

The QFT estimate for any tranquilizer or sedative is presented because the NAMCS and NHAMCS do
not allow estimation for these drug categories separately.

265

Table 9.1 Comparison of Summary Data for Pain Relievers from the 2012 NSDUH Questionnaire Field Test and the 2010 National
Ambulatory Medical Survey and the 2010 National Hospital Ambulatory Medical Care Survey
NSDUH QFT,1
Number in
NSDUH QFT,
NSDUH QFT,1
NHAMCS Hospital
Thousands (SE)
Number in
Number in
NAMCS, Number of Outpatient, Number of
Thousands (SE) Past Year Use But Thousands (SE)
Mentions
Mentions
Not Misuse3
Any Past Year Use2
Past Year Misuse2 in Thousands (SE)4
in Thousands (SE)4
1

266

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits
(NAMCS/NHAMCS)
Any Prescription Pain Reliever5/Any Narcotic
Analgesic6
Vicodin®, Lortab®, Lorcet®, or Hydrocodone7
OxyContin®, Percocet®, Percodan®, Tylox®,
or Oxycodone8
Darvocet®, Darvon®, or Propoxyphene7
Ultram®, Ultram® ER, Ultracet®, Ryzolt®, or
Tramadol7
Tylenol® with Codeine 3 or 4, or Codeine
Pills7
Avinza®, Kadian®, MS Contin®, Oramorph®
SR, or Morphine
Actiq®, Duragesic®, Fentora®, or Fentanyl
Suboxone®, Subutex®, or Buprenorphine
Demerol®
Dilaudid®
Methadone
Opana® or Opana® ER
Talacen®, Talwin®, or Talwin® NX
Any Other Prescription Pain Reliever

94,036 (5,617)
61,084 (4,412)

79,423 (4,800)
51,839 (3,807)

14,613 (1,894)
9,174 (1,313)

77,194 (6,493)
35,868 (3,520)

8,744 (1,161)
2,890 (378)

30,249 (2,884)
5,074 (1,092)

25,192 (2,622)
4,765 (1,059)

4,986
310

(811)
(181)

13,517 (1,543)
7,944 (1,158)

1,957
600

(284)
(142)

15,332 (2,037)

12,873 (1,777)

2,388

(631)

11,690 (1,563)

1,548

(198)

27,734 (2,653)

23,547 (2,426)

4,117

(728)

3,185

(476)

444

(86)

9,562 (1,472)
2,203 (645)
2,354 (588)
1,660 (363)
2,113 (536)
1,453 (413)
675 (211)
142 (101)
21,019 (2,079)

8,564 (1,409)
2,033 (649)
1,391 (513)
1,540 (351)
1,486 (494)
817 (304)
199 (121)
81
(81)
20,433 (2,065)

998
169
963
120
627
636
475
60
527

(347)
(120)
(305)
(90)
(190)
(262)
(173)
(60)
(202)

1,408
1,848
1,535*
310*
858
1,518
39*
117*

(272)
(325)
(650)
(154)
(218)
(341)
(25)
(93)
N/A

405 (120)
1,026* (372)
88*
(32)
343* (251)
106* (36)
146
(38)
5*
(4)
0*
(0)
N/A

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
N/A = not applicable (NSDUH) or not available (NAMCS/NHAMCS); NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Care
Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
7
For NAMCS/NHAMCS: generic or generic with acetaminophen.
8
For NAMCS/NHAMCS: generic, generic with acetaminophen, or generic with aspirin.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012; National Ambulatory
Medical Care Survey (NAMCS), 2010; National Hospital Ambulatory Medical Care Survey (NHAMCS), 2010.

Table 9.2 Comparison of Summary Data for Tranquilizers and Sedatives from the 2012 NSDUH Questionnaire Field Test and the 2010
National Ambulatory Medical Survey and the 2010 National Hospital Ambulatory Medical Care Survey
NSDUH QFT,
Number in
Thousands (SE)
Any Past Year Use2
36,934 (3,494)
17,610 (1,993)

NSDUH QFT,1
Number in
Thousands (SE)
Past Year Use But
Not Misuse3
31,212 (3,147)
15,724 (1,782)

46,607 (3,857)
27,943 (2,950)

41,019 (3,470)
22,883 (2,672)

6,819 (1,021)
5,060 (876)

15,157 (2,040)
6,513 (1,018)
6,586 (1,138)
6,194 (1,221)
254 (161)
107
(76)
203 (131)
0*
(0)
852 (505)
1,766 (636)
12,967 (1,816)
1,044 (496)

11,489 (1,784)
5,277 (907)
5,307 (1,019)
4,761 (1,077)
207 (154)
107
(76)
203 (131)
0*
(0)
852 (505)
1,573 (615)
11,417 (1,681)
984 (493)

3,668
1,237
1,279
1,433
47
0*
0*
0*
0*
193
1,550
60

267

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits
(NAMCS/NHAMCS)
Any Tranquilizer5
Any Sedative5
Any Tranquilizer or Any Sedative6/Any
Anxiolytic, Sedative, Hypnotic, or Muscle
Relaxant7
Any Benzodiazepine
Xanax®, Xanax® XR, Alprazolam, or
Extended-Release Alprazolam8
Ativan® or Lorazepam8
Klonopin® or Clonazepam8
Valium® or Diazepam8
Librium®8
Tranxene®8
Oxazepam (also known as Serax®)8
Dalmane® or Flurazepam9
Halcion® or Triazolam9
Restoril® or Temazepam9
Flexeril® or Soma®
Buspirone (also known as BuSpar®)
Hydroxyzine (also known as Atarax® or
Vistaril®)
Meprobamate (also known as Equanil® or
Miltown®)
See notes at end of table.

NSDUH QFT,1
NHAMCS Hospital
Number in
NAMCS, Number of Outpatient, Number
Thousands (SE)
Mentions
of Mentions
Past Year Misuse2 in Thousands (SE)4 in Thousands (SE)4
N/A
5,722 (917)
N/A
1,886 (535)
N/A
N/A

(676)
(361)
(445)
(403)
(47)
(0)
(0)
(0)
(0)
(160)
(393)
(60)

1,486

(576)

1,417

(572)

(69)

(60)

(0)

(60)

114,180 (8,913)
54,334 (4,534)
18,498 (1,808)
13,022 (1,447)
11,814 (1,578)
6,096 (841)
430* (212)
201*
(99)
164*
(61)
12*
(12)
97* (60)
2,333 (368)
11,442 (1,373)
2,330 (365)
3,649
114*

(700)
(61)

13,078 (1,745)
6,906 (1,139)
1,711
1,716
1,455
461
18*
5*
17*
32*
9*
313*
1,318
312

(289)
(368)
(241)
(100)
(12)
(5)
(17)
(26)
(5)
(97)
(188)
(64)

676 (123)
0*

(0)
(continued)

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits
(NAMCS/NHAMCS)
Ambien®, Ambien® CR, Zolpidem, or
Extended-Release Zolpidem
Lunesta®
Sonata® or Zaleplon
Butisol®, Seconal®, or Phenobarbital/
Barbiturates10
Any Other Prescription Tranquilizer
Any Other Prescription Sedative

NSDUH QFT,
Number in
Thousands (SE)
Any Past Year Use2
14,080 (1,949)
2,555 (746)
1,186 (597)
705
4,206
2,898

(401)
(863)
(666)

NSDUH QFT,1
Number in
Thousands (SE)
Past Year Use But
Not Misuse3
12,351 (1,690)
2,263 (709)
1,029 (577)
599
4,206
2,845

(394)
(863)
(665)

NSDUH QFT,1
NHAMCS Hospital
Number in
NAMCS, Number of Outpatient, Number
Thousands (SE)
Mentions
of Mentions
Past Year Misuse2 in Thousands (SE)4 in Thousands (SE)4
1,729 (528)
292 (230)
156 (156)
105 (77)
0* (0)
47 (47)

17,051 (1,757)
2,365 (519)
125* (53)
673

(177)
N/A
N/A

1,312 (192)
119* (47)
42* (20)
72 (16)
N/A
N/A

268

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
N/A = not applicable (NSDUH) or not available (NAMCS/NHAMCS); NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Care
Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
Created from NSDUH QFT summary measures for any tranquilizer and any sedative use or misuse.
7
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
8
Benzodiazepine that is included in the NSDUH tranquilizers module.
9
Benzodiazepine that is included in the NSDUH sedatives module.
10
NSDUH asks specifically about Butisol®, Seconal®, and phenobarbital. NAMCS and NHAMCS include a category for any barbiturates.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012; National Ambulatory
Medical Care Survey (NAMCS), 2010; National Hospital Ambulatory Medical Care Survey (NHAMCS), 2010.

Table 9.3 Comparison of Summary Data for Stimulants from the 2012 NSDUH Questionnaire Field Test and the 2010 National
Ambulatory Medical Survey and the 2010 National Hospital Ambulatory Medical Care Survey

269

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits
(NAMCS/NHAMCS)
Any Prescription Stimulant5/Any Central
Nervous System Stimulant6
Adderall®, Adderall® XR, Dexedrine®,
Dextroamphetamine, or AmphetamineDextroamphetamine Combinations
Ritalin®, Ritalin® SR, Ritalin® LA, Concerta®,
Daytrana®, Metadate® CD, Metadate® ER,
Focalin®, Focalin® XR, Methylphenidate, or
Dexmethylphenidate
Didrex® or Benzphetamine
Diethylpropion
Phendimetrazine
Phentermine
Provigil®
Tenuate®
Vyvanse®
Any Other Prescription Stimulant

NSDUH QFT,1
NSDUH QFT,1
Number in
Number in
NSDUH QFT,1
NHAMCS Hospital
Thousands (SE) Thousands (SE)
Number in
NAMCS, Number of Outpatient, Number of
Any Past Year Past Year Use But Thousands (SE)
Mentions
Mentions
Use2
Not Misuse3
Past Year Misuse2 in Thousands (SE)4
in Thousands (SE)4
14,512 (1,548)

9,332 (1,180)

5,180 (936)

7,908 (1,115)

4,039 (750)

3,828 (748)

4,860

(762)

351 (60)

3,676
123
60
374
1,882
181
0*
1,794
2,569

2,242
123
0*
374
1,775
181
0*
1,164
2,391

1,434 (364)
0* (0)
60 (60)
0* (0)
107 (76)
0* (0)
0* (0)
589 (222)
177 (177)

3,637
3*
0*
48*
1,157*
792
389*
1,142

(664)
(3)
(0)
(48)
(515)
(209)
(279)
(279)
N/A

521 (120)
6* (5)
0* (0)
6* (6)
111* (36)
73* (24)
19* (13)
130* (41)
N/A

(635)
(87)
(60)
(374)
(562)
(145)
(0)
(562)
(620)

(485)
(87)
(0)
(374)
(527)
(145)
(0)
(500)
(594)

17,054 (2,731)

1,437 (240)

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
N/A = not applicable (NSDUH) or not available (NAMCS/NHAMCS); NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Care
Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012; National Ambulatory
Medical Care Survey (NAMCS), 2010; National Hospital Ambulatory Medical Care Survey (NHAMCS), 2010.

•

The four most commonly reported groups of prescription pain relievers in outpatient
clinic visits in 2010 in Table 9.1 for the NAMCS were Vicodin®, Lortab®, Lorcet® or
hydrocodone (35.9 million mentions); OxyContin®, Percocet®, Percodan®, Tylox®, or
oxycodone (13.5 million mentions); Ultram®, Ultram® ER, Ultracet®, Ryzolt®, or
tramadol (11.7 million mentions); and Darvocet®, Darvon, or propoxyphene
(7.9 million mentions). The first three of these also were the three most commonly
reported groups of pain relievers in the NHAMCS.

•

The two most commonly used groups of prescription tranquilizers or sedatives in the
past year for the QFT in Table 9.2 were Xanax®, Xanax® XR, alprazolam, or
extended-release alprazolam (15.2 million persons); and Ambien®, Ambien® CR,
zolpidem, or extended-release zolpidem (14.1 million persons). These were the same
two most commonly reported groups of prescription tranquilizers or sedatives in
outpatient clinic visits in 2010 for the NAMCS (18.5 million and 17.1 million
mentions, respectively). In the NHAMCS, however, there were more mentions of
Ativan® or lorazepam and Klonopin® or clonazepam than for sedatives containing
zolpidem. Differences in the characteristics and medical needs of patients in a general
outpatient clinic setting and those in outpatient hospital clinics could explain these
results.

One possible explanation for the difference in order of the mentions for the broader
categories in the QFT and in the NAMCS and NHAMCS data is that the estimates for the
outpatient datasets were specifically for narcotic analgesics such as those explicitly included in
the QFT. In contrast, the estimate of past year use in the QFT was for prescription pain relievers,
including past year use of "any other prescription pain reliever" besides the specific drugs
included in the pain relievers screener. As shown in Tables 9.1 and L-1, an estimated
21.0 million persons aged 12 or older in the QFT (8.7 percent) were past year users of any other
prescription pain reliever, which was greater than most of the estimates for pain relievers.
However, other pain relievers could include drugs such as ibuprofen (e.g., Motrin®) that may be
available in dosages that require a prescription but are not psychoactive. Only about
500,000 persons aged 12 or older (0.2 percent) reported past year misuse of other pain relievers.
Relative to the estimated 21.0 million persons who were past year users of other pain relievers,
this number who misused other pain relievers comprised about 3 percent of those who reported
any use of other pain relievers. This estimate for past year misuse of other pain relievers also was
lower than the most commonly reported pain relievers that were misused.
An additional issue to consider for these comparisons is that the prescription drug reports
in the NAMCS and NHAMCS are roughly 2 years older than the estimates for the QFT.
For example, one of the most commonly mentioned groups of pain relievers in these 2010 data
was the group containing propoxyphene, which has since been removed from the market in the
United States. Although the mentions of drugs in these datasets do not correspond directly to
actual use or numbers of prescriptions, it could be worthwhile to see how these estimates look
when the NAMCS and NHAMCS data become available for 2012.
These findings also may suggest analytic limitations in presenting estimates of any past
year use in NSDUH reports following the 2015 partial redesign. Asking about past year use may
aid respondents in the cognitive tasks of identifying which prescription drugs they used and then

270

identifying which ones of those they misused. Data on any past year use also provide a
denominator for estimating the percentages of past year misusers among persons who have used
prescription drugs in the past year. However, issues such as which prescription drugs
respondents are thinking of when they report past year use of "any other" pain reliever suggest
that it also will be important to consider any limitations in measurement of any past year use
before these estimates are included as a regular component of national reports, along with
estimates of misuse.
9.2.3

Monitoring the Future

MTF includes questions for 8th, 10th, and 12th graders about their misuse in the past
12 months of the pain relievers Vicodin® and OxyContin®, prescription tranquilizers,
amphetamines, and the stimulants Adderall® and Ritalin®. Misuse of prescription drugs is
defined as use "not under a doctor's orders." Where drug use measures have been similar
between NSDUH and MTF, MTF estimates historically have been higher than corresponding
NSDUH estimates. Despite differences in the sizes of estimates, both surveys show similar
trends for substance use (CBHSQ, 2012e).
Published MTF data from the survey that was administered to 8th, 10th, and 12th graders
in the spring of 2011 were available for comparison with QFT estimates (Johnston, O'Malley,
Bachman, & Schulenberg, 2012a). Combined data for adolescents in these three grades are
shown in Table L-4 in Appendix L, along with QFT estimates for adolescents aged 12 to 20 who
reported that they were in the 8th, 10th, or 12th grades.
Published MTF estimates from 2011 also were available for young adults aged 19 to 24
(Johnston, O'Malley, Bachman, & Schulenberg, 2012b). These data and corresponding QFT
estimates are shown in Table L-5. In addition to the prescription drug estimates described
previously for adolescents, MTF data in Table L-5 for young adults include estimates for misuse
of narcotics other than heroin (corresponding to the QFT measure for pain relievers), the
stimulant Provigil®, and sedatives (barbiturates). Since 2002, questions in MTF about narcotics
other than heroin have included Vicodin®, OxyContin®, and Percocet® as examples of these
types of drugs (Johnston et al., 2012b).40
Standard errors are not included for these published MTF estimates. Consequently,
testing was not conducted to identify statistically significant differences between the QFT and
MTF estimates. Terms in this section such as "greater than," "less than," "more likely," or "less
likely" are used to indicate the relative magnitude of the estimates but do not indicate statistical
significance. Readers are advised not to infer that any differences in estimates described in this
section are statistically significant.

Examples of narcotics other than heroin in the MTF questions prior to 2002 were Talwin®, laudanum,
and paregoric, each of which had negligible rates of use by 2001 (Johnston et al., 2012b).

271

9.2.4

Prescription Drug Misuse in the QFT and Monitoring the Future
9.2.4.1

8th, 10th, and 12th Graders

Highlights of QFT and MTF estimates for 8th, 10th, and 12th graders include the
following:
•

The QFT estimates for past year misuse of Vicodin® and OxyContin® among 8th,
10th, and 12th graders (1.5 and 0.8 percent, respectively) were lower than
corresponding MTF estimates for the specific drugs (5.1 and 3.4 percent). However,
the QFT estimates for past year misuse of Vicodin®, Lortab®, Lorcet®, or
hydrocodone (3.0 percent) and for OxyContin®, Percocet®, Percodan®, Tylox®, or
oxycodone (1.4 percent) were closer to the MTF estimates for the single prescription
drugs.

•

QFT and MTF estimates for past year misuse of tranquilizers were similar for
adolescents in these three grades (2.8 and 3.9 percent), given the size of the standard
error for the QFT estimate (1.12 percent).

•

The QFT estimate for past year misuse of prescription stimulants (0.7 percent) was
considerably lower than the MTF estimate for amphetamines (5.9 percent). However,
there were no QFT respondents in the 8th, 10th, or 12th grades who reported past
year misuse of Ritalin®. In comparison, the MTF estimate for past year misuse of
Ritalin® was 2.1 percent.

9.2.4.2

Young Adults

Highlights of QFT and MTF estimates for young adults include the following:
•

The QFT estimates for past year misuse of prescription pain relievers among young
adults were in the direction of being greater than the MTF estimates for misuse of
narcotics other than heroin. For example, the QFT estimate of past year misuse of
pain relievers among young adults aged 19 to 20 was 15.9 percent, and the MTF
estimate for narcotics other than heroin was 7.7 percent.

•

Estimates for past year misuse of OxyContin® among young adults were similar for
the QFT and MTF. Among young adults aged 19 to 20, for example, the QFT
estimate was 3.6 percent, and the MTF estimate was 3.3 percent.

•

The QFT estimate of past year misuse of Vicodin® among young adults aged 21 to 22
(2.9 percent) was lower than corresponding MTF estimate (7.1 percent). As for
adolescents, however, the QFT estimate among adults aged 21 to 22 for any misuse of
Vicodin®, Lortab®, Lorcet®, or hydrocodone (7.4 percent) was similar to the MTF
estimate.

•

Based on the sizes of the standard errors for the QFT estimates, the QFT and MTF
estimates for young adults were similar for past year misuse of tranquilizers and
prescription stimulants/amphetamines. Among adults aged 23 to 24, estimates of past
year misuse of sedatives/barbiturates also were similar between the QFT (3.7 percent)
and MTF (3.5 percent).

272

•

Estimates of past year misuse of Adderall® were similar for the QFT and MTF, based
on the sizes of the standard errors for the QFT. For adults aged 21 to 22, the QFT
estimate was 7.6 percent, and the MTF estimate was 9.4 percent.

On the one hand, findings of higher estimates of prescription drug misuse among 8th,
10th, and 12th graders in MTF than in the QFT are consistent with patterns for NSDUH and
MTF that have been observed for other drugs (CBHSQ, 2012a, 2012e). However, these estimates
of misuse tended to converge when QFT data included misuse of any drugs with the same active
ingredient as these two specific drugs. This result could indicate that reports of misuse of
"Vicodin" and "OxyContin" in the MTF refer to misuse of any drugs that MTF respondents
recognize by these brand names, such as pain relievers other than Vicodin® that contain
hydrocodone.
The generally higher QFT estimates among young adults for past year misuse of any pain
relievers compared with MTF estimates for narcotics other than heroin is consistent with the
different structure and content of these questionnaires. Specifically, QFT respondents can report
use and then subsequent misuse in the past year of up to 40 possible pain relievers. In contrast,
MTF respondents are provided with only three examples of narcotics other than heroin:
Vicodin®, OxyContin®, and Percocet®. Furthermore, as shown in Table L-1 in Appendix L, QFT
estimates of past year misuse among persons aged 12 or older for generic hydrocodone, generic
oxycodone, Tylenol® with codeine 3 or 4, and any pain relievers containing tramadol were
similar to or greater than the estimates for some of these pain relievers that are provided to MTF
respondents as examples of narcotics other than heroin.
Limitations of these comparisons include the small QFT sample size, especially when the
sample sizes are reduced further to limit the estimates to 8th, 10th, and 12th graders or to young
adults in 2-year age groupings, and the unavailability of exact information on the precision of
estimates in MTF based on combined data for 8th, 10th, and 12th graders or for young adults.
However, the combined MTF sample in 2011 consisted of nearly 47,000 students from these
three grades. In addition, 95 percent confidence intervals for past year prevalence estimates
among adolescents in the individual grades provide some indication of the potential precision of
estimates when data from all three grades are combined (Johnston et al., 2012a). For the followup surveys of young adults, a cohort of approximately 2,400 persons who participated in the
survey as 12th graders is followed longitudinally at 2-year intervals (Johnston et al., 2012b).41
Because of the smaller QFT sample sizes when the data were further subdivided for
comparison with the MTF estimates, the estimate of Adderall® misuse in the QFT for 8th, 10th,
and 12th graders was unreliable. No QFT respondents were estimated to be past year misusers of
Ritalin® for 8th, 10th, and 12th graders or for young adults aged 19 or 20. Similarly, no young
adults aged 19 to 24 in the QFT reported past year misuse of Provigil®. Combining data from the
QFT and DR would be expected to improve the precision of these estimates.
41

More detailed information about the design for the longitudinal follow-up is provided in the 2011 MTF
report for college students and adults aged 19 to 50 (Johnston et al., 2012b). A weighted sample size of
approximately 5,500 adults aged 19 to 30 was reported for the 2011 data collection. The unweighted number of
respondents was not specified but will be larger because the stratum of drug users from high school is oversampled
for follow-up and therefore contributes less to the weighted number.

273

9.3

Estimates for Selected Health and Demographic Items

The National Health Interview Survey (NHIS) was chosen as a benchmark survey for
evaluating two new NSDUH survey measures—persons living in households with no telephone
or only cellular telephone service and the number of visits to health care professionals in the past
year. In addition, NHIS estimates on family income and highest level of education for adults
were compared with estimates from NSDUH. Although the question text for education
(item QD11) remained the same, the response categories were changed to reflect the concept of
educational attainment rather than years of education. For example, response categories with
types of degrees have replaced years of college and there are separate categories for a high
school diploma versus "12TH GRADE, NO DIPLOMA." Although the NSDUH questions on
family income will remain mostly unchanged in the redesigned questionnaire, the questions will
be administered in audio computer-assisted self-interviewing (ACASI) rather than through
computer-assisted personal interviewing (CAPI), and the change in mode could produce
differences in estimates.
The purpose of the NHIS is to monitor the health of the U.S. population through data
collection and analysis on a broad range of health topics. The NHIS covers the civilian,
noninstitutionalized population residing in the United States at the time of the interview.
Excluded populations include patients in long-term care facilities; persons on active duty with
the armed forces (though their dependents are included); persons incarcerated in the prison
system; and U.S. nationals living in foreign countries. As such, the population covered by the
NHIS is similar to the NSDUH population. For these comparisons, only data from NHIS
interviews that were conducted in English have been included. However, NHIS public use files
do not contain geographic identifiers that would allow for excluding data from Alaska and
Hawaii. In addition, the most recent NHIS data files were only from 2011. NHIS estimates in
Table L-6 in Appendix L were calculated using SUDAAN (RTI International, 2008) and by
following the procedures described in the NHIS documentation of variance estimation
procedures (NCHS, 2012c).
Comparisons of estimates between the QFT and the 2011 NHIS for selected health and
demographic items are shown in Table L-6 in Appendix L. Except for education, all of the
estimates shown in this table are for persons aged 12 or older.
•

The QFT estimate of 1.4 percent for persons living in a household without any
telephone service is very similar to the 2011 NHIS estimate of 1.2 percent. Trend data
from the NHIS has shown that the percentage of persons living in a household with
only wireless service has been steadily increasing since 2003 (Blumberg & Luke,
2013). The QFT estimate for the proportion of adults living in a household either
without phone service or only with cellular telephone service was 35.9 percent, which
was slightly higher than the NHIS estimate of 31.5 percent. The NHIS estimate
increased from 32.0 to 38.4 percent between the first 6 months of 2011 and the last
6 months of 2012. For children over the same time period, the percentage increased
from 38.1 to 46.9 percent. Given that trajectory, some of the difference between the
QFT estimate and the NHIS estimate could have resulted from this trend. Consistent
with this explanation, the QFT estimate for having at least one telephone at the
address that was not a cellular telephone was 64.1 percent, which was lower than the
274

NHIS estimate of 68.1 percent. Likewise, for anyone at the address having a working
cellular telephone, the QFT estimate of 92.3 percent was slightly higher than the
NHIS estimate of 90.4 percent.
•

Compared with the NHIS, the QFT sample had lower proportions of persons 12 or
older who had no visits to a health care professional in the past 12 months
(15.5 percent in the QFT vs. 17.2 percent in the NHIS) and also lower percentages of
persons with 10 or more visits (10.6 vs. 13.1 percent). Differences between the QFT
and NHIS questions on visits to doctors or other health care professionals may
contribute to differences in the estimates. The NHIS question asks respondents to
exclude certain types of visits that may be reported in other questions, such as
hospital visits, emergency room visits, and dental visits, while the QFT item does not.
This difference would presumably lead to higher estimates of visits for the QFT than
the NHIS. Also, the QFT question refers to more types of health care professionals
("a doctor, nurse, physician assistant or nurse practitioner") than the NHIS question
("doctor or other health care professional").

•

The QFT data estimate of 9.7 percent of persons who stayed overnight in a hospital in
the past year was higher than the NHIS estimate of 8.3 percent. This was consistent
with results from a comparison of reports on overnight hospital visits for persons
18 or older between the 2006 NSDUH and the 2006 NHIS reported in a NSDUH data
review (Pemberton, Bose, Kilmer, Kroutil, Forman-Hoffman, & Gfroerer, 2013).
The NSDUH estimate was 11.1 percent, while the NHIS estimate was 8.8 percent.

•

The QFT estimate of 26.5 percent for persons aged 12 or older who made an
emergency room visit in the past year was higher than the estimate from the 2011
NHIS (20.3 percent). The NSDUH data review reported a similarly large difference
for persons aged 18 or older (28.8 vs. 20.4 percent) and for persons aged 12 to 17
(31.9 vs. 17.8 percent) (Pemberton et al., 2013). The NSDUH data review also noted
that the NHIS question mentions "hospitals," while the NSDUH question does not
specifically mention "hospitals"; it may be that NSDUH respondents are including
emergency visits to trauma or urgent care centers that are not associated with hospital
emergency rooms.

•

A new series of questions added to the QFT questionnaire asked respondents whether
a doctor or other health care professional had ever told them whether they had one or
more of nine health conditions, as shown in Table L-6 in Appendix L. The QFT and
2011 NHIS estimates were generally similar for most of these health conditions, but
significant differences were observed for a few conditions with QFT estimates being
lower than NHIS estimates. Estimates from the QFT and 2011 NHIS were very
similar for any kind of heart condition or heart disease, diabetes or sugar diabetes, and
kidney disease.42 For hepatitis and asthma, the QFT estimates appeared to be slightly
lower than the 2011 NHIS estimates. QFT estimates were significantly lower than the
comparable 2011 NHIS estimates for the following conditions: chronic bronchitis,

The NHIS does not contain a question on ever having been told by a doctor or health professional about
kidney disease. The estimate for the QFT response category of "Kidney disease, not including bladder infection or
incontinence" was compared with the estimate from the NHIS item that asked about "Weak or failing kidneys? Do not include kidney stones, bladder infections or incontinence (past 12 months)."

275

emphysema, or chronic obstructive pulmonary disease (COPD)43; cirrhosis of the
liver; cancer or a malignancy of any kind; and hypertension or high blood pressure.
In relative terms, hypertension or high blood pressure had the greatest difference
between the QFT estimate (17.8 percent) and the 2011 NHIS estimate (30.3 percent)
among all conditions. One key difference between the QFT and NHIS instruments
could have contributed to these observed differences in estimates for health
conditions. In the QFT instrument, the health conditions were treated as response
categories in a "code all that apply" format, whereas in the NHIS instrument the
parallel categories were administered as separate, individual items.
•

Another new series of questions added to the QFT instrument asked respondents
whether they had any of six types of disabilities or physical limitations. The QFT and
2011 NHIS estimates were similar for three types of disabilities or physical
limitations, but slightly different for the other three types. Estimates from the QFT
and 2011 NHIS were very similar for being deaf or having serious hearing difficulty,
being blind or having serious difficulty seeing, and having serious difficulty
concentrating, remembering, or making decisions. QFT estimates appeared to be
significantly lower than the comparable 2011 NHIS estimates for the following
disabilities or physical limitations: having serious difficulty walking or climbing
stairs, having difficulty dressing or bathing, and having difficulty doing errands alone,
such as visiting a doctor's office or shopping. In relative terms, having serious
difficulty walking or climbing stairs had the greatest difference between the QFT
estimate (6.4 percent) and the 2011 NHIS estimate (9.0 percent) among all
conditions.

•

Relative to the NHIS sample, family incomes in the QFT sample were generally
lower. In the QFT data, 31.0 percent of persons aged 12 or older had a family income
of greater than or equal to $75,000 compared with 35.6 percent in the NHIS sample.
With respect to education, the QFT distribution for adults aged 18 or older was
similar to the distribution from the 2011 NHIS. The observed differences in income
levels for the QFT sample could have been a factor in explaining differences between
the QFT versus other data sources, such as the 2011 and 2012 quarters 3 and 4
comparisons samples, for items that were the most highly correlated with income.
Section 9.4.3 provides a more detailed discussion of benchmarking QFT estimates for
income levels to other surveys, and Section 9.4.5 provides a more detailed discussion
of benchmarking QFT estimates for education levels to other surveys.

The QFT questionnaire included questions on height and weight, which was the first time
these questions have been fielded in a NSDUH data collection since the mid-1990s.
QFT estimates for height and weight were compared with three sources:
•

2011 NHIS estimates,

•

2009-2010 National Health and Nutrition Examination Survey (NHANES)
self-reported estimates, and

The estimate based on the QFT response category "Chronic bronchitis, emphysema, chronic obstructive
pulmonary disease, also called COPD" was compared with an NHIS estimate based on lifetime reports of
emphysema and past 12 month reports of chronic bronchitis.

276

•

2009-2010 NHANES directly measured estimates.

In addition, because coding of NHIS height and weight data includes specific lower and upper
bounds, the QFT estimates for height and weight were calculated both unbounded and bounded
following NHIS criteria. The second calculation provided a more equivalent comparison between
the QFT and 2011 NHIS data. The summary statistics for height presented in Table L-7 and the
summary statistics for weight presented in Table L-8 in Appendix L provided some sense of how
the QFT statistics for these new questionnaire items compared with other national surveys.
•

Both the unbounded QFT mean height estimate (66.8 inches) and the NHIS-bounded
QFT mean height estimate (66.4 inches) were very similar to the NHIS mean height
estimate (66.8 inches) and the NHANES directly measured mean height estimate
(66.5 inches). The NHANES self-reported mean height estimate (67.1 inches)
appeared to be slightly higher than the other four estimates, but not appreciably so.
Additional summary statistics revealed some anomalies in height reports that were
allowed in the QFT questionnaire. For example, implausible minimum and maximum
unbounded height values were accidentally provided by some QFT respondents, and
the computer-assisted interviewing program allowed these values to be entered.

•

Both the unbounded QFT mean weight estimate (179.0 pounds) and the NHISbounded QFT mean weight estimate (178.1 pounds) were very similar to the
NHANES directly measured mean weight estimate (179.2 pounds) and the NHANES
self-reported mean weight estimate (177.8 pounds). The NHIS mean weight estimate
(171 pounds) was somewhat lower than the other four estimates. Anomalous
reporting of weight data in the QFT appeared to be less common than for the height
reports, and minimum and maximum weight reports were fairly similar to the NHIS
and NHANES data. One possible explanation for this is that height appeared first in
the questionnaire, so QFT respondent could have learned from the height screens how
to more accurately enter their data on the weight screens.

Overall, the QFT height and weight estimates aligned closely to estimates from the 2011
NHIS and 2009-2010 NHANES, both self-reported and directly measured. Some observed
anomalies among QFT respondents in reporting height figures suggests range checks could be
applied to these questions and editing rules developed for these items to avoid having
implausible values in the NSDUH data. For the DR, the ranges for height data in feet and inches
will be edited for accuracy for the height question, and the upper limit for the weight question
will be increased.

9.4

Estimates for Additional Demographic and Household Items

Based on results showing significant differences between QFT estimates and 2011 and
2012 quarters 3 and 4 estimates, benchmarking further demographic and household items to
other national surveys was undertaken. This benchmarking was intended to determine whether
the QFT estimates also differed from other national survey estimates with the same target
population and comparable survey items. The following QFT items were benchmarked to other
national surveys:
•

received income and participation in government assistance programs,
277

•

health insurance coverage,

•

income,

•

employment status and unemployment rates, and

•

education.

Given that all of these items were moved from CAPI to ACASI administration in the QFT and
two sets of these items—health insurance and income—allow for proxy reports, this section
highlights the implications of the benchmarking results for the DR and 2015 partial redesign.
9.4.1

Received Income and Participation in Government Assistance Programs

In Tables L-9 through L-12 in Appendix L, QFT estimates for five types of received
income or participation in government assistance programs for all persons aged 12 or older and
three separate age groups are presented with parallel estimates from the 2011 comparison
sample, the 2012 quarters 3 and 4 comparison sample, the 2011 American Community Survey
(ACS), and the 2011 NHIS. The three separate age groups are persons aged 12 to 17, 18 to 25,
and 26 or older. Estimates for all data sources are provided in both percentages and thousands of
persons, with standard errors in parentheses. Several notable comparisons can be observed from
these tables:
•

For all persons aged 12 or older (Table L-9), estimates for receipt of social security
were very similar across all five survey data sources at about 27 percent. Estimates
for social security were also similar across these data sources for the three separate
age groups (Tables L-10 through L-12).

•

The QFT estimate for receipt of wages for all persons aged 12 and older
(68.6 percent) was significantly lower than the estimates from the four other data
sources, which were all close to 80 percent. This pattern held for receipt of wages
across all three separate age groups.

•

For supplemental security income (SSI), the QFT estimate for all persons aged 12 or
older (9.4 percent) was generally higher than the estimates from most of the other
data sources. Estimates for SSI from the other surveys ranged from 5.0 percent in the
2011 NHIS to 7.6 percent in the 2012 quarters 3 and 4 comparison sample. This
pattern for receipt of SSI was very similar across the three separate age groups.

•

The QFT estimate for participation in food stamp44 programs for all persons aged
12 or older (17.6 percent) was also generally higher than the estimates from the four
other data sources. Estimates for food stamp receipt from the other surveys ranged
from 13.0 percent in the 2011 NHIS to 15.6 percent in the 2012 quarters 3 and 4
comparison sample. This pattern for receipt of food stamps was very similar across
the three separate age groups.

•

For receipt of welfare payments, such as those from Temporary Assistance for Needy
Families (TANF), the QFT estimate for all persons aged 12 or older (3.6 percent) was

Food stamp programs are now more commonly known as the Supplemental Nutrition Assistance
Program (SNAP).

278

higher than the estimates from the 2011 comparison sample (2.5 percent) and the
2012 quarters 3 and 4 comparison sample (2.3 percent), but it was similar to the 2011
ACS estimate (3.3 percent) and the 2011 NHIS estimate (3.2 percent). The pattern for
receipt of welfare payments generally held across the three separate age groups, with
the QFT estimates being somewhat higher than the 2011 and 2012 quarters 3 and 4
comparison estimates, but similar to the 2011 ACS and 2011 NHIS estimates.
Benchmarking QFT estimates for five types of received income or participation in
government assistance programs to both recent NSDUH data and other national survey data
revealed mixed results. Estimates for receipt of social security payments were quite similar
across all five surveys. The QFT estimate for receipt of wages was substantially lower than the
estimates from the other four survey sources. For receipt of welfare payments, QFT estimates
were generally similar to the 2011 ACS and 2011 NHIS estimates, but higher than the 2011 and
2012 quarters 3 and 4 comparison estimates.

Estimates of participation in two programs—SSI and food stamps—appeared to be
clearly greater for the QFT sample than in the other four surveys. This finding suggests that QFT
respondents were either somewhat lower overall in socioeconomic status or that QFT
respondents were more likely to report participation in these programs in ACASI mode than
other survey respondents were in an interviewer-administered mode. Similar to the discussion in
Section 9.3 on lower income and education levels among the QFT sample, these findings suggest
that QFT respondents had a somewhat lower socioeconomic status than the 2011 and 2012
quarters 3 and 4 comparisons samples. This difference could have accounted for some of the
observed differences between the QFT estimates and the 2011 and 2012 quarters 3 and 4
comparison estimates for those items that were the most highly correlated with socioeconomic
status.
9.4.2

Health Insurance Coverage

In Tables L-13 through L-16 in Appendix L, QFT estimates for four types of health
insurance coverage for all persons aged 12 or older and three separate age groups are presented
with parallel estimates from the 2011 comparison sample, the 2012 quarters 3 and 4 comparison
sample, the 2011 ACS, and the 2011 NHIS. The three separate age groups are persons aged 12 to
17, 18 to 25, and 26 or older. A few notable comparisons can be observed from these tables:
•

For all persons aged 12 or older (Table L-13), estimates for the first three types of
health insurance coverage—Medicare, Medicaid, and TRICARE, CHAMPUS, or
other military health care sources—were generally similar across all five survey data
sources. This pattern generally held for these three types of health insurance coverage
across the three separate age groups (Tables L-14 through L-16).

•

Two exceptions to the general pattern noted above were observed. First, the QFT
estimate for Medicaid coverage for all persons aged 12 or older (13.4 percent) was
slightly higher than the parallel estimates from the 2011 comparison sample
(11.6 percent), the 2012 quarters 3 and 4 comparison sample (11.5 percent), and the
2011 NHIS (10.6 percent), but it was similar to the 2011 ACS estimate (12.9 percent).
This difference appeared to be driven mostly by the estimate for persons aged 12 to

279

17 (Table L-14), where the QFT estimate was at least 5 percent higher than the
estimates from the other four data sources.
•

In addition, the 2011 NHIS estimate for health insurance coverage via TRICARE,
CHAMPUS, or other military health care sources for all persons aged 12 or older
(3.5 percent) was lower than the estimates from the other four data sources, which
were all close to 5 percent. This difference appeared to be driven mostly by the
estimate for persons aged 12 to 17 (Table L-14), where the 2011 NHIS estimate of
3.9 percent was higher than the estimates from the other four data sources, which
ranged from 5.2 to 5.6 percent.

•

For all persons aged 12 or older, the QFT estimate (62.1 percent) for private health
insurance was lower than the estimates from the other four data sources, which
ranged from 67.1 to 68.7 percent. Although this pattern generally held for private
health insurance across the three separate age groups, differences in estimates
between the QFT and the other four surveys were somewhat more pronounced for
persons aged 12 to 17 (Table L-14) and persons aged 18 to 25 (Table L-15).

Benchmarking QFT estimates for four types of health insurance coverage to both recent NSDUH
data and other national survey data revealed mixed results. Across all age groups, the largest and
most consistent differences between QFT estimates and estimates from the other four data
sources were observed for private health insurance. Differences between QFT estimates and
estimates from the other four data sources for the other three types of health insurance coverage
were generally smaller and less consistent across age groups.
9.4.3

Family Income

In Tables L-17 through L-20, QFT estimates for three income categories for all persons
aged 12 or older and three separate age groups are presented with parallel estimates from the
2011 comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2011 NHIS.
The three separate age groups are persons aged 12 to 17, 18 to 25, and 26 or older. Two notable
comparisons can be observed from these tables:
•

For all persons aged 12 or older (Table L-17), the QFT estimate for family income of
$49,999 (52.1 percent) or less was only slightly higher than the 2011 and 2012
quarters 3 and 4 comparison estimate, but it was significantly higher than the 2011
NHIS estimate (46.5 percent). Correspondingly, the QFT estimates for a family
income of $50,000 to $74,999 and a family income of $75,000 or greater were lower
than estimates for the 2011 comparison sample, the 2012 quarters 3 and 4 comparison
sample, and the 2011 NHIS. QFT estimates for these two income categories were
somewhat closer to the 2011 and 2012 quarters 3 and 4 comparison estimates than to
the 2011 NHIS estimates.

•

This pattern generally held for the three separate age groups (Tables L-14 through
L-16), although the differences between the QFT estimates and the other three
sources were most pronounced for persons aged 12 to 17 (Table L-18). This finding
suggests that proxy and self-reports of income from QFT respondents aged 12 to 17
contributed the most to the observed differences in estimates for all persons compared
with the other three surveys.
280

Overall, the QFT estimates resulted in higher proportions of persons at lower income levels and
lower proportions at higher income levels, compared to three other sources of survey data. This
difference could have accounted for some of the observed differences between QFT estimates
and the 2011 and 2012 quarters 3 and 4 comparison estimates for those items that were the most
highly correlated with income level.
9.4.4

Employment Status and Unemployment Rates

In Tables L-21 through L-23, QFT estimates for four employment categories for all
persons aged 18 or older and two separate age groups are presented with parallel estimates from
the 2011 comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2012
quarters 3 and 4 Current Population Survey (CPS). The two separate age groups are persons aged
18 to 25 and those aged 26 or older. A few notable comparisons can be observed from these
tables:
•

For all persons aged 18 or older (Table L-21), the QFT estimate of persons employed
full time (52.0 percent) was slightly higher than the 2011 comparison estimate
(49.7 percent) and the 2012 quarters 3 and 4 CPS estimate (49.2 percent), but it was
similar to the 2012 quarters 3 and 4 comparison estimate (51.3 percent). A similar
pattern was observed for adults aged 26 or older (Table L-23), but the differences
between the QFT and three other survey estimates of full-time employment were
more pronounced for adults aged 18 to 25 (Table L-22). This finding suggest that
reports of full-time employment from QFT respondents aged 18 to 25 contributed the
most to the observed differences in estimates for all persons compared with the other
three surveys.

•

For all persons aged 18 or older, the QFT estimate of persons employed part time
(14.2 percent) was slightly higher than the 2012 quarters 3 and 4 CPS estimate
(11.2 percent), but it was similar to the 2011 comparison estimate (14.1 percent) and
the 2012 quarters 3 and 4 comparison estimate (13.9 percent). A similar pattern was
observed for both adults aged 18 to 25 and for adults aged 26 or older.

•

The QFT estimate for being unemployed for all persons aged 18 or older (5.5 percent)
was slightly higher than the 2012 quarters 3 and 4 CPS estimate (4.9 percent), but it
was similar to the 2011 comparison estimate (5.8 percent) and the 2012 quarters 3
and 4 comparison estimate (5.5 percent). A similar pattern was observed for both
adults aged 18 to 25 and for adults aged 26 or older, although the difference between
the QFT and the 2012 quarters 3 and 4 CPS estimate for being unemployed among
adults aged 18 to 25 was larger than the difference among adults aged 26 or older.

•

For all persons aged 18 or older, the QFT estimate of persons with an employment
status of other (28.3 percent), such as being retired or otherwise not in the labor force,
was lower than the 2012 quarters 3 and 4 CPS estimate (34.7 percent), but it was
similar to the 2011 comparison estimate (30.4 percent) and the 2012 quarters 3 and 4
comparison estimate (29.3 percent). A similar pattern was observed for adults aged
26 or older, but the differences between the QFT and three other survey estimates for
persons with an employment status of other were more pronounced for adults aged
18 to 25. This finding suggest that reports of an employment status of other from QFT

281

respondents aged 18 to 25 contributed the most to the observed differences in
estimates for all persons compared with the other three surveys.
In addition, Table L-24 provides calculated unemployment rate estimates among persons
aged 18 or older for three age groups for the QFT, the 2011 comparison sample, the 2012
quarters 3 and 4 comparison sample, and the 2012 quarters 3 and 4 CPS. QFT unemployment
rate estimates were similar to the 2012 quarters 3 and 4 comparison sample and the 2012 quarters
3 and 4 CPS for all persons aged 18 or older and for persons aged 18 to 25. Unemployment rate
estimates for the 2011 comparison sample were higher than the other three surveys for all
persons aged 18 or older and for persons aged 18 to 25. These differences in estimates from the
lone 2011 source and the three 2012 sources could simply reflect a trend of declining
unemployment rates for adults aged 18 to 25. For adults aged 26 or older, unemployment rate
estimates were similar across all four surveys.
Overall, comparisons between the QFT and three other sources of survey data on
employment status and unemployment rates showed significant differences mostly for adults
aged 18 to 25. Observed differences for all adults and adults aged 26 or older were relatively
small. These results could be attributable to either differences in reporting employment status
among respondents aged 18 to 25 in the QFT sample or the impact of actual trends in
employment for adults aged 18 to 25 from 2011 to 2012.
9.4.5

Education

In Tables L-25 through L-27, QFT estimates for four education categories for all persons
aged 18 or older and two separate age groups are presented with parallel estimates from the 2011
comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2011 NHIS. The two
separate age groups are persons aged 18 to 25 and those aged 26 or older. A few notable
comparisons can be observed from these tables:
•

For all persons aged 18 or older (Table L-25), estimates for less than a high school
education and having a college degree were similar across the four surveys.

•

QFT estimates differed from the three other survey data sources for the two education
categories—high school graduate and some college. The QFT estimate for persons
aged 18 or older being high school graduates (26.6 percent) was lower than the
estimates for the 2011 comparison sample (30.3 percent) and the 2012 quarters 3 and
4 comparison sample (30.1 percent), but it was similar to the 2011 NHIS estimate
(27.8 percent). Similarly, the QFT estimate for persons aged 18 or older having some
college (32.1 percent) was higher than the estimates for the 2011 comparison sample
(27.4 percent) and the 2012 quarters 3 and 4 comparison sample (27.7 percent), but it
was similar to the 2011 NHIS estimate (31.3 percent).

•

Differences in estimates between the QFT sample and the 2011 and 2012 quarters 3
and 4 comparison samples for the high school graduate and some college categories
were more pronounced among adults aged 26 or older (Table L-27). Among adults
aged 18 to 25, QFT estimates for the high school graduate and some college
categories were actually very similar to the 2011 and 2012 quarters 3 and 4
comparison estimates.

282

•

In contrast, differences in estimates between the QFT sample and the 2011 NHIS for
the high school graduate and some college categories were more pronounced among
adults aged 18 to 25 (Table L-26). Among adults aged 26 or older, QFT estimates for
the high school graduate and some college categories were similar the 2011 NHIS
estimates.

Overall, comparisons between the QFT and three other data sources of survey data on
education level differed for two categories—high school graduate and some college. Although
for all adults aged 18 or older the QFT estimates were more similar to the 2011 NHIS estimates
than to the 2011 and 2012 quarters 3 and 4 comparison samples, differences among the four data
sources for the high school graduate and some college categories varied across the two age
groups of adults aged 18 to 25 and adults aged 26 or older. These mixed results suggest that
differences in the education level of QFT respondents versus the 2011 and 2012 quarters 3 and 4
comparison samples likely had a minimal impact, if any, on observed differences between
estimates for items correlated with education.
Based partly on the results for the demographic and household items discussed in
Section 9.4, the following changes to these questions will be made for the DR:
•

reordering the list of potential sources of household income in the introductory item
INTRTINN;

•

editing the wording of item QI03N on the receipt of SSI for accuracy;

•

deleting item QI05N on income from wages or pay, and adding this to the list of
potential sources of household income in the introductory item INTRTINN;

•

editing the wording of item QI07N on the receipt of food stamps for accuracy;

•

removing the "Help" instructions in item QHI06 on private health insurance, and
moving key terms into the question itself;

•

editing the "Help" instructions for several employment questions;

•

deleting the question about size of workplace; and

•

further revising of the consistency check questions to be consistent with the
categories in item QD11 on educational attainment.

In addition, see Appendix R for more details on data quality issues for the demographic and
household items discussed in this section that were moved from CAPI to ACASI administration
for the QFT.

283

284

10. Summary and Implications
As noted in Chapter 1, the primary goal of the 2012 Questionnaire Field Test (QFT) was
to measure, using multiple indicators, the total effect on National Survey on Drug Use and
Health (NSDUH) estimates from the full set of changes to the protocol planned for the 2015
partial redesign. This chapter summarizes key findings from the various indicators examined in
Chapters 4 through 9 to inform the likely impact on the protocol planned for both the 2013 Dress
Rehearsal (DR) and 2015 partial redesign. Two kinds of implications of the QFT results are
discussed for the DR and the partial redesign:
•

areas where the QFT findings suggest changes to the field test protocol should be
considered for the DR data collection, or

•

areas where the QFT findings suggest further scrutiny is warranted in the DR analysis
to determine the full implications of these findings for the partial redesign.

Where appropriate, decisions made on changes to the field test protocol for the DR are noted.
Section 10.1 highlights key outcomes of the QFT data collection related to data quality
(Chapter 4), including screening and interview response rates, variable imputation rates and item
missingness rates, interview timing results, and other data quality indicators. Conclusions from
specific assessments of the redesigned protocol in Chapter 5—including field observations,
responses to field interviewer (FI) debriefing questions, new equipment surveys, and focus
groups with FIs—are summarized in Section 10.2. Section 10.3 discusses key findings from
comparing QFT estimates with main study estimates for substance use items other than
methamphetamine and prescription drugs (Chapter 6); Section 10.4 focuses on key findings
from comparing QFT estimates only for methamphetamine and prescription drug with main
study estimates for these two set of items (Chapter 7); and Section 10.5 presents key findings
from comparing QFT estimates for noncore survey items with the parallel main study estimates
(Chapter 8). Key findings from comparisons of selected QFT estimates with other survey
estimates, as presented in Chapter 9, are summarized in Section 10.6. Finally, Section 10.7
provides a summary list of QFT questionnaire items identified as needing careful reexamination
in the DR analysis because the item missingness rate was significantly higher than the rates for
the comparison data, the estimate produced from the item differed significantly from comparison
estimates, or both types of outcomes occurred.

10.1 Data Collection Outcomes and Data Quality Assessment
As detailed in Chapter 4, data quality for the QFT was examined through the following
four types of indicators, which were compared with the 2011 and 2012 quarters 3 and 4
comparison samples where appropriate:
•

Screening and interview response rates. The overall response rates were lower for
the QFT than for the 2011 and 2012 quarters 3 and 4 comparison samples, primarily
due to lower interview response rates and a shorter data collection period. The lack of
ability to complete screenings and interviews in Spanish and reduced flexibility in

285

assigning cases among available interviewers appeared to limit QFT response rates.
QFT interviews were also less likely to be completed on the first interview visit to a
dwelling unit. Nonetheless, the distribution of visits made for completing QFT
screenings and interviews was similar overall to the 2011 and 2012 quarters 3 and 4
comparison samples. The available evidence indicates that the lower QFT response
rate had a minimal impact on most estimates in comparison with the 2011 and 2012
quarters 3 and 4 comparison samples.
•

Variable imputation rates and item missingness rates. Comparing imputation rates
for QFT data with the 2011 and 2012 quarters 3 and 4 comparison data generally
indicated similarly low rates of imputation for most items. For some variables,
including several substance use estimates and health insurance items, QFT imputation
rates were significantly higher than the 2011 and 2012 quarters 3 and 4 comparison
data because of higher item missingness rates. Missingness rates for QFT items
(including those that were new, revised, or moved in the QFT questionnaire) were
generally low and followed similar patterns as the 2011 and 2012 quarters 3 and 4
comparison data. For example, certain health insurance and income items that had
relatively high missingness rates in the QFT data had similarly high missingness rates
in the 2011 and 2012 quarters 3 and 4 comparison data. Despite this general pattern,
a number of notable differences in missingness rates were observed between the QFT
data and the 2011 and 2012 quarters 3 and 4 comparison data. Although QFT
missingness rates were actually lower for two sets of items—workplace alcohol and
drug use policies and health insurance coverage for treatment of alcohol abuse,
substance abuse, or mental health—the most notable differences in QFT rates were
those that were significantly higher than the 2011 and 2012 quarters 3 and 4
comparison data. Several sets of items that were moved to audio computer-assisted
self-interviewing (ACASI) administration in the QFT questionnaire produced
significantly higher missingness rates than the 2011 and 2012 quarters 3 and 4
comparison data administered via computer-assisted personal interviewing (CAPI),
including the following:
– marital status, moves home in the past year, and State of residence 1 year ago;
– full-time or part-time student status, missing school due to illness or injury, and
skipping school days;
– work at a job or business at any time in the past week;
– recent employment history, missing workdays, and size of employing
organization;
– private health insurance coverage;
– receipt of various sources of income and participation in government assistance
programs; and
– two of the items on family income.

An investigation of the data quality for items moved to ACASI administration with relatively
high missingness rates is first discussed in Section 4.4.1 in Chapter 4 and is elaborated on in
Section 9.4 in Chapter 9. In addition, a detailed analysis of the impact of the higher item

286

missingness rates observed for several items that were moved from CAPI to ACASI
administration in the QFT instrument is included in Appendix R of this report.
•

Interview timing results. The overall mean interview time for the QFT interviews
was actually lower than the mean times for the 2011 and 2012 quarters 3 and 4
comparison interviews. Despite these lower mean interview times for the full QFT
interviews, additions and revisions to the hallucinogens, inhalants, and prescription
drug sections in the partially redesigned QFT questionnaire contributed to higher
administration times for the core substance use modules compared with the 2011 and
2012 quarters 3 and 4 comparison interviews. As expected, the redesigned
prescription drug modules led to greater QFT administration times for these modules,
but this difference was primarily attributable to the pain relievers module. Lower
mean times for several back-end demographic sections (including employment,
income, and administrative residual times) for the QFT interviews contributed
significantly to the lower overall interview times compared with the 2011 and 2012
quarters 3 and 4 comparison interviews.

Similar to the 2011 and 2012 quarters 3 and 4 comparison interviews, higher interview
administration times were observed in the QFT for respondents aged 12 to 17, 50 to 64, and
65 or older. In addition, more extreme overall interview times of less than 30 minutes or more
than 240 minutes were observed in the QFT data than in the 2011 and 2012 quarters 3 and 4
comparison interviews. The overall mean interview time for QFT respondents aged 65 or older
was higher than the time recorded for those aged 65 or older in the 2011 and 2012 quarters 3 and
4 interviews. Average time to complete the redesigned prescription drug modules contributed
significantly to the higher administration times among QFT respondents aged 65 or older. As a
result, the impact for respondents aged 65 or older was an increase of 8 minutes in mean
interview timing in the QFT compared with the current instrument.
•

Other data quality indicators, including hard errors and patterned responses.
These outcomes observed in the QFT data raised the possibility that two steps could
be considered to improve the interview for the DR or the 2015 partial redesign:
– alerting respondents that responses of "1" or "2" in the prescription drug screening
questions do not necessarily mean "yes" or "no," and
– capturing information about potential initiation of prescription drug misuse more
than 12 months ago for those respondents who reported past year initiation of all
prescription drugs in a category that they misused in the past year.

The first change will not made for the DR, but the second change will be made in the DR
questionnaire. Results from the DR data collection could lead to further examination of these
changes for the 2015 partial redesign.

10.2 Assessments of the Redesigned Protocol
As described in Chapter 5, four field-related efforts were used to assess the partially
redesigned questionnaire and protocol used in the QFT. Overall, these assessments provided
some assurance that the revised questionnaire and protocol will facilitate continued high quality
and efficiency in NSDUH data collection when the partial redesign is implemented in 2015.

287

Based on these assessments and discussions between the Substance Abuse and Mental Health
Services Administration (SAMHSA) and RTI, several protocol changes will be implemented for
the DR for the screening, the computer-assisted interview, the interviewer training and field
materials, and the data collection equipment and tools. Appendix Q provides a comprehensive
list of protocol changes considered for the DR and indicates whether the change will be
implemented.
Key results from the four field-related assessments are highlighted below, with
comparisons to the 2011 and 2012 quarters 3 and 4 comparison data where appropriate:
•

Observations of QFT interviewers. The majority of FIs displayed positive behaviors
when conducting QFT screenings and interviews. The types and pattern of errors
observed among QFT interviewers were not specifically related to the QFT protocol
and could have been observed on the main study. Overall, the results from QFT field
observations suggested that relatively few specific changes to the protocol are needed
for the DR or the 2015 partial redesign.

•

FI debriefing items. Responses to the QFT FI debriefing items provided some
evidence of how respondents reacted to the partially redesigned protocol. One
important finding was that recall of the redesigned lead letter appeared to be
associated with willingness to do the interview, although it cannot be determined
whether this can be attributed to the fact that the letter increases cooperation or that
recall of the letter is a reflection of the respondent's willingness to cooperate. No
problems were revealed regarding several changes in the data collection protocol,
including the use of electronic calendars and having proxy respondents reply through
ACASI rather than CAPI. FI responses to the debriefing items indicated that a
majority of respondents who were selected in households and completed the
interview recalled seeing the lead letter. Data from the debriefing items also
corroborated findings that respondents aged 65 or older—who generally took longer
to complete the QFT interview—were more likely to report that the interview took
too much time to complete. In addition, QFT respondents with less than a high school
education compared with respondents with higher levels of education also reported
that the interview was too long. These results suggest that these two subgroups of
respondents might face greater cognitive burdens than other respondents. The finding
that QFT respondents aged 65 or older had significantly longer overall interview
times was consistent with timing data from the 2011 and 2012 quarters 3 and 4
comparison interviews. Data on interview timing by education level was not produced
for the QFT interviews, the 2011 comparison interviews, and the 2012 quarters 3 and
4 comparison interviews. The results from QFT FI debriefing items do not suggest
any specific changes to the protocol that could be implemented for the DR or the
2015 partial redesign.

•

New equipment surveys of QFT interviewers. To assess a new tablet device that is
planned to be implemented for the 2015 NSDUH and was used for the QFT
household screening, surveys of QFT FIs were conducted before data collection
began and as data collection was ending. The results of these surveys indicated that
the tablet was generally well received by FIs for use as a screening device.
Comments from FIs suggested enhancements to specific features and additional
288

functionality, which were considered for implementation in the DR, including the
following:
– revisions to symbols available on the primary keyboard,
– improve calendar usability, and
– ability to continuously highlight the selected case on the select case screen.
Only the calendar usability item will be implemented for the DR.
•

Focus groups conducted with QFT interviewers. Three focus groups were conducted
with QFT FIs at locations where relatively high numbers of FIs worked. In general,
FIs expressed mostly positive sentiments about the QFT training program, the
revisions made to the lead letter and the question and answer (Q&A) brochure, and
using the tablet device for screenings. As indicated in Table 5.42 in Chapter 5,
participants in these focus groups echoed comments made in the equipment surveys
about additional functionality they would like to have on the tablet device. FIs also
noted the following concerns about using the QFT protocol, the first two of which led
to changes for the DR protocol:
– a number of FIs indicated they did not like the portfolio, which resulted in a new
portfolio being selected for the DR;
– FIs noticed that the Q&A brochure included a picture of an interview taking place
with the paper version of the reference calendar visible, which led to replacement
of this image in the Q&A brochure to be used in the DR;
– FIs noted that some members of sampled households mistakenly thought they
represented social services when the Department of Health and Human Services
was mentioned; and
– some FIs expressed concerns about including county/parish/district in the
salutation of the lead letter.

10.3 QFT Estimates Compared with NSDUH Estimates: Substance Use
Items Other than Methamphetamine and Prescription Drugs
Findings from the QFT data and the 2011 and 2012 quarters 3 and 4 comparison datasets
detailed in Chapter 6 indicate that most prevalence rates for core substances appeared to remain
similar for most of these substances, including the use of cigarettes, alcohol, marijuana, cocaine,
and heroin. These results generally held for recency of use and age groups, with some notable
exceptions.
•

Estimates of lifetime use for persons aged 12 or older of any hallucinogen, lysergic
acid diethylamide (LSD), phencyclidine (PCP), and Ecstasy did not differ between
the QFT and the 2011 or 2012 comparison data. However, use of hallucinogens was
greater for 12 to 17 year olds in the QFT data compared with the 2011 and 2012
quarters 3 and 4 comparison data. Including noncore hallucinogens data produced
estimates for any hallucinogen among 12 to 17 year olds that were more similar

289

across the QFT, 2011 comparison data, and 2012 quarters 3 and 4 comparison
datasets and were not statistically different.
•

Addition of new types of inhalants in the QFT instrument, including felt-tip markers
and computer cleaners, led to an expected increase in reported lifetime use of
inhalants, overall and across the age groups for most comparisons. Past year and past
month use of inhalants did not differ between the QFT and the 2011 and 2012
quarters 3 and 4 comparison data for all respondents aged 12 or older, although for
adolescents aged 12 to 17 the QFT estimate of past year use of inhalants was greater
than the estimate for the 2012 comparison data.

•

Among female respondents in the QFT, estimates of binge alcohol drinking were
greater than in the 2011 and 2012 quarters 3 and 4 comparison datasets. Lowering the
threshold for females from five to four drinks per occasion appeared to affect the QFT
estimates in the expected direction.

An additional noteworthy finding from these analyses is that moving the questions for the
hallucinogens called ketamine, tryptamines,45 and Salvia divinorum from the special drugs
module to the core hallucinogens module did not appear to affect lifetime reporting because of
their earlier placement in the QFT. Specifically, earlier placement of these questions in the QFT
could yield increased reports of lifetime use. In the main survey, later placement of these
questions could result in some lifetime users of these substances reporting nonuse if they have
learned by that point in the interview that answering questions about lifetime drug use as "yes"
leads to additional questions and that answering these questions as "no" leads to fewer questions.
However, the effect of this change in the placement of these questions could warrant further
investigation in the DR and in preliminary data from the 2015 survey (e.g., from the first two
quarters) to verify that these results from the QFT are not simply a function of the smaller
sample size that received the QFT questionnaire. Given that most estimates for use of substances
other than methamphetamine and prescription drugs did not differ between the QFT and
corresponding main study data (except where noted), the results did not suggest specific changes
to the instrument or protocol for the DR or the 2015 partial redesign for these core drug modules.

10.4 QFT Estimates Compared with NSDUH Estimates: Methamphetamine
and Prescription Drug Items
Chapter 7 presented findings on methamphetamine use and prescription drug misuse
from the comparison data for 2011 and 2012 quarters 3 and 4 and from the QFT data. As noted at
the beginning of this chapter, considerable changes were made to the methamphetamine and
prescription drug modules for the QFT. These changes are planned for implementation in the
partially redesigned NSDUH questionnaire in 2015 and seem likely to affect estimates of
methamphetamine use and misuse of prescription drugs starting in 2015. Comparing QFT data
with the 2011 and 2012 quarters 3 and 4 data revealed significant differences for the following
substances:

Tryptamines include dimethyltryptamine (DMT), alpha-methyltryptamine (AMT), and N, N-diisopropyl5-methoxytryptamine (5-MeO-DIPT) or "Foxy."

290

•

The lifetime estimate for methamphetamine use among persons aged 12 or older was
higher (or in the direction of being higher) in the QFT than in the comparison data.
This difference appeared to be driven by higher prevalence rates among adults aged
18 or older in the QFT than in the 2011 and 2012 quarters 3 and 4 comparison data.

•

Prescription drug estimates for lifetime misuse among all persons aged 12 or older
were lower in the QFT data than in the 2011 and 2012 quarters 3 and 4 comparison
data for pain relievers and tranquilizers. These differences were not statistically
significant for every age group.

•

Estimates of past year misuse for pain relievers, OxyContin®, and sedatives among
persons aged 12 or older were higher for the QFT than for the 2011 and 2012 quarters
3 and 4 comparison data.

•

For stimulants, past year misuse and past month misuse among persons aged 12 or
older typically were higher in the QFT data than in the 2011 and 2012 quarters 3 and
4 comparison datasets. These differences between the QFT and main study estimates
were essentially eliminated when data from noncore questions on the misuse of
Adderall® were included in estimates from the main study comparison data.

If trends in past year and past month use of methamphetamine continue to remain fairly
stable based on NSDUH data for 2012 to 2014, then moving the methamphetamine questions to
a separate module in 2015 might not disrupt the trend data for past year and past month use.
Advance monitoring of estimates of methamphetamine use from the 2015 survey (e.g., based on
the first two quarters of data) will be important for anticipating potential disruptions in the trend
data because of the changes to the methamphetamine questions in 2015.
For prescription drugs, the QFT findings support starting a new baseline in 2015 for
trends in prescription drug misuse. It might also be useful to consider whether to discontinue
reporting trend data for lifetime misuse of prescription drugs after 2014 because of questions
about the accuracy of respondent self-reports of misuse of prescription drugs more than
12 months prior to the interview.

10.5 QFT Estimates Compared with NSDUH Estimates: Noncore Items
Comparisons between QFT estimates and the 2011 comparison estimates and the 2012
quarters 3 and 4 comparison estimates for several types of noncore items were presented in
Chapter 8. These estimates included substance dependence and abuse (Section 8.2), needle use
(Section 8.3), medical marijuana reports (Section 8.4), demographic and household items
(Section 8.5), and QFT items potentially subject to context effects due to the redesigned
questionnaire (Section 8.6).
10.5.1 Substance Dependence and Abuse
QFT estimates of dependence, abuse, or dependence or abuse for persons aged 12 or
older (as shown in Table K-1 in Appendix K) were not significantly different from corresponding
estimates in the 2011 or 2012 comparison data. No significant differences in estimates of illicit
drug dependence, illicit drug abuse, or illicit drug dependence or abuse were observed among
persons in each of the age groups. Some notable differences were observed for specific age
291

groups, however. Estimates for adolescents aged 12 to 17 were lower in the QFT than in the
2011 comparison data for prescription drug dependence or abuse, pain reliever dependence or
abuse, and dependence or abuse for illicit drugs other than marijuana. In addition, for adults aged
26 or older, estimates were lower in the QFT than in the 2012 comparison data for prescription
drug dependence, dependence for illicit drugs other than marijuana, and dependence or abuse for
illicit drugs other than marijuana. Given the higher estimates of past year misuse of these
substances in the QFT, these lower QFT estimates for any prescription drug and pain relievers
for some age groups relative to estimates in the comparison data can be viewed as
counterintuitive. Two possible explanations of these findings are as follows:
•

The smaller QFT sample size and its effect on the numbers of respondents who
reported sufficient numbers of problems to be classified with dependence or abuse for
prescription drugs could have contributed to the observed differences within age
groups.

•

The respondent burden involved in answering the questions about past year misuse of
prescription drugs in the QFT could have suppressed reporting of dependence or
abuse symptoms for prescription drugs.

If similar findings for illicit drug dependence or abuse estimates are observed in the DR data,
then these findings would suggest that questionnaire changes for 2015 will not appreciably affect
substance use dependence or abuse trends for any illicit drug. However, if substance use
disorders for prescription drugs—especially prescription pain relievers—contribute more
substantially to estimates of substance use disorders for illicit drugs other than marijuana, then
changes to the prescription drug module in 2015 could affect dependence or abuse trends for
illicit drugs other than marijuana. In addition to the DR data, analysis of data from the first two
quarters of 2015 could also assist in anticipating any effects on dependence or abuse trends for
illicit drugs other than marijuana and for prescription drugs.
10.5.2 Needle Use
As shown in Table K-5 in Appendix K, lifetime estimates of needle use among persons
aged 12 or older were similar between the QFT and the 2011 and 2012 comparison data.
The findings for needle use suggest that planned changes to the questionnaire in 2015 will not
affect the 2-year trends for heroin, cocaine, or methamphetamine between 2014 and 2015.
However, changes to the questions for injection of stimulants could require creation of new trend
data for 2002 to 2015 for lifetime use of a needle to inject cocaine, heroin, or methamphetamine.
If prevalence estimates for past year injection of stimulants are presented in NSDUH detailed
tables based on the redesigned questionnaire, a new baseline would need to be established in
2015.
10.5.3 Medical Marijuana
QFT responses to a new question on the medical use of marijuana (added to the blunts
module) were used to examine how reports of using marijuana for medical purposes aligned with
the current State laws. The data for this examination were quite limited because only 15 QFT
respondents reported that at least some of their marijuana use in the past year was allegedly
recommended by a doctor. Of these 15 respondents, 7 respondents reported living in a State that
292

had a medical marijuana law in effect in 2012; the other 8 respondents did not live in States that
had a medical marijuana law in effect in 2012. These inconsistencies in reports could have been
explained by either (1) respondents referring to prior use in the past year in a different State with
a medical marijuana law in effect, or (2) respondents referring to past year use where they
accessed marijuana in neighboring States that had a medical marijuana law. Early review of the
2013 main study data will examine the alignment between reports of using marijuana for medical
purposes with the current State laws where respondents report use for a larger number of
respondents and States.
10.5.4 Demographic and Household Items
A notable change in the QFT instrument was moving questions on health insurance
coverage and family income from interviewer administration using CAPI to self-administration
using ACASI. As presented in Appendix K in Tables K-6 through K-9, the primary pattern of
differences for demographic and household items between the QFT and the 2011 or 2012
comparison datasets were higher estimates for the following items:
•

participation in government assistance programs,

•

receiving supplemental security income, and

•

participating in food stamp programs.

These observed differences in estimates indicated a pattern tending toward lower
socioeconomic status among the QFT sample, although this result cannot be disentangled from
the impact of moving these questions to ACASI administration on how QFT respondents
answered these questions. In addition, the relatively smaller sample size for the QFT makes it
difficult to predict whether estimates of participation in government programs and receipt of
specific types of income will change significantly when the partially redesigned instrument and
protocol are implemented in 2015. If similar patterns in demographic and household
characteristics are observed in the 2015 data, the QFT findings suggest some estimates that are
most strongly correlated with these demographic and household characteristics could be affected.
10.5.5 Selected Items Potentially Subject to Context Effects
The introduction of new items in the questionnaire may lead to changes in estimates
because of context effects. As noted in Section 8.6 in Chapter 8, items were selected for analysis
of context effects where a change introduced to the first (or contextual) item could affect the
response process for the subsequent (target) item. The potential presence of such effects could
not be distinguished from changes in estimates because of the full set of changes made to the
QFT survey protocol and questionnaire. Comparisons between the QFT sample and the 2011 and
2012 quarters 3 and 4 samples for relevant items are shown in Tables K-11 to K-18. Overall, few
differences were observed between the QFT and the 2011 or 2012 comparison samples for the
items examined.
One notable difference was that the average number of years since last use for
hallucinogens in the QFT sample was lower than in the 2011 comparison sample. One
explanation for this difference is that the 2011 comparison data did not take into account reports

293

of lifetime use of ketamine, DMT/AMT/"Foxy," or Salvia divinorum from the noncore special
drugs module.
Statistically significant differences were also observed for some mental health measures.
For example, past month serious psychological distress (SPD) among adults 18 years or older
was lower in the QFT sample than in both the 2011 and the 2012 comparison samples. Given
that the QFT questionnaire did not include any new items or substantial changes to the items
immediately preceding the Kessler-6 (K6) items, it is not clear why some QFT respondents
would have interpreted the K6 items differently compared with respondents in the 2011 and 2012
quarters 3 and 4 comparison samples. For the DR and the 2015 partial redesign, further
monitoring of these estimates seems warranted to understand whether estimates of SPD might
change with the redesigned questionnaire and protocol. Additional analysis could examine which
demographic and other variables might contribute to changes in SPD between the QFT data and
the two comparison datasets.

10.6 Selected QFT Estimates Compared with Other Survey Estimates
Section 9.2 in Chapter 9 presented comparisons of QFT estimates of prescription drug
use and misuse with estimates of prescription drugs that were mentioned in outpatient visits in
the 2010 National Ambulatory Medical Care Survey (NAMCS) and the 2010 National Hospital
Ambulatory Medical Care Survey (NHAMCS). In addition, this section and Tables L-1 to L-3 in
Appendix L presented data on past year use and misuse for all of the individual prescription
drugs that were included in the QFT. Section 9.2 also presented comparisons of QFT estimates
of prescription drug misuse with estimates from Monitoring the Future (MTF) for adolescents in
the 8th, 10th, and 12th grades and for young adults aged 19 to 24. Notable findings included the
following:
•

For pain relievers, tranquilizers, and sedatives in the QFT, most past year use was
accounted for by use that did not involve misuse. In comparison, misuse appeared to
be fairly common among some past year users of stimulants.

•

The two most commonly used groups of prescription pain relievers in the past year
for the QFT (Vicodin®, Lortab®, Lorcet® or hydrocodone; and OxyContin®,
Percocet®, Percodan®, Tylox®, or oxycodone) also were the two most commonly
mentioned groups of narcotic analgesics in the 2010 NAMCS and the 2010
NHAMCS.

•

The two most commonly used groups of prescription tranquilizers or sedatives in the
past year for the QFT (Xanax®, Xanax® XR, alprazolam, or extended-release
alprazolam; and Ambien®, Ambien® CR, zolpidem, or extended-release zolpidem)
also were the two most commonly mentioned groups of prescription tranquilizers or
sedatives in outpatient clinic visits for the 2010 NAMCS.

•

Among adolescents in the 8th, 10th, and 12th grades, QFT estimates for past year
misuse of prescription drugs tended to be lower than corresponding estimates from
the MTF. This pattern was consistent with prior comparisons of substance use
estimates in NSDUH and MTF for adolescents. However, some QFT estimates that
were based on the misuse of any prescription drug with the same active ingredient

294

started to converge with MTF estimates that were based on questions about misuse of
a specific drug.
•

Among young adults, QFT estimates of past year use of prescription pain relievers
tended to be higher than MTF estimates for misuse of narcotics other than heroin.
This was consistent with the differences between the two questionnaires, particularly
the much greater number of examples of prescription pain relievers in the QFT.

On the one hand, low estimates in the QFT—particularly for past year misuse—could be
informative to SAMHSA for identifying prescription drugs that could be dropped for the 2015
partial redesign without seriously sacrificing the validity of prevalence estimates and while also
helping to reduce respondent burden and fatigue. However, other considerations besides
prevalence in deciding whether to keep or drop a particular prescription drug for the partial
redesign include (a) the potential number of respondents who would report misuse of that drug in
the larger sample for the redesign; (b) the length of time that a prescription drug has been on the
market; and (c) public health considerations for misuse of certain prescription drugs, such as
extended-release drugs with higher overall dosages. Furthermore, data in Tables L-1 to L-3 in
Appendix L for specific prescription drugs and patterns for estimates of past year misuse of pain
relievers among young adults suggest that the number of examples of individual prescription
drugs that are presented to respondents can encourage more complete reporting of misuse.
Although respondents may report the name of a drug they recognize despite it not being
the actual drug that they took, misreporting of the exact drug that they used or misused in the
past year may be less critical for analysis and reporting purposes. If respondents can correctly
recall that they used or misused a prescription drug that had a particular active ingredient, then
these self-reports and the associated estimates still would be accurate, even if respondents cannot
perfectly recall exactly which prescription drug it was.
These issues will not affect the content of questions about the use or misuse of specific
prescription drugs in the DR questionnaire. Changing the content of the DR questions in this
manner would affect the comparability of the DR data for prescription drugs relative to the QFT
data and could affect the ability to analyze combined QFT and DR data for English-language
interviews to improve the precision of estimates. Analyzing combined QFT and DR data for the
prescription drug modules also would be useful for evaluating whether to change the content of
questions about specific prescription drugs for the 2015 partial redesign. For example, observing
a low prevalence of use or misuse for certain prescription drugs in combined QFT and DR data
could provide further justification for dropping these drugs from the questionnaire for 2015.
DR data also will be useful for examining whether issues of name recognition for brand-name
drugs instead of the generic equivalent that were observed in the QFT for certain prescription
drugs continue to be observed in the DR. In addition, a plan will be developed for identifying
important changes in prescription drugs in the United States for application in the 2015 NSDUH
and later years.

295

10.7 Summary of QFT Questionnaire Items Identified as Needing
Reexamination in the DR Analysis
As detailed in Chapters 4, 8, and 9, and noted in previous sections of Chapter 10, the
QFT analysis identified a number of questionnaire items that will be need to reexamined
carefully as part of the DR analysis. For these items, either minor changes or no changes will be
made in the DR questionnaire, so DR results could lead to consideration of changes to these
items in the 2015 main study instrument. For example, some sets of items moved from CAPI to
ACASI administration in the QFT instrument could be administered in CAPI in 2015. These
questionnaire items will be reexamined in the DR analysis for one or both of the following two
criteria:
•

the item missingness rate was significantly higher than the rates for the 2011 and
2012 quarters 3 and 4 comparison samples, and/or

•

the estimate produced from the item differed significantly from the estimates from the
2011 comparison data, the 2012 quarters 3 and 4 comparison data, or comparison data
from other surveys.

Table 10.1 provides lists of QFT estimates and questionnaire items and indicates which of the
two criteria were observed in the analysis. A few important points are worth noting about the
estimates and items listed in this table:
•

Although differences were observed for QFT estimates and the 2011 comparison
data, the 2012 quarters 3 and 4 comparison data, or comparison data from other
surveys, some of these observed differences were based on relatively small sample
sizes. Combining the QFT data with the DR data might improve the statistical power
for some of these estimates, but for other estimates statistical power might remain
limited in the DR analysis.

•

In addition, some differences observed between the QFT data and comparison data
were found only among specific age groups. Table 10.1 does not note each of the
specific age groups where differences were observed for each estimate or item
because the observed differences were considered sufficient to add the estimate to this
list. Detailed findings for item missingness rates were presented in Chapter 4, and
comparisons of estimates were presented in Chapters 6 through 9. However, some of
these significant differences between the QFT and comparison data occurred because
no QFT respondents in these age groups reported the characteristic of interest (e.g.,
past year or past month cocaine use); such estimates typically would be suppressed
because of low precision. If the DR sample also yields no English-language
respondents in these subgroups who reported the characteristic of interest, then
apparent significant differences between the combined QFT and DR data and the
comparison datasets could be an artifact of the small sample sizes in both field tests.

•

A number of the questionnaire items on this list were new in the QFT instrument,
significantly revised in the QFT instrument, or moved from one part of the instrument
to another (either being moved to a different module or moved from CAPI to ACASI
administration). For reference, Table C-1 in Appendix C indicates the type of change
for new, revised, or moved items and provides a brief description of each change.
296

Table 10.1 Questionnaire Items Identified from the QFT Analysis as Needing Reexamination in the
DR Analysis

QFT Estimate or Questionnaire Item1,2
Past year cocaine use
Past month cocaine use
Past year heroin use
Past month heroin use
Lifetime inhalants use
Past year smokeless tobacco use
Past month smokeless tobacco use
Lifetime use of any prescription drug
Past year use of any prescription drug
Past month serious psychological distress (SPD)
Are you now married, widowed, divorced, or
separated, or have you never married? (QD07)
Is anyone in your immediate family currently serving
in the U.S. military? (QD10d)
How many times in the past 12 months have you
moved? (QD13)
In what State did you live in 1 year ago today?
(QD13a)
Are you a full-time student or a part-time student?
(QD19)
During the past 30 days, how many whole days of
school did you miss because you were sick or
injured? (QD20)
During the past 30 days, how many whole days of
school did you miss because you skipped or "cut" or
just didn't want to be there? (QD21)
Did you work at a job or business at any time last
week? (QD26)
Did you work at a job or business at any time during
the past 12 months? (QD33)
How many different employers have you had in the
past 12 months? (QD36)
During the past 12 months, was there ever a time
when you did not have at least one job or business?
(QD37)
In how many weeks during the past 12 months did
you not have at least one job or business? (QD38)
In what year did you last work at a job or business?
(QD39a)
During the past 30 days, how many whole days of
work did you miss because you were sick or injured?
(QD40)

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4
No
No
No
No
No
No
No
No
No
No

See notes at end of table.

Estimate Was
Significantly Different
from Comparison Data5,6
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes

Yes7

N/A

Yes

N/A

Yes

N/A

Yes

No
(continued)

297

Table 10.1 Questionnaire Items Identified from the QFT Analysis as Needing Reexamination in the
DR Analysis (continued)

QFT Estimate or Questionnaire Item1,2
During the past 30 days, how many whole days of
work did you miss because you just didn't want to be
there? (QD41)
How many people work for your employer out of this
office, store, etc.? (QD42)
Currently covered by private health insurance?
(QHI06)
In [YEAR], did you receive Supplemental Security
Income or SSI? (QI03N)
In [YEAR], did you receive food stamps? (QI07N)
At any time during [YEAR], even for 1 month, did
you receive any cash assistance from a State or
county welfare program such as [TANFFILL]?
(QI08N)
In [YEAR], because of low income, did you receive
any other kind of nonmonetary welfare or public
assistance? (QI10N)
Before taxes and other deductions, was your total
personal income from all sources during [YEAR]
more or less than $20,000? (QI20N)
Of these income groups, which category best
represents [SAMPLE MEMBER] total personal
income during [YEAR]? (QI21A)

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Estimate Was
Significantly Different
from Comparison Data5,6

Yes

DR = dress rehearsal; N/A = not applicable; Q = question; QFT = Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Missing data include selection of responses of either "don't know" or "refused" for the question.
4
Item missingness rates for QFT questionnaire items were compared only with the 2011 main study data and the
2012 quarters 3 and 4 main study comparison data.
5
QFT estimates were compared with estimates from other survey data sources based on the comparability of the
survey design and questions. As detailed in Chapter 9, the other data sources used for comparing estimates
included the 2011 National Survey on Drug Use and Health (NSDUH) main study, the 2012 quarters 3 and 4
NSDUH main study, the 2010 National Ambulatory Medical Care Survey (NAMCS), the 2010 National Hospital
Ambulatory Medical Care Survey (NHAMCS), the 2011 Monitoring the Future (MTF), the 2011 National Health
Interview Survey (NHIS), the 2009-2010 National Health and Nutrition Examination Survey (NHANES), the 2011
American Community Survey (ACS), and the 2012 quarters 3 and 4 Current Population Survey (CPS).
6
Items marked N/A in this column indicate those for which the estimate from the item was not compared with any
of the other data sources listed in footnote 5. Item QD10 was a new question in the QFT; therefore, no estimates are
available from the 2011 NSDUH main study or the 2012 quarters 3 and 4 NSDUH main study for comparison.
Given the units of analysis reported for items QD13a and QD39a, indicators were not developed to compare QFT
estimates with any of the other data sources.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health,
2012.

298

References
Aldworth, J., Chromy, J. R., Foster, M., Heller, D., & Novak, S. (2005). Serious psychological
distress report. In 2004 National Survey on Drug Use and Health: Methodological resource book
(Section 16, prepared for the Substance Abuse and Mental Health Services Administration,
Office of Applied Studies, under Contract No. 283-03-9028, Deliverable No. 26,
RTI/0208726.187.013). Research Triangle Park, NC: RTI International.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders
(DSM-IV) (4th ed.). Washington, DC: Author.
Biemer, P., & Wang, K. H. (2006). National Survey on Drug Use and Health: Adjusting for
nonignorable nonresponse in the NSDUH using interview callback models (prepared for the
Substance Abuse and Mental Health Services Administration). Research Triangle Park, NC:
RTI International.
Blumberg, S. J., & Luke, J. V. (2012, December). Wireless substitution: Early release of
estimates from the National Health Interview Survey, January-June 2012. Retrieved from
http://www.cdc.gov/nchs/nhis/releases.htm
Center for Behavioral Health Statistics and Quality. (2011). Results from the 2010 National
Survey on Drug Use and Health: Summary of national findings (HHS Publication No. SMA
11-4658, NSDUH Series H-41). Rockville, MD: Substance Abuse and Mental Health Services
Administration.
Center for Behavioral Health Statistics and Quality. (2012a). Comparing and evaluating youth
substance use estimates from the National Survey on Drug Use and Health and other surveys
(HHS Publication No. SMA 12-4727, Methodology Series M-9). Rockville, MD: Substance
Abuse and Mental Health Services Administration.
Center for Behavioral Health Statistics and Quality. (2012b). Drug Abuse Warning Network,
2010: National estimates of drug-related emergency department visits (HHS Publication No.
SMA 12-4733, DAWN Series D-38). Rockville, MD: Substance Abuse and Mental Health
Services Administration.
Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health
Services Administration. (2012c). National Survey on Drug Use and Health: 2011 public use file
and codebook. Retrieved from http://www.icpsr.umich.edu/icpsrweb/SAMHDA/studies/34481
Center for Behavioral Health Statistics and Quality. (2012d). Results from the 2011 National
Survey on Drug Use and Health: Detailed tables. Rockville, MD: Substance Abuse and Mental
Health Services Administration.

299

Center for Behavioral Health Statistics and Quality. (2012e). Results from the 2011 National
Survey on Drug Use and Health: Summary of national findings (HHS Publication No. SMA
12-4713, NSDUH Series H-44). Rockville, MD: Substance Abuse and Mental Health Services
Administration.
Chang, L., & Krosnick, J. A. (2010). Comparing oral interviewing with self-administered
computerized questionnaires: An experiment. Public Opinion Quarterly, 74(1), 154-167.
doi: 10.1093/poq/nfp090
Chen, P., Cribb, D., Dai, L., Gordek, H., Laufenberg, J., Sathe, N., & Westlake, M. (2013).
Person-level sampling weight calibration. In 2011 National Survey on Drug Use and Health:
Methodological resource book (Section 12, prepared for the Substance Abuse and Mental Health
Services Administration under Contract No. HHSS283200800004C, Phase II, Deliverable No.
39, RTI/0211838.207.004). Research Triangle Park, NC: RTI International.
Currivan, D., Peytcheva, E., Rodriguez, S., Stolzenberg, S., Leiman, J., & Obscura, A. (2009).
Report on the contact materials focus group findings and recommendations (prepared for the
Substance Abuse and Mental Health Services Administration). Research Triangle Park, NC:
RTI International.
de Leeuw, E., Hox, J., & Kef, S. (2003). Computer-assisted self-interviewing tailored for special
populations and topics. Field Methods, 15(3), 223-251. doi: 10.1177/1525822X03254714
Folsom, R.E., & Singh, A. C. (2000). The generalized exponential model for sampling weight
calibration for extreme values, nonresponse, and poststratification. In Proceedings of the 2000
Joint Statistical Meetings, American Statistical Association, Survey Research Methods Section,
Indianapolis, IN (pp. 598-603). Alexandria, VA: American Statistical Association.
Gfroerer, J., Eyerman, J., & Chromy, J. (Eds.). (2002). Redesigning an ongoing national
household survey: Methodological issues (HHS Publication No. SMA 03-3768). Rockville, MD:
Substance Abuse and Mental Health Services Administration, Office of Applied Studies.
Gfroerer, J., Wright, D., & Kopstein, A. (1997). Prevalence of youth substance use: The impact
of methodological differences between two national surveys. Drug and Alcohol Dependence, 47,
19-30.
Grant, B. F., & Dawson, D. A. (2006). Introduction to the National Epidemiologic Survey on
Alcohol and Related Conditions. Alcohol Research & Health, 29(2), 74-78.
Jaspan, H. B., Flisher, A.J., Myer, L., Mathews, C., Seebregts, C., Berwick, J. R., Wood, R., &
Bekker, L. G. (2007). Brief report: Methods for collecting sexual behaviour information from
South African adolescents—A comparison of paper versus personal digital assistant
questionnaires. Journal of Adolescence, 30(2), 353-359.
Johnson, T., Fendrich, M., & Mackesy-Amiti, M. E. (2010). Computer literacy and the accuracy
of substance use reporting in an ACASI survey. Social Science Computer Review, 28, 515-523.

300

Johnston, L. D., O'Malley, P. M., Bachman, J. G., & Schulenberg, J. E. (2012a). Monitoring the
Future national survey results on drug use, 1975-2011. Volume I: Secondary school students.
Ann Arbor, MI: Institute for Social Research, University of Michigan.
Johnston, L. D., O'Malley, P. M., Bachman, J. G., & Schulenberg, J. E. (2012b). Monitoring the
Future national survey results on drug use, 1975-2011. Volume II: College students and adults
ages 19-50. Ann Arbor, MI: Institute for Social Research, University of Michigan.
Krosnick, J. A. (1991). Response strategies for coping with the cognitive demands of attitude
measures in surveys. Applied Cognitive Psychology, 5(3), 213-236. doi:
10.1002/acp.2350050305
Kroutil, L. A., Handley, W., & Bradshaw, M. R. (2012a). General principles and procedures for
editing drug use data in the 2010 NSDUH computer-assisted interview. In 2010 National Survey
on Drug Use and Health: Methodological resource book (Section 10, prepared for the Substance
Abuse and Mental Health Services Administration under Contract No. HHSS283200800004C,
Deliverable No. 39, RTI/0211838.107.003). Research Triangle Park, NC: RTI International.
Kroutil, L. A., Handley, W., Bradshaw, M. R., Chien, C., & Felts, B. J. (2012b). Procedures for
editing supplementary self-administered data in the 2010 NSDUH computer-assisted interview.
In 2010 National Survey on Drug Use and Health: Methodological resource book (Section 10,
prepared for the Substance Abuse and Mental Health Services Administration under Contract
No. HHSS283200800004C, Deliverable No. 39, RTI/0211838.107.003). Research Triangle Park,
NC: RTI International.
Kroutil, L., Vorburger, M., & Aldworth, J. (2007). Evaluation of the effects of new noncore drug
data on prevalence estimates in the National Survey on Drug Use and Health (NSDUH). In 2006
National Survey on Drug Use and Health: Methodological resource book (Section 18, prepared
for the Substance Abuse and Mental Health Services Administration under Contract No. 2832004-00022, Deliverable No. 39, RTI/0209009). Research Triangle Park, NC: RTI International.
Kroutil, L. A., Vorburger, M., Aldworth, J., & Colliver, J. D. (2010). Estimated drug use based
on direct questioning and open-ended questions: Responses in the 2006 National Survey on Drug
Use and Health. International Journal of Methods in Psychiatric Research, 19(2), 74-87.
doi: 10.1002/mpr.302
Langhaug, L. F., Cheung, Y. B., Pascoe, S. J. S., Mavhu, W., Chirawu, P., & Cowan, F. M.
(2007, July 30). Comparing 4 questionnaire delivery methods for collection of self reported
sexual behaviour data in rural Zimbabwean youth. Presented at the 17th meeting of the
International Society for Sexually Transmitted Diseases Research (ISSTDR), Seattle, WA.
Langhaug, L. F., Sherr, L., & Cowan, F. M. (2010). How to improve the validity of sexual
behavior reporting: Systematic review of questionnaire delivery modes in developing countries.
Tropical Medicine & International Health, 15(3), 362-381.
Lara, D., Strickler, J., Olavarrieta, C. D., & Ellertson, C. (2004). Measuring induced abortion in
Mexico: A comparison of four methodologies. Sociological Methods & Research, 32(4),
529-558. doi: 10.1177/0049124103262685
301

Morton, K. B., Martin, P. C., Shook-Sa, B. E., Chromy, J. R., & Hirsch, E. L. (2012). Sample
design report. In 2011 National Survey on Drug Use and Health: Methodological resource book
(Section 2, prepared for the Substance Abuse and Mental Health Services Administration under
Contract No. HHSS283200800004C, Phase II, Deliverable No. 8, RTI/0211838.203.004).
Research Triangle Park, NC: RTI International.
National Center for Health Statistics. (2012a). 2010 NAMCS micro-data file documentation.
Retrieved from http://www.cdc.gov/nchs/
National Center for Health Statistics. (2012b). 2010 NHAMCS micro-data file documentation.
Retrieved from http://www.cdc.gov/nchs/
National Center for Health Statistics. (2012c, June 12). Variance estimation and other analytic
issues, NHIS 2006-2011. Retrieved from http://www.cdc.gov/nchs/data/nhis/2006var.pdf
Pemberton, M., Bose, J., Kilmer, G., Kroutil, L., Forman-Hoffman, V., & Gfroerer, J. (2013).
CBHSQ Data Review: Comparison of NSDUH health and health care utilization estimates to
other national data sources. Rockville, MD: U.S. Department of Health and Human Services,
Substance Abuse and Mental Health Services Administration, Center for Behavioral Health
Statistics and Quality.
Peytchev, A., Conrad, F. G., Couper, M. P., & Tourangeau, R. (2010). Increasing respondents'
use of definitions in web surveys. Journal of Official Statistics, 26, 633-650.
Plummer, M. L., Ross, D. A., Wight, D., Changalucha, J., Mshana, G., Wamoyi, J., Todd, J.,
Anemona, A., Mosha, F. F., Obasi, A. I., & Hayes, R. J. (2004a). "A bit more truthful":
The validity of adolescent sexual behaviour data collected in rural northern Tanzania using five
methods. Sexually Transmitted Infections, 80(Suppl. 2), ii49-ii56.
Plummer, M. L., Wight, D., Ross, D. A., Balira, R., Anemona, A., Todd, J., Salamba, Z., Obasi,
A. I., Grosskurth, H., Changalunga, J., & Hayes, R. J. (2004b). Asking semi-literate adolescents
about sexual behaviour: The validity of assisted self-completion questionnaire (ASCQ) data in
rural Tanzania. Tropical Medicine & International Health, 9, 737-754.
Potdar, R., & Koenig, M. A. (2005). Does audio-CASI improve reports of risky behavior?
Evidence from a randomized field trial among young urban men in India. Studies in Family
Planning, 36(2), 107-116.
RTI International. (2008). SUDAAN®, Release 10.0 [computer software]. Research Triangle
Park, NC: Author.
Tourangeau, R., & Smith, T. W. (1996). Asking sensitive questions: The impact of data
collection mode, question format, and question context. Public Opinion Quarterly, 60(2),
275-304.
Tourangeau, R., & Yan, T. (2007). Sensitive questions in surveys. Psychological Bulletin,
133(5), 859-883.

302

Turner, C. F., Ku, L., Rogers, S. M., Lindberg, L. D., Pleck, J. H., & Sonenstein, F. L. (1998).
Adolescent sexual behavior, drug use, and violence: Increased reporting with computer survey
technology. Science, 280, 867-873. doi: 10.1126/science.280.5365.867
U.S. Census Bureau. (2013). U.S. Census 2010: Interactive population map. Retrieved from
http://www.census.gov/2010census/popmap/
U.S. Food and Drug Administration. (2013). Drugs. Retrieved from
http://www.fda.gov/Drugs/default.htm
van den Brakel, J. A., Vis-Visschers, R., & Schmeets, J. J. G. (2013). An experiment with data
collection modes and incentives in the Dutch Family and Fertility Survey for Young Moroccans
and Turks. BMC Medical Research Methodology, 13(11), 321-334. doi:
10.1177/1525822X06287533
Yan, T., Curtin, R., & Jans, M. (2010). Trends in income nonresponse over two decades. Journal
of Official Statistics, 26(1), 145-164.

303

304

Appendix A: Redesigned NSDUH Questionnaire and
Redesigned Contact Materials for the 2015 Partial Redesign

Overview of Changes in the Redesigned NSDUH Questionnaire
Module

Design 2015 Changes

Core Demographics




Beginning ACASI Section
Tutorial
Calendar
Tobacco
Alcohol
Marijuana
Cocaine
Crack
Heroin
Hallucinogens
Inhalants
*Methamphetamine
Pain Relievers
Tranquilizers
Stimulants
Sedatives

Special Drugs

Risk/Availability
Blunts
Substance Dependence and
Abuse
Special Topics
Market Information for
Marijuana
Prior Substance Use
Drug Treatment
Health Care
Adult Mental Health
Service Utilization
Social Environment

New military veterans questions added
Detailed education categories added
No changes

Combined and deleted variables to streamline the module

New electronic version introduced after ACASI Tutorial.

Combined smokeless sections

Changed binge definition to 4 or more drinks for females
No changes
No changes
No changes
No changes

Added Ketamine/Special K, DMT/AMT/Foxy, and Salvia divinorum

Added markers and air duster

New Methamphetamine module modeled after cocaine

New prescription drug modules created, including screeners

New prescription drug modules created, including screeners

New prescription drug modules created, including screeners

New prescription drug modules created, including screeners

Removed all Meth questions except SD10a and SD10b

Removed "Desoxyn, or Methedrine" from SD10a and SD10b

Removed Ketamine/Special K, DMT/AMT/Foxy, and Salvia Divinorum,
Ambien, Adderall

Included GHB

Changed SD10c to "any other drug"

Replaced all instances of "not prescribed for you or that you took only for the
experience or feeling it caused" with "not prescribed for you"

Added an introduction to SD05: "The computer recorded that you have used a
needle …"
No changes

Added medical marijuana questions

Revised stimulant questions to reflect separate methamphetamine and
prescription stimulant modules
No changes

Dropped entire module



Dropped all PD questions.
Revised methamphetamine questions to refer to stand-alone methamphetamine
module.

Dropped "which came first" questions
No changes

Added new extended module

Note – overall health question remained in Core Demographics.
No changes


Dropped SEN04 - # of times moved in past 5 years

A-1

Module
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health Service
Utilization
Adolescent Depression
Consumption of Alcohol
Back-End Demographics:
Education
Employment
*New: Back End ACASI
Household Roster
Proxy
information/decision
*Proxy Tutorial
Health Insurance

Design 2015 Changes
No changes

Dropped YE04 - # of times moved in past 5 years
No changes
No changes
No changes
No changes

Dropped all prescription drugs (Meth should remain) from "used with alcohol"
question (CA09)

Dropped 4+ binge questions for females
No changes

New disability items added before the education items and module

Moved to ACASI section

Moved to ACASI section

Dropped I&O questions


Dropped step relationships item

No changes


Verification
MHSS Recruitment Screens

Created new module to introduce proxy respondent to CAI program
No changes, but moved to ACASI section

Moved to ACASI section

Top response category revised

New cell phone/land line question added
No changes

Eliminated because no MHSS recruitment occurred as part of the QFT

FI Observation Questions



Income



Moved to tablet screening device

A-2

Redesigned NSDUH Lead Letter Questionnaire
UNITED STATES DEPARTMENT OF HEALTH & HUMAN SERVICES
ROCKVILLE, MD 20857

[NAME County/Parish/District] Resident at:
1234 Main Street
Anywhere, XX 12345
Dear [NAME County/Parish/District] Resident:
The U.S. Department of Health and Human Services is conducting a study called the National Survey on
Drug Use and Health. This study asks questions about use or non-use of alcohol, tobacco and other
substances. The study also asks about mental health and other health-related topics relevant for all people.
Since 1971, this information has been used by local, state and national agencies for planning and
providing treatment and prevention programs.
Your address was randomly chosen, through scientific methods, along with more than 200,000 others
across the country. RTI International, a nonprofit organization, was selected to conduct this study. Soon,
an RTI interviewer will be in your neighborhood to give you more information. The interviewer will carry
an identification card like the example shown below.
First, the interviewer will ask a few general questions. Then the interviewer may ask one or two members
of your household to complete the full interview. It is possible no one will be chosen to be interviewed. If
anyone is chosen and completes the full interview, he or she will receive $30 in cash.
By Federal law*, the answers you give will be kept confidential and will be used only for statistical
purposes.
Please share this information with any others in your household. Feel free to ask the interviewer any
questions you have about this study. More information is also available on the study website at:
http://nsduhweb.rti.org or you may contact us at 1-800-848-4079.
Your help is very important to this study’s success. Thank you for your cooperation.
Sincerely,

Joel Kennet, Ph.D.
National Study Director, DHHS

Ilona S. Johnson
National Field Director, RTI
You will be contacted by: ___________________________________________
Interviewer Name

*Confidentiality protected by the Confidential Information Protection and Statistical Efficiency Act of 2002 (PL 107-347)
Authorized by the U.S. Congress as part of Section 505 of the Public Health Service Act (42 USC 290aa4)
Approved by Office of Management and Budget (OMB Approval No. XXXX-XXXX)

XX10010052

A-3

A-4

A-5

A-6

Appendix B: Questionnaire Field Test Screening and
Interview Response Rates, by Sample Release and Age
Group and for Each State

Table B-1 2012 Questionnaire Field Test Weighted Screening and Interview Response Rates, by Sample Release and Age Group

B-1

Sample Release/Age Group
Overall
Sample Release
Wave 1 (9/1/12)
Wave 2 (9/28/12)
Added DUs
Age Group
12-17
18-25
26-34
35-49
50+
Sample Release x Age Group
Wave 1, 12-17
Wave 2, 12-17
Added, 12-17
Wave 1, 18-25
Wave 2, 18-25
Added, 18-25
Wave 1, 26-34
Wave 2, 26-34
Added, 26-34
Wave 1, 35-49
Wave 2, 35-49
Added, 35-49
Wave 1, 50+
Wave 2, 50+
Added, 50+

Weighted
DU
Screening
Response
Rate
83.58%

Total
Selected
DUs
5,358

Total
Eligible
DUs
4,623

Weighted
DU
Eligibility
Rate
86.24%

4,902
415
41

4,222
368
33

86.09%
88.48%
80.98%

3,548
259
30

84.59%
71.02%
91.34%

2,614
187
22

N/A
N/A
N/A
N/A
N/A

N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A

DU = dwelling unit; N/A = not applicable.

Total
Completed
Screeners
3,837

Weighted
Interview
Response
Rate
69.04%

Weighted
Overall
Response
Rate
57.71%

1,904
125
15

69.46%
63.78%
65.21%

58.75%
45.30%
59.56%

663
667
451
557
485

544
505
307
369
319

82.25%
75.26%
68.91%
66.32%
65.43%

N/A
N/A
N/A
N/A
N/A

616
46
1
620
33
14
417
31
3
513
42
2
448
35
2

508
35
1
471
25
9
285
19
3
341
27
1
299
19
1

82.62%
77.06%
100.00%
75.34%
78.29%
64.47%
69.06%
63.71%
100.00%
66.38%
66.40%
45.26%
66.22%
55.89%
50.00%

N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A

Total
Selected
2,823

Total
Respondents
2,044

Table B-2 2012 Questionnaire Field Test Unweighted Screening and Interview Response Rates, by Sample Release and Age Group

B-2

Total
Eligible
DUs
4,623

DU
Eligibility
Rate
86.28%

4,902
415
41

4,222
368
33

86.13%
88.67%
80.49%

3,548
259
30

84.04%
70.38%
90.91%

2,614
187
22

N/A
N/A
N/A
N/A
N/A

N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A

DU = dwelling unit; N/A = not applicable.

Total
Completed
Screeners
3,837

DU
Screening
Response
Rate
83.00%

Total
Selected
DUs
5,358

Interview
Response
Rate
72.41%

Overall
Response
Rate
60.09%

1,904
125
15

72.84%
66.84%
68.18%

61.21%
47.05%
61.98%

663
667
451
557
485

544
505
307
369
319

82.05%
75.71%
68.07%
66.25%
65.77%

N/A
N/A
N/A
N/A
N/A

616
46
1
620
33
14
417
31
3
513
42
2
448
35
2

508
35
1
471
25
9
285
19
3
341
27
1
299
19
1

82.47%
76.09%
100.00%
75.97%
75.76%
64.29%
68.35%
61.29%
100.00%
66.47%
64.29%
50.00%
66.74%
54.29%
50.00%

N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A

Total
Selected
2,823

Total
Respondents
2,044

Table B-3 2012 Questionnaire Field Test Weighted Screening and Interview Response Rates, by
State

State
Overall
AL
AZ
CA
CO
CT
FL
GA
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MT
NC
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
SC
TN
TX
UT
VA
WA
WI
WV

Total
Selected
DUs
5,358
127
72
533
124
108
450
137
230
170
30
85
140
107
75
46
207
72
47
22
102
84
28
155
20
51
326
254
119
16
308
64
112
284
102
190
162
132
67

Total
Eligible
DUs
4,623
85
66
482
117
93
364
125
189
127
28
67
117
103
71
42
186
65
44
21
87
75
28
134
16
45
277
210
100
15
278
53
99
260
85
185
139
98
47

Weighted
DU
Eligibility
Rate
86.24%
66.96%
91.66%
90.56%
94.34%
86.43%
80.31%
91.23%
82.15%
75.42%
92.75%
78.77%
83.66%
96.58%
94.67%
90.99%
89.85%
90.27%
93.56%
95.45%
85.30%
89.25%
100.00%
86.46%
80.00%
88.24%
84.98%
82.97%
83.48%
93.75%
90.28%
82.86%
88.38%
91.68%
83.55%
97.24%
85.80%
71.93%
70.15%

Total
Completed
Screeners
3,837
70
48
347
73
78
288
105
136
110
26
63
104
82
56
39
154
61
39
19
77
69
23
123
16
41
197
187
86
15
242
46
88
233
79
169
114
90
44

Weighted
DU
Screening
Response
Rate
83.58%
82.32%
72.82%
71.61%
62.31%
83.46%
79.01%
84.06%
71.85%
86.41%
92.18%
93.99%
88.91%
79.39%
78.88%
94.59%
82.81%
93.78%
88.63%
90.48%
88.47%
92.09%
82.14%
91.82%
100.00%
91.11%
71.08%
89.17%
86.28%
100.00%
87.07%
86.67%
88.92%
89.57%
92.87%
91.46%
82.03%
91.39%
93.61%

DU = dwelling unit.

B-3

Total
Selected
2,823
60
26
262
54
60
219
74
97
79
29
38
75
53
34
19
122
46
29
20
60
41
14
76
5
40
177
129
60
11
179
40
71
203
72
115
53
51
30

Total
Respondents
2,044
45
14
170
33
41
169
60
72
63
19
28
66
33
32
12
86
36
16
16
50
25
11
52
4
33
105
103
40
8
121
31
51
146
63
83
46
38
23

Weighted
Interview
Response
Rate
69.04%
66.68%
48.31%
59.99%
53.11%
56.67%
71.63%
81.55%
68.04%
74.13%
68.94%
72.09%
86.13%
64.88%
93.95%
63.02%
72.57%
76.87%
58.84%
82.53%
82.97%
52.86%
85.12%
72.13%
79.55%
85.79%
57.98%
73.94%
67.31%
69.91%
65.52%
82.07%
65.53%
65.90%
84.60%
69.95%
87.62%
70.17%
71.01%

Weighted
Overall
Response
Rate
57.71%
54.89%
35.18%
42.96%
33.09%
47.29%
56.60%
68.55%
48.88%
64.05%
63.55%
67.76%
76.58%
51.51%
74.11%
59.61%
60.09%
72.09%
52.15%
74.67%
73.40%
48.68%
69.92%
66.24%
79.55%
78.17%
41.21%
65.94%
58.08%
69.91%
57.05%
71.13%
58.27%
59.03%
78.56%
63.98%
71.88%
64.12%
66.47%

Table B-4 2012 Questionnaire Field Test Unweighted Screening and Interview Response Rates, by
State

State
Overall
AL
AZ
CA
CO
CT
FL
GA
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MT
NC
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
SC
TN
TX
UT
VA
WA
WI
WV

Total
Selected
DUs
5,358
127
72
533
124
108
450
137
230
170
30
85
140
107
75
46
207
72
47
22
102
84
28
155
20
51
326
254
119
16
308
64
112
284
102
190
162
132
67

Total
Eligible
DUs
4,623
85
66
482
117
93
364
125
189
127
28
67
117
103
71
42
186
65
44
21
87
75
28
134
16
45
277
210
100
15
278
53
99
260
85
185
139
98
47

DU
Eligibility
Rate
86.28%
66.93%
91.67%
90.43%
94.35%
86.11%
80.89%
91.24%
82.17%
74.71%
93.33%
78.82%
83.57%
96.26%
94.67%
91.30%
89.86%
90.28%
93.62%
95.45%
85.29%
89.29%
100.00%
86.45%
80.00%
88.24%
84.97%
82.68%
84.03%
93.75%
90.26%
82.81%
88.39%
91.55%
83.33%
97.37%
85.80%
74.24%
70.15%

Total
Completed
Screeners
3,837
70
48
347
73
78
288
105
136
110
26
63
104
82
56
39
154
61
39
19
77
69
23
123
16
41
197
187
86
15
242
46
88
233
79
169
114
90
44

DU
Screening
Response
Rate
83.00%
82.35%
72.73%
71.99%
62.39%
83.87%
79.12%
84.00%
71.96%
86.61%
92.86%
94.03%
88.89%
79.61%
78.87%
92.86%
82.80%
93.85%
88.64%
90.48%
88.51%
92.00%
82.14%
91.79%
100.00%
91.11%
71.12%
89.05%
86.00%
100.00%
87.05%
86.79%
88.89%
89.62%
92.94%
91.35%
82.01%
91.84%
93.62%

DU = dwelling unit.

B-4

Total
Selected
2,823
60
26
262
54
60
219
74
97
79
29
38
75
53
34
19
122
46
29
20
60
41
14
76
5
40
177
129
60
11
179
40
71
203
72
115
53
51
30

Total
Respondents
2,044
45
14
170
33
41
169
60
72
63
19
28
66
33
32
12
86
36
16
16
50
25
11
52
4
33
105
103
40
8
121
31
51
146
63
83
46
38
23

Interview
Response
Rate
72.41%
75.00%
53.85%
64.89%
61.11%
68.33%
77.17%
81.08%
74.23%
79.75%
65.52%
73.68%
88.00%
62.26%
94.12%
63.16%
70.49%
78.26%
55.17%
80.00%
83.33%
60.98%
78.57%
68.42%
80.00%
82.50%
59.32%
79.84%
66.67%
72.73%
67.60%
77.50%
71.83%
71.92%
87.50%
72.17%
86.79%
74.51%
76.67%

Overall
Response
Rate
60.09%
61.76%
39.16%
46.71%
38.13%
57.31%
61.06%
68.11%
53.41%
69.07%
60.84%
69.29%
78.22%
49.57%
74.23%
58.65%
58.36%
73.44%
48.90%
72.38%
73.75%
56.10%
64.54%
62.80%
80.00%
75.17%
42.19%
71.10%
57.33%
72.73%
58.84%
67.26%
63.85%
64.45%
81.32%
65.93%
71.18%
68.43%
71.77%

Appendix C: Missing Data Rates for New, Moved, or
Revised Items in the 2012 Questionnaire Field Instrument

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older

QFT Instrument Item1,2
Race (QD055)
Are you currently serving full-time in a
Reserve component? (V2b)
Have you ever served on active duty in
the United States Armed Forces or
Reserve components? (QD10a)
When did you serve on active duty in
the United States Armed Forces or
Reserve components? (QD10b15)
Did you ever serve on active duty in the
U.S. Armed Forces or Reserve
components in a military combat zone
or an area where you drew imminent
danger pay or hostile fire pay?
(QD10c)
What is the highest grade or year of
school you have completed? (QD11)
Previously served as a proxy for another
respondent? (PREVCOM)
Previously completed any part of this
interview yourself, including
answering questions on behalf of a
member of your household?
(PREVCOM2)
Use of "smokeless" tobacco such as
snuff, dip, chewing tobacco, or
"snus." (CG25)
How old were you the first time you
used "smokeless" tobacco? (CG26)
Did you first use "smokeless" tobacco
in [YEAR] or [YEAR]? CG26a
Did you first use "smokeless" tobacco
in [YEAR ]? (CG26b)
In what month in [CURRENT YEAR]
did you first use "smokeless" tobacco?
(CG26c)
In what month in [YEAR FROM
CG26a or CG26b] did you first use
"smokeless" tobacco? (CG26d)
During the past 30 days, have you used
"smokeless" tobacco? (CG27)
How long has it been since you last used
"smokeless" tobacco? (CG28)
See notes at end of table.

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Type of
Question
Data4
3
(unweighted)
(weighted)
Change
Description of Change (unweighted)
Added response
categories for Guamanian
R
or Chamorro and Samoan.
2,044
4
0.1
Added two questions
about serving in reserve
N
components.
4
0
0.0*
Added three questions
about active-duty U.S.
N
military service.
115
0
0.0
Added three questions
about active-duty U.S.
N
military service.
83
0
0.0*

Added three questions
about active-duty U.S.
military service.
Changed response
categories.
Added two questions to
determine if R had
previously served as a
proxy.

Added two questions to
determine if R had
previously served as a
proxy.

N
R

Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.

Edited to include all forms
of smokeless tobacco.

R
R
R

Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.

C-1

0.0*

2,044

0.0

1,351

0.0

0.0*

2,044

0.0

332

0.0

0.0*

2.2*

332

0.0

233

0.1
(continued)

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 30 days, on how many
days did you use "smokeless"
tobacco? (CG29)
During the past 30 days, on how many
days did you have [Insert #] or more
drinks on the same occasion? (AL08)

Type of
Change3
R
R

Ever used Ketamine (LS01i)

Ever used DMT, AMT, or Foxy (LS01j)

Ever used Salvia divinorum (LS01k)

How long has it been since you last used
Ketamine? (LS33)

How long has it been since you last used
DMT, AMT, or Foxy? (LS34)

How long has it been since you last used
Salvia divinorum? (LS35)

Have you ever, inhaled felt-tip pens,
felt-tip markers, or magic markers for
kicks or to get high? (IN01h1)
Have you ever inhaled computer
keyboard cleaner, also known as air
duster, for kicks or to get high?
(IN01ii)
Have you ever used
methamphetamine? (ME01)
How old were you the first time you
used methamphetamine? (ME02)
Did you first use methamphetamine in
[YEAR]? (ME03a)
In what month in [YEAR] did you first
use methamphetamine? (ME03c)
In what month in [YEAR] did you first
use methamphetamine? (ME03d)

N
N
N
N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)
Edited to include all forms
of smokeless tobacco.
Changed question
wording for women to "4
or more drinks."
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added these items to
measure time since last
use of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.
Added these items to
measure time since last
use of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.
Added these items to
measure time since last
use of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.
Added question to
measure use of felt-tip
pens, felt-tip markers, or
magic markers.
Added question to
measure use computer
keyboard cleaner, also
known as air duster.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.

See notes at end of table.

0.8*

916

0.7

2,044

0.2

2,044

0.2

2,044

0.3

0.0*

4.1*

0.0*

2,044

0.0

2,044

0.0

2,044

0.1

112

0.0

0.0*

0.0*
(continued)

C-2

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How long has it been since you last
used methamphetamine?
(MELAST3)
How many days you’ve used
methamphetamine during the past 12
months. (MEFRAME3)
How many days in the past 12 months
did you use methamphetamine?
(MEYRAVE)
How many days did you use
methamphetamine each month during
the past 12 months? (MEMONAVE)
How many days did you use
methamphetamine each week during
the past 12 months? (MEWKAVE)
During the past 30 days, on how many
days did you use methamphetamine?
(ME06)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR015)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR025)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR035)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR045)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR055)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR065)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR075)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR085)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR095)
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR105)
In the past 12 months, have you used
any other prescription pain reliever?
(PR11)
Have you ever used any prescription
pain reliever?
(PR12)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)

Added to measure use of
methamphetamine.

112

0.0

Added to measure use of
methamphetamine.

0.0*

Added to measure use of
methamphetamine.

0.0*

Added to measure use of
methamphetamine.

0.0*

Added to measure use of
methamphetamine.

0.0*

2,044

0.6

2,044

0.4

2,044

0.4

2,044

0.4

2,044

0.4

2,044

0.3

2,044

0.3

2,044

0.3

2,044

0.3

2,044

0.3

2,044

0.3

1,311

0.9

N
N
N
N
N
N
N
N
N
N
N
N
N

Added to measure use of
methamphetamine.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.
Added questions to
indicate use of
prescription pain relievers.

See notes at end of table.

(continued)

C-3

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, which, if any, of
these tranquilizers have you used?
(TR015)
In the past 12 months, which, if any, of
these tranquilizers have you used?
(TR025)
In the past 12 months, which, if any, of
these tranquilizers have you used?
(TR035)
In the past 12 months, which, if any, of
these tranquilizers have you used?
(TR045)
In the past 12 months, which, if any, of
these tranquilizers have you used?
(TR055)
In the past 12 months, have you used
any other prescription tranquilizer?
(TR06)
Have you ever, even once, used any
prescription tranquilizer?
(TR07)
In the past 12 months, which, if any, of
these stimulants have you used?
(ST015)
In the past 12 months, which, if any, of
these stimulants have you used?
(ST025)
In the past 12 months, which, if any, of
these stimulants have you used?
(ST035)
In the past 12 months, which, if any, of
these stimulants have you used?
(ST045)
In the past 12 months, which, if any, of
these stimulants have you used?
(ST055)
In the past 12 months, have you used
any other prescription stimulant?
(ST06)
Have you ever, even once, used any
prescription stimulant? (ST07)
In the past 12 months, which, if any, of
these sedatives have you used?
(SV015)
In the past 12 months, which, if any, of
these sedatives have you used?
(SV025)
In the past 12 months, which, if any, of
these sedatives have you used?
(SV035)
In the past 12 months, which, if any, of
these sedatives have you used?
(SV045)

Type of
Change3
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)
Added questions to
indicate use of
prescription tranquilizers.
2,044
11
0.2
Added questions to
indicate use of
prescription tranquilizers.
2,044
10
0.2
Added questions to
indicate use of
prescription tranquilizers.
2,044
10
0.2
Added questions to
indicate use of
prescription tranquilizers.
2,044
10
0.2
Added questions to
indicate use of
prescription tranquilizers.
2,044
11
0.2
Added questions to
indicate use of
prescription tranquilizers.
2,044
2
0.3
Added questions to
indicate use of
prescription tranquilizers.
1,763
6
0.2
Added questions to
indicate use of
prescription stimulants.
2,044
11
0.2
Added questions to
indicate use of
prescription stimulants.
2,044
11
0.2
Added questions to
indicate use of
prescription stimulants.
2,044
10
0.2
Added questions to
indicate use of
prescription stimulants.
2,044
11
0.2
Added questions to
indicate use of
prescription stimulants.
2,044
12
0.3
Added questions to
indicate use of
prescription stimulants.
2,044
6
0.4
Added questions to
indicate use of
prescription stimulants.
1,885
4
0.1
Added questions to
indicate use of
prescription sedatives.
2,044
11
0.2
Added questions to
indicate use of
prescription sedatives.
2,044
10
0.2
Added questions to
indicate use of
prescription sedatives.
2,044
10
0.2
Added questions to
indicate use of
prescription sedatives.
2,044
9
0.2

See notes at end of table.

(continued)

C-4

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, which, if any, of
these sedatives have you used?
(SV055)
In the past 12 months, have you used
any other prescription sedative?
(SV06)
Have you ever used any prescription
sedative? (SV07)
Have you ever, even once, used any
prescription pain reliever in any way
a doctor did not direct you to use it?
(PRL01)
In the past 12 months, did you use
Vicodin in any way a doctor did not
direct you to use it? (PRY01)
How old were you when you first used
Vicodin in a way a doctor did not
direct you to use it? (PRY01a)
Did you first use Vicodin in a way a
doctor did not direct you to use it in
[YEAR]? (PRY01b)
Did you first use Vicodin in a way a
doctor did not direct you to use it in
[YEAR? (PRY01c)
In what month in [PRYFU1] did you
first use Vicodin in a way a doctor did
not direct you to use it? (PRY01d)
In the past 12 months, did you use
Lortab in a way a doctor did not
direct you to use it? (PRY02)
How old were you when you first used
Lortab in a way a doctor did not
direct you to use it? (PRY02a)
Did you first use Lortab in a way a
doctor did not direct you to use it in
[YEAR]? (PRY02b)
Did you first use Lortab in a way a
doctor did not direct you to use it in
[YEAR]? (PRY02c)
In what month in [PRYFU2] did you
first use Lortab in a way a doctor did
not direct you to use it? (PRY02d)
In the past 12 months, did you use
Lorcet in any way a doctor did not
direct you to use it? (PRY03)
How old were you when you first used
Lorcet in a way a doctor did not direct
you to use it? (PRY03a)
In what month in [PRYFU3] did you
first use Lorcet in a way a doctor
did not direct you to use it? (PRY03d)

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Type of
Question
Data4
3
(unweighted)
(weighted)
Change
Description of Change
(unweighted)
Added questions to
indicate use of prescription
N
sedatives.
2,044
10
0.2
Added questions to
indicate use of prescription
N
sedatives.
2,044
3
0.3
Added questions to
indicate use of prescription
N
sedatives.
1,913
8
0.2

N
N
N
N
N
N
N
N
N
N
N
N
N
N

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

See notes at end of table.

431

0.0

243

0.0

1.6*

0.0*

107

0.5

1.9*

0.0*

0.0*
(continued)

C-5

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
hydrocodone in any way a doctor did
not direct you to use it? (PRY04)
How old were you when you first used
hydrocodone in a way a doctor did
not direct you to use it? (PRY04a)
Did you first use hydrocodone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY04b)
Did you first use hydrocodone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY04c)
In what month in [PRYFU4] did you
first use hydrocodone in a way a
doctor did not direct you to use it?
(PRY04d)
In the past 12 months, did you use
OxyContin in any way a doctor did
not direct you to use it? (PRY05)
How old were you when you first used
OxyContin in a way a doctor did not
direct you to use it? (PRY05a)
Did you first use OxyContin in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY05b)
Did you first use OxyContin in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY05c)
In what month in [PRYFU5] did you
first use OxyContin in a way a doctor
did not direct you to use it? (PRY05d)
In the past 12 months, did you use
Percocet in any way a doctor did not
direct you to use it? (PRY06)
How old were you when you first used
Percocet in a way a doctor did not
direct you to use it? (PRY06a)
Did you first use Percocet in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY06b)
In what month in [PRYFU6] did you
first use Percocet in a way a doctor
did not direct you to use it? (PRY06d)
In the past 12 months, did you use
Percodan in any way a doctor did not
direct you to use it? (PRY07)
How old were you when you first used
Percodan in a way a doctor did not
direct you to use it? (PRY07a)

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

12.8*

0.0*

13.4*

132

0.4

0.0*

9.2*

0.0*

Type of
Change3
N

Added questions to
indicate misuse of
prescription pain relievers.

N
N

N
N
N
N

See notes at end of table.

(continued)

C-6

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Did you first use Percodan in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY07b)
In what month in [PRYFU7] did you
first use Percodan in a way a doctor
did not direct you to use it? (PRY07d)
In the past 12 months, did you use
Tylox in any way a doctor did not
direct you to use it? (PRY08)
How old were you when you first used
Tylox in a way a doctor did not direct
you to use it? (PRY08a)
In the past 12 months, did you use
oxycodone in any way a doctor did
not direct you to use it? (PRY09)
How old were you when you first used
oxycodone in a way a doctor did not
direct you to use it? (PRY09a)
Did you first use oxycodone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY09b)
Did you first use oxycodone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY09c)
In what month in [PRYFU9] did you
first use oxycodone in a way a doctor
did not direct you to use it? (PRY09d)
How old were you when you first used
Darvocet in a way a doctor did not
direct you to use it? (PRY10)
How old were you when you first used
Darvocet in a way a doctor did not
direct you to use it? (PRY10a)
In the past 12 months, did you use
Darvon in any way a doctor did not
direct you to use it? (PRY11)
In the past 12 months, did you use
propoxyphene in any way a doctor
did not direct you to use it? (PRY12)
How old were you when you first used
propoxyphene in a way a doctor did
not direct you to use it? (PRY12a)
In the past 12 months, did you use
Ultram in any way a doctor did not
direct you to use it? (PRY13)
How old were you when you first used
Ultram in a way a doctor did not
direct you to use it? (PRY13a)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Added questions to
indicate misuse of
prescription pain relievers.

N
N
N
N
N

N
N
N
N
N
N
N
N
N

See notes at end of table.

0.0*

128

0.4

0.0*

24.7*

0.0*

1.3*

0.0*
(continued)

C-7

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Did you first use Ultram in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY13b)
Did you first use Ultram in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY13c)
In what month in [PRYFU13] did you
first use Ultram in a way a doctor did
not direct you to use it? (PRY13d)
In the past 12 months, did you use
Ultram ER in any way a doctor did
not direct you to use it? (PRY14)
In the past 12 months, did you use
Ultracet in any way a doctor did not
direct you to use it? (PRY15)
How old were you when you first used
Ultracet in a way a doctor did not
direct you to use it? (PRY15a)
Did you first use Ultracet in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY15b)
In what month in [PRYFU15] did you
first use Ultracet in a way a doctor
did not direct you to use it? (PRY15d)
In the past 12 months, did you use
Ryzolt in any way a doctor did not
direct you to use it? (PRY16)
In the past 12 months, did you use
tramadol in any way a doctor did not
direct you to use it? (PRY17)
How old were you when you first used
tramadol in a way a doctor did not
direct you to use it? (PRY17a)
Did you first use tramadol in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY17b)
In what month in [PRYFU17] did you
first use tramadol in a way a doctor
did not direct you to use it? (PRY17d)
In the past 12 months, did you use
Tylenol with codeine 3 or 4 in any
way a doctor did not direct you to use
it? (PRY18)
How old were you when you first used
Tylenol with codeine 3 or 4 in a way
a doctor did not direct you to use
it? (PRY18a)

Type of
Change3

N
N
N
N
N

0.0*

35.4*

0.0*

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

11.3*

Added questions to
indicate misuse of
prescription pain relievers.

234

0.9

Added questions to
indicate misuse of
prescription pain relievers.

2.4*

See notes at end of table.

(continued)

C-8

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Did you first use Tylenol with codeine
3 or 4 in a way a doctor did not direct
you to use it in [CURRENT YEAR –
1] or [CURRENT YEAR]? (PRY18b)
Did you first use Tylenol with codeine
in a way a doctor did not direct you to
use it in [CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY18c)
In what month in [PRYFU18] did you
first use Tylenol with codeine in a
way a doctor did not direct you to use
it? (PRY18d)
In the past 12 months, did you use
codeine pills in any way a doctor did
not direct you to use them? (PRY19)
How old were you when you first used
codeine pills in a way a doctor did not
direct you to use them? (PRY19a)
Did you first use codeine pills in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY19b)
In what month in [PRYFU19] did you
first use codeine pills in a way a
doctor did not direct you to use it?
(PRY19d)
In the past 12 months, did you use
Avinza in any way a doctor did not
direct you to use it? (PRY20)
In the past 12 months, did you use
Kadian in any way a doctor did not
direct you to use it? (PRY21)
How old were you when you first used
Kadian in a way a doctor did not
direct you to use it? (PRY21a)
In the past 12 months, did you use MS
Contin in any way a doctor did not
direct you to use it? (PRY22)
In the past 12 months, did you use
morphine in any way a doctor did not
direct you to use it? (PRY24)
How old were you when you first used
morphine in a way a doctor did not
direct you to use it? (PRY24a)
Did you first use morphine in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY24b)
In what month in [PRYFU24] did you
first use morphine in a way a doctor
did not direct you to use it? (PRY24d)
In the past 12 months, did you use
Actiq in any way a doctor did not
direct you to use it? (PRY25)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

12.1*

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

N
N

N
N
N
N
N
N
N

N
N
N

See notes at end of table.

(continued)

C-9

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
Duragesic in any way a doctor did not
direct you to use it? (PRY26)
In the past 12 months, did you use
Fentora in any way a doctor did not
direct you to use it? (PRY27)
In the past 12 months, did you use
fentanyl in any way a doctor did not
direct you to use it? (PRY28)
How old were you when you first used
fentanyl in a way a doctor did not
direct you to use it? (PRY28a)
Did you first use fentanyl in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY28b)
In what month in [PRYFU28] did you
first use fentanyl in a way a doctor
did not direct you to use it? (PRY28d)
In the past 12 months, did you use
Suboxone in any way a doctor did not
direct you to use it? (PRY29)
How old were you when you first used
Suboxone in a way a doctor did not
direct you to use it? (PRY29a)
Did you first use Suboxone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY29b)
In what month in [PRYFU29] did you
first use Suboxone in a way a doctor
did not direct you to use it? (PRY29d)
In the past 12 months, did you use
Subutex in any way a doctor did not
direct you to use it? (PRY30)
How old were you when you first used
Subutex in a way a doctor did not
direct you to use it? (PRY30a)
Did you first use Subutex in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY30b)
Did you first use Subutex in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (PRY30c)
In what month in [PRYFU30] did you
first use Subutex in a way a doctor
did not direct you to use it? (PRY30d)
In the past 12 months, did you use
buprenorphine in any way a doctor
did not direct you to use it? (PRY31)

Type of
Change3
N
N
N
N

N
N
N
N

Added questions to
indicate misuse of
prescription pain relievers.

N
N
N

See notes at end of table.

0.0*

0.0*
(continued)

C-10

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How old were you when you first used
buprenorphine in a way a doctor did
not direct you to use it? (PRY31a)
In the past 12 months, did you use
Demerol in any way a doctor did not
direct you to use it? (PRY32)
How old were you when you first used
Demerol in a way a doctor did not
direct you to use it? (PRY32a)
Did you first use Demerol in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY32b)
In what month in [PRYFU32] did you
first use Demerol in a way a doctor
did not direct you to use it? (PRY32d)
In the past 12 months, did you use
Dilaudid in any way a doctor did not
direct you to use it? (PRY33)
How old were you when you first used
Dilaudid in a way a doctor did not
direct you to use it? (PRY33a)
Did you first use Dilaudid in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY33b)
In what month in [PRYFU33] did you
first use Dilaudid in a way a doctor
did not direct you to use it? (PRY33d)
In the past 12 months, did you use
methadone in any way a doctor did
not direct you to use it? (PRY34)
How old were you when you first used
methadone in a way a doctor did not
direct you to use it? (PRY34a)
Did you first use methadone in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY34b)
In what month in [PRYFU34] did you
first use methadone in a way a doctor
did not direct you to use it? (PRY34d)
In the past 12 months, did you use
Opana in any way a doctor did not
direct you to use it? (PRY35)
How old were you when you first used
Opana in a way a doctor did not direct
you to use it? (PRY35a)
Did you first use Opana in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY35b)

Type of
Change3
N
N
N

N
N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)
Added questions to
indicate misuse of
prescription pain relievers.
1
0
0.0*
Added questions to
indicate misuse of
prescription pain relievers.
14
0
0.0*
Added questions to
indicate misuse of
prescription pain relievers.
2
0
0.0*
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

Added questions to
indicate misuse of
prescription pain relievers.

N
N
N

See notes at end of table.

0.0*

0.0*
(continued)

C-11

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In what month in [PRYFU35] did you
first use Opana in a way a doctor
did not direct you to use it? (PRY35d)
In the past 12 months, did you use
Opana ER in any way a doctor did not
direct you to use it? (PRY36)
How old were you when you first used
Opana ER in a way a doctor did not
direct you to use it? (PRY36a)
Did you first use Opana ER in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY36b)
In what month in [PRYFU36] did you
first use Opana ER in a way a doctor
did not direct you to use it? (PRY36d)
In the past 12 months, did you use
Talwin in any way a doctor did not
direct you to use it? (PRY38)
How old were you when you first used
Talwin in a way a doctor did not
direct you to use it? (PRY38a)
Did you first use Talwin in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (PRY38b)
In what month in [PRYFU38] did you
first use Talwin in a way a doctor
did not direct you to use it? (PRY38d)
In the past 12 months, did you use
Talwin NX in any way a doctor did
not direct you to use it? (PRY39)
In the past 12 months, did you use any
prescription pain reliever in a way a
doctor did not direct you to use it?
(PRY40)
How old were you when you first used
any prescription pain reliever in a
way a doctor did not direct you to use
it? (PRY40a)
Did you first use any prescription pain
reliever in a way a doctor did not
direct you to use it in [CURRENT
YEAR - 1] or [CURRENT YEAR]?
(PRY40b)
Did you first use any prescription pain
reliever in a way a doctor did not
direct you to use it in [CURRENT
YEAR - 2] or [CURRENT YEAR –
1]? (PRY40c)
In what month in [PRYFU40] did you
first use any prescription drug in a
way a doctor did not direct you to use
it? (PRY40d)

Type of
Change3
N
N
N

N
N
N
N

0.0*

100.0*

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

180

0.7

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

N
N

See notes at end of table.

(continued)

C-12

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Have you ever used any prescription
pain reliever in any way a doctor did
not direct you to use it? (PRL02)
In the past 30 days, did you use
[PRNAMEFILL] in any way a doctor
did not direct you to use? (PRM01)
During the past 30 days, on how many
days did you use [PRNAMEFILL] in
any way a doctor did not direct you to
use? (PRM02)
During the past 30 days, did you use
[PRNAMEFILL] in any way a doctor
did not direct you to use while you
were drinking alcohol or within a
couple of hours of drinking?(PRM03)
Which of these statements describe
your use of [PRNAMEFILL] at any
time in the past 12 months? (PRY415)
Which of these pain relievers did you
use the last time? (PRY42A)
What were the reasons you used
[PRLASTFILL2] that time?
(PRYMOTIV5)
Which was the main reason you used
[PRLASTFILL2] that time?
(PRYMOT1)
How did you get the [PRLASTFILL]?
(PRY42B)
How did your friend or relative get the
[PRLASTFILL]? (PRY42C)
Have you ever, even once, used any
prescription tranquilizer in any way a
doctor did not direct you to use it?
(TRL01)
In the past 12 months, did you use
Xanax in any way a doctor did not
direct you to use it? (TRY01)
How old were you when you first used
Xanax in a way a doctor did not direct
you to use it? (TRY01a)
Did you first use Xanax in a way a
doctor did not direct you to use it in
[YEAR]? (TRY01b)
In what month in [TRYFU1] did you
first use Xanax in a way a doctor did
not direct you to use it? (TRY01d)
In the past 12 months, did you use
Xanax XR in a way a doctor did not
direct you to use it? (TRY02)

Added questions to
indicate misuse of
prescription pain relievers.

Type of
Change3
N

N
N
N
N
N
R
R

N
N
N
N
N
N

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added "fill" and moved
from the noncore prior
substance use module.
Added "fill" and moved
from the noncore prior
substance use module.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

See notes at end of table.

0.0*

156

3.2

2.6*

149

1.2

0.0*

156

1.2

137

0.0

102

0.0

0.0*

11.1*

0.0*
(continued)

C-13

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How old were you when you first used
Xanax XR in a way a doctor did not
direct you to use it? (TRY02a)
In the past 12 months, did you use
alprazolam in any way a doctor did
not direct you to use it? (TRY03)
How old were you when you first used
alprazolam in a way a doctor did not
direct you to use it? (TRY03a)
Did you first use alprazolam in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY03b)
In what month in [TRYFU3] did you
first use alprazolam in a way a doctor
did not direct you to use it? (TRY03d)
In the past 12 months, did you use
extended-release alprazolam in any
way a doctor did not direct you to use
it? (TRY04)
How old were you when you first used
extended-release alprazolam in a way
a doctor did not direct you to use it?
(TRY04a)
Did you first use extended-release
alprazolam in a way a doctor did not
direct you to use it in [CURRENT
YEAR - 1] or [CURRENT YEAR]?
(TRY04b)
In what month in [TRYFU4] did you
first use extended-release alprazolam
in a way a doctor did not direct you to
use it? (TRY04d)
In the past 12 months, did you use
Ativan in any way a doctor did not
direct you to use it? (TRY05)
How old were you when you first used
Ativan in a way a doctor did not
direct you to use it? (TRY05a)
In what month in [TRYFU5] did you
first use Ativan in a way a doctor
did not direct you to use it? (TRY05d)
In the past 12 months, did you use
Klonopin in any way a doctor did not
direct you to use it? (TRY06)
How old were you when you first used
Klonopin in a way a doctor did not
direct you to use it? (TRY06a)
Did you first use Klonopin in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY06b)

Type of
Change3
N
N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

100*

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

N
N
N
N
N

See notes at end of table.

24.7*

0.0*

(continued)

C-14

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In what month in [TRYFU6] did you
first use Klonopin in a way a doctor
did not direct you to use it? (TRY06d)
In the past 12 months, did you use
lorazepam in any way a doctor did not
direct you to use it? (TRY07)
How old were you when you first used
lorazepam in a way a doctor did not
direct you to use it? (TRY07a)
Did you first use lorazepam in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY07b)
In what month in [TRYFU7] did you
first use lorazepam in a way a doctor
did not direct you to use it? (TRY07d)
In the past 12 months, did you use
clonazepam in any way a doctor did
not direct you to use it? (TRY08)
How old were you when you first used
clonazepam in a way a doctor did not
direct you to use it? (TRY08a)
Did you first use clonazepam in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY08b)
In what month in [TRYFU8] did you
first use clonazepam in a way a doctor
did not direct you to use it? (TRY08d)
In the past 12 months, did you use
Valium in any way a doctor did not
direct you to use it? (TRY09)
How old were you when you first used
Valium in a way a doctor did not
direct you to use it? (TRY09a)
Did you first use Valium in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY09b)
Did you first use Valium in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (TRY09c)
In what month in [TRYFU9] did you
first use Valium in a way a doctor
did not direct you to use it? (TRY09d)
In the past 12 months, did you use
Librium in any way a doctor did not
direct you to use it? (TRY10)
How old were you when you first used
Librium in a way a doctor did not
direct you to use it? (TRY10a)

Type of
Change3
N
N
N

N
N
N
N

Added questions to
indicate misuse of
prescription tranquilizers.

N
N
N

N
N
N
N

See notes at end of table.

0.0*

0.0*
(continued)

C-15

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
Tranxene in any way a doctor did not
direct you to use it? (TRY11)
In the past 12 months, did you use
diazepam in any way a doctor did not
direct you to use it? (TRY12)
How old were you when you first used
diazepam in a way a doctor did not
direct you to use it? (TRY12a)
Did you first use diazepam in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY12b)
In what month in [TRYFU12] did you
first use diazepam in a way a doctor
did not direct you to use it? (TRY12d)
In the past 12 months, did you use
oxazepam, also known as Serax, in
any way a doctor did not direct you to
use it? (TRY13)
In the past 12 months, did you use
Flexeril in any way a doctor did not
direct you to use it? (TRY14)
How old were you when you first used
Flexeril in a way a doctor did not
direct you to use it? (TRY14a)
Did you first use Flexeril in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY14b)
Did you first use Flexeril in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (TRY14c)
In what month in [TRYFU14] did you
first use Flexeril in a way a doctor
did not direct you to use it? (TRY14d)
In the past 12 months, did you use
Soma in any way a doctor did not
direct you to use it? (TRY15)
How old were you when you first used
Soma in a way a doctor did not direct
you to use it? (TRY15a)
Did you first use Soma in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY15b)
In what month in [TRYFU15] did you
first use Soma in a way a doctor did
not direct you to use it? (TRY15d)

Type of
Change3
N
N
N

N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

N
N

N
N
N
N

N
N

See notes at end of table

0.0*

13.7*
(continued)

C-16

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
buspirone, also known as BuSpar, in
any way a doctor did not direct you to
use it? (TRY16)
How old were you when you first used
buspirone, also known as BuSpar, in
a way a doctor did not direct you to
use it? (TRY16a)
Did you first use buspirone, also known
as BuSpar, in a way a doctor did not
direct you to use it in [CURRENT
YEAR - 1] or [CURRENT YEAR]?
(TRY16b)
In what month in [TRYFU16] did you
first use buspirone, also known as
BuSpar, in a way a doctor did not
direct you to use it? (TRY16d)
In the past 12 months, did you use
hydroxyzine, also known as Atarax or
Vistaril, in any way a doctor did not
direct you to use it? (TRY17)
How old were you when you first used
hydroxyzine, also known as Atarax or
Vistaril, in a way a doctor did not
direct you to use it? (TRY17a)
In the past 12 months, did you use
meprobamate, also known as Equanil
or Miltown, in any way a doctor did
not direct you to use it? (TRY18)
How old were you when you first used
meprobamate, also known as Equanil
or Miltown, in a way a doctor did not
direct you to use it? (TRY18a)
Did you first use meprobamate, also
known as Equanil or Miltown, in a
way a doctor did not direct you to use
it in [CURRENT YEAR - 1] or
[CURRENT YEAR]? (TRY18b)
In what month in [TRYFU18] did you
first use meprobamate, also known as
Equanil or Miltown, in a way a doctor
did not direct you to use it? (TRY18d)
In the past 12 months, did you use any
prescription tranquilizer in a way a
doctor did not direct you to use it?
(TRY19)
How old were you when you first used
any prescription tranquilizer in a way
a doctor did not direct you to use
it? (TRY19a)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

100.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

100.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

See notes at end of table.

(continued)

C-17

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Have you ever, even once, used any
prescription tranquilizer in any way a
doctor did not direct you to use it?
(TRL02)
In the past 30 days, did you use
[TRNAMEFILL] in any way a doctor
did not direct you to use? (TRM01)
During the past 30 days, on how many
days did you use [TRNAMEFILL] in
any way a doctor did not direct you to
use? (TRM02)
During the past 30 days, did you use
[TRNAMEFILL] in any way a doctor
did not direct you to use while you
were drinking alcohol or within a
couple of hours of drinking? (TRM03)
Which of these statements describe
your use of [TRNAMEFILL] at any
time in the past 12 months? (TRY205)
Which of these tranquilizers did you
use the last time? (TRY21A)
What were the reasons you used
[TRLASTFILL2] that time?
(TRYMOTIV5)
Which was the main reason you used
[TRLASTFILL2] that time?
(TRYMOT1)
Please type in the other way you got the
[TRLASTFILL3] (TRY21B)
How did your friend or relative get the
[TRLASTFILL]? (TRY21C)
Have you ever, even once, used any
prescription stimulant in any way a
doctor did not direct you to use it?
(STL01)
In the past 12 months, did you use
Adderall in any way a doctor did not
direct you to use it? (STY01)
How old were you when you first used
Adderall in a way a doctor did not
direct you to use it? (STY01a)
Did you first use Adderall in a way a
doctor did not direct you to use it in
[YEAR]? (STY01b)
Did you first use Adderall in a way a
doctor did not direct you to use it in
[YEAR]? (STY01c)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

N
N
N
N
N
R
R

N
N
N
N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added "fill" and moved
from the noncore prior
substance use module.
Added "fill" and moved
from the noncore prior
substance use module.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

209

0.0

0.0*

2.1*

2.3*

0.0*

2.8*

0.0*

0.0*
(continued)

C-18

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In what month did you first use
Adderall in a way a doctor did not
direct you to use it? (STY01d)
In the past 12 months, did you use
Adderall XR in any way a doctor did
not direct you to use it? (STY02)
How old were you when you first used
Adderall XR in a way a doctor did not
direct you to use it? (STY02a)
Did you first use Adderall XR in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY02b)
Did you first use Adderall XR in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (STY02c)
In what month in [STYFU2] did you
first use Adderall XR in a way a
doctor did not direct you to use it?
(STY02d)
In the past 12 months, did you use
Dexedrine in any way a doctor did
not direct you to use it? (STY03)
How old were you when you first used
Dexedrine in a way a doctor did not
direct you to use it? (STY03a)
Did you first use Dexedrine in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY03b)
In what month in [STYFU3] did you
first use Dexedrine in a way a
doctor did not direct you to use it?
(STY03d)
In the past 12 months, did you use
dextroamphetamine in any way a
doctor did not direct you to use it?
(STY04)
How old were you when you first used
dextroamphetamine in a way a doctor
did not direct you to use it? (STY04a)
Did you first use dextroamphetamine in
a way a doctor did not direct you to
use it in [CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY04b)
In what month in [STYFU4] did you
first use dextroamphetamine in a way
a doctor did not direct you to use it?
(STY04d)

Added questions to
indicate misuse of
prescription stimulants.

8.6*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Type of
Change3
N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

See notes at end of table.

(continued)

C-19

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
mixed amphetamine
dextroamphetamine pills other than
Adderall in any way a doctor did not
direct you to use them? (STY05)
How old were you when you first used
mixed amphetamine
dextroamphetamine pills other than
Adderall in a way a doctor did not
direct you to use them? (STY05a)
Did you first use mixed amphetamine
dextroamphetamine pills other than
Adderall in a way a doctor did not
direct you to use it in [CURRENT
YEAR - 1] or [CURRENT YEAR]?
(STY05b)
In what month in [STYFU5] did you
first use mixed amphetamine
dextroamphetamine pills other than
Adderall in a way a doctor did not
direct you to use it? (STY05d)
In the past 12 months, did you use
Ritalin in any way a doctor did not
direct you to use it? (STY06)
How old were you when you first used
Ritalin in a way a doctor did not
direct you to use it? (STY06a)
Did you first use Ritalin in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY06b)
In what month in [STYFU6] did you
first use Ritalin in a way a doctor did
not direct you to use it? (STY06d)
In the past 12 months, did you use
Ritalin SR or Ritalin LA in any way a
doctor did not direct you to use it?
(STY07)
How old were you when you first used
Ritalin SR or Ritalin LA in a way a
doctor did not direct you to use it?
(STY07a)
In what month in [STYFU7] did you
first use Ritalin SR or Ritalin LA in a
way a doctor did not direct you to use
it? (STY07d)
In the past 12 months, did you use
Concerta in any way a doctor did not
direct you to use it? (STY08)
How old were you when you first used
Concerta in a way a doctor did not
direct you to use it? (STY08a)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

N
N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

N
N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

(continued)

C-20

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Did you first use Concerta in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY08b)
In what month in [STYFU8] did you
first use Concerta in a way a doctor
did not direct you to use it? (STY08d)
In the past 12 months, did you use
Daytrana in any way a doctor did not
direct you to use it? (STY09)
How old were you when you first used
Daytrana in a way a doctor did not
direct you to use it? (STY09a)
Did you first use Daytrana in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY09b)
In what month in [STYFU9] did you
first use Daytrana in a way a doctor
did not direct you to use it? (STY09d)
In the past 12 months, did you use
methylphenidate in any way a doctor
did not direct you to use it? (STY10)
How old were you when you first used
methylphenidate in a way a doctor did
not direct you to use it? (STY10a)
In what month in [STYFU10] did you
first use methylphenidate in a way a
doctor did not direct you to use it?
(STY10d)
In the past 12 months, did you use
Metadate CD in any way a doctor did
not direct you to use it? (STY11)
In the past 12 months, did you use
Metadate ER in any way a doctor did
not direct you to use it? (STY12)
In the past 12 months, did you use
Focalin in any way a doctor did not
direct you to use it? (STY13)
How old were you when you first used
Focalin in a way a doctor did not
direct you to use it? (STY13a)
Did you first use Focalin in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY13b)
In what month in [STYFU13] did you
first use Focalin in a way a doctor did
not direct you to use it? (STY13d)
In the past 12 months, did you use
Focalin XR in any way a doctor did
not direct you to use it? (STY14)

Type of
Change3

N
N
N
N

N
N
N
N
N

N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

0.0*

100.0*

0.0*

0.0*
(continued)

C-21

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How old were you when you first used
Focalin XR in a way a doctor did not
direct you to use it? (STY14a)
Did you first use Focalin XR in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY14b)
In what month in [STYFU14] did you
first use Focalin XR in a way a doctor
did not direct you to use it? (STY14d)
In the past 12 months, did you use
dexmethylphenidate in any way a
doctor did not direct you to use it?
(STY15)
How old were you when you first used
dexmethylphenidate in a way a doctor
did not direct you to use it? (STY15a)
In what month in [STYFU15] did you
first use dexmethylphenidate in a way
a doctor did not direct you to use it?
(STY15d)
In the past 12 months, did you use
benzphetamine in any way a doctor
did not direct you to use it? (STY16)
In the past 12 months, did you use
Didrex in any way a doctor did not
direct you to use it? (STY17)
In the past 12 months, did you use
diethylpropion in any way a doctor
did not direct you to use it? (STY18)
How old were you when you first used
diethylpropion in a way a doctor did
not direct you to use it? (STY18a)
In the past 12 months, did you use
phendimetrazine in any way a doctor
did not direct you to use it? (STY19)
In the past 12 months, did you use
phentermine in any way a doctor did
not direct you to use it? (STY20)
How old were you when you first used
phentermine in a way a doctor did not
direct you to use it? (STY20a)
Did you first use phentermine in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY20b)
In what month in [STYFU20] did you
first use phentermine in a way a
doctor did not direct you to use it?
(STY20d)
In the past 12 months, did you use
Provigil in any way a doctor did not
direct you to use it? (STY21)

Type of
Change3
N

N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

N
N
N
N
N
N
N

N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

0.0*

0.0*
(continued)

C-22

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In the past 12 months, did you use
Tenuate in any way a doctor did not
direct you to use it? (STY22)
How old were you when you first used
Tenuate in a way a doctor did not
direct you to use it? (STY22a)
In the past 12 months, did you use
Vyvanse in any way a doctor did not
direct you to use it? (STY23)
How old were you when you first used
Vyvanse in a way a doctor did not
direct you to use it? (STY23a)
Did you first use Vyvanse in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (STY23b)
Did you first use Vyvanse in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (STY23c)
In what month in [STYFU23] did you
first use Vyvanse in a way a doctor
did not direct you to use it? (STY23d)
In the past 12 months, did you use any
prescription stimulant in a way a
doctor did not direct you to use it?
(STY24)
How old were you when you first used
any prescription stimulant in a way a
doctor did not direct you to use it?
(STY24a)
Have you ever, even once, used any
prescription stimulant in any way a
doctor did not direct you to use it?
(STL02)
In the past 30 days, did you use
[STNAMEFILL] in any way a doctor
did not direct you to use? (STM01)
During the past 30 days, on how many
days did you use [STNAMEFILL’] in
any way a doctor did not direct you to
use? (STM02)
During the past 30 days, did you use
[STNAMEFILL] in any way a doctor
did not direct you to use while you
were drinking alcohol or within a
couple of hours of drinking?(STM03)
Which of these statements describe
your use of [STNAMEFILL] at any
time in the past 12 months? (STY255)
At any time in the past 12 months, did
you ever use a needle to inject
[STNAMEFILL]? (STY25a)

Added questions to
indicate misuse of
prescription stimulants.

Type of
Change3
N
N
N

0.0*

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

0.0*

Added questions to
indicate misuse of
prescription stimulants.

0.0*

100

0.4

0.0*

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

N
N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

(continued)

C-23

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How long has it been since you last
used a needle to inject
[STNAMEFILL]? (STY25b)
Which of these stimulants did you use
the last time? (STY26a)
What were the reasons you used
[STLASTFILL2] that time?
(STYMOTIV5)
Which was the main reason you used
[STLASTFILL2] that time?
(STYMOT1)
How did you get the [STLASTFILL]?
(STY26b)
How did your friend or relative get the
[STLASTFILL]? (STY26c)
Have you ever, even once, used any
prescription sedative in any way a
doctor did not direct you to use it?
(SVL01)
In the past 12 months, did you use
Ambien in any way a doctor did not
direct you to use it? (SVY01)
How old were you when you first used
Ambien in a way a doctor did not
direct you to use it? (SVY01a)
Did you first use Ambien in a way a
doctor did not direct you to use it in
[YEAR]? (SVY01b)
In what month in did you first use
Ambien in a way a doctor did not
direct you to use it? (SVY01d)
In the past 12 months, did you use
Ambien CR in a way a doctor did not
direct you to use it? (SVY02)
How old were you when you first used
Ambien CR in a way a doctor did not
direct you to use it? (SVY02a)
Did you first use Ambien CR in a way a
doctor did not direct you to use it in
[YEAR]? (SVY02b)
In what month in did you first use
Ambien CR in a way a doctor did not
direct you to use it? (SVY02d)
In the past 12 months, did you use
zolpidem in any way a doctor did not
direct you to use it? (SVY03)
How old were you when you first used
zolpidem in a way a doctor did not
direct you to use it? (SVY03a)

Type of
Change3
N
N
N
N
R
R

N
N
N
N
N
N
N
N
N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)
Added questions to
indicate misuse of
prescription stimulants.
1
0
0.0*
Added questions to
indicate misuse of
prescription stimulants.
33
2
4.4*
Added questions to
indicate misuse of
prescription stimulants.
58
0
0.0*
Added questions to
indicate misuse of
prescription stimulants.
26
0
0.0*
Added "fill" and moved
from the noncore prior
substance use module.
59
1
1.3*
Added "fill" and moved
from the noncore prior
substance use module.
29
1
1.8*
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

See notes at end of table.

112

0.0

0.0*

0.0*
(continued)

C-24

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Did you first use zolpidem in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (SVY03b)
In what month in [SVYFU3] did you
first use zolpidem in a way a doctor
did not direct you to use it? (SVY03d)
In the past 12 months, did you use
extended-release zolpidem in any way
a doctor did not direct you to use it?
(SVY04)
How old were you when you first used
extended-release zolpidem in a way a
doctor did not direct you to use it?
(SVY04a)
In the past 12 months, did you use
Lunesta in any way a doctor did not
direct you to use it? (SVY05)
How old were you when you first used
Lunesta in a way a doctor did not
direct you to use it? (SVY05a)
Did you first use Lunesta in a way
a doctor did not direct you to use it in
[CURRENT YEAR - 2] or
[CURRENT YEAR - 1]? (SVY05c)
In what month in [SVYFU5] did you
first use Lunesta in a way a doctor
did not direct you to use it?
(SVY05d)
In the past 12 months, did you use
Sonata in any way a doctor did not
direct you to use it? (SVY06)
How old were you when you first used
Sonata in a way a doctor did not
direct you to use it? (SVY06a)
In the past 12 months, did you use
Dalmane in any way a doctor did not
direct you to use it? (SVY08)
In the past 12 months, did you use
Halcion in any way a doctor did not
direct you to use it? (SVY09)
How old were you when you first used
Halcion in a way a doctor did not
direct you to use it? (SVY09a)
In the past 12 months, did you use
triazolam in any way a doctor did not
direct you to use it? (SVY11)
In the past 12 months, did you use
Restoril in any way a doctor did not
direct you to use it? (SVY12)
How old were you when you first used
Restoril in a way a doctor did not
direct you to use it? (SVY12a)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

N
N

N
N
N
N
N
N
N
N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

See notes at end of table.

0.0*

0.0*
(continued)

C-25

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
In what month in [SVYFU12] did you
first use Restoril in a way a doctor
did not direct you to use it? (SVY12d)
In the past 12 months, did you use
temazepam in any way a doctor did
not direct you to use it? (SVY13)
In the past 12 months, did you use
Butisol in any way a doctor did not
direct you to use it? (SVY14)
How old were you when you first used
Butisol in a way a doctor did not
direct you to use it? (SVY14a)
Did you first use Butisol in a way a
doctor did not direct you to use it in
[CURRENT YEAR - 1] or
[CURRENT YEAR]? (SVY14b)
In what month in [SVYFU14] did you
first use Butisol in a way a doctor
did not direct you to use it? (SVY14d)
In the past 12 months, did you use
Seconal in any way a doctor did not
direct you to use it? (SVY15)
How old were you when you first used
Seconal in a way a doctor did not
direct you to use it? (SVY15a)
In the past 12 months, did you use
phenobarbital in any way a doctor did
not direct you to use it? (SVY16)
How old were you when you first used
phenobarbital in a way a doctor did
not direct you to use it? (SVY16a)
In the past 12 months, did you use any
prescription sedative in a way a
doctor did not direct you to use it?
(SVY17)
How old were you when you first used
any prescription sedative in a way a
doctor did not direct you to use it?
(SVY17a)
Have you ever, even once, used any
prescription sedative in any way a
doctor did not direct you to use it?
(SVL02)
In the past 30 days, did you use
[SVNAMEFILL] in any way a doctor
did not direct you to use ? (SVM01)
During the past 30 days, on how many
days did you use [SVNAMEFILL] in
any way a doctor did not direct you to
use? (SVM02)

Type of
Change3
N
N
N
N

Added questions to
indicate misuse of
prescription sedatives.

0.2*

Added questions to
indicate misuse of
prescription sedatives.

0.0*

112

0.0

0.0*

N
N
N
N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

See notes at end of table.

0.0*

(continued)

C-26

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 30 days, did you use
[SVNAMEFILL] in any way a doctor
did not direct you to use while you
were drinking alcohol or within a
couple of hours of drinking?(SVM03)
Which of these statements describe
your use of [SVNAMEFILL] at any
time in the past 12 months? (SVY185)

Type of
Change3

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.
Added "fill" and moved
from the noncore prior
substance use module.
Added "fill" and moved
from the noncore prior
substance use module.
QFT SD15 is similar to
2012 SD10c, with edits to
the wording to ask about
any other drug and to
remove "only for the
experience or feeling that
it caused."

0.0*

2.2*

0.0*

2,044

0.0

New medical marijuana
questions in blunts module

344

0.0

New medical marijuana
questions in blunts module

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

Which of these sedatives did you use
the last time? (SVY19a)
What were the reasons you used
[SVLASTFILL2] that time?
(SVYMOTIV5)
Which was the main reason you used
[SVLASTFILL] that time?
(SVYMOT1)

How did you get the [SVLASTFILL]?
(SVY19B)

How did your friend or relative get the
[SVLASTFILL]? (SVY19C)

Have you ever, even once, used a
needle to inject any drug that was not
prescribed for you? (SD15)
Was any of your marijuana use in the
past 12 months recommended by a
doctor? (MJMM)
Was all of your marijuana use in the
past 12 months recommended by a
doctor? (MJMM01)
During the past 12 months, was there a
month or more when you spent a lot
of your time getting or using
methamphetamine? (DRME01)
During the past 12 months, was there a
month or more when you spent a lot
of your time getting over the effects
of the methamphetamine you used?
(DRME02)
During the past 12 months, did you try
to set limits on how often or how
much methamphetamine you would
use? (DRME04)
Were you able to keep to the limits you
set, or did you often use
methamphetamine more than you
intended to? (DRME05)

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

N
N

See notes at end of table.

(continued)

C-27

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 12 months, did you
need to use more methamphetamine
than you used to in order to get the
effect you wanted? (DRME06)
During the past 12 months, did you
notice that using the same amount of
methamphetamine had less effect on
you than it used to? (DRME07)
During the past 12 months, did you
want to or try to cut down or stop
using methamphetamine? (DRME08)
During the past 12 months, were you
able to cut down or stop using
methamphetamine every time you
wanted to or tried to? (DRME09)
During the past 12 months, have you
felt kind of blue or down when you
cut down or stopped using
methamphetamine? (DRME10)
During the past 12 months, have you
felt kind of blue or down when you
cut down or stopped using
methamphetamine? (DRME10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped using
methamphetamine? (DRME11)
During the past 12 months, did you
have 2 or more of these symptoms at
the same time that lasted for longer
than a day after you cut back or
stopped using methamphetamine?
(DRME12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of
methamphetamine? (DRME13)
Did you continue to use
methamphetamine even though you
thought it was causing you to have
problems with your emotions, nerves,
or mental health? (DRME14)
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of
methamphetamine? (DRME15)
During the past 12 months, did using
methamphetamine cause you to give
up or spend less time doing these
types of important activities?
(DRME17)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)
New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine
New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

See notes at end of table.

(continued)

C-28

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 12 months, did using
methamphetamine cause you to have
serious problems either at home,
work, or school? (DRME18)
During the past 12 months, did you
regularly use methamphetamine and
then do something where using
methamphetamine might have put
you in physical danger? (DRME19)
During the past 12 months, did using
methamphetamine cause you to do
things that repeatedly got you in
trouble with the law? (DRME20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused by
your use of methamphetamine?
(DRME21)
Did you continue to use
methamphetamine even though you
thought it caused problems with
family or friends? (DRME22)
During the past 12 months, was there a
month or more when you spent a lot
of your time getting or using
prescription stimulants? (DRST01)
During the past 12 months, was there a
month or more when you spent a lot
of your time getting over the effects
of the prescription stimulants you
used? (DRST02)
During the past 12 months, did you try
to set limits on how often or how
much prescription stimulants you
would use? (DRST04)
Were you able to keep to the limits you
set, or did you often use prescription
stimulants more than you intended to?
(DRST05)
During the past 12 months, did you
need to use more prescription
stimulants than you used to in order
to get the effect you wanted?
(DRST06)
During the past 12 months, did you
notice that using the same amount of
prescription stimulants had less effect
on you than it used to? (DRST07)
During the past 12 months, did you
want to or try to cut down or stop
using prescription stimulants?
(DRST08)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine

0.0*

New questions about
dependence and abuse of
methamphetamine
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

See notes at end of table.

(continued)

C-29

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 12 months, were you
able to cut down or stop using
prescription stimulants every time
you wanted to or tried to? (DRST09)
During the past 12 months, did you cut
down or stop using prescription
stimulants at least one time?
(DRST10)
During the past 12 months, have you
felt kind of blue or down when you
cut down or stopped using
methamphetamine? (DRME10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped using
prescription stimulants? (DRST11)
During the past 12 months, did you
have 2 or more of these symptoms at
the same time that lasted for longer
than a day after you cut back or
stopped using prescription
stimulants? (DRST12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of prescription
stimulants? (DRST13)
Did you continue to use prescription
stimulants even though you thought
this was causing you to have
problems with your emotions, nerves,
or mental health? (DRST14)
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of prescription
stimulants? (DRST15)
Did you continue to use prescription
stimulants even though this was
causing you to have physical
problems? (DRST16)
During the past 12 months, did using
prescription stimulants cause you to
give up or spend less time doing these
types of important activities?
(DRST17)
During the past 12 months, did using
prescription stimulants cause you to
have serious problems either at home,
work, or school? (DRST18)

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

Type of
Change3

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

See notes at end of table.

(continued)

C-30

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
During the past 12 months, did you
regularly use prescription stimulants
and then do something where using
prescription stimulants might have
put you in physical danger?(DRST19)
During the past 12 months, did using
prescription stimulants cause you to
do things that repeatedly got you in
trouble with the law? (DRST20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused by
your use of prescription stimulants?
(DRST21)
Did you continue to use prescription
stimulants even though you thought
this caused problems with family or
friends? (DRST22)

Type of
Change3

How old were you the last time you
used any methamphetamine for kicks
or to get high? (LU17)

Did you last use methamphetamine for
kicks or to get high in [YEAR]?
(LU17a)

Did you last use methamphetamine for
kicks or to get high in [YEAR]?
(LU17b)

In what month did you last use
methamphetamine for kicks or to get
high? (LU17c)

In what month in did you last use
methamphetamine for kicks or to get
high? (LU17d)

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
deleted from this module.
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
deleted from this module.
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
deleted from this module.
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
deleted from this module.
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
deleted from this module.

See notes at end of table.

0.0*

103

1.3

0.0*

0.0*
(continued)

C-31

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2

Type of
Change3

Height (HLTH04)
About how tall are you, without shoes
in feet? (HLTH05)
About how tall are you, without shoes
in inches? (HLTH06)
About how tall are you, without shoes
in meters? (HLTH07)
About how tall are you, without shoes
in centimeters? (HLTH08)

Weight (HLTH09)
About how much do you weigh in
pounds? (HLTH10)
About how much do you weigh in
kilograms? (HLTH12)
About how much did you weigh before
you got pregnant in pounds?
(HLTH13)
During the past 12 months, how many
times have you visited a doctor,
nurse, physician assistant or nurse
practitioner about your own health at
a doctor’s office, a clinic, or some
other place? (HLTH19)
During the past 12 months, did any
doctor or other health care
professional ask, either in person or
on a form, if you smoke cigarettes or
use any other tobacco products?
(HLTH20a)
During the past 12 months, did any
doctor or other health care
professional ask, either in person or
on a form, if you drink alcohol?
(HLTH20b)
During the past 12 months, did any
doctor or other health care
professional ask, either in person or
on a form, if you use illegal drugs?
(HLTH20c)
During the past 12 months, did any
doctor or other health care
professional advise you to quit
smoking cigarettes or quit using any
other tobacco products? (HLTH21)

N
N
N
N

N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)
New questions about
height and weight.
2,043
17
0.5
New questions about
height and weight.
1,926
5
0.1
New questions about
height and weight.
1,991
11
0.3
New questions about
height and weight.
20
1
3.1*
New questions about
height and weight.
29
2
3.6*
New questions about
height and weight.
2,043
25
0.9
New questions about
height and weight.
1,978
16
0.8
New questions about
height and weight.
14
1
4.4*

New questions about
height and weight.

0.0*

New questions about
health.

2,043

2.1

New questions about
health.

1,696

0.7

New questions about
health.

1,696

0.8

New questions about
health.

1,696

1.2

New questions about
health.

996

0.1

See notes at end of table.

(continued)

C-32

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Choose the statement or statements
below that describe any discussions
you may have had in person with a
doctor or other health professional
about your alcohol use. (HLTH225)
During the past 12 months, did any
doctor or other health care
professional talk to you about your
use of marijuana, cocaine, crack,
Heroin, inhalants, hallucinogens, or
methamphetamine? (HLTH23)
During the past 12 months, did you
have a sexually transmitted disease
such as chlamydia, gonorrhea, herpes
or syphilis? (HLTH24)
Conditions that a doctor or other health
care professional has ever told you
that you had (HLTH255)
What kind of cancer was it? (HLTH265)
How old were you when your blood
cancer was first diagnosed?
(HLTH28a)
How old were you when your bone
cancer was first diagnosed?
(HLTH28b)
How old were you when your brain
cancer was first diagnosed?
(HLTH28c)
How old were you when your breast
cancer was first diagnosed?
(HLTH28d)
How old were you when your cervical
cancer was first diagnosed?
(HLTH28e)
How old were you when your colon
cancer was first diagnosed?
(HLTH28f)
How old were you when your
esophageal cancer was first
diagnosed? (HLTH28g)
How old were you when your kidney
cancer was first diagnosed?
(HLTH28i)
How old were you when your leukemia
was first diagnosed? (HLTH28k)
How old were you when your lung
cancer was first diagnosed?
(HLTH28m)
How old were you when your
lymphoma was first diagnosed?
(HLTH28n)

Type of
Change3

Number of Case
Asked the
Number of Cases
Question with Missing Data Missing Data4
(weighted)
Description of Change
(unweighted)
(unweighted)

New questions about
health.

1,053

1.5

New questions about
health.

297

0.0

New questions about
health.

2,043

0.2

2,043

0.4

0.0*

New questions about
health.
New questions about
health.

New questions about
health.

82.1*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

See notes at end of table.

(continued)

C-33

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How old were you when your
melanoma was first diagnosed?
(HLTH28o)
How old were you when your ovarian
cancer was first diagnosed?
(HLTH28q)
How old were you when your
pancreatic cancer was first
diagnosed? (HLTH28r)
How old were you when your prostate
cancer was first diagnosed?
(HLTH28s)
How old were you when your skin [not
melanoma] cancer was first
diagnosed? (HLTH28u)
How old were you when your skin
cancer was first diagnosed?
(HLTH28v)
How old were you when your thyroid
cancer was first diagnosed?
(HLTH28aa)
How old were you when your uterine
cancer was first diagnosed?
(HLTH28bb)
How old were you when the type of
cancer listed below was first
diagnosed? (HLTH28cc)
Did you have cancer during the past 12
months? (HLTH29)
How old were you when your heart
condition or heart disease was first
diagnosed? (HLTH30)
Did you have any kind of heart
condition or heart disease in the past
12 months? (HLTH31)
How old were you when your diabetes
or sugar diabetes was first diagnosed?
(HLTH32)
How old were you when your chronic
bronchitis, emphysema, or chronic
obstructive pulmonary disease, also
called COPD were first diagnosed?
(HLTH33)
How old were you when your cirrhosis
of the liver was first diagnosed?
(HLTH34)
How old were you when your hepatitis
was first diagnosed? (HLTH35)
How old were you when your kidney
disease was first diagnosed?
(HLTH36)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

0.0*

New questions about
health.

124

1.4

New questions about
health.

118

0.8

New questions about
health.

109

2.1

New questions about
health.

0.4*

0.0*

New questions about
health.
New questions about
health.

3.7*

New questions about
health.

0.0*

See notes at end of table.

(continued)

C-34

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
How old were you when your asthma
was first diagnosed? (HLTH37)
Do you still have asthma? (HLTH38)
Are you currently taking prescription
medicine for your high blood
pressure? (HLTH40)
How old were you when your high
blood pressure was first diagnosed?
(HLTH41)
How many times in the past 12 months
have you moved? (QD13)
In what state did you live in one year
ago today? (QD13a)
Were you born in the United States?
(QD14)
Have you lived in the United States for
at least one year? (QD16a)
For how many years have you lived in
the United States? (QD16b)
For how many months have you lived
in the United States? (QD16c)
Are you now attending or are you
currently enrolled in school? (QD17)
What grade or year of school are you
now attending? (QD18)
Are you a full-time student or a part
time student? (QD19)
During the past 30 days, how many
whole days of school did you miss
because you were sick or injured?
(QD20)
During the past 30 days, how many
whole days of school did you miss
because you skipped or "cut" or just
didn't want to be there? (QD21)
Are you now married, widowed,
divorced or separated, or have you
never married? (QD07)
How many times have you been
married? (QD08)
Is anyone in your immediate family
currently serving in the United States
military? (QD10d)
See notes at end of table.

Type of
Change3
N
N
N

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)
New questions about
health.
256
24
5.2
New questions about
health.
256
7
1.4
New questions about
health.

199

0.0

153

5.9

2,043

0.8

618

0.7

2,043

0.0

239

0.3

227

0.0

19.7*

2,043

0.1

804

0.5

804

New questions about
health.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

690

1.4

Administered in ACASI
instead of CAPI.

597

1.5

Administered in ACASI
instead of CAPI.

1,778

0.4

859

0.2

2,043

N
M
M
M
M
M
M
M
M

M
N

Administered in ACASI
instead of CAPI.
New question on
immediate family serving
in the military.

C-35

0.9
(continued)

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Which member or members of your
immediate family are currently in the
United States military? (QD10e5)
Did you work at a job or business at
any time last week? (QD26)
Even though you did not work at any
time last week, did you have a job or
business? (QD27)
How many hours did you work last
week at all jobs or businesses?(QD28)
Do you usually work 35 hours or more
per week at all jobs or businesses?
(QD29)
Which one of these reasons best
describes why you did not work last
week? (QD30)
Which one of these reasons best
describes why you did not have a job
or business last week? (QD31)
During the past 30 days, did you make
specific efforts to find work? (QD32)
Did you work at a job or business at
any time during the past 12 months?
(QD33)
How many different employers have
you had in the past 12 months?
(QD36)
During the past 12 months, was there
ever a time when you did not have at
least one job or business? (QD37)
In how many weeks during the past 12
months did you not have at least one
job or business? (QD38)
In what year did you last work at a job
or business? (QD39a)
In what month did you last work at a
job or business? (QD39b)
During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)
During the past 30 days, how many
whole days of work did you miss
because you just didn’t want to be
there? (QD41)
Thinking about the location where you
work, how many people work for
your employer out of this office,
store, etc.? (QD42)

Type of
Change3
N
M

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change (unweighted)
New question on
immediate family serving
in the military.
143
20
8.9
Administered in ACASI
instead of CAPI.
1,778
6
0.2
747

0.5

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

1,025

0.3

Administered in ACASI
instead of CAPI.

1,129

0.2

Administered in ACASI
instead of CAPI.

104

0.0

643

0.8

156

0.0

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

649

0.6

Administered in ACASI
instead of CAPI.

1,066

0.8

Administered in ACASI
instead of CAPI.

1,129

0.3

249

4.3

643

5.2

175

0.7

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

1,129

0.6

Administered in ACASI
instead of CAPI.

1,129

0.5

Administered in ACASI
instead of CAPI.

1,129

1.1

M
M

See notes at end of table.

(continued)

C-36

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
At your workplace, is there a written
policy about employee use of alcohol
or drugs? (QD43)
Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)
At your workplace, have you ever been
given any educational information
regarding the use of alcohol or drugs?
(QD45)
Through your workplace, is there
access to any type of employee
assistance program or other type of
counseling program for employees
who have alcohol or drug-related
problems? (QD46)
Does your workplace ever test its
employees for alcohol use? (QD47)
Does your workplace ever test its
employees for drug use? (QD48)
Does your workplace test its employees
for drug or alcohol use as part of the
hiring process? (QD49)
Does your workplace test its employees
for drug or alcohol use on a random
basis? (QD50)
According to the policy at your
workplace, what happens to an
employee the first time he or she tests
positive for illicit drugs? (QD51)
Would you be more or less likely to
want to work for an employer that
tests its employees for drug use as
part of the hiring process? (QD52)
Would you be more or less likely to
want to work for an employer that
tests its employees for drug or alcohol
use on a random basis? (QD53)
How well do you speak English?
(QD55)
Are you deaf or do you have serious
difficulty hearing? (QD56)
Are you blind or do you have serious
difficulty seeing, even when wearing
glasses? (QD57)
Because of a physical, mental or
emotional condition, do you have
serious difficulty concentrating,
remembering, or making decisions?
(QD58)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Administered in ACASI
instead of CAPI.

1,129

3.0

Administered in ACASI
instead of CAPI.

858

0.4

Administered in ACASI
instead of CAPI.

1,129

0.4

1,129

7.7

1,129

3.2

1,129

3.0

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

530

1.2

Administered in ACASI
instead of CAPI.

530

3.7

Administered in ACASI
instead of CAPI.

530

11.3

Administered in ACASI
instead of CAPI.

1,129

0.5

Administered in ACASI
instead of CAPI.

1,129

0.3

New questions.

2,043

0.0

New questions.

2,043

0.1

New questions.

2,043

0.1

New questions.

2,043

0.2

M
M

See notes at end of table.

(continued)

C-37

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
Do you have serious difficulty walking
or climbing stairs? (QD59)
Do you have difficulty dressing or
bathing? (QD60)
Because of a physical, mental or
emotional condition, do you have
difficulty doing errands alone such as
visiting a doctors’ office or shopping?
(QD61)
[SAMPLE MEMBER A] covered by
Medicare? (QHI01)
You have indicated that [SAMPLE
MEMBER B] covered by Medicare,
which is a health insurance program
for persons aged 65 and older and for
certain disabled persons. Is this
correct? (QHI01v)
[SAMPLE MEMBER A] covered by
Medicaid? (QHI02)
You have indicated that [SAMPLE
MEMBER B] covered by Medicaid,
which is a public assistance program
that pays for medical care for low
income and disabled persons. Is this
correct? (QHI02v)
Is [SAMPLE MEMBER A] currently
covered by [CHIPFILL]? (QHI02a)
Is [SAMPLE MEMBER A] currently
covered by TRICARE, or
CHAMPUS, CHAMPVA, the VA, or
military health care? (QHI03)
Is [SAMPLE MEMBER A] currently
covered by private health insurance?
(QHI06)
Was [MEMBER] private health
insurance obtained through work,
such as through an employer, union,
or professional association? (QHI07)
Does [MEMBER] private health
insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse? (QHI09)
Does [MEMBER] private health
insurance include coverage for
treatment for mental or emotional
problems? (QHI10)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

New questions.

2,043

0.1

New questions.

2,043

0.0

New questions.
Administered in ACASI
instead of CAPI.

1,778

0.1

2,042

0.6

1.1*

2,042

0.8

M
M

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

0.0*

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

663

3.8

Administered in ACASI
instead of CAPI.

2,042

0.6

Administered in ACASI
instead of CAPI.

2,042

0.7

Administered in ACASI
instead of CAPI.

1,148

0.1

Administered in ACASI
instead of CAPI.

1,148

322

26.4

Administered in ACASI
instead of CAPI.

1,148

330

27.6

Administered in ACASI
instead of CAPI.

1,148

209

18.2

See notes at end of table.

(continued)

C-38

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)

QFT Instrument Item1,2
[MEMBER] currently covered by any
kind of health insurance, including
Indian Health Insurance? (QHI11)
During the past 12 months, was there
any time when [MEMBER] did not
have any kind of health insurance or
coverage? (QHI13)
During the past 12 months, about how
many months without any kind of
health insurance or coverage?
(QHI14)
About how long has it been since
[MEMBER] last had any kind of
health care coverage? (QHI15)
Which of these reasons is the main
reason why [MEMBER] stopped
being covered by health insurance?
(QHI17)
Which of these reasons describe why
[SAMPLE MEMBER] never had
health insurance coverage? (QHI185)
In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)
In [YEAR], did you receive
Supplemental Security Income or
SSI? (QI03N)
In [YEAR], did you receive income
from wages or pay earned while
working at a job or business?
(QI05N)
In [YEAR], did you receive food
stamps? (QI07N)
At any time during [YEAR], did you
receive any cash assistance from a
state or county welfare program such
as [TANFFILL]? (QI08N)
In [YEAR], because of low income, did
you receive any other kind of non
monetary welfare or public
assistance? (QI10N)
For how many months in [YEAR] did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance? (QI12AN)
For how many months in [YEAR] did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance, not including food
stamps? (QI12BN)
Before taxes and other deductions, was
your total personal income from all
sources during [YEAR] more or less
than 20,000 dollars? (QI20N)

Type of
Change3

Number of
Number of Cases
Cases Asked the with Missing
Missing Data4
Question
Data4
(unweighted)
(weighted)
Description of Change
(unweighted)

Administered in ACASI
instead of CAPI.

412

0.0

Administered in ACASI
instead of CAPI.

1,685

0.2

Administered in ACASI
instead of CAPI.

155

1.1

Administered in ACASI
instead of CAPI.

325

0.8

Administered in ACASI
instead of CAPI.

258

1.6

Administered in ACASI
instead of CAPI.

0.6*

New item.

2,042

Administered in ACASI
instead of CAPI.

2,042

1.5

2,042

1.1

2,042

0.5

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

2,042

Administered in ACASI
instead of CAPI.

2,042

0.6

Administered in ACASI
instead of CAPI.

3.6*

Administered in ACASI
instead of CAPI.

114

5.1

Administered in ACASI
instead of CAPI.

2,042

3.7

See notes at end of table.

(continued)

C-39

Table C-1 Item Missing Rates for New, Moved, or Revised Items in the 2012 Questionnaire Field Test among
Persons Aged 12 or Older (continued)
Number of Case
Asked the
Number of Cases
with Missing Data Missing Data4
Question
(weighted)
Description of Change
(unweighted)
(unweighted)

Type of
QFT Instrument Item1,2
Change3
Of these income groups, which
category best represents [MEMBER]
total personal income during
Administered in ACASI
[YEAR]? (QI21A)
M
instead of CAPI.
1,196
46
4.6
Of these income groups, which
category best represents [MEMBER]
total personal income during
Administered in ACASI
[YEAR]? (QI21B)
M
instead of CAPI.
769
24
3.6
Before taxes and other deductions, was
the total combined family income
during [YEAR] more or less than
Administered in ACASI
20,000 dollars? (QI22)
M
instead of CAPI.
1,131
91
9.5
Of these income groups, which
category best represents your total
combined family income during
Administered in ACASI
[YEAR]. (QI23A)
M
instead of CAPI.
365
27
9.7
Of these income groups, which
category best represents your total
combined family income during
Administered in ACASI
[YEAR] (QI23B)
M
instead of CAPI.
1,328
87
6.1
Is there at least one telephone at this
address that is not a cell phone?
(CELL1)
N
New item.
2,042
10
0.3
Do you or anyone at this address have a
working cell phone? (CELL2)
N
New item.
2,042
5
0.1
* Low precision; estimate would be suppressed due to not meeting the NSDUH sample size (N < 100) suppression rule.
ACASI = audio computer-assisted self-interviewing; CAPI = computer-assisted personal interviewing; QFT = Questionnaire Field Test;
R = respondent.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Changes to questionnaire items fall under three categories: N = new item, R= revised item, and M= no changes to item but moved to
another place in the questionnaire or moved from being interviewer-administered to self-administered.
4
Missing data include selection of responses of either "don't know" or "refused" for the question. "Missing Data (weighted)" denotes the
weighted percentage of missing data. Denominators for these percentages were based on the total number of cases (i.e., respondents)
who were asked the question.
5
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not asked
the question if all response options were coded as "blank" (e.g., 98 for 2-digit variables).
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health, 2012.

C-40

Appendix D: QFT Field Observation Materials –
Screening Checklist, QFT Field Observation Interview
Checklist, and Field Observer Reference Sheet

Revised 8/29/12

QFT Screening Observation Checklist
Directions: Complete one QFT Screening Observation Checklist for each screening you observe that ends in a code 22,
25, 26, 30, 31, or 32. For each screening procedure and summary item listed below, place a mark in the "Correct," "Error,"
or "N/A" column. For each Error or N/A response, provide a brief description in the space just below that item. If you
observe an error that does not fit any of the categories below, describe that error in item 21. You should complete this
checklist in hard copy using a clipboard or hard binder while at the household observing a screening. Within 24 hours you
should enter this information into the QFT Reporting Spreadsheet and email the spreadsheet to Jenna Gasperson.

Screening Case ID:

Date of Observation:

FI Name: ______________________________________________________________ FI ID:
Observer Name: __________________________________________________ Observer Title: _______________
SCREENING PROCEDURES OBSERVED
1. Displayed ID Badge prominently when knocking on door
2. On Tablet "Study Introduction" screen when reached door
3. Included all required information in introduction (Mark each item when spoken by FI)
FI Name
RTI International
U.S. Department of Health and Human Services
Lead Letter
4. If R didn't recall Lead Letter, FI offered one to R (gave QFT version of LL)
5. Confirmed SR was an adult resident of SDU (FI does not need to confirm age when it is
obvious SR is 18 or older)
6. Verified that he/she was at the correct address
7. Gave QFT Study Description to R
8. Read Tablet "Informed Consent" screen to R
9. Checked for missed DUs by reading the correct Tablet screen verbatim (This screen should
not be read at apartments/condos)
10. Asked all roster questions verbatim (Describe each roster question not read verbatim)
11. Recorded race based on R answer, not FI observation (If the SR refuses to answer for the
householder, the FI can record an answer based on his/her observation of the race of the SR)

D-1

Correct

Error

N/A

SCREENING PROCEDURES OBSERVED

Correct

Error

N/A

Correct

Error

N/A

12. Obtained all screening information directly from the SR (Not by observation or a proxy)
13. Confirmed accuracy & completeness of roster data w/ SR
14. For codes 22, 25, 26, or 30, correctly followed verification procedures
15. For code 31 or 32, presented project and interview information accurately
16. For code 31 or 32, demonstrated flexibility in scheduling interview time
SCREENING PROCEDURES OBSERVED (continued)
17. For code 31 or 32, left appropriate information about future interview (If R asks questions
or would like more information about the interview)
18. For code 31 or 32, made attempts to begin interview right away
19. Provided R with the correct QFT materials (did not substitute main study versions)
20. Answered R questions correctly and thoroughly, referencing the appropriate QFT details
[e.g., RTI International, DHHS, did not mention QFT or field test, sample size, pay or
payment (should use give or receive), etc.]
21. OTHER PROCEDURAL VIOLATION NOT NOTED ON THIS CHECKLIST:

SCREENING SUMMARY
22. Did the presentation flow well? If NO, describe:
23. Was visibility an issue when using the Tablet? If YES, describe:
24. Were there any issues with the equipment (Tablet, Tablet case)? If YES, describe:
25. Was there any difficulty using the Tablet keyboard? If YES, describe:
26. Was there any respondent confusion due to something the FI said or did? If YES, describe:
27. Was there any respondent confusion due to a procedure OR to the Tablet screening program itself? If YES, describe:
28. Was there any FI confusion due to the Tablet or screening program itself? If YES, describe:

D-2

SCREENING PROCEDURES OBSERVED

Correct

Error

N/A

Correct

Error

N/A

29. Were there any respondent comments on the contact materials?
30. Did the respondent make any comments about specific screening questions?

ADDITIONAL OBSERVER COMMENTS:

SEGMENT MAPS AND LISTS PROCEDURES OBSERVED
M1. Had segment maps readily available for reference while in the field (Either in the car or
located with screening and interviewing materials) NOTE: If you are unsure, wait until the
END of the observation and then ask the FI if he/she has the maps
M2. [IF THIS IS FI's FIRST VISIT TO THE DWELLING UNIT(s)] Used segment maps to
locate sample dwelling unit(s)
M3. [IF THIS IS FI's FIRST VISIT TO THE DWELLING UNIT(s)] Used the segment maps
and either the printed list of SDUs or the original list of dwelling units to check for missed
DUs in the interval between the SDU and the next listed dwelling unit
M4. [IF A MISSED DU IS FOUND] Used segment map and original list of dwelling units to
make sure the missed DU was not already listed

D-3

Revised 8/29/12

QFT Interviewing Observation Checklist
Directions: Complete one QFT Interviewing Observation Checklist for each interview you observe. For each interview
procedure and summary item listed below, place a mark in the "Correct," "Error," or "N/A" column. For each Error or
N/A response, provide a brief description in the space just below that item. If you observe an error that does not fit any of
the categories below, describe that error in item 14. You should complete this checklist in hard copy using a clipboard or
hard binder while at the household observing an interview. Within 24 hours you should enter this information into the
QFT Reporting Spreadsheet and email the spreadsheet to Jenna Gasperson.
Interview Case ID:

A / B (please circle A or B)

Date of Observation:
FI Name: _________________________________________________________ FI ID:
Observer Name: ______________________________________________ Observer Title: _______________
INTERVIEWING PROCEDURES OBSERVED
1. If IR was a minor, FI first obtained consent from parent or legal guardian
2. If IR was not SR, explained purpose of study and visit thoroughly
3. If IR was not SR, handed QFT STUDY DESCRIPTION to the respondent
4. Read INTRO TO CAI from QFT Showcard Booklet verbatim to respondent
5. Chose the most private available location
6. Set up equipment efficiently
7. Explained HEADPHONE usage, offered headphones to IR, and plugged in
8. Kept ACASI portion private (did not read ACASI), but remained attentive
9. Read all screens verbatim (Record the ID number of all questions not read verbatim
below)
10. Presented QFT SHOWCARDS when prompted by the CAI
11. Followed the proper QFT Quality Control Form and Incentive procedures
12. Answered IR questions correctly and thoroughly, referencing the appropriate QFT
details [e.g., RTI International, DHHS, did not mention QFT or field test, sample size,
pay or payment (should use give or receive), etc.]
13. Provided IR with the correct QFT materials (did not substitute main study versions)

D-4

Correct

Error

N/A

14. OTHER PROCEDURAL VIOLATION NOT NOTED ON THIS CHECKLIST:

INTERVIEWING SUMMARY
15. Did the respondent have trouble understanding any questions asked during the interview? If YES, describe:

16. Were there any issues with transition between the screening and the interview? If YES, describe:

17. Were there any issues with transition between the ACASI and CAPI sections of the interview? If YES, describe:

18. Was there any respondent confusion due to something the FI said or did? If YES, describe:

19. Was there any respondent confusion due to a procedure OR to the CAI instrument itself? If YES, describe:

20. Was there any FI confusion due to the CAI instrument? If YES, describe:

21. If a proxy was used, was there any confusion regarding their role, the equipment, adjusting the volume, etc.? If
YES, describe:

22. If a proxy was used, was there any difficulty understanding the ACASI tutorial? If YES, describe:

23. Was there any confusion when the FI was completing the debriefing questions on the Tablet?

D-5

24. Did the respondent or proxy make any comments about specific interview questions?

25. Did the respondent or FI make any comments about the length of the interview?

ADDITIONAL OBSERVER COMMENTS

D-6

Revised
8/03/12

NSDUH QFT Field Observations: Field Observer
Reference Sheet
QFT Field Observer Task List (Task number 0211838.102.003.006)
Please follow these steps while planning and conducting field observation trips. It is not necessary to actually
complete or submit this form; it is designed as a helpful tool so you do not skip any protocol steps.
Enter a check mark in the space provided as you complete each item.
A. TRAVEL PREPARATION
____

Receive Field Observation Assignment.

____

Contact the FI's Field Supervisor. Send the FS an email to obtain the FI's contact
information and other information that will be pertinent to planning your trip. In the email
request the following information:
_____ a) FI contact information (FI phone numbers can also be found in the FI Lookup
form the General Information link on the CMS)
_____ b) Location of segment and distance between FI segments
_____ c) Any other information the FS feels is significant
You should also request that the FS send a copy of the QFT FI Field Observations
Instructions to the FI and notify him/her that you will soon be in contact.

____

Contact the Field Interviewer. Call each FI and make plans for the observation. You will
need to discuss the following:
_____ a) Date most convenient for observation (Must be completed before September 17th)
_____ b) Workload – For how long will the FI have work?
_____ c) Segment information – Location of segment, type of attire needed
_____ d) Other information – Suggested hotels, coordinating transportation to segment
You should also confirm that the FS has sent a copy of the QFT FI instructions and tell the
FI that you will be spending the whole workday in the field with him/her. Let him/her know
that it is necessary to observe an interview and encourage him/her to set up an appointment
in advance of your arrival.

____

Once the date of observation has been determined, email your observation plans to Jenna
Gasperson, copying Gretchen McHenry, the managing FS, RS, and your supervisor. In the
email, include the dates you will observe each FI and any trip details associated with the
observation (dates you will fly, drive, return, etc.).

D-7

Are flight or hotel arrangements necessary?



YES (flights) continue with 6.
8.

YES (hotels)

continue with



NO Skip to Field Preparation.

Make flight and rental car arrangements with Carlson Wagonlit Travel (online or by
phone) at least 14 days prior to scheduled trip. You will need your Bank of America
number and task number (0211838.102.003.006) ready when calling. Before calling Carlson
Wagonlit, review flight options on Expedia and select the best and most reasonable flight in
terms of costs and time.

Immediately after booking your flight, send completed General Travel Information Form to
the NSDUH Secretaries, Jenna Gasperson and Gretchen McHenry, copying your
supervisor. A copy of the General Travel Information Form can be found on the
Downloadable Project Forms and Report Shells on the CMS.

Determine the government per diem and lodging rates for the area by clicking the 'US
Gov't Per Diems' link on the General Information page of the CMS. Please keep costs in
mind when identifying a hotel and when expensing meals.

Make hotel reservations at or under the given per diem. When looking for a place to stay,
search the internet for hotels in the area and/or gather FS and FI suggestions. You cannot
pay more than the official government rate. It is imperative that you verify the government
rate on the 'US Gov't Per Diems' link after the hotel tells you what their government rate is.
You should also try to find a hotel that includes free parking and internet. Call the hotel to
confirm these details before booking.

10.

Update the CMS travel Calendar (with dates of travel, hotel, and contact information), SRD
travel calendar, and your Outlook Calendar.

B. FIELD PREPARATION
____

Print the QFT forms from the email sent by the FO Manager:

____

QFT Field Observation FI Instructions Form: You should hand a copy of this form
to the FI when you meet him/her in the field. It contains the script the FI is to read
to the respondent when introducing you and your role as the observer.

____

QFT Field Observer Reference Sheet: This form outlines your role and
responsibilities as the observer.

____

NSDUH QFT Screening Scripts: Print and read through this file before going to the
field. Use the script while observing an FI conducting a screening so you can check
whether he/she reads the tablet screens verbatim. Note that there is an HU script
and a GQU script within this file.

____

NSDUH QFT CAI Script: Print and read through this file before going to the field.
Use the script to while observing an FI conducting an interview so you can check
whether he/she reads the CAI screens verbatim.
QFT Screening Observation Checklist: One copy of this form must be completed for
each screening case you observe than ends in a code 22, 25, 26, 30, 31, or 32. You
should complete this checklist in hard copy using a clipboard or hard binder while
at the household observing a screening. You should print at least 8 of these
checklists per FI to be observed.

D-8

____

QFT Interviewing Observation Checklist: One copy of this form must be completed
for each completed interview you observe. You should complete this checklist in
hard copy using a clipboard or hard binder while at the household observing an
interview. You should print at least 4 of these checklists per FI to be observed.

Make sufficient copies of both the screening and interviewing checklists before going into
the field (we recommend printing 8 screening checklists and 4 interviewing checklists per
FI).

C. AFTER THE OBSERVATION
____

Enter data from your checklists into the QFT Screening and Interview Report spreadsheets.
Please enter the results of all cases observed for all FIs in one screening and one interview
spreadsheet and e-mail to the FO Manager, Jenna Gasperson, within 24 hours of
completing all QFT FO assignments.

____

Send an e-mail to the FS, copying the RS, RD, and [NSDUH] QFT Field Observations
([email protected]), sharing positive feedback about the FI's performance
within 24 hours of completing your observation.

____

As soon as you have completed all of the field observations you will be conducting for the
quarter, please ship all completed hardcopy field observation checklists via United States
Postal Service or intra-office mail to Jenna Gasperson at RTI.

D-9

D-10

Appendix E: QFT Field Interviewer Debriefing Questions

Document Format:
• Screen names bolded
• Screen/question/instructional text designated by black and red text and non-italicized text in
parenthesis (Upper-lower black text to be read, red text is instructions to FI)
• Fills designated by parentheses and italics
• Logic designated by brackets
• Text of instructional message boxes provided in bracketed logic
• Response categories underlined
QFTDBF1 [IF SCREENING CALL RECORD = RESULT CODE 30, 31 or 32]
THESE QUESTIONS ARE FOR YOU TO ANSWER. DO NOT READ TO THE R.
Did the respondent remember receiving the Lead Letter?
YES
NO
Next [QFTDBF2]
QFTDBF2 [IF QFTDBF1 NE BLANK]
What comments, if any, did the respondent make about the Lead Letter or in response to the Lead
Letter? Check all that apply
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.

THE RESPONDENT DID NOT MAKE ANY COMMENTS ABOUT THE LEAD LETTER
R WAS LOOKING FORWARD TO YOUR VISIT/BEEN WAITING FOR YOU
R WAS INTERESTED IN THE STUDY
R WOULD LIKE TO PARTICIPATE IN THE STUDY
R DO NOT BELIEVE THE GOVERNMENT IS PAYING $30/WASTE OF TAX DOLLARS
THE LETTER ANSWERED THE R’S QUESTIONS/CONCERNS
R DID NOT WANT SOMEONE COMING TO MY HOME WITHOUT MY PERMISSION
R WAS CONFUSED BY THE LETTER
THE LETTER DID NOT ANSWER ALL OF THE R’S QUESTIONS/CONCERNS
R DOES NOT BELIEVE THE SURVEY IS CONFIDENTIAL
R THOUGHT THIS WAS A SCAM
R DOES NOT OPEN ANYTHING ADDRESSED TO “RESIDENT”
OTHER

Next [RECORD OF CALLS]
INTERVIEW DEBRIEFING QUESTIONS:
THESE QUESTIONS ARE FOR YOU TO ANSWER. DO NOT READ TO THE R.
QFTDBF3 [IF INTERVIEW A CALL RECORD OR INTERVIEW B CALL RECORD = RESULT
CODE 70]
When did you give the respondent (or parent/guardian of youth respondent) the Q&A Brochure?

E-1

1. BEFORE THE INTERVIEW
2. DURING THE INTERVIEW
3. AT THE END OF THE INTERVIEW
Next [QFTDBF3a]
QFTDBF3a [IF QFTDB3 NE BLANK]
What comments, if any, did the respondent (or parent/guardian) make about the Q&A Brochure?
Check all that apply
1. THERE WERE NO COMMENTS ABOUT THE Q&A BROCHURE
2. THE BROCHURE DID NOT ANSWER ALL OF THE RESPONDENT’S QUESTIONS ABOUT
THE STUDY.
3. THE BROCHURE ADDRESSED THE RESPONDENT’S QUESTIONS
4. RESPONDENT WAS CONFUSED BY THE BROCHURE.
5. THE BROCHURE ENCOURAGED THE RESPONDENT TO PARTICIPATE.
6. OTHER
Next [QFTDBF4]
QFTDBF4 [IF QFTDBF3a NE BLANK]
Did you conduct this interview at the respondent’s home, either inside or outside?
YES
NO
Next [IF QFTDBF4=YES, GO TO QFTDBF6]
QFTDBF5 [IF QFTDBF4=NO]
Where did you conduct this interview?
1. AT THE RESPONDENT’S WORKPLACE
2. AT THE HOME OF THE RESPONDENT’S RELATIVE OR FRIEND
3. IN SOME TYPE OF CONFERENCE ROOM IN A RESIDENCE HALL, SCHOOL OR
APARTMENT COMPLEX
4. AT A LIBRARY
5. IN SOME TYPE OF COMMON AREA, SUCH AS A LOBBY, HALLWAY, STAIRWELL, OR
LAUNDRY ROOM
6. SOME OTHER PLACE
Next [IF QFTDBF5=6, GO TO QFTDBF5a]
QFTDBF5a [IFQFTDBF5=6]
Where did the interview take place?

E-2

ALLOW 140 CHARACTERS
Next [QFTDBF6]
QFTDBF6 [IF QFTDBF4=YES; OR QFTDBF5=1, 2, 3, 4, OR 5; OR QFTDBF5a NE BLANK]
Please indicate how private the interview was. Do not count yourself or a project observer as another
person in the room.
1. COMPLETELY PRIVATE – NO ONE WAS IN THE ROOM OR COULD OVERHEAR ANY
PART OF THE INTERVIEW
2. MINOR DISTRACTIONS – PERSON(S) IN THE ROOM OR LISTENING LESS THAN 1/3 OF
THE TIME
3. PERSON(S) IN THE ROOM OR LISTENING ABOUT 1/3 OF THE TIME
4. SERIOUS INTERRUPTIONS OF PRIVACY MORE THAN HALF THE TIME
5. CONSTANT PRESENCE OF OTHER PERSON(S)
Next [IF QFTDBF6=1, GO TO QFTDBF9; IF QFTDBF6 NE1, GO TO QFTDBF7]
QFTDBF7 [IF QFTDBF6 NE1]
Not including yourself or project observers, other people present or listening to the interview were:
Check all that apply
1.
2.
3.
4.
5.
6.
7.

PARENT(S)
SPOUSE
LIVE-IN PARTNER/BOYFRIEND/GIRLFRIEND
OTHER ADULT RELATIVE(S)
OTHER ADULT(S)
CHILD(REN) UNDER 15
OTHER

Next [IF QFTDBF7=1, 2, 3, 4, 5, OR 6, GO TO QFTDBF9]
QFTDBF8 [IF QFTDBF7=7]
Please enter a description of the other person(s) present or listening to the interview. This description may
be relationship to the respondent if you have this information, or simply the gender and estimated age.
ALLOW 140 CHARACTERS
Next [QFTDBF9]
QFTDBF9 [IF QFTDBF6=1; OR IF QFTDBF7=1, 2, 3, 4, 5, OR 6; OR IF QFTDBF8 NE BLANK]
Did the respondent make any comments about the interview being too long?

E-3

YES
NO
Next [QFTDBF10]
QFTDBF10 [IF QFTDBF9 NE BLANK]
Did the respondent have any questions or comments about the Prescription Drug questions in the ACASI
section of the questionnaire?
YES
NO
Next [IF QFTDBF10 =NO, GO TO QFTDBF11]
QFTDBF10a [IF QFTDBF10= YES]
Please describe the respondent’s comments about the Prescription Drug questions.
ALLOW 140 CHARACTERS
Next [QFTDBF11]
QFTDBF11 [IF QFTDBF10 = NO OR QFTDBF10a NE BLANK]
Did the respondent have any questions or comments about the on-screen calendars in the ACASI section
of the questionnaire? If the respondent asked how to access the calendar at any time during the ACASI
portion of the interview, select “YES.”
YES
NO
Next [IF QFTDBF11=NO, GO TO QFTDBF12]
QFTDBF11a [IF QFTDBF11 = YES]
What comments did the respondent make about the on-screen calendars? Check all that apply
1.
2.
3.
4.
5.
6.

THE RESPONDENT ASKED HOW TO ACCESS THE CALENDAR.
THE RESPONDENT ASKED HOW TO CLOSE THE CALENDAR.
THE RESPONDENT DID NOT SEE THE REFERENCE DATES ON THE CALENDAR.
THE CALENDAR HELPED THE RESPONDENT ANSWER THE QUESTION.
THE CALENDAR COVERED THE QUESTIONS OR THE IMAGES ON THE SCREEN.
OTHER

Next [QFTDBF12]

E-4

QFTDBF12 [IF QFTDBF11=NO; OR IF QFTDBF11a NE BLANK]
Did the respondent have trouble understanding any other questions asked during the interview?
YES
NO
Next [IF QFTDBF12=NO, GO TO QFTDBF13]
QFTDBF12a [IF QFTDBF12=YES]
Enter the screen name and a brief description of what the respondent found confusing. If you do not
know the screen name, please provide as much information as possible.
ALLOW 140 CHARACTERS
Next [QFTDBF13]
QFTDBF13 [IF QFTDBF12=NO OR QFTDBF12a NE BLANK]
Was a proxy used for the income and health insurance questions?
YES
NO
Next [IF QFTDBF13=NO, GO TO RECORD OF CALLS]
QFTDBF14 [IF QFTDBF13=YES]
Did the respondent have any questions or concerns about his/ her answers being revealed to the proxy?
YES
NO
Next [QFTDBF15]
QFTDBF15 [IF QFTDBF14 NE BLANK]
Did the respondent have any other questions or comments about the proxy interview?
YES
NO
Next [IF QFTDBF15 =NO, GO TO QFTDBF16]
QFTDBF15a [IF QFTDBF15=YES]

E-5

Please describe the other questions or comments the respondent had about the proxy interview.
ALLOW 140 CHARACTERS
Next [QFTDBF16]

QFTDBF16 [IF QFTDBF15 =NO; OR QFTDBF15a NE BLANK]
Were there any problems with the proxy’s understanding of the ACASI tutorial?
YES
NO
Next [IF QFTDBF16 =NO, GO TO QFTDBF17]
QFTDBF16a [IF QFTDBF16=YES]
Which of the following describes the problems with the proxy’s understanding of the tutorial?
Check all that apply
1. THE PROXY DID NOT UNDERSTAND HOW TO ANSWER THE QUESTIONS.
2. THE PROXY DID NOT KNOW WHY HE/SHE WAS ASKED TO ANSWER THESE
QUESTIONS
3. OTHER
Next [IF QFTDBF16a=1 OR 2, GO TO QFTDBF17]
QFTDBF16b [IF QFTDBF16a=3]
Please describe the other problems with the proxy’s understanding of the tutorial.
ALLOW 140 CHARACTERS
Next [QFTDBF17]
QFTDBF17 [IF QFTDBF16a=1 OR 2; OR QFTDBF16b NE BLANK]
Were there any problems with the proxy’s use of ACASI to answer the income and health insurance
questions?
YES
NO
Next [IF QFTDBF17= NO, GO TO RECORD OF CALLS]
QFTDBF17a [IF QFTDBF17=YES]

E-6

Which of the following describes the problems with the proxy’s use of ACASI in answering the income
and health insurance questions? Check all that apply.
1.
2.
3.
4.

THE PROXY DID NOT KNOW THE ANSWERS TO THE QUESTIONS
THE PROXY DID NOT KNOW HOW TO ENTER HIS/HER ANSWERS TO THE QUESTIONS
THE PROXY REFUSED TO ANSWER SOME QUESTIONS
THE PROXY DID NOT KNOW WHY HE/SHE WAS ASKED TO ANSWER THESE
QUESTIONS
5. OTHER
Next [RECORD OF CALLS]

E-7

E-8

Appendix F: Complete Results from the
QFT New Equipment User Satisfaction Survey

The following tables provide field interviewer (FI) responses to each of the usability items compared between the August
2012 survey before the Questionnaire Field Test (QFT) data collection and the October 2012 survey after the QFT data
collection. Six FIs did not complete the second survey because they did not pass training, dropped out of the QFT after
training or did not work any QFT cases in the field. One FI was on medical leave at the time of the second survey
administration and was unable to complete the survey.
Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

58% (93)

26% (42)

14% (23)

1% (1)

160

QFT FI Survey 2

54% (83)

22% (34)

18% (27)

4% (6)

2% (3)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

50% (80)

39% (62)

9% (14)

2% (3)

1% (1)

160

QFT FI Survey 2

55% (84)

33% (50)

6% (9)

1% (1)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

38% (61)

40% (64)

15% (24)

6% (10)

1% (1)

160

QFT FI Survey 2

56% (85)

32% (49)

8% (12)

4% (6)

1% (1)

153

Q4. I like the layout of the
screening program.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

40% (64)

47% (75)

11% (17)

2% (3)

1% (1)

160

QFT FI Survey 2

44% (67)

36% (55)

9% (14)

8% (13)

3% (4)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

45% (72)

43% (68)

9% (15)

3% (4)

1% (1)

160

QFT FI Survey 2

62% (95)

31% (48)

4% (6)

2% (3)

1% (1)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

49% (79)

42% (67)

8% (13)

1% (1)

0% (0)

160

QFT FI Survey 2

63% (96)

32% (49)

4% (6)

1% (1)

153

Q1. I would like using the tablet on
a regular basis for my fieldwork.

Q2. The tablet is easy to use.

Q3. I can use the tablet without
needing technical assistance.

Q5. I learned to use the tablet
quickly.

Q6. I am able to efficiently
complete screenings using the
tablet.

F-1

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

42% (67)

41% (65)

15% (24)

3% (4)

0% (0)

160

QFT FI Survey 2

49% (75)

35% (54)

12% (18)

3% (5)

1% (1)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

45% (72)

44% (70)

9% (15)

1% (2)

1% (1)

160

QFT FI Survey 2

61% (94)

31% (48)

6% (9)

1% (1)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

34% (54)

43% (68)

13% (20)

11% (17)

1% (1)

160

QFT FI Survey 2

47% (72)

37% (57)

10% (16)

4% (6)

1% (2)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

17% (27)

36% (58)

41% (65)

6% (9)

1% (1)

160

QFT FI Survey 2

29% (45)

25% (38)

38% (58)

5% (8)

3% (4)

153

Q11. I can easily type ROC notes or
comments using the keyboard on the
tablet.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

38% (60)

48% (77)

11% (17)

3% (5)

1% (1)

160

QFT FI Survey 2

46% (71)

34% (52)

9% (14)

7% (11)

3% (5)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

11% (18)

23% (36)

31% (49)

30% (48)

6% (9)

160

QFT FI Survey 2

12% (19)

10% (15)

36% (55)

35% (53)

7% (11)

153

Q7. I find the tablet intuitive,
in that it's clear what I need to do.

Q8. I feel confident using the tablet.

Q9. I think veteran interviewers
will be able to use the tablet without
much training.

Q10. I think the tablet will work
well in a variety of weather
conditions such as sunshine,
rain and snow.

Q12. I prefer to move through
the screening program using
swipe gestures rather than the
Next or Previous buttons

F-2

Q13. I prefer to tap the screen
with my finger rather than use
a stylus.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

14% (22)

13% (21)

23% (37)

41% (66)

9% (14)

160

QFT FI Survey 2

16% (25)

8% (12)

20% (31)

43% (66)

12% (19)

153

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

29% (46)

49% (79)

17% (27)

4% (7)

1% (1)

160

QFT FI Survey 2

35% (53)

40% (61)

13% (20)

10% (15)

3% (4)

153

Q15. I am satisfied with the
design of the carrying case
provided for the tablet.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

QFT FI Survey 1

36% (57)

44% (70)

17% (24)

5% (8)

1% (1)

160

QFT FI Survey 2

35% (53)

37% (57)

15% (23)

9% (14)

4% (6)

153

Q14. The weight of the tablet
is suitable for screening at the
door.

F-3

The following tables provide FI responses to questions on QFT training from the August 2012 survey before the QFT data
collection.
QFT FI Survey 1
(August 2012)

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

1. Reading the QFT FI Handbook
helped me prepare for training.

53% (85)

43% (68)

4% (6)

1% (1)

0% (0)

160

2. Completing the QFT iLearning
course helped prepare me for training.

57% (91)

38% (60)

4% (6)

2% (3)

0% (0)

160

3. The overall pace of the QFT
Training Session was just right for me.

45% (72)

37% (59)

13% (20)

4% (7)

1% (2)

160

4. I feel ready to properly conduct
QFT screenings using the tablet.

60% (96)

36% (58)

4% (6)

0% (0)

160

5. I feel ready to properly conduct
QFT interviews using the tablet.

62% (99)

33% (53)

5% (8)

0% (0)

160

6. Overall, the training program has
prepared me to properly complete my QFT
tasks.

59% (94)

39% (62)

3% (4)

0% (0)

160

7. I enjoyed attending the QFT
Training Session.

59% (95)

7% (11)

0% (0)

160

34% (54)

QFT FI Survey 1
(August 2012)

Never

Rarely, When
Unusual
Situations Arise

2-3 Times a
Week

Each Day
with QFT
Work

Total

8. During the next month as you complete your
QFT work, how often do you think you will
reference the QFT FI Handbook?

0% (0)

41% (65)

19% (30)

160

F-4

The following tables provide FI responses to questions on QFT training from the October 2012 survey after the QFT data
collection.
QFT FI Survey 2
(October 2012)
1. The amount of background
information provided about the
QFT was just right.

Strongly
Agree
38% (58)

Agree

Neutral

Disagree

Strongly
Disagree

Total

50% (76)

10% (16)

1% (2)

1% (1)

153

8% (12)

5% (7)

2% (3)

153

2. The amount of training on the tablet was
just right.

39% (60)

3. The amount of training on transmission was
just right.

41% (63)

48% (73)

5% (8)

4% (6)

2% (3)

153

34% (52)

44% (67)

14% (21)

7% (11)

1% (2)

153

14% (22)

12% (19)

2% (3)

153

8% (13)

1% (1)

0% (0)

153

4. The amount of training on
equipment troubleshooting was
just right.
5. The amount of training on
administrative tasks (ePTEs,
ePTE Summary data in tablet, etc.)
was just right.
6. Overall, the QFT training program prepared
me to conduct my QFT tasks.

QFT FI Survey 2
(October 2012)
7. During the time since training
as you completed your QFT work,
how often did you reference the
QFT FI Handbook?

30% (46)

50% (77)

46% (71)

41% (63)

41% (62)

Never

Rarely, When
Unusual
Situations Arise

2-3 Times a
Week

Each Day
with QFT
Work

Total

20% (31)

65% (99)

12% (18)

3% (5)

153

F-5

The following two tables provide verbatim comments from FIs from the August 2012 survey before the QFT data collection and the October 2012
survey after the QFT data collection.
No.

Comments QFT FI Survey 1 (August 2012)

F-6

General Comments about Tablet/Screening Program
1
I really like the new tablet. It is user friendly & modern equipment material that wil enhance data collection in the field.
2
It seems to be fine, but have to try it out first on real cases
3
very nice, screen is clear
4
I love the new tablet and am looking forward to working with it soon
5
so far it seems to be ok, I will further test it next week
6
Great tool for in field use. Look forward to using it on a regular basis.
7
I feel that is more accurate, it gives feed back that I was not able to see in iPaq, or don't know how
8
Easy to work, more visible sign of cases information
9
Great step forward, seems more efficient.
10
I feel confortable using the tablet and I feel more efficient.
It is about the as large as a screening device should be, any larger and it woulb combersom. connecting to laptop is very easy. with a little more
11
practice would greatly improve the comfort of using
12
User friendly Less likly to make a mistake (- jump to wrong case) no velcro!
13
I like the size of the font. the ipac is way too small
14
Love the larger #'s and print.
I like that the text is easier to read due to larger screen area/font. It's easy to use and the case is well designed. I especially appreciate the lack of
15
velcro
I like that the tablet is large enough to write ROC's w/o hitting wrong keys. I am not totally comfortable with the tablet yet to feel competant, but am
16
confident that will come with using it.
17
I like the way I can see better because the tablet has larger print.
18
It is easier to read and to enter data because of the large size.
19
love the larger print that you can see the ROC record w/out opening
20
well lit screen, characters are larger, better for myself.
I appreciate that it is very easy to read the script on the tablet. The organization of the case listing screens is far friendlier (lines not so close
21
together) than the same on the iPac.
22
i like the tablet a lot and the carrin g case is so light i think this will be a benefit to the program.
23
IPAQ does everything the tablet can do. Should use tablet for both scrn and ivw
24
it seems to be very easy in handling it and better features that other devices
i think it will be much easier to see the tablet with the size, however not sure at this time about using it in different weather environments since we
25
have not tested it. I wish we could switch now!!

No.
26
27
28
29
30

F-7

31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48

Comments QFT FI Survey 1 (August 2012)
I wish the tablet were a little skinnier, would be easier to enter notes, like on a smartphone - using my thumbs. Its a little too wide. Older or
technology challenged FI's will have more trouble.
I will need to get used to the size of the tablet, my wrist hurt on the first morning after holding for several hours. The neck strap is too wide for my
use. I prefer the width of the IPAQ strap.
still undecided about use of touch vs stylus. would prefer touch only, but stylus may prove better on some screens, and i do not want to go back and
forth, so may end up using just the stylus.
As a lefty my thumb hits the volume button on the rightside even with the case covering it, also there is no way to "teach" the tablet my input style
like on the ipaq and newton
easy to get to wrong screen.....very sensitive
I love it. I think it has smooth transition from scrn to scrn.I like the fact you can use your finger or stylus,so far it seems comfortable to hold.LOVE
IT
i have none but others have used hem and enjoyed those
a little more time consuming switching back and forth letters and numbers on keyboard. should resolve itself with practice :)
navagation of the tablet is somewhat confusing but may get less with use, I like most of the features of the tablet
Just need some practice on the tablet to know how to move from one screen to next.
I will practice a lot to be more comfortable. I believe you can teach an old dog new tricks Old dogs just have to practice over and over
I think it will be advantageous to use in the field, and that the IPAQ is becoming obselete.
THE TABLET IS A GOOD LEARNING EXPERIENCE FOR ME BECAUSE I HAVE NEVER USED A TOUCH SCREEN BEFORE I USED
ONE AT THE TRAINING.
I really like it and the fact that the SR can also see the sereen as well to know what I am entering when I screen hem and when someone in their
household comes up for Interviews.
im concerned about tablet in inclimate weather snow/ cover over tablet bulky but will probable get used to it
I don't think it's going to be as physically easy to transport and use as the ipaq, but i'm open to the new experience.
a bit bulky
It's heavier than ipaq; Must memorize the main touch screen conventions for accessing items. The symbols are new (hover descriptions would be
helpful for new-to-touch screens) Would like $,apostophe
May be a bit heavy carrying around neck, will have to see.
holding tablet for some time hurt my wrist/didn't fit as easily & neatly into plam of my hand. I might have issues w holding the tablet and tryi g to
pull paprs out to had to Rs
it is a little large (not heavy) to hold. in general though it is much better than the ipaq.
Need field exp b4 commenting on case and ease at door. Seems cumbersome compared to ipaq which was quickly at hand when hung frm neck.Tab
may b too hvy
Tablet is easy to use. Practice is very important

No.
49
50
51
52
53
54

72
73

Comments QFT FI Survey 1 (August 2012)
somewhat drastic departure from iPaq, so tough for us oldsters to master
Just getting aquainted
It ws easier than I thought it would be to learn to use the tablet.
Changes on interview are really good. Use of tablet will be easier for screening
I think the use of a tablet will give a more professional and up to date impressions to the respondant.
would love to see interview process done on some similar tool
Use of the tablet is great. However, it would be lot better of the interview was also included on the tablet. Maybe a seperate pas-code protected file
would allow responses to be kept seperated.
was easy to use self explanatory easy to foolow directions
excellent choice, the new tablet is great.
One bonus of the ipaq was that there was little to no theft risk. Now working in sketchier areas, that becomes more of a concern.
tablet - keys are too narrow for fat stylus tip. Also, the shift key acts like a cap lock many times and I have to select it again to get out of cap lock.
Need numbers to be on same keyboard screeen.
like it think it will work well
I love the tocuh screen option, it is great to be somewhat current with technology, thank you!
easy to use, professional looking
I feel it will be a good change but I really will not know until try in field
I am so EXCITED about using the tablet!!! Laptops are very heavy and I hope we are moving towards getting away from them and maybe have just
one device???
I think it is not only helpful to the FI's to have an updated device, but it also appears more professional and clean when screening with up-to-date
technology at the selected dwelling units.
impressed so far!
Tablet is very user friendly. Much improvement over IPAQ
Such an improvement over Ipaq...it's early yet, may discover new and better features and usability as I use it more - OR may find problems and
issues - seems great at this time.
at this time I realy like the way the tablet works, I look foreward to tring it in the field and hope to have the same results
I find with a quick tutorial, most people will be abel to use the tablet with ease. People with no exposure to technical gadgets, may need a bit more
help
I have to exit the screening program and get to the view cases screen to get the case id number. It is not on every screen like the ipaq; a little
inconvenient but not a big deal.
Typing answers & navigating the keyboard still allows for mistakes & lag time in relation to the lack of sensitivity; the amount of time it takes to
press a button, and the time the letters appear.
Technelogically advanced, very positive change,

It is much easier to use than the ipaq

55
56
57
58

F-8

59
60
61
62
63
64
65
66
67
68
69
70
71

No.

Comments QFT FI Survey 1 (August 2012)

Specific Screener Functions/Features
75
Would like to see Case ID at top of ROC w/o going to another screen or tap
76
Should have distribution of calls
77
the way cases are formatted on the screen (being able to view codes) might pose to be a problem with time effciency ahile in the field.
From the main menu we are not able to see the total # of cases. When completing a transmission it is helpful to know # of cases added or removed.
78
This is no longer available.
The one thing that is something I'd have to get used to is staying on the R screen if completing an OTS INT Ld ltr debriefing pops up. extra steps to
79
get bk to QID Screen
do not like that you must do debriefing questions before the eroc...I like to put in int appointments on spot..defriefing should be AFTER you commit
80
screening or at least after Eroc
I would like to see added the the feature in the tablet where you can see if you receive new cases or they were taken away. Also to see the number of
81
cases you have.
82
the done button is on left and commit buttom is on right, will need to pay attention and hit correct one, and not mistakenly hit the cancel button
I have to exit the screening program and get to the view cases screen to get the case id number. It is not on every screen like the ipaq; a little
83
inconvenient but not a big deal.

F-9

Accessories - Carrying Case, Stylus
I always have the stylus and a pen handy when using the iPaq using the one holder on the case (yes, both do fit) The holder on this case will only
84
hold the stylus
85
I wish I could get a left handed version of the case
86
I will definitely need a backup stylus because of nails; am able to do very little with finger tips; can only use knuckle on some functions
87
The case doesn't look as sturdy and I worry about the tablet slipping out of the bottomI
88
You should check and see it Otterbox makes a case for the Tablet, I think the provide the best protection for smartphones.
istylist rather than fingers- errors using fingers. easier & more consitant to use next button rather than swipping.- this way was too inconsitant.
89
locating added D.U.'s- frustrating..no pen holder
concerned that the tablet wont fall out from the bottom after much use/movement. SCRN: after removing all reference to SR from roster the tablet
90
allowed me to move on and sel an int in a 1 person hh
91
The Flap at the bottom of the case is annoying when open and trying to close the cover
stylus holder for left handed FIs..allow screen to rotate when using keyboard, add option to view only one segment, always have entire line number
92
including segment on all screens
Stylus has a tiny hole where it could be attached to a cord to hang on FIs neck so it won't disappear if dropped. Would like to have some support in
93
doing this.
94
I find it a little cumberson pointing the stylus and getting the selection screen I need. It seems it appears sometimes fast and sometimes slow.

No.
95
96
97
98
99
Training
100
101
102

Comments QFT FI Survey 1 (August 2012)
The cover flap on the case is a bit cumbersome. Would like to be able to get a message on the tablet after transmission about added or deleted cases.
Like that you can view comments on select case scr
carring case is not the most ideal
So glad the vel cro is gone!
The strap on the tablet case is wide and alittle cumbersome,
Strap appears too wide; after using in class, not sure it is going to be comfortable around my neck given the extra weight of the tablet. the actual
case is outstanding, no velcro to catch on clothes
Training needs will depend on the abilities of the FIs.
use kid gloves when training older fi's. you do not want to lose them as they are respected by community and keep nsduh productivity good.
younger fi's walking up the door with the tablet-R will think
More instruction should be given regarding double checking of household roster correctly added, or have access to show the entire entry at once

F-10

No.

Comments QFT FI Survey #2 (October 2012)

F-11

General Comments about Tablet/Screening Program
1
very easy to use in the field
2
I think the tablet ia great and easy to use.
3
I like the bigger screen. It is easier to read. I like the carrying case because it allows me to wear the tablet around my neck.
this is a wondeful tablet, however I noticed that the some of the keyboard symbols like quotations marks,asterisk were not available.there were on
4
the keyboard, just not functioning
the respondnets also enjoyed being able to read along with the screener, especially when I asked for verification information they where able to
5
read along.
6
I really enjoyed the experience of using the tablet. It's lighter than the Ipaq,The larger screen and larger addresses are a plus
7
I LOVE the fact that the print is larger on the tablet. It is easier to see & use
8
the tablet was great. was able to see screen better cause it is larger print
9
seems more efficient & responsive than the ipaq. I like that I can see the time always on the screen.
10
There is a need for the $ sign on the keyboard as I frequently use it. That seems to be the only deficiency I had found. Otherwise it's great!
too big. difficult to carry, too easy to open wrong case or press wrong buttons. other functions open accidentally. brightness didnt always adjust
11
correctly
very partial to the old ipaq,especially it's size and the way it fits my hand. sometimes have trouble getting the tablet to respond-maybe bcuz i
12
always use syllus.
13
wish tablet was a little smaller/ I worry about snow and rain
14
unfortunately the device is more cumbersom, due to the increased size over the ipaq.
15
rather sensative to touch when holding it, you have to watch where your thump is or it can change the field your in
16
The tablet is very sensitive. It jumps for no reason. It will jump to another screen without touching the tablet. I don't like this.
The tablet doesn't fit as easily in my hand as the ipaq did, it's way too wide, and the screen is bright and colorful, but not neccesary for just
17
screening. Also the volume button is badly placed
18
The tablet becomes very heavy after a couple of hours. Also very difficult to use when in the rain and sunny days.
tablet is too big to fit easily in hand AND allow that hand to be useful for things such as holding/handling papers; once tablet is in hand that hand
19
is completly immobilized from anything else
the surface is too sensitive; changes screens at the slightest inadvertant tap. Also, more difficult to type on than ipaq; have to change numeric to
20
alphabeticd screens and bigger isn't better
The only problem with the tablet is it is so sensitive. Sometimes you accidently hit something and it goes to a wrong screen. You have to take time
21
to getr back to the correct screen.
It's a bit sensative to accidental touch (screen) making you go to a different screen. Have to "back-out" sometimes when transitioning from car to
22
front door, or while waiting for someone to answer
virtual keyboard is v poor; much better r (eg, SwiftKey) avail. text entry time consuming, missing/non-working characters. roc comments
23
sometimes dont show up. have to log-in just in order to log out

No.
24
25
26
27
28
29
30
31
32

F-12

33
34
35
36
37
38
39
40
41
42
43

Comments QFT FI Survey #2 (October 2012)
too sensitive, slighest touch, the screen changes.w/ seg info , materials. tablet at door, it can be difficult to manage, esp when you have a du
description, tto heavy. traps not useful w/ so much
Not sure that it stays charged very long and takes time to charge up.
My only issue is that when it is in an air conditioned car and then step out into the heat the screen fogs up. Other than that I love the tablet and it
works great. Hope to get to use it all the time
Much much better than the iPaq and easy to use it.
Learning curve navigating between screens and entering text.Can not swipe all screens so I use stylus and next icon to navigate all screens for
screening
I found the tablet is easier to use when typing notes vs the IPAC I also like the fact that it didn't have to be reset all the time the fact that when a
case is closed is a good feature also
I find the tablet to be far more effective than the ipaq, in terms of presenting more information on the select case screen. In severe cold, not sure
how it will do, as well as extreme rain.
I find the size of the tablet to be difficult for the size of my hands. I prefer the i-Pac but I am sure I will figure out how to use the tablet more
efficiently as time passes. Screen changes if bump
I feel that it does not keep the charge sufficiently
I enjoy using the tablet because it was a learning experience for me. The tablet is cumbersome I wish we could use a tablet that is the same size as
our ipaqs!!!
I am yet to work on the field uisng new new SG Tablet
Having to swtich between using a swiping motion and the next button, means I always use the next button. It's not as hardy or as lightweight as the
ipaq and I think it's more of a theft risk.
it would be easier to enter notes if the "swype" keypad was installed...when I am at the case screen, it gets confusing because I see a little more
than just the case ID, I did not like seeing codes
Have not used Tablet in all weather conditions, neutral. The carrying case doesn't have a slot for a pen, just stylus. Tablet is more sensative to
touch so we have to be extra careful inputtng info.
Either the swipe feature or screen sensitivity cause case migration. You think you're entering a ROC for one case but end up with another. Some
method is needed to fix the selected case
EASY TO HIT THINGS YOU DO NOT WANT TO......
eaily read;tablet too heavy to have about neck;over shoulder necessary;all pending cases disappeared while infield,reappeared upon re-boot,no
calendar,not happy in heat
Do not like that tablet does not show incoming transm. info. Laptop says transmitted successfully only some of the time. Stylus tip has partially
worn off. Ints have ranged from.50 to 2 hrs.
Compared to the IPAQ the Tablet seems to be much more tactile. As well as the bigger screen is much easier to read and clearer. Like can see
selected R on ROC records without having to go in case
At first it felt heavy, but I got used to it. The only time the weight really bothered me was when my carpal tunnel flared up, as it sometimes does
after a lot of driving.

No.
44
45

Comments QFT FI Survey #2 (October 2012)
Along with the tablet, I suggest a car charger
Would like to have punctuation and numbers with the letters on the keyboard but maybe just because that's what I'm used to having.

F-13

Specific Screener Functions/Features
46
Would like to look at finalize more easlily.Got stuck trying to return.
47
Would like final cases taken off main screen rather than have all appear
would like call distribution like the ipac had, easier to know wxactly WHEN to visit an area. control costs better, and tells me when NOT to visit
48
an area
Wish we would change the case listing to see codes easier-maybe table form. wish case allowed touching of sides without interrupting tablet ops.
49
Lov font size and big buttons
50
When the tablet is ready to go into the field on a regular basis there needs to be a way to see when letters to refusals have been mailed.
51
when screening and a end at verifying screen, you want to change age on a member ,the choice is age range,can that be looked into
there were several instances in which the tablet would revert to the case list after I had selected the next DU to screen, even after having selected
52
the physical description of the DU.
The tablet does not show when the conversion letters have been sent out. It really is a must have feature when doing refusal conversions. Liked the
53
ability to see who was the IR right at the bottom
the commit entry is on top right, and on some screens it is the cancel entry, which i did occasionally tap on cancel by mistake. done and commit
54
should be on same side
Tablet should keep HIGHLIGHT on current line (eg., during interview with power off; sometimes stylus activation is slow or delayed; tablet
55
battery seems to have short life—Intv off, 20% power
Obtaining the Case ID by tapping the screen is difficult. It requires several taps before appearing. The case ID constantly appearing as on the IPAC
56
is preferred.
57
It would be nice to have the number of cases at the top in a particular segment since only a few cases are shown
58
It would be nice to have the case ID displayed on the selection screen.
it would be nice to be able to edit roc codes once they are saved, before they are transmitted to RTI. this was possible using the ipaq, but with the
59
tablet, you can only edit the notes for rocs
it would be better if the numbers were on the same screen as the letters so that I wouldn't have to keep switching back and forth between screens
60
when I need to type a number.
i have only had a few unsuccessful transmissions and sometime trying to transmit the screen says the screening software is still running, when it
61
clearly is not and on the rainbow screen.
learned today about ROC discrepancies due to editing eROC later when at home; fix so both original eroc time plus time of editing (when done
62
later) registers.
layout too sparse for large datasets. dislike that i cannot keep placeholder of case i was at last. v hrad to count results,review status of cases.
63
designed 4 1 case at a time, not friendly case mgmt

No.
64
65
66
67
68
69
70

Comments QFT FI Survey #2 (October 2012)
ifinger accidently touched ref on ver screen lost phone number. On ver screen, when put comments, top buttons disappear. must press button to get
back, but goes too quickly must try couple times to do
HATE!!!: codes aren't lined up to far right. HATE!!!: have to keep switching between keyboards. HATE!!: cannot switch roc codes (if mistake)
Hate: final screen b4 selection doesn't show ALL demog
Choosen line in tablet needs to keep highlighted even after touching it. When opening a line is a lot faster to tap twice than to hold the stylus.
Bold address not the case #. sorting combos-keep segments separate. "HUMAN SERVICES" made folks think we're welfare in my state.
"International" made some think we're from a foreign country or state.
I find the layout difficult to work with because it is hard to distinguish between cases. There is too much information for each line that is not really
necessary, such as having city and zip code on
you cant amend the code on a roc w/o deleting- then yu must renter this amends the time of the roc.=there is no way to tell when refusal letters are
sent= like help button w definitions of roc codes
Can't figure out how to find out when and what type of letter was sent to DU. l love the way we complete the comments to an interview on the
tablet instead of the laptop.

F-14

Accessories - Carrying Case, Stylus
I like using the stylus, but he stylus is to short making it a little awkward to use. It would be better for me if it were the same size as the old stylus,
71
like a pen
the styles is too short & is hadr ti place it in its holder i droped the styles several times. the screen is dificalt to start you can press too hard with the
72
styles the tablet is balkey the screen is
73
the stylus is in the way of the on/off switch can not tell if letters have been sent must call FS she's very busy fs
Stylus is in the way when I use the power on button;Problems trying to transmit; much more focus on using the tablet in training and less on the
74
interview -making mistakes and learning to fix them
75
stylus holder for lefties/extra pen-allow screen to rotate for larger keyboard-"sleep mode" faster76
stylus does not easily fit into side loop, too slippery also
77
the case is a little hard to hold by design. placing your hand underneath the straps is not comfortable.
tablet cover gets in the way a little; stylus holder could use a plastic opening at top to ease replacement of stylus; "other" in lead letter feedback
78
should allow comments; trans'n done ????
79
sun glare difficult to see, strap on carrying case too wide/bulky, would like to be able to go to next line in ROC w/o going to end, car charger?
Strap for the carrying case too thick; SRs are much more interested in the tablet vs the iPaq; can't edit codes after committing; cases don't stay
80
highlighted (apartment complex - all addresses same)
81
snap closure difficult to use. constantly moving it around to find the snap. the cover for the cord hook up is annoying.
82
Screen glare is difficult, needs an additional loop for a pen
need a pen / pencil at the SR door (appointment cards) carrying case needs a place to put a pen and have it handy. Press and hold to select case
83
keeps screen from moving to wrong line accidentlly.

No.
84
85
86
87
88
89
90
91

Comments QFT FI Survey #2 (October 2012)
My tablet doesn't respond quickly to my fingers; so i always use the stylus. the strap is cumbersome. Prefer to put my fore arm through the back
holds it securely and is good for me as a lefty.
its easiest 2 use stylist rather than finger- its more accurate. i dont swipe-the swipe commands r not consistant. using next button is always
consistent. wish there was place to hold pen for apt. x's
The tablet could be attached inside a portfolio holding our printed materials. We would only have one thing to carry to the door. It would make us
look more professional and less like meter-readers.
i think the tablet case should have a stylus holder and a pen holder, one on each side. If I need to fill out the simy or appointment card it would be
nice to have a holder for both pen and stylus.
I like the carrying case with the snap rather than the velcro closure and the flap that covers the connection.. Screening program was very easy to
use and the ability to make corrects extremely easy..
carrying case a little cumbersome could enable swpye for typing this would be easier
can not close the snap on the screen cover when the tablet is connected to cable when charging or when connected to the laptop. A cover designed
for left handed FIs would be nice
I think the case is too bulky

F-15

F-16

Appendix G: Moderator's Guide for QFT Focus Groups
with Field Interviewers

SECTION I: Introduction (5 minutes)
MODERATOR: PARTICIPANTS SHOULD BE SITTING AROUND THE TABLE WITH THE
SEAT AT THE HEAD OF THE TABLE RESERVED FOR THE MODERATOR. PARTICIPANTS
SHOULD BE ASKED TO WRITE THEIR NAMES ON BOTH SIDES OF A "NAME TENT" AND
PLACE IT SO IT CAN BE SEEN FROM THE FRONT OF THE ROOM.
INTRODUCTION OF MODERATOR AND NOTE TAKER: Hello, and thank you for attending this
group discussion. My name is [MODERATOR'S NAME] from [MODERATOR'S AFFILIATION].
This is [NOTE TAKER'S NAME] from [NOTE TAKER'S AFFILIATION].
This group discussion is intended to gather feedback from all of you on your experiences
completing data collection for the 2012 Questionnaire Field Test (QFT). As you know, several
changes to the NSDUH questionnaire, procedures, equipment and materials were tested during
this field test. We plan to examine the data collected using the QFT interview questionnaire and
procedures to assess how well they performed in the field. However, we cannot gather all of the
information we need just by analyzing the survey data. Therefore, we are hoping you can share
your experiences with administering the QFT interview, including what sorts of feedback you
received from respondents, and what types of issues you encountered that could be improved in
the future. A summary of the feedback you provide in this discussion will be included in the QFT
report provided to SAMHSA and will inform potential changes to the protocol changes in the
future. I will be leading today's discussion and [NOTETAKER'S NAME] will be taking notes.
We just have a few ground rules for our discussion:

•

We are video recording the session and also have a note taker so we don't miss
anything that is said, and so that those who cannot observe this discussion can review
the recording.

•

Please avoid side conversations among yourselves. Only one person should speak at a
time. This serves two purposes. First, it lets the whole group hear the remarks someone
makes. Second, it ensures that the recording will be clear.

•

To get the best benefit from this group, we want to hear from everyone in the room. Like
any group, I imagine some of you like to talk while others may be quieter. So if I haven't
heard from you, I may call on you. This allows us to hear from everyone several times
throughout the discussion. If you'd rather not answer a particular question, you can just
tell me that you would like to "pass."

•

There are no right or wrong answers to the questions I will be asking. Everyone's input is
equally important and helpful. We are interested in all your ideas, comments, and
suggestions. It is OK to disagree with what someone says, but we ask that you do so
respectfully.

•

Please take a minute now to turn off your cell phones so we aren't interrupted.

•

If you need to take a break or use the restroom, please leave the room quietly.

Before we begin, let's briefly introduce ourselves, starting to my left (or right).

G-1

SECTION II: Reactions to the Redesigned Contact Materials (15-20 minutes)
1. When you sent lead letters to the households in your QFT assignment, did you expect the
letter to have a similar impact on cooperation among members of sampled households, a
greater impact, or less impact? [PROBES: Tell me more about that. What do others think?]
2. How often did members of sampled QFT households mention to you that they had seen the
lead letter? Do you think members of sampled QFT households mentioned seeing the letter
about as often as main study households you have recently screened, more often, or less
often? [PROBES: Tell me more about that. What do others think?]
3. How often did members of QFT households make comments or ask questions about the
lead letter? Did members of sampled QFT households make comments or ask questions
about the letter about as often as main study households you have recently screened, more
often, or less often?
4. [IF APPLICABLE] When members of sampled QFT households made comments about the
lead letter, did they focus on the content of the letter, on the appearance or layout of the
letter, or a mix of both? [PROBE: Please provide examples of any comments on the content
or appearance of the letter that you can recall.]
5. [IF APPLICABLE] When members of sampled QFT households asked questions referring to
the lead letter, what kinds of questions did they ask you? [PROBE: Please provide
examples of any questions about the letter that you can recall.]
6. How often did members of sampled QFT households make comments or ask questions
about the question and answer brochure? Did members of sampled QFT households
make comments or ask questions about the brochure about as often as main study
respondents you have recently interviewed, more often, or less often? [PROBES: Tell me
more about that. What do others think?]
7. [IF APPLICABLE] When members of sampled QFT households made comments about the
question and answer brochure, did they focus on the content of the brochure, the
appearance or layout of the brochure, or a mix of both? [PROBE: Please provide examples
of any comments on the content or appearance of the brochure that you can recall.]
8. [IF APPLICABLE] When members of sampled QFT households asked questions referring to
the question and answer brochure, what kinds of questions did they ask you? [PROBE:
Please provide examples of any questions about the brochure that you can recall.]
9. Overall, do you think QFT sample members reactions to the lead letter and question and
answer brochure were similar to the reactions you receive to the current main study
contact materials, or were they different somehow? [FOR ANY WHO INDICATE
REACTIONS THEY RECEIVED WERE DIFFERENT FOR QFT HOUSEHOLDS, ASK: What
were the main ways that QFT sample members' reactions to the contact materials were
different than the reactions you receive to the main study letter and brochure?]

G-2

SECTION III: Administering Household Screenings and Using the Tablet (15-20 minutes)
1. Do you feel the QFT training provided you with a thorough understanding of the purpose
and goals of the QFT? [FOR ANY WHO INDICATE THE TRAINING DID NOT A
THOROUGH UNDERSTANDING OF THE QFT PURPOSE AND GOALS, ASK: What are
the main ways you would recommend to improve training about the purpose and goals of
the QFT?]
2. Do you feel that the new portfolio met your needs for organizing your field materials? [FOR
ANY WHO INDICATE THE PORTFOLIO DID NOT MEET THEIR NEEDS, ASK: What kind
of portfolio would be more useful to you for organizing field materials?]
3. Do you feel that the QFT training provided good instruction on how to use the tablet to
conduct household screenings? [FOR ANY WHO DO NOT THINK THE INSTRUCTION
WAS GOOD: What are the main ways you would recommend to improve training on using
the tablet for household screenings?]
4. Do you feel that the QFT training provided sufficient time for you to learn how to use the
tablet and get comfortable using it? [FOR ANY WHO DO NOT THINK THE TRAINING TIME
WAS SUFFICIENT: How much time do you think would be sufficient to learn how to use the
tablet and get comfortable using it?]
5. How long did it take you to feel fully comfortable using the tablet computer to conduct QFT
screenings? [FOR THOSE WHO INDICATE NOT QUICKLY FEELING COMFORTABLE
USING THE TABLET, ASK: What do you think were the greatest challenges you faced in
getting comfortable using the tablet to conduct screeners in the QFT?]
6. Do you feel that the size and weight of the tablet was appropriate for conducting screeners
on doorsteps? [PROBES, ESPECIALLY FOR ANY WHO RAISE CONCERNS: Tell me more
about that. What do others think?]
7. Do you feel that the design and usability of the tablet carrying case met your needs for
transporting and using the tablet in the field? [FOR ANY WHO RAISE CONCERNS
ABOUTHE DESIGN OR USABILITY OF THE CARRYING CASE, ASK: How do you think
the carrying case could be altered to make it work better for you in the field?]
8. How did respondents react to the use of US Department of Health and Human Services, as
opposed to the US Public Health Service? Were reactions positive or negative? Did this
cause any confusion among respondents?
9. Did you experience any difficulties typing in ROC notes or comments using the keyboard
on the tablet? [FOR ANY WHO INDICATE HAVING DIFFICULTY TYPING ROC NOTES OR
COMMENTS, ASK: How often did you encounter problems typing in ROC notes or
comments using the keyboard on the tablet? How were you able to overcome this
challenge?]
10. Did you encounter any problems completing the observation questions on the tablet?
[FOR ANY WHO INDICATE HAVING PROBLEMS COMPLETING THE OBSERVATION
QUESTIONS: Please tell us more about that problem. How were you able to resolve this?]

G-3

11. Did you ever ask for technical assistance with the tablet at any point during the QFT data
collection? [FOR ANY WHO INDICATE REQUESTING TECHNICAL ASSISTANCE WITH
THE TABLET, ASK: Can you tell me why you asked for assistance with the tablet? Was
assistance provided quickly enough for you to continue with your QFT assignment as
planned?]
12. Did you wish that the tablet had additional capabilities available to you, such as copy and
paste, predictive typing, or rotating between landscape and portrait display? [FOR ANY
WHO INDICATE WANTING ADDITIONAL CAPABILITIES, ASK: What capabilities would
you like to have on the tablet? How would this improve the usability of the tablet for
completing household screenings?]
13. Compared to the iPAQ you use for the main study, would you say the tablet was about as
easy to use as for screening households, easier to use, or not as easy to use? [FOR ANY
WHO INDICATE THE TABLET WAS NOT AS EASY TO USE AS THE IPAQ, ASK: What
are the main reasons why you feel the tablet was not as easy to use as the iPAQ?]
14. Compared to the iPAQ, were there any screening functions that you would have liked to
have had on the tablet for the QFT, such as having finalized cases disappear from the select
case screen? [FOR ANY WHO INDICATE WANTING FUNCTIONS CURRENTLY ON THE
IPAQ, ASK: What iPAQ functions would you like to have on the tablet? How would this
improve the usability of the tablet for completing household screenings?]
15. Please share any comments you had about transmitting your work using the new
equipment.
SECTION IV: Administering the Redesigned Questionnaire and Protocol (30-35 minutes)
1. How often did QFT respondents make comments or ask questions about using the
computerized version of the reference date calendar? Would you say QFT respondents
made comments or asked questions about as often as main study respondents using the
paper version of the calendar, less often, or more often? [PROBES: Tell me more about
that. What do others think?]
2. [IF APPLICABLE] What kinds of feedback or questions did you receive from respondents
about the computerized version of the reference date calendar? Please provide examples of
any comments or questions that you can recall.
3. Did you expect the computerized version of the reference date calendar to be as easy
for QFT respondents to use as the paper version of the calendar, easier to use, or harder to
use? [PROBES: Tell me more about that. What do others think?]
4. How often did QFT respondents or proxy respondents make comments or ask questions
about specific questions or modules when completing either the items you administered
to them or completing the ACASI portion of the interview protocol themselves? Would you
say QFT respondents made comments or asked questions on any specific questions or
modules about as often as main study respondents, less often, or more often? [PROBES:
Tell me more about that. What do others think?]

G-4

5. Did you expect QFT respondents (or proxy respondents) to make comments or ask
questions about specific questions or modules as often as main study respondents, less
often, or more often? [PROBES: Tell me more about that. What do others think?]
6. Did QFT respondents make any comments or ask any questions about the new module
introducing proxy respondents to ACASI? Please provide examples of any comments or
questions that you can recall.
7. How often did QFT proxy respondents have trouble hearing the audio for questions in the
second ACASI portion of the interview? Did you ever have to adjust the volume for proxy
respondents?
8. Did QFT respondents make any comments or ask any questions about any other specific
questions or features of the protocol when completing any of the modules (except for the
prescription drug module)? [PROBE: Please provide examples of any comments or
questions on specific questions or features of the protocol that you can recall.]
SECTION V: Reactions to the Redesigned Prescription Drug Module (15-20 minutes)
1. How often did QFT respondents make comments or react specifically to the burden
required to answer the questions in the prescription drug module? [PROBE: Please provide
examples of any comments or reactions to the burden of the prescription drug questions you
can recall.]
2. How often did QFT respondents make comments or react specifically to the length of time
required to complete the prescription drug module? [PROBE: Please provide examples of
any comments or reactions to the length of the prescription drug module you can recall.]
3. Did you expect QFT respondents to react specifically to either the burden or length of time
required to complete the prescription drug module as often as main study respondents, less
often, or more often? [PROBES: Tell me more about that. What do others think?]
4. How often did QFT respondents make comments or react specifically to the electronic pill
images in the prescription drug module? [PROBE: Please provide examples of any
comments or reactions to the electronic pill images in the prescription drug module you can
recall.]
5. Did you expect QFT respondents to react specifically to the electronic pill images as often
as main study respondents do to the showcard pill images, less often, or more often?
[PROBES: Tell me more about that. What do others think?]
6. How often did QFT respondents make comments or react specifically to the questions
designed to capture misuse of prescription drugs? [PROBE: Please provide examples of
any comments or reactions to the questions on misuse of prescription drugs you can recall.]
7. Did you expect QFT respondents to react specifically to the questions designed to capture
misuse of prescription drugs as often as main study respondents do with the current
questions, less often, or more often? [PROBES: Tell me more about that. What do others
think?]

G-5

8. Did QFT respondents make any comments or ask any questions about any other specific
aspects of the prescription drug module? [PROBE: Please provide examples of any
comments or questions about the prescription drug module that you can recall.]
SECTION VI: Overall Reactions to the Redesigned Questionnaire (15-20 minutes)
1. How often did QFT respondents make comments or react specifically to the burden
required to answer any of the other interview questions? Would you say QFT
respondents commented on the burden of the interview questions about as often as main
study respondents, less often, or more often? [PROBES: Tell me more about that. What do
others think?]
2. [IF APPLICABLE] When QFT respondents made comments or reacted specifically to the
burden of the interview questions, were the comments or reactions mostly positive,
mostly negative, or a mix of both? [PROBE: Please provide examples of any comments or
reactions that you can recall.]
3. How often did QFT respondents make comments or react specifically to the length of time
required to complete the entire interview protocol? Would you say QFT respondents
commented on the interview length about as often as main study respondents, less often, or
more often? [PROBES: Tell me more about that. What do others think?]
4. [IF APPLICABLE] When members of sampled QFT households made comments or reacted
specifically to the length of time to complete the entire interview protocol, were the
comments or reactions mostly positive, mostly negative, or a mix of both? [PROBE: Please
provide examples of any comments or reactions that you can recall.]
5. Did you expect QFT respondents to react specifically to either the burden or length of time
required to complete the entire interview protocol as often as main study respondents, less
often, or more often? [PROBES: Tell me more about that. What do others think?]
6. Did QFT respondents raise any other specific concerns when completing the questions
you administered to them or completing the ACASI portion of the interview protocol
themselves? [PROBE: Please provide examples of any concerns that you can recall.]
7. Did you expect QFT respondents raise any other specific concerns when completing the
questions you administered to them or completing the ACASI portion of the interview as
often as main study respondents, less often, or more often? [PROBES: Tell me more about
that. What do others think?]
8. If a video containing a 20-30 second video clip of the annual press conference were added
to the tablet, do you think this would be a useful tool for gaining cooperation from
respondents at the doorstep? Why or why not?
SECTION VII: Conclusion (5 minutes)
Are there any final comments or any questions on any of the topics we discussed, or other
topics on the QFT data collection?
I want to thank you all again for your active participation.
THE RTI NOTETAKER WILL NOW TURN OFF THE VIDEO CAMERA.

G-6

Appendix H:
Selected Notes on Analysis Variables for the QFT

1. Key Illicit Drug Measures in Appendices I and J
Measure
Use of Any Illicit Drug, Standard Definition

•
•
•
•
•
•
•

Substances Included
Marijuana
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants2
Methamphetamine3
Prescription Drugs3

–
–
–
–
Use of Any Illicit Drug, Alternate Definition 1

Use of Any Illicit Drug, Alternate Definition 2

Use of Illicit Drugs Other Than Marijuana, Standard
Definition

•
•
•
•
•
•
•
•
•
•
•
•
•
•

Marijuana
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants
Marijuana
Cocaine (including crack)
Heroin
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants2
Methamphetamine3
Prescription Drugs3

–
–
–
–
Use of Illicit Drugs Other Than Marijuana,
Alternate Definition

•
•
•
•

Pain Relievers
Tranquilizers
Stimulants3
Sedatives

Cocaine
Heroin
Hallucinogens1
Inhalants2

For the 2011 and 2012 comparison data, estimates are based on the use of any of the following hallucinogens: LSD, also called
"acid"; PCP, also called "angel dust" or phencyclidine; peyote; mescaline; psilocybin; or "Ecstasy," also called MDMA; or any
other hallucinogen. QFT estimates are based on the use of any of the hallucinogens from the 2011 and 2012 comparison data,
plus the following: ketamine, also called "Special K" or "Super K"; DMT, AMT, or 5-MeO-DIPT ("Foxy"); or Salvia
divinorum.
2
Lifetime estimates of inhalant use for the 2011 and 2012 comparison data are based on the use of any of the following: amyl
nitrite, "poppers," locker room odorizers, or "rush"; correction fluid, degreaser, or cleaning fluid; gasoline or lighter fluid; glue,
shoe polish, or toluene; halothane, ether, or other anesthetics; lacquer thinner or other paint solvents; lighter gases, such as
butane or propane; nitrous oxide or "whippits"; spray paints; other aerosol sprays; or any other inhalant. QFT estimates of
lifetime use of inhalants are based on the use of any of the inhalants from the 2011 and 2012 comparison data, plus the
following: felt-tip pens, felt-tip markers, or magic markers; and computer cleaner, also known as air duster.
3
Estimates of any prescription drug misuse, stimulant misuse, and methamphetamine use for the 2011 and 2012 comparison data
include data from the new methamphetamine items added in 2005 and 2006 (i.e., core plus noncore data). Estimates of stimulant
misuse for the QFT vary according to whether they include data from the separate core methamphetamine module.

H-1

2. Stimulant Misuse:
•

The standard definition for the 2011 and 2012 comparison data and the QFT includes
use of methamphetamine and misuse of prescription stimulants. Estimates for the
2011 and 2012 comparison data also include data from the new methamphetamine
items added in 2005 and 2006 (i.e., core plus noncore data).

•

The QFT definition includes data only for misuse of prescription stimulants. A
corresponding measure is not available for the 2011 and 2012 comparison data.

3. Binge Alcohol Use – For the 2011 and 2012 comparison data, binge alcohol use is defined
for both males and females as drinking at least five or more drinks on the same occasion (i.e.,
at the same time or within a couple of hours of each other) on at least 1 day in the past 30
days. For the QFT, binge alcohol use is defined for males as drinking five or more drinks on
the same occasion and for females as drinking four or more drinks on the same occasion on
at least 1 day in the past 30 days. Estimates in the QFT for persons aged 12 or older and by
age group (i.e., regardless of gender) also take into account the lower threshold for females.
4. Methamphetamine Dependence – For the QFT sample, respondent s were classified with
past year methamphetamine dependence if they reported three of the following problems in
the past year because of their use of methamphetamine:
•

spent a great deal of time over a period of a month getting, using, or getting over the
effects of methamphetamine (METHLOTTM=1 or METHGTOVR=1, corresponding
to questions DRME01 and DRME02);

•

used methamphetamine more often than intended or was unable to keep set limits on
methamphetamine use (METHKPLMT=2, corresponding to DRME05);

•

needed to use methamphetamine more than before to get desired effects or noticed
that same amount of methamphetamine use had less effect than before
(METHNDMOR=1 or METHLSEFX=1, corresponding to DRME06 and DRME07);

•

inability to cut down or stop using methamphetamine every time tried or wanted to
(METHCUTEV=2, corresponding to DRME09);

•

continued to use methamphetamine even though it was causing problems with
emotions, nerves, mental health, or physical problems (METHEMCTD=1 or
METHPHCTD=1, corresponding to DRME14 and DRME16);

•

methamphetamine use reduced or eliminated involvement or participation in
important activities (METHLSACT=1, corresponding to DRME17); or

•

reported feeling blue or down when trying to stop or cut down using
methamphetamine (METHFLBLU=1, corresponding to DRME10a), as well as
experiencing two or more additional methamphetamine withdrawal symptoms at the
same time that lasted longer than a day after methamphetamine use was cut back or
stopped. Symptoms include (i) feeling tired or exhausted, (ii) having bad dreams, (iii)
having trouble sleeping or sleeping more than normal, (iv) feeling hungry more often,
and (v) feeling either very slowed down or could not sit still (METHWDSMT=1,
corresponding to DRME12).

H-2

5. Methamphetamine Abuse – For the QFT sample, respondents were classified with past year
abuse of methamphetamine if they had not been classified with past year methamphetamine
dependence and if they reported one or more of the following problems in the past year
because of their use of methamphetamine:
•

serious problems at home, work, or school caused by using methamphetamine, such
as
– neglecting their children,
– missing work or school,
– doing a poor job at work or school,
– losing a job or dropping out of school
(METHSERPB=1, corresponding to DRME18);

•

used methamphetamine regularly and then did something that might have put you in
physical danger (METHPDANG=1, corresponding to DRME19);

•

use of methamphetamine caused you to do things that repeatedly got you in trouble
with the law (STMLAWTR=1, corresponding to DRME20); and

•

problems with family or friends probably caused by using methamphetamine
(METHMFPB=1 corresponding to DRME21) and continued to use
methamphetamine even though you thought that using methamphetamine caused
these problems (METHFMCTD=1, corresponding to DRME22).

6. In the QFT sample, a respondent was classified as having illicit drug dependence
(DEPNDILL) if he or she was classified as having dependence on any of the following:
marijuana, hallucinogens, inhalants, tranquilizers, cocaine, heroin, pain relievers, stimulants,
sedatives, or methamphetamine.
7. In the QFT sample, a respondent was classified as having illicit drug abuse (ABUSEILL) if
he or she was not classified as having illicit drug dependence (DEPNDILL = 0) and met
abuse criteria for any of the following: marijuana, hallucinogens, inhalants, tranquilizers,
cocaine, heroin, pain relievers, stimulants, sedatives, or methamphetamine.
8. The following measures involving new survey items for comparisons between the QFT
sample and the 2011 National Health Interview Survey (NHIS) were based on the raw survey
measures, as follows:
Measure
Living in a household with only cellular or no
telephone service
Number of visits to doctor or other health care
professional, past 12 months (none; 1; 2 to 3; 4 to 9;
10 or more)
Has been in a hospital overnight, past 12 months?
Emergency room visit in past 12 months?

QFT Survey Questions
CELL1 = 2
HLTH19, HLTH19a
HLTH17
HLTH16

H-3

H-4

Appendix I: Detailed Tables for Core Substance Use Items
Other than Methamphetamine and Prescription Drugs in the
2011 and 2012 Comparison Data and the QFT

Table I-1 Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime among
Persons Aged 12 or Older: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Substance
ILLICIT DRUGS,
Alternate Definition 14
Alternate Definition 25
Marijuana and Hashish
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
ILLICIT DRUGS OTHER
THAN MARIJUANA,
Alternate Definition4
Cocaine or Heroin6
CIGARETTES
SMOKELESS TOBACCO7
ALCOHOL

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 65,928)1

2012
Comparison
(n = 31,213)1,2

2012 QFT
(n = 2,044)1,3

45.3
44.0
43.6
14.8
3.3
1.7
14.8
9.4
2.5
5.9
8.2a

45.9
44.7
44.5
14.7
3.5
1.8
15.0
9.5
2.6
6.2
8.3a

47.5
45.0
44.7
14.2
4.1
1.9
16.2
10.7
2.9
6.4
11.1

-2.2 (1.79)
-1.1 (1.78)
-1.1 (1.76)
0.5 (1.20)
-0.8 (0.69)
-0.2 (0.42)
-1.4 (1.33)
-1.4 (1.10)
-0.4 (0.60)
-0.4 (0.72)
-2.8 (0.87)

-1.5 (1.84)
-0.3 (1.87)
-0.2 (1.85)
0.5 (1.18)
-0.6 (0.67)
-0.0 (0.42)
-1.2 (1.34)
-1.2 (1.16)
-0.3 (0.62)
-0.1 (0.74)
-2.8 (0.84)

22.4
14.9
63.9
18.8
83.2

22.3
14.8
63.2
18.4
83.4

25.0
14.3
62.5
17.4
81.8

-2.6 (1.46)
0.5 (1.20)
1.3 (1.55)
1.4 (1.07)
1.4 (1.30)

-2.7 (1.46)
0.5 (1.18)
0.6 (1.66)
1.0 (1.10)
1.5 (1.25)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or
inhalants but do not include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit
Drugs Other Than Marijuana in this definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include
hallucinogens, inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
6
Cocaine use includes crack.
7
Smokeless tobacco refers to snuff or chewing tobacco (2011 and 2012 comparison data), or snuff, dip, chewing
tobacco, or "snus" (QFT). For the 2011 and 2012 comparison data, estimates are based on responses to separate
sets of questions about use of snuff and use of chewing tobacco. Estimates for the QFT are based on responses to
questions about use of any smokeless tobacco product.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-1

Table I-2 Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime among
Persons Aged 12 to 17: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 22,419)1

2012
Comparison
(n = 10,465)1,2

2012 QFT
( n =541)1,3

22.3a
17.6
17.5
1.3a
0.3
0.3
3.7a
0.9
0.3
2.4
7.5a

20.0a
16.5
16.4
1.2a
0.2
0.3
3.2a
1.1
0.4
1.9
5.7a

26.7
19.2
19.0
0.2
0.2
0.2
6.5
1.0
1.0
2.9
11.7

-4.5 (2.10)
-1.7 (1.80)
-1.5 (1.75)
1.1 (0.23)
0.1 (0.21)
0.0 (0.24)
-2.7 (1.32)
-0.1 (0.46)
-0.7 (0.45)
-0.5 (0.77)
-4.3 (1.48)

-6.7 (2.14)
-2.8 (1.87)
-2.6 (1.82)
1.0 (0.24)
-0.0 (0.21)
0.1 (0.25)
-3.3 (1.37)
0.1 (0.47)
-0.5 (0.45)
-1.0 (0.78)
-6.1 (1.46)

10.3a
1.4a
19.2
6.9
34.6

8.2a
1.3a
16.4
6.4
31.4

16.3
0.5
19.1
8.3
33.5

-6.0 (1.90)
1.0 (0.33)
0.1 (2.17)
-1.3 (1.38)
1.1 (2.09)

-8.1 (1.87)
0.9 (0.36)
-2.7 (2.23)
-1.9 (1.47)
-2.1 (2.04)

I-2

Table I-3 Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime among
Persons Aged 18 to 25: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 21,662)1

2012
Comparison
(n = 10,336)1,2

2012 QFT
(n = 504)1,3

54.5
53.1
53.0
12.6
2.1
1.8
18.1
6.2
1.1
12.6
9.2

54.2
53.0
52.9
12.3
2.0
2.1
18.0
6.1
1.0
13.1
7.9a

56.0
52.2
52.2
10.5
1.8
2.4
19.4
7.5
0.7
11.0
11.7

-1.4 (2.35)
1.0 (2.46)
0.9 (2.46)
2.0 (1.57)
0.3 (0.61)
-0.6 (0.70)
-1.3 (2.26)
-1.3 (1.67)
0.3 (0.39)
1.6 (1.53)
-2.5 (1.75)

-1.7 (2.58)
0.8 (2.63)
0.7 (2.63)
1.7 (1.52)
0.1 (0.63)
-0.3 (0.69)
-1.4 (2.32)
-1.3 (1.66)
0.2 (0.38)
2.1 (1.54)
-3.7 (1.69)

24.0
12.7
61.4
21.0
84.6

23.6a
12.4
58.9
20.2
85.2

28.8
10.5
61.6
20.7
82.6

-4.8 (2.54)
2.2 (1.58)
-0.2 (2.98)
0.3 (2.28)
2.0 (1.99)

-5.2 (2.56)
1.9 (1.52)
-2.7 (3.18)
-0.5 (2.31)
2.6 (2.04)

I-3

Table I-4 Substance Use Other Than Methamphetamine or Prescription Drugs in Lifetime among
Persons Aged 26 or Older: Percentages, Differences, and Standard Error of Differences,
2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 21,847)1

2012
Comparison
(n = 10,412)1,2

2012 QFT
(n = 999)1,3

46.7
45.8
45.3
16.9
3.9
1.9
15.7
11.0
3.0
5.2
8.2a

47.9
46.9
46.7
16.9
4.2
2.0
16.0
11.2
3.2
5.6
8.7a

48.7
47.1
46.7
16.7
5.0
2.0
16.9
12.6
3.5
6.0
10.9

-2.0 (2.26)
-1.4 (2.23)
-1.4 (2.22)
0.2 (1.55)
-1.0 (0.88)
-0.1 (0.52)
-1.3 (1.58)
-1.5 (1.40)
-0.5 (0.78)
-0.8 (0.84)
-2.7 (1.05)

-0.8 (2.37)
-0.2 (2.37)
-0.1 (2.35)
0.2 (1.54)
-0.8 (0.86)
-0.0 (0.53)
-0.9 (1.58)
-1.3 (1.46)
-0.3 (0.79)
-0.4 (0.86)
-2.2 (1.03)

23.7
17.0
70.1
20.0
89.3

23.9
17.0
70.0
19.6
89.8

25.5
16.8
68.4
18.0
88.0

-1.8 (1.71)
0.2 (1.54)
1.8 (1.78)
1.9 (1.31)
1.3 (1.55)

-1.6 (1.73)
0.2 (1.53)
1.6 (1.92)
1.6 (1.34)
1.8 (1.51)

I-4

Table I-5 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 65,928)1

2012
Comparison
(n = 31,213)1,2

2012 QFT
(n = 2,044)1,3

12.8
12.3
12.0
1.5
0.2
0.3
1.6
0.3
0.0
1.0
0.7

12.9
12.5
12.1
1.7
0.3
0.2
1.6
0.4
0.1
1.0
0.6

13.5
12.7
12.4
1.5
0.4
0.2
2.1
0.5
0.1
1.0
0.9

-0.7 (1.21)
-0.4 (1.14)
-0.5 (1.10)
0.0 (0.34)
-0.1 (0.15)
0.1 (0.07)
-0.5 (0.43)
-0.1 (0.15)
-0.0 (0.04)
-0.0 (0.23)
-0.2 (0.19)

-0.6 (1.18)
-0.2 (1.11)
-0.4 (1.07)
0.3 (0.35)
-0.1 (0.16)
0.1 (0.07)
-0.5 (0.43)
-0.0 (0.16)
-0.0 (0.04)
-0.0 (0.24)
-0.3 (0.20)

3.2
1.6
26.5
4.7a
67.1

3.3
1.8
26.1
4.7a
67.6

3.5
1.5
28.0
6.8
66.8

-0.4 (0.56)
0.1 (0.36)
-1.5 (1.73)
-2.1 (0.67)
0.3 (1.71)

-0.3 (0.57)
0.3 (0.37)
-1.9 (1.81)
-2.1 (0.67)
0.8 (1.65)

I-5

Table I-6 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 12 to 17: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 22,419)1

2012
Comparison
(n = 10,465)1,2

2012 QFT
(n = 541)1,3

15.8
13.9
13.8
0.9a
0.1a
0.2
2.4
0.6a
0.2
1.5
3.0

14.2a
12.7
12.6
0.7a
0.1
0.1
2.1
0.6a
0.2
1.1
2.1a

18.2
15.3
15.1
0.0*
0.0*
0.2
3.6
0.2
0.5
1.6
4.1

-2.4 (1.82)
-1.4 (1.61)
-1.3 (1.55)
0.9 (0.09)
0.1 (0.03)
-0.0 (0.24)
-1.1 (1.01)
0.5 (0.16)
-0.3 (0.29)
-0.1 (0.60)
-1.1 (0.93)

-4.0 (1.89)
-2.6 (1.67)
-2.4 (1.62)
0.7 (0.12)
0.1 (0.05)
-0.1 (0.24)
-1.4 (1.04)
0.5 (0.19)
-0.3 (0.29)
-0.6 (0.62)
-2.0 (0.90)

5.3
1.0a
12.9
4.4
27.2

4.2a
0.8
10.6
3.7
24.3

7.0
0.2
12.5
5.6
25.7

-1.7 (1.35)
0.7 (0.25)
0.4 (1.70)
-1.2 (1.18)
1.4 (1.82)

-2.8 (1.36)
0.5 (0.28)
-1.9 (1.77)
-2.0 (1.25)
-1.4 (1.85)

I-6

Table I-7 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 18 to 25: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 21,662)1

2012
Comparison
(n = 10,336)1,2

2012 QFT
(n = 504)1,3

32.6
31.8
31.4
4.5
0.3
0.7
6.8
1.7
0.2
4.1
1.5

33.1
32.5
31.9
4.6
0.5
0.8
6.5
1.8
0.1
4.1
1.2

32.9
30.3
29.9
3.5
0.4
1.0
7.4
2.3
0.2
4.1
1.4

-0.3 (2.60)
1.5 (2.53)
1.5 (2.54)
1.0 (0.97)
-0.0 (0.27)
-0.3 (0.45)
-0.5 (1.59)
-0.6 (0.74)
-0.1 (0.23)
-0.0 (1.03)
0.0 (0.62)

0.2 (2.69)
2.1 (2.60)
2.0 (2.61)
1.1 (0.93)
0.1 (0.27)
-0.1 (0.46)
-0.8 (1.61)
-0.4 (0.75)
-0.1 (0.24)
-0.0 (1.05)
-0.2 (0.59)

9.7
4.8
42.7
9.5
77.5

9.7
4.8
40.9
9.1
78.5

10.4
3.8
42.7
8.7
76.9

-0.6 (1.79)
1.0 (1.00)
-0.1 (2.93)
0.8 (1.49)
0.6 (2.20)

-0.7 (1.79)
1.0 (0.96)
-1.8 (2.93)
0.5 (1.50)
1.6 (2.33)

I-7

Table I-8 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 26 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
Comparison
(n = 21,847)1

2012
Comparison
(n = 10,412)1,2

8.9
8.6
8.3
1.0
0.3
0.2a
0.6
0.1
0.0
0.3
0.3

9.1
8.9
8.5
1.4
0.3
0.1a
0.7
0.1
0.0
0.4
0.3

1.7
1.1
25.4
3.9a
70.5

2.0
1.4
25.5
4.0a
71.3

2012 QFT
(n = 999)1,3
9.4
9.2
9.0
1.3
0.4
0.0*
1.0
0.2
0.0*
0.4
0.4

1.9
1.3
27.4
6.6
70.3

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

-0.6 (1.23)
-0.6 (1.17)
-0.7 (1.15)
-0.3 (0.37)
-0.1 (0.20)
0.2 (0.03)
-0.4 (0.33)
-0.1 (0.13)
0.0 (0.01)
-0.0 (0.18)
-0.1 (0.21)

-0.3 (1.24)
-0.3 (1.18)
-0.5 (1.16)
0.1 (0.39)
-0.1 (0.21)
0.1 (0.03)
-0.3 (0.33)
-0.0 (0.14)
0.0 (0.01)
0.0 (0.18)
-0.1 (0.21)

-0.2 (0.47)
-0.2 (0.37)
-2.0 (2.10)
-2.7 (0.78)
0.1 (2.11)

0.1 (0.50)
0.1 (0.40)
-1.9 (2.21)
-2.6 (0.79)
0.9 (2.10)

I-8

Table I-9 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 12 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
Comparison
(n = 65,928)1

2012
Comparison
(n = 31,213)1,2

2012 QFT
(n = 2,044)1,3

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

7.7
7.5
7.3
0.5
0.1a
0.1
0.4
0.1
0.0
0.2a
0.2

7.6
7.4
7.2
0.5
0.1a
0.1
0.4
0.1
0.0
0.2
0.2

7.8
7.6
7.4
0.3
0.0
0.1
0.4
0.1
0.1
0.1
0.3

-0.1 (0.86)
-0.1 (0.86)
-0.1 (0.82)
0.2 (0.14)
0.1 (0.03)
0.1 (0.04)
-0.0 (0.13)
-0.1 (0.07)
-0.1 (0.04)
0.1 (0.06)
-0.0 (0.10)

-0.2 (0.88)
-0.2 (0.88)
-0.2 (0.84)
0.2 (0.15)
0.1 (0.03)
0.0 (0.05)
-0.0 (0.14)
-0.0 (0.07)
-0.1 (0.04)
0.1 (0.06)
-0.1 (0.10)

1.1
0.6
22.5
3.4a
53.0
22.3

1.0
0.6
22.2
3.5a
53.4
22.9

1.0
0.4
24.2
5.2
51.6
23.9

0.2 (0.22)
0.3 (0.16)
-1.8 (1.57)
-1.8 (0.59)
1.4 (1.79)
-1.6 (1.24)

0.0 (0.23)
0.2 (0.17)
-2.0 (1.65)
-1.7 (0.58)
1.8 (1.80)
-1.1 (1.31)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or
inhalants but do not include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit
Drugs Other Than Marijuana in this definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include
hallucinogens, inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
6
Cocaine use includes crack.
7
Smokeless tobacco refers to snuff or chewing tobacco (2011 and 2012 comparison data), or snuff, dip, chewing
tobacco, or "snus" (QFT). For the 2011 and 2012 comparison data, estimates are based on responses to separate
sets of questions about use of snuff and use of chewing tobacco. Estimates for the QFT are based on responses to
questions about use of any smokeless tobacco product.
8
Binge Alcohol Use in the 2011 and 2012 comparison data is defined as drinking five or more drinks on the same
occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days.
Binge Alcohol Use in the QFT is defined for males as drinking five or more drinks on the same occasion and for
females as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-9

Table I-10 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 12 to 17: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
Comparison
(n = 22,419)1

2012
Comparison
(n = 10,465)1,2

2012 QFT
(n = 541)1,3

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

8.5
7.7
7.7
0.3a
0.0
0.1
0.8
0.1
0.0
0.4
0.8

7.2
6.6
6.6
0.1a
0.0
0.0*
0.5
0.1
0.0*
0.2
0.5

8.1
6.7
6.7
0.0*
0.0*
0.0*
1.2
0.2
0.3
0.3
1.0

0.4 (1.23)
1.0 (1.09)
1.0 (1.08)
0.3 (0.05)
0.0 (0.01)
0.1 (0.03)
-0.4 (0.50)
-0.0 (0.16)
-0.3 (0.25)
0.1 (0.25)
-0.2 (0.48)

-0.9 (1.28)
-0.1 (1.12)
-0.1 (1.12)
0.1 (0.03)
0.0 (0.01)
0.0 (0.00)
-0.7 (0.51)
-0.0 (0.14)
-0.3 (0.25)
-0.1 (0.24)
-0.5 (0.48)

1.7
0.3a
7.8
2.1
13.4a
6.9

1.0
0.1a
6.1
2.2
11.6
6.2

1.7
0.0*
6.1
3.7
10.3
5.6

0.1 (0.61)
0.3 (0.06)
1.7 (1.18)
-1.6 (1.02)
3.1 (1.28)
1.3 (1.01)

-0.6 (0.61)
0.1 (0.03)
-0.1 (1.22)
-1.5 (1.03)
1.3 (1.22)
0.6 (0.98)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or
inhalants but do not include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit
Drugs Other Than Marijuana in this definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include
hallucinogens, inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
6
Cocaine use includes crack.
7
Smokeless tobacco refers to snuff or chewing tobacco (2011 and 2012 comparison data), or snuff, dip, chewing
tobacco, or "snus" (QFT). For the 2011 and 2012 comparison data, estimates are based on responses to separate
sets of questions about use of snuff and use of chewing tobacco. Estimates for the QFT are based on responses to
questions about use of any smokeless tobacco product.
8
Binge Alcohol Use in the 2011 and 2012 comparison data is defined as drinking five or more drinks on the same
occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days.
Binge Alcohol Use in the QFT is defined for males as drinking five or more drinks on the same occasion and for
females as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-10

Table I-11 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 18 to 25: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

2011
Comparison
(n = 21,662)1

2012
Comparison
(n = 10,336)1,2

2012 QFT
(n = 504)1,3

19.9
19.6
19.2
1.3a
0.1
0.3
1.7
0.3
0.0
0.9
0.4

19.5
19.2
18.9
1.0
0.1
0.3
1.6
0.4
0.0
0.9
0.3

18.2
17.8
17.8
0.4
0.1
0.4
2.0
0.5
0.2
0.5
0.6

1.7 (2.18)
1.8 (2.19)
1.4 (2.18)
0.9 (0.35)
-0.1 (0.14)
-0.1 (0.30)
-0.3 (0.76)
-0.2 (0.32)
-0.2 (0.23)
0.5 (0.35)
-0.2 (0.37)

1.2 (2.16)
1.4 (2.16)
1.1 (2.16)
0.6 (0.33)
-0.1 (0.15)
-0.1 (0.30)
-0.5 (0.79)
-0.1 (0.35)
-0.2 (0.23)
0.4 (0.36)
-0.3 (0.37)

3.1
1.5a
34.0
5.6
61.4
39.3

2.7
1.2
31.8
5.7
61.8
39.6

3.1
0.7
33.7
4.8
60.9
41.5

0.0 (0.87)
0.9 (0.43)
0.2 (2.63)
0.8 (1.26)
0.6 (2.82)
-2.2 (3.15)

-0.4 (0.90)
0.5 (0.42)
-1.9 (2.67)
0.9 (1.26)
0.9 (3.05)
-1.8 (3.21)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or
inhalants but do not include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit
Drugs Other Than Marijuana in this definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include
hallucinogens, inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
6
Cocaine use includes crack.
7
Smokeless tobacco refers to snuff or chewing tobacco (2011 and 2012 comparison data), or snuff, dip, chewing
tobacco, or "snus" (QFT). For the 2011 and 2012 comparison data, estimates are based on responses to separate
sets of questions about use of snuff and use of chewing tobacco. Estimates for the QFT are based on responses to
questions about use of any smokeless tobacco product.
8
Binge Alcohol Use in the 2011 and 2012 comparison data is defined as drinking five or more drinks on the same
occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days.
Binge Alcohol Use in the QFT is defined for males as drinking five or more drinks on the same occasion and for
females as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-11

Table I-12 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 26 or Older: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
Comparison
(n = 21,847)1

2012
Comparison
(n = 10,412)1,2

5.4
5.3
5.1
0.4
0.1a
0.1a
0.1
0.0
0.0
0.1a
0.1

5.5
5.4
5.2
0.5
0.1a
0.1a
0.2
0.0
0.0
0.1a
0.1

0.7
0.5
22.3
3.1a
56.7
21.4

0.7
0.5
22.6
3.3a
57.4
22.1

2012 QFT
(n = 999)1,3
5.9
5.9
5.7
0.3
0.0*
0.0*
0.1
0.1
0.0*
0.0*
0.1

0.5
0.3
24.9
5.5
55.4
23.2

QFT vs. 2011
Comparison,
Difference
(SE)

QFT vs. 2012
Comparison,
Difference
(SE)

-0.5 (0.92)
-0.6 (0.92)
-0.6 (0.88)
0.1 (0.18)
0.1 (0.02)
0.1 (0.02)
0.0 (0.06)
-0.0 (0.06)
0.0 (0.00)
0.1 (0.02)
0.1 (0.09)

-0.4 (0.96)
-0.4 (0.96)
-0.5 (0.93)
0.1 (0.19)
0.1 (0.04)
0.1 (0.03)
0.1 (0.07)
-0.0 (0.06)
0.0 (0.01)
0.1 (0.04)
0.0 (0.10)

0.2 (0.21)
0.1 (0.18)
-2.6 (1.91)
-2.3 (0.69)
1.3 (2.16)
-1.9 (1.37)

0.2 (0.22)
0.2 (0.19)
-2.3 (2.00)
-2.2 (0.70)
2.0 (2.19)
-1.2 (1.52)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or
inhalants but do not include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit
Drugs Other Than Marijuana in this definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include
hallucinogens, inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
6
Cocaine use includes crack.
7
Smokeless tobacco refers to snuff or chewing tobacco (2011 and 2012 comparison data), or snuff, dip, chewing
tobacco, or "snus" (QFT). For the 2011 and 2012 comparison data, estimates are based on responses to separate
sets of questions about use of snuff and use of chewing tobacco. Estimates for the QFT are based on responses to
questions about use of any smokeless tobacco product.
8
Binge Alcohol Use in the 2011 and 2012 comparison data is defined as drinking five or more drinks on the same
occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days.
Binge Alcohol Use in the QFT is defined for males as drinking five or more drinks on the same occasion and for
females as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-12

Table I-13 Specific Hallucinogen Use in Lifetime, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Hallucinogen/Age Group
Hallucinogens, Aged 12 or
Older
Ketamine4,5
DMT, AMT, or 5-MeODIPT ("Foxy")4
Salvia divinorum4
Other Hallucinogens6
Hallucinogens, Aged 12 to 17
Ketamine4,5
DMT, AMT, or 5-MeODIPT ("Foxy")4
Salvia divinorum4
Other Hallucinogens6
Hallucinogens, Aged 18 to 25
Ketamine4,5
DMT, AMT, or 5-MeODIPT ("Foxy")4
Salvia divinorum4
Other Hallucinogens6
Hallucinogens, Aged 26 or
Older
Ketamine4,5
DMT, AMT, or 5-MeODIPT ("Foxy")4
Salvia divinorum4
Other Hallucinogens6

2011
Comparison
vs. QFT,
Difference
(SE)

2012
Comparison
vs. QFT,
Difference
(SE)

2011
Comparison
(n = 65,928)1

2012
Comparison
(n = 31,213)1,2

2012 QFT
(n = 2,044)1,3

14.8
1.0

15.0
1.1

16.2
1.4

-1.4 (1.33)
-0.4 (0.31)

-1.2 (1.34)
-0.3 (0.32)

0.4
2.1
1.6a
3.7a
0.4

0.7
2.0
1.6a
3.2a
0.2

0.6
2.4
0.6
6.5
0.6

-0.2 (0.18)
-0.3 (0.46)
1.0 (0.18)
-2.7 (1.32)
-0.2 (0.35)

0.1 (0.20)
-0.4 (0.46)
1.1 (0.19)
-3.3 (1.37)
-0.4 (0.35)

0.3
1.5
1.0
18.1
1.5

0.4
0.8
1.0
18.0
1.7

0.7
2.0
0.8
19.4
1.6

-0.4 (0.40)
-0.5 (0.68)
0.2 (0.39)
-1.3 (2.26)
-0.1 (0.62)

-0.3 (0.41)
-1.2 (0.67)
0.2 (0.41)
-1.4 (2.32)
0.1 (0.62)

1.5
9.1
3.8a

2.2
7.9
3.4a

1.2
8.0
1.7

0.2 (0.49)
1.1 (1.78)
2.1 (0.59)

0.9 (0.51)
-0.1 (1.79)
1.8 (0.67)

15.7
0.9

16.0
1.1

16.9
1.4

-1.3 (1.58)
-0.5 (0.38)

-0.9 (1.58)
-0.3 (0.39)

0.3
1.0
1.2a

0.4
1.1
1.4a

0.5
1.5
0.3

-0.2 (0.21)
-0.5 (0.44)
0.9 (0.19)

-0.0 (0.24)
-0.4 (0.44)
1.0 (0.20)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
AMT = alpha-methyltryptamine; DMT = dimethyltryptamine; 5-MeO-DIPT = 5-methoxy-diisopropyltryptamine; QFT
= Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Asked in the hallucinogens module in the QFT and in the special drugs module in the 2011 and 2012 comparison
data.
5
Ketamine is also known as "Special K" or "Super K."
6
For the 2011 and 2012 comparison data, use of any other hallucinogens besides the following: LSD, also called
"acid"; PCP, also called "angel dust" or phencyclidine; peyote; mescaline; psilocybin; or "Ecstasy," also called
MDMA. For the QFT, use of any other hallucinogens besides the ones in the 2011 and 2012 comparison data, plus
the following: ketamine; DMT, AMT, or 5-MeO-DIPT ("Foxy"); or Salvia divinorum.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-13

Table I-14 Specific Inhalant Use in Lifetime, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Inhalant/Age Group
Inhalants, Aged 12 or Older
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays4
Other Inhalants5
Inhalants, Aged 12 to 17
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays4
Other Inhalants5
Inhalants, Aged 18 to 25
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays4
Other Inhalants5
Inhalants, Aged 26 or Older
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays4
Other Inhalants5

2011
Comparison
(n = 65,928)1
8.2a
N/A

2012
Comparison
(n = 31,213)1,2
8.3a
N/A

2012 QFT
(n = 2,044)1,3
11.1
3.3

2011
Comparison
vs. QFT,
Difference
(SE)
-2.8 (0.87)
N/A

2012
Comparison
vs. QFT,
Difference
(SE)
-2.8 (0.84)
N/A

N/A
0.9
0.5
7.5a
N/A

N/A
0.8
0.5
5.7a
N/A

1.2
1.0
0.5
11.7
9.4

N/A
-0.1 (0.24)
0.0 (0.19)
-4.3 (1.48)
N/A

N/A
-0.1 (0.24)
-0.1 (0.19)
-6.1 (1.46)
N/A

N/A
1.6
1.6
9.2
N/A

N/A
1.2
1.2
7.9a
N/A

1.1
1.0
0.8
11.7
5.8

N/A
0.6 (0.48)
0.8 (0.44)
-2.5 (1.75)
N/A

N/A
0.1 (0.48)
0.3 (0.45)
-3.7 (1.69)
N/A

N/A
1.8a
0.8a
8.2a
N/A

N/A
1.5a
0.7a
8.7a
N/A

2.4
0.7
0.1
10.9
2.0

N/A
1.1 (0.37)
0.7 (0.16)
-2.7 (1.05)
N/A

N/A
0.8 (0.35)
0.6 (0.17)
-2.2 (1.03)
N/A

N/A
0.6
0.4

N/A
0.7
0.3

1.0
1.0
0.6

N/A
-0.4 (0.30)
-0.2 (0.24)

N/A
-0.3 (0.30)
-0.2 (0.25)

I-14

Table I-15 Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group and
Gender: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Gender
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2011
Comparison
(n = 65,928)1
53.0
57.3
49.1
13.4a
13.3
13.6
61.4
63.9
58.9
56.7
62.2
51.7

2012
Comparison
(n = 31,213)1,2
53.4
57.3
49.8
11.6
11.5
11.7
61.8
65.2
58.4
57.4
62.2
53.0

2012 QFT
(n = 2,044)1,3
51.6
55.3
48.2
10.3
11.1
9.5
60.9
67.2
54.6
55.4
59.2
51.8

2011
Comparison
vs. QFT,
Difference
(SE)
1.4 (1.79)
2.0 (2.40)
0.9 (2.38)
3.1 (1.28)
2.2 (1.84)
4.0 (2.09)
0.6 (2.82)
-3.3 (4.23)
4.4 (3.09)
1.3 (2.16)
3.0 (2.98)
-0.1 (2.96)

2012
Comparison
vs. QFT,
Difference
(SE)
1.8 (1.80)
2.0 (2.30)
1.6 (2.50)
1.3 (1.22)
0.4 (1.74)
2.2 (2.05)
0.9 (3.05)
-2.1 (4.32)
3.8 (3.39)
2.0 (2.19)
2.9 (2.85)
1.1 (3.14)

I-15

Table I-16 Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011
Comparison
(n = 65,928)1

2012
Comparison
(n = 31,213)1,2

2011
2012
Comparison vs. Comparison vs.
QFT, Difference QFT, Difference
(SE)
(SE)

2012 QFT
Age Group/Gender
(n = 2,044)1,3
BINGE ALCOHOL USE, CORE
ONLY4
22.3
22.9
23.9
-1.6 (1.24)
-1.1 (1.31)
Aged 12 or Older
Male
29.3
30.4
30.1
-0.8 (2.00)
0.3 (2.07)
Female
15.8
15.8
18.2
-2.4 (1.33)
-2.4 (1.37)
6.9
6.2
5.6
1.3 (1.01)
0.6 (0.98)
Aged 12 to 17
Male
7.3
6.4
5.1
2.2 (1.30)
1.3 (1.23)
Female
6.4
5.9
6.1
0.3 (1.46)
-0.3 (1.40)
39.3
39.6
41.5
-2.2 (3.15)
-1.8 (3.21)
Aged 18 to 25
Male
45.7
46.5
48.1
-2.4 (4.58)
-1.6 (4.46)
Female
33.0
32.8
34.9
-1.9 (3.24)
-2.0 (3.34)
21.4
22.1
23.2
-1.9 (1.37)
-1.2 (1.52)
Aged 26 or Older
Male
29.4
30.7
30.2
-0.9 (2.30)
0.5 (2.48)
Female
14.0
14.1
16.8
-2.8 (1.62)
-2.7 (1.68)
BINGE ALCOHOL USE, CORE
PLUS NONCORE5
24.9
25.4
23.9
0.9 (1.25)
1.5 (1.32)
Aged 12 or Older
Male
29.3
30.4
30.1
-0.8 (2.00)
0.3 (2.07)
Female
20.7
20.8
18.2
2.5 (1.36)
2.6 (1.38)
7.5
6.8
5.6
1.9 (1.02)
1.2 (0.98)
Aged 12 to 17
Male
7.3
6.4
5.1
2.2 (1.30)
1.3 (1.23)
Female
7.8
7.1
6.1
1.7 (1.47)
1.0 (1.40)
42.4
43.0
41.5
1.0 (3.16)
1.5 (3.25)
Aged 18 to 25
Male
45.7
46.5
48.1
-2.4 (4.58)
-1.6 (4.46)
Female
39.2
39.5
34.9
4.3 (3.23)
4.6 (3.41)
24.0
24.8
23.2
0.8 (1.37)
1.5 (1.52)
Aged 26 or Older
Male
29.4
30.7
30.2
-0.9 (2.30)
0.5 (2.48)
Female
19.1
19.3
16.8
2.3 (1.64)
2.5 (1.70)
* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Binge Alcohol Use in the 2011 and 2012 comparison data based on only core alcohol module data is defined as drinking five or
more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past
30 days. Binge Alcohol Use in the QFT is defined for males as drinking five or more drinks on the same occasion and for
females as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
5
Binge Alcohol Use in the 2011 and 2012 comparison data based on core plus noncore data is defined for males as drinking five
or more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past
30 days. The measure for females in the 2011 and 2012 comparison data is defined as drinking five or more drinks on the same
occasion on at least 1 day in the past 30 days or usually having four drinks on those days when respondents drank alcohol in the
past 30 days based on the core alcohol module data, or drinking four or more drinks on the same occasion on at least 1 day in
the past 30 days (including the last occasion of alcohol use) based on the noncore consumption of alcohol module data. QFT
data for binge alcohol use based on the core alcohol module data are repeated in these rows.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and
2012.

I-16

Table I-17 Lifetime Use of Felt-Tip Pens, Computer Cleaners, or Other Inhalants, by Age Group
and Past Year Use of Inhalants according to Types of Inhalants Used in Lifetime among
Persons Aged 12 or Older: Percentages, 2012 Questionnaire Field Test

Inhalant/Age Group
LIFETIME USE
Felt-Tip Pens or Computer Keyboard
Cleaner3
Other Inhalants, Excluding Felt-Tip Pens or
Computer Keyboard Cleaner4
PAST YEAR USE
Among Lifetime Users of Felt-Tip Pens or
Computer Keyboard Cleaner3
Among Lifetime Users of Other Inhalants,
Excluding Users of Felt-Tip Pens or
Computer Keyboard Cleaner4

Aged 12 or
Older
(n = 2,044)1,2

Aged 12 to 17 Aged 18 to 25
(n = 541)1,2
(n = 504)1,2

Aged 26 or
Older
(n = 999)1,2

4.1

10.0

7.4

2.8

7.0

1.8

4.3

8.1

12.8

—

5.0

—

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
— Estimate not made because of small sample size.
NOTE: Denominators for lifetime use estimates consist of the total QFT sample for persons aged 12 or older or
within the specific age groups. Denominators for past year use estimates among persons aged 12 or older
consist of lifetime users of inhalants aged 12 or older who reported use of felt-tip pens or computer keyboard
cleaner (n = 128) or who reported lifetime use of other inhalants but not these two specific inhalants (n =
115).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Estimates could include lifetime use of other inhalants in addition to lifetime use of felt-tip pens, felt-tip markers, or
magic markers; or computer cleaner, also known as air duster.
4
Other inhalants in the QFT include the following: amyl nitrite, "poppers," locker room odorizers, or "rush";
correction fluid, degreaser, or cleaning fluid; gasoline or lighter fluid; glue, shoe polish, or toluene; halothane, ether,
or other anesthetics; lacquer thinner or other paint solvents; lighter gases, such as butane or propane; nitrous oxide or
"whippits"; spray paints; other aerosol sprays, or other inhalants besides those that were listed.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
2011 and 2012.

I-17

Table I-18 Use of Hallucinogens in Lifetime among Persons Aged 12 or Older with or without
Noncore Hallucinogen Data, by Age Group: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
Comparison
vs. QFT,
Difference
(SE)

2012
Comparison
vs. QFT,
Difference
(SE)

14.8

15.0

16.2

-1.4 (1.33)

-1.2 (1.34)

15.4

15.5

16.2

-0.9 (1.34)

-0.7 (1.34)

3.7a

3.2a

6.5

-2.7 (1.32)

-3.3 (1.37)

6.5

-2.0 (1.33)

-2.8 (1.36)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore

4.5

3.6

18.1

18.0

19.4

-1.3 (2.26)

-1.4 (2.32)

20.3

19.8

19.4

0.9 (2.27)

0.4 (2.31)

15.7

16.0

16.9

-1.3 (1.58)

-0.9 (1.58)

15.9

16.3

16.9

-1.0 (1.58)

-0.7 (1.58)

Aged 18 to 25
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data) 4
4

Core Plus Noncore

I-18

Table I-19 Use of Hallucinogens in the Past Year among Persons Aged 12 or Older with or without
Noncore Hallucinogen Data, by Age Group: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
Comparison
vs. QFT,
Difference
(SE)

2012
Comparison
vs. QFT,
Difference
(SE)

1.6

2.1

-0.5 (0.43)

1.9

1.8

2.1

-0.2 (0.43)

-0.3 (0.43)

2.4

2.1

3.6

-1.1 (1.01)

-1.4 (1.04)

2.9

2.4

3.6

-0.7 (1.02)

-1.2 (1.04)

6.8

6.5

7.4

-0.5 (1.59)

-0.8 (1.61)

7.9

7.0

7.4

0.5 (1.60)

-0.3 (1.61)

0.6

0.7

1.0

-0.4 (0.33)

-0.3 (0.33)

0.7

0.8

1.0

-0.3 (0.33)

-0.2 (0.33)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 18 to 25

Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data) 4
4

Core Plus Noncore

I-19

Table I-20 Use of Hallucinogens in the Past Month among Persons Aged 12 or Older with or
without Noncore Hallucinogen Data, by Age Group: Percentages, Differences, and
Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
Comparison
vs. QFT,
Difference
(SE)

2012
Comparison
vs. QFT,
Difference
(SE)

0.4

-0.0 (0.13)

-0.0 (0.14)

0.5

0.4

0.0 (0.13)

-0.0 (0.14)

0.8

0.5

1.2

-0.4 (0.50)

-0.7 (0.51)

1.0

0.6

1.2

-0.2 (0.50)

-0.6 (0.51)

1.7

1.6

2.0

-0.3 (0.76)

-0.5 (0.79)

1.9

1.7

2.0

-0.1 (0.76)

-0.4 (0.79)

0.1

0.2

0.1

0.0 (0.06)

0.1 (0.07)

0.1

0.2

0.1

0.1 (0.06)

0.1 (0.07)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 18 to 25

Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data) 4
4

Core Plus Noncore

I-20

Appendix J: Detailed Tables for Methamphetamine and
Prescription Drug Items in the 2011 and
2012 Comparison Data and the QFT

Table J-1 Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons Aged 12 or
Older: Percentages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test
2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
a
20.5
21.0
17.9
2.6 (1.37)
3.1 (1.29)

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse

13.6
a

14.4a

12.0

1.6 (1.05)

2.4 (1.00)

5.6

3.2 (0.80)

3.8 (0.77)

Tranquilizer Misuse

8.8

9.3

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

3.0

3.3

3.4

-0.4 (0.58)

-0.1 (0.56)

8.2

8.3

9.0

-0.7 (1.05)

-0.7 (0.98)

N/A

3.9

N/A

6.5

-1.7 (0.88)

-1.7 (0.82)

Stimulant Misuse, QFT Definition

Methamphetamine Use

4.8

4,5,8

48.6

49.3

50.1

-1.4 (1.72)

-0.8 (1.77)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

45.4

46.0

47.5

-2.1 (1.79)

-1.4 (1.84)

30.2

30.4

30.9

-0.7 (1.56)

-0.5 (1.55)

Illicit Drugs, Standard Definition
9

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the
2011 and 2012 comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core
plus noncore data).
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is
included as a stimulant and a prescription drug for the 2011 and 2012 comparison data, but is not included for the 2012
QFT.
6
Estimate for the 2012 QFT includes data for methamphetamine and misuse of prescription stimulants.
7
Estimate for the 2012 QFT includes data only for misuse of prescription stimulants.
8
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack),
heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics that was misused. For the 2012 QFT, both
measures also included methamphetamine.
9
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because
methamphetamine is included as a stimulant in the 2011 and 2012 comparison data, methamphetamine users in these
data by definition also are misusers of stimulants and psychotherapeutics. However, comparison data respondents who
misused psychotherapeutics but did not use methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011
and 2012.

J-1

Table J-2 Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons Aged 12 to
17: Percentages, Differences, and Standard Error of Differences, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 22,419)1 (n = 10,465)1,2
10.1
9.8

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 541)1,3 Difference (SE)
(SE)
7.7
2.5 (1.28)
2.2 (1.21)

Pain Reliever Misuse

8.6

8.2

6.4

2.2 (1.11)

1.8 (1.08)

Tranquilizer Misuse

2.8

2.9

2.4

0.5 (0.79)

0.5 (0.81)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

0.6

0.7

0.3

0.3 (0.22)

0.3 (0.23)

2.1

2.2

-0.2 (0.65)

-0.1 (0.68)

N/A

1.9

N/A

0.8

0.7

Stimulant Misuse, QFT Definition

Methamphetamine Use

Illicit Drugs, Standard Definition
9

4,5,8

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

25.5
22.4

16.0

0.5

0.3 (0.30)

0.2 (0.30)

23.4

28.5

-3.0 (2.14)

-5.1 (2.19)

20.1

26.7

-4.4 (2.10)

-6.7 (2.14)

14.1a

19.1

-3.1 (2.10)

-5.1 (2.05)

J-2

Table J-3 Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons Aged 18 to
25: Percentages, Differences, and Standard Error of Differences, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,662)1 (n = 10,336)1,2
27.9
27.9

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse

22.7

22.2

Tranquilizer Misuse

12.7

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6
Stimulant Misuse, QFT Definition

Methamphetamine Use

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 504)1,3 Difference (SE)
(SE)
26.6
1.3 (2.24)
1.2 (2.26)
19.9

2.7 (2.14)

2.2 (2.12)

8.8

3.9 (1.51)

4.1 (1.60)

2.6

-1.2 (0.78)

-1.5 (0.76)

12.9

1.4

1.1

9.5

13.1

-3.6 (1.94)

-3.6 (1.90)

N/A

11.0

N/A

3.4

2.9

4.1

-0.7 (0.92)

-1.2 (0.93)

4,5,8

58.0

58.2

58.6

-0.6 (2.37)

-0.4 (2.61)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

54.6

54.3

56.0

-1.4 (2.36)

-1.7 (2.58)

35.3

35.4

37.0

-1.7 (2.62)

-1.6 (2.66)

Illicit Drugs, Standard Definition
9

J-3

Table J-4 Misuse of Prescription Drugs or Methamphetamine in Lifetime among Persons Aged 26 or
Older: Percentages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,847)1 (n = 10,412)1,2
20.5
21.2a

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse

12.7
a

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 999)1,3 Difference (SE)
(SE)
17.7
2.8 (1.64)
3.5 (1.59)

13.8a

11.3

1.4 (1.20)

2.5 (1.18)

5.4

3.4 (0.91)

4.1 (0.88)

Tranquilizer Misuse

8.8

9.5

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

3.6

4.1

3.9

-0.3 (0.74)

0.1 (0.72)

8.8

8.9

9.1

-0.3 (1.25)

-0.2 (1.18)

N/A

2.9

N/A

Stimulant Misuse, QFT Definition

Methamphetamine Use

5.6

7.7

-2.1 (1.13)

-2.1 (1.04)

4,5,8

50.0

51.1

51.4

-1.4 (2.22)

-0.3 (2.31)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

46.8

48.0

48.7

-1.9 (2.27)

-0.7 (2.37)

31.1

31.6

31.4

-0.2 (1.88)

0.2 (1.91)

Illicit Drugs, Standard Definition
9

J-4

Table J-5 Misuse of Prescription Drugs or Methamphetamine in the Past Year among Persons Aged
12 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
a
a
5.7
5.9
8.1
-2.3 (0.84)
-2.1 (0.82)

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse
OxyContin® Misuse6

4.3a

4.4a

6.0

-1.7 (0.76)

-1.6 (0.76)

0.6

0.5

1.1

-0.4 (0.35)

-0.6 (0.36)

Tranquilizer Misuse

2.0

2.3

Sedative Misuse
Stimulant Misuse, Standard
Definition4,7

0.2

Stimulant Misuse, QFT Definition
4

Methamphetamine Use

-0.3 (0.39)

-0.1 (0.39)

0.8

-0.6 (0.22)

1.2a

2.1

-1.0 (0.40)

-0.9 (0.39)

0.2

1.1a
8

2.4

N/A

1.8

N/A

0.4

0.5

-0.1 (0.20)

-0.2 (0.20)

4,5,9

15.2

15.6

17.1

-1.9 (1.26)

-1.5 (1.23)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,9

12.8

12.9

13.7

-0.8 (1.21)

-0.7 (1.18)

7.4a

7.8a

9.7

-2.3 (0.95)

-2.0 (0.95)

Illicit Drugs, Standard Definition
10

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the
2011 and 2012 comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core
plus noncore data).
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is
included as a stimulant and a prescription drug for the 2011 and 2012 comparison data, but is not included for the 2012
QFT.
6
Lifetime and Past Month misuse of OxyContin® are not shown because these estimates cannot be produced from the
2012 QFT.
7
Estimate for the 2012 QFT includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the 2012 QFT includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack),
heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics that was misused. For the 2012 QFT, both
measures also included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because
methamphetamine is included as a stimulant in the 2011 and 2012 comparison data, methamphetamine users in these
data by definition also are misusers of stimulants and psychotherapeutics. However, comparison data respondents who
misused psychotherapeutics but did not use methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011
and 2012.

J-5

Table J-6 Misuse of Prescription Drugs or Methamphetamine in the Past Year among Persons Aged
12 to 17: Percentages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 22,419)1 (n = 10,465)1,2
6.8
6.1

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 541)1,3 Difference (SE)
(SE)
6.6
0.3 (1.25)
-0.5 (1.26)

Pain Reliever Misuse
OxyContin® Misuse6

5.8

4.9

5.0

0.8 (1.05)

-0.2 (1.08)

0.8

0.5

0.8

0.0 (0.45)

-0.2 (0.45)

Tranquilizer Misuse

1.9

1.7

2.0

-0.2 (0.76)

-0.3 (0.78)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,7

0.3

0.0 (0.22)

-0.0 (0.22)

1.2

1.4

-0.2 (0.50)

-0.2 (0.51)

N/A

1.2

N/A

0.4

0.3

Stimulant Misuse, QFT Definition

Methamphetamine Use

Illicit Drugs, Standard Definition
10

4,5,9

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,9

0.2

0.3 (0.16)

0.2 (0.17)

16.6

20.6

-2.1 (1.92)

-4.0 (1.98)

15.9

14.2

18.2

-2.3 (1.82)

-4.0 (1.89)

9.9

8.3

11.6

-1.7 (1.74)

-3.3 (1.75)

18.5

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the
2011 and 2012 comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core
plus noncore data).
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is
included as a stimulant and a prescription drug for the 2011 and 2012 comparison data, but is not included for the 2012
QFT.
6
Lifetime and Past Month misuse of OxyContin® are not shown because these estimates cannot be produced from the
2012 QFT.
7
Estimate for the 2012 QFT includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the 2012 QFT includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack),
heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics that was misused. For the 2012 QFT, both
measures also included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because
methamphetamine is included as a stimulant in the 2011 and 2012 comparison data, methamphetamine users in these
data by definition also are misusers of stimulants and psychotherapeutics. However, comparison data respondents who
misused psychotherapeutics but did not use methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011
and 2012.

J-6

Table J-7 Misuse of Prescription Drugs or Methamphetamine in the Past Year among Persons Aged
18 to 25: Percentages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,662)1 (n = 10,336)1,2
13.0a
13.2a

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse
OxyContin® Misuse6

10.0a

9.3a

15.2

-5.2 (1.95)

-5.9 (1.96)

1.9

1.4

Tranquilizer Misuse

4.6

Sedative Misuse
Stimulant Misuse, Standard
Definition4,7

0.4

3.2a

Stimulant Misuse, QFT Definition

Methamphetamine Use

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 504)1,3 Difference (SE)
(SE)
22.8
-9.8 (2.27)
-9.6 (2.31)
2.9

-1.0 (0.86)

-1.5 (0.85)

4.9

7.8

-3.2 (1.34)

-2.9 (1.37)

0.3

1.8

-1.5 (0.71)

-1.6 (0.70)

3.8a

9.1

-5.9 (1.66)

-5.3 (1.66)

N/A

8.9

N/A

0.6

0.8

0.7

-0.0 (0.35)

0.2 (0.37)

4,5,9

35.9

36.8

39.1

-3.2 (2.74)

-2.4 (2.87)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,9

32.6

33.2

32.9

-0.2 (2.60)

0.3 (2.69)

17.7a

17.9a

25.3

-7.6 (2.57)

-7.5 (2.63)

Illicit Drugs, Standard Definition
10

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the
2011 and 2012 comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core
plus noncore data).
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is
included as a stimulant and a prescription drug for the 2011 and 2012 comparison data, but is not included for the 2012
QFT.
6
Lifetime and Past Month misuse of OxyContin® are not shown because these estimates cannot be produced from the
2012 QFT.
7
Estimate for the 2012 QFT includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the 2012 QFT includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack),
heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics that was misused. For the 2012 QFT, both
measures also included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because
methamphetamine is included as a stimulant in the 2011 and 2012 comparison data, methamphetamine users in these
data by definition also are misusers of stimulants and psychotherapeutics. However, comparison data respondents who
misused psychotherapeutics but did not use methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011
and 2012.

J-7

Table J-8 Misuse of Prescription Drugs or Methamphetamine in the Past Year among Persons Aged
26 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,847)1 (n = 10,412)1,2
4.3
4.6

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 999)1,3 Difference (SE)
(SE)
5.7
-1.4 (0.86)
-1.0 (0.84)

Pain Reliever Misuse
OxyContin® Misuse6

3.1

3.5

4.5

-1.4 (0.80)

-1.0 (0.80)

0.4

0.3

0.8

-0.4 (0.42)

-0.5 (0.43)

Tranquilizer Misuse

1.6

1.9

1.4

0.1 (0.37)

0.5 (0.38)

0.6

-0.5 (0.25)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,7

0.1

Stimulant Misuse, QFT Definition

Methamphetamine Use

0.1

0.7

1.0

-0.3 (0.34)

-0.3 (0.33)

N/A

0.6

N/A

0.4

0.3

0.6

-0.2 (0.26)

-0.3 (0.26)

4,5,9

11.1

11.7

12.7

-1.6 (1.31)

-1.0 (1.32)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,9

8.9

9.2

9.7

-0.8 (1.25)

-0.5 (1.25)

5.3

5.9

6.7

-1.5 (0.94)

-0.8 (0.93)

Illicit Drugs, Standard Definition
10

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the
2011 and 2012 comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core
plus noncore data).
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is
included as a stimulant and a prescription drug for the 2011 and 2012 comparison data, but is not included for the 2012
QFT.
6
Lifetime and Past Month misuse of OxyContin® are not shown because these estimates cannot be produced from the
2012 QFT.
7
Estimate for the 2012 QFT includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the 2012 QFT includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack),
heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics that was misused. For the 2012 QFT, both
measures also included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because
methamphetamine is included as a stimulant in the 2011 and 2012 comparison data, methamphetamine users in these
data by definition also are misusers of stimulants and psychotherapeutics. However, comparison data respondents who
misused psychotherapeutics but did not use methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011
and 2012.

J-8

Table J-9 Misuse of Prescription Drugs or Methamphetamine in the Past Month among Persons Aged
12 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
2.4
2.4
3.2
-0.8 (0.47)
-0.8 (0.46)

Drug Measure
Prescription Drug Misuse4,5
Pain Reliever Misuse

1.7

2.0

-0.3 (0.37)

-0.4 (0.37)

Tranquilizer Misuse

0.7

0.8

0.9

-0.1 (0.23)

-0.1 (0.24)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

0.1

0.3

-0.2 (0.15)

0.4a

0.4

0.8

-0.4 (0.22)

-0.4 (0.21)

N/A

0.5

N/A

Stimulant Misuse, QFT Definition

Methamphetamine Use

0.2

0.1

0.4

-0.3 (0.17)

4,5,8

8.9

9.8

-0.8 (0.98)

-0.9 (0.98)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

7.7

7.6

8.0

-0.3 (0.87)

-0.4 (0.89)

3.1

3.7

-0.6 (0.49)

-0.7 (0.48)

Illicit Drugs, Standard Definition
9

J-9

Table J-10 Misuse of Prescription Drugs or Methamphetamine in the Past Month among Persons
Aged 12 to 17: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 22,419)1 (n = 10,465)1,2
2.7a
2.5a

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 541)1,3 Difference (SE)
(SE)
1.3
1.3 (0.48)
1.1 (0.50)

Pain Reliever Misuse

2.2a

2.0a

0.6

1.5 (0.33)

1.4 (0.34)

Tranquilizer Misuse

0.6

0.5

0.4

0.2 (0.28)

0.1 (0.29)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

0.1

-0.1 (0.15)

-0.0 (0.15)

0.4

0.5

-0.1 (0.27)

-0.0 (0.27)

N/A

0.3

N/A

Stimulant Misuse, QFT Definition

Methamphetamine Use

0.1

0.2

-0.0 (0.16)

4,5,8

9.8

8.6

8.5

1.3 (1.23)

0.1 (1.31)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

8.5

7.2

8.1

0.4 (1.23)

-0.9 (1.28)

4.0a

3.2

2.5

1.5 (0.70)

0.7 (0.71)

Illicit Drugs, Standard Definition
9

J-10

Table J-11 Misuse of Prescription Drugs or Methamphetamine in the Past Month among Persons
Aged 18 to 25: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,662)1 (n = 10,336)1,2
5.0
4.9

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 504)1,3 Difference (SE)
(SE)
7.4
-2.3 (1.25)
-2.4 (1.29)

Pain Reliever Misuse

3.6

3.4

4.6

-1.1 (1.01)

-1.3 (1.03)

Tranquilizer Misuse

1.6

1.3

2.2

-0.6 (0.67)

-0.8 (0.66)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

0.1

-0.0 (0.15)

1.0a

2.7

-1.7 (0.72)

Stimulant Misuse, QFT Definition

Methamphetamine Use

N/A

2.4

N/A

0.2

0.3

0.5

-0.3 (0.31)

-0.2 (0.31)

4,5,8

21.7

21.4

22.7

-0.9 (2.27)

-1.3 (2.24)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

20.0

19.5

18.4

1.6 (2.17)

1.1 (2.15)

7.0

6.6

9.0

-2.0 (1.32)

-2.4 (1.32)

Illicit Drugs, Standard Definition
9

J-11

Table J-12 Misuse of Prescription Drugs or Methamphetamine in the Past Month among Persons
Aged 26 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
2012
Comparison Comparison
(n = 21,847)1 (n = 10,412)1,2
1.8
1.9

Drug Measure
Prescription Drug Misuse4,5

2012
2011
Comparison
Comparison vs.
vs. QFT,
2012 QFT
QFT,
Difference
(n = 999)1,3 Difference (SE)
(SE)
2.7
-0.8 (0.54)
-0.7 (0.53)

Pain Reliever Misuse

1.3

1.8

-0.4 (0.46)

Tranquilizer Misuse

0.6

0.7

-0.1 (0.25)

0.0 (0.26)

Sedative Misuse
Stimulant Misuse, Standard
Definition4,6

0.1

0.0

0.3

-0.3 (0.19)

0.3

0.5

-0.3 (0.23)

-0.2 (0.23)

N/A

0.2

N/A

Stimulant Misuse, QFT Definition

Methamphetamine Use

0.2

0.1

0.4

-0.3 (0.22)

-0.3 (0.21)

4,5,8

6.5

6.7

7.7

-1.1 (1.07)

-0.9 (1.10)

Alternate Definition 3
Illicit Drugs Other Than Marijuana,
Standard Definition4,5,8

5.4

5.5

6.1

-0.7 (0.94)

-0.6 (0.99)

2.3

2.4

3.0

-0.7 (0.56)

-0.5 (0.55)

Illicit Drugs, Standard Definition
9

J-12

Table J-13 Misuse of Stimulants in Lifetime among Persons Aged 12 or Older with or without
Noncore Adderall® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

2011
2012
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
2012
Comparison vs. Comparison vs.
QFT,
QFT,
Difference
Difference
(SE)
(SE)

8.2

8.3

9.0

-0.7 (1.05)

-0.7 (0.98)

9.7

9.9

9.0

0.8 (1.05)

1.0 (0.97)

N/A

3.9

N/A

2.1

2.2

-0.2 (0.65)

-0.1 (0.68)

Aged 12 to 17
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
QFT Definition6

3.6a

3.5

2.2

1.4 (0.66)

1.3 (0.68)

N/A

1.9

N/A

9.5

13.1

-3.6 (1.94)

-3.6 (1.90)

15.4

16.0

13.1

2.3 (1.97)

2.9 (1.93)

N/A

11.0

N/A

8.8

8.9

9.1

-0.3 (1.25)

-0.2 (1.18)

9.5

9.7

9.1

0.4 (1.24)

0.6 (1.17)

N/A

2.9

N/A

Aged 18 to 25
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition
Aged 26 or Older

Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

J-13

Table J-14 Misuse of Stimulants in the Past Year among Persons Aged 12 or Older with or without
Noncore Adderall® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

2011
2012
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
2012
Comparison vs. Comparison vs.
QFT,
QFT,
Difference
Difference
(SE)
(SE)

1.1a

1.2a

2.1

-1.0 (0.40)

-0.9 (0.39)

1.8

1.9

2.1

-0.3 (0.40)

-0.2 (0.40)

N/A

1.8

N/A

1.2

1.4

-0.2 (0.50)

-0.2 (0.51)

2.2

2.0

1.4

0.8 (0.50)

0.6 (0.51)

N/A

1.2

N/A

Aged 12 to 17
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
QFT Definition6
Aged 18 to 25
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

3.2a

3.8a

9.1

-5.9 (1.66)

-5.3 (1.66)

6.3

7.0

9.1

-2.8 (1.67)

-2.2 (1.69)

N/A

8.9

N/A

0.7

1.0

-0.3 (0.34)

-0.3 (0.33)

1.0

0.0 (0.34)

-0.0 (0.34)

N/A

0.6

N/A

Aged 26 or Older
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

J-14

Table J-15 Misuse of Stimulants in the Past Month among Persons Aged 12 or Older with or without
Noncore Adderall® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

2011
2012
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
2012
Comparison vs. Comparison vs.
QFT,
QFT,
Difference
Difference
(SE)
(SE)

0.4a

0.4

0.8

-0.4 (0.22)

-0.4 (0.21)

0.6

0.8

-0.2 (0.22)

-0.2 (0.21)

N/A

0.5

N/A

0.4

0.5

-0.1 (0.27)

-0.0 (0.27)

0.7

0.5

0.3 (0.26)

0.2 (0.27)

N/A

0.3

N/A

Aged 12 to 17
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
QFT Definition6
Aged 18 to 25
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

1.0a

2.7

-1.7 (0.72)

1.9

2.0

2.7

-0.9 (0.73)

-0.8 (0.76)

N/A

2.4

N/A

0.3

0.5

-0.3 (0.23)

-0.2 (0.23)

0.3

0.4

0.5

-0.2 (0.24)

-0.1 (0.23)

N/A

0.2

N/A

Aged 26 or Older
Standard Definition4
Standard Definition, Plus Noncore
Adderall®5
6

QFT Definition

J-15

Table J-16 Misuse of Sedatives in Lifetime among Persons Aged 12 or Older with or without Noncore
Ambien® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
3.0
5.0

3.3

5.1

3.4

-0.4 (0.58)

-0.1 (0.56)

3.4

1.7 (0.58)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore

0.6

0.3

0.3 (0.22)

0.3 (0.23)

0.3

1.2 (0.23)

1.2 (0.25)

1.4

1.1a

2.6

-1.2 (0.78)

-1.5 (0.76)

4.1

3.7

2.6

1.4 (0.77)

1.1 (0.78)

3.6

4.1

3.9

-0.3 (0.74)

0.1 (0.72)

3.9

1.7 (0.74)

1.9 (0.75)

1.5

0.7

1.5

Aged 18 to 25
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data)4
4

Core Plus Noncore

5.7

5.8

J-16

Table J-17 Misuse of Sedatives in the Past Year among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
0.2a

0.2a

0.8

-0.6 (0.22)

0.9

0.7

0.8

0.1 (0.21)

-0.0 (0.23)

0.3

0.0 (0.22)

-0.0 (0.22)

0.7

0.3

0.5 (0.22)

0.4 (0.22)

0.4a

0.3a

1.8

-1.5 (0.71)

-1.6 (0.70)

1.4

1.1

1.8

-0.5 (0.71)

-0.8 (0.71)

0.1a

0.6

-0.5 (0.25)

0.8

0.7

0.6

0.2 (0.25)

0.0 (0.26)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore

0.8

Aged 18 to 25
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data)4
4

Core Plus Noncore

J-17

Table J-18 Misuse of Sedatives in the Past Month among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group: Percentages, Differences, and Standard Error of
Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)4
4

Core Plus Noncore

2012
2011
Comparison
Comparison vs.
vs. QFT,
2011
2012
QFT,
Difference
2012 QFT
Comparison Comparison
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3 Difference (SE)
0.1

0.1

0.3

-0.2 (0.15)

0.3

0.1

0.3

-0.0 (0.15)

-0.2 (0.15)

0.1

-0.1 (0.15)

-0.0 (0.15)

0.2

0.1

0.1 (0.15)

0.1 (0.16)

0.1

-0.0 (0.15)

0.4

0.3

0.1

0.2 (0.15)

0.1 (0.16)

0.1

0.0

0.3

-0.3 (0.19)

0.2

0.1

0.3

-0.1 (0.19)

-0.2 (0.19)

Aged 12 to 17
Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 18 to 25

Core Only (without Noncore Data)4
4

Core Plus Noncore
Aged 26 or Older

Core Only (without Noncore Data)4
4

Core Plus Noncore

J-18

Appendix K: Detailed Tables for Noncore Estimates in the
2011 and 2012 Comparison Data and the QFT

Table K-1 Substance Dependence or Abuse in the Past Year among Persons Aged 12 or Older, by
Survey Protocol: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Dependence or Abuse Measure
DEPENDENCE
Illicit Drugs4
Marijuana
Hallucinogens
Inhalants
Prescription Drugs5
Pain Relievers
Stimulants Among
Methamphetamine Users
Methamphetamine
Illicit Drugs Other Than Marijuana4
Illicit Drugs Excluding Marijuana6
ABUSE
Illicit Drugs4
Marijuana
Hallucinogens
Inhalants
Prescription Drugs5
Pain Relievers
Illicit Drugs Other Than Marijuana4
Illicit Drugs Excluding Marijuana6
DEPENDENCE OR ABUSE
Illicit Drugs4
Marijuana
Hallucinogens
Inhalants
Prescription Drugs5
Pain Relievers
Illicit Drugs Other Than Marijuana4
Illicit Drugs Excluding Marijuana6

QFT vs. 2011 QFT vs. 2012
Comparison, Comparison,
2012
2011
Difference
Comparison Comparison 2012 QFT Difference
(SE)
(SE)
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3
1.8
1.1
0.1a
0.0
0.6
0.6

2.0
1.0
0.0
0.0
0.8
0.6

1.5
0.9
0.0
0.0
0.5
0.4

0.3 (0.26)
0.2 (0.20)
0.0 (0.02)
0.0 (0.02)
0.1 (0.16)
0.2 (0.13)

0.4 (0.28)
0.1 (0.20)
0.0 (0.02)
-0.0 (0.02)
0.2 (0.18)
0.2 (0.15)

0.1
N/A
0.9
0.8

0.1
N/A
1.1
1.0

N/A
0.0
0.8
0.7

N/A
N/A
0.2 (0.20)
0.1 (0.19)

N/A
N/A
0.3 (0.21)
0.3 (0.19)

0.8
0.6
0.1
0.0
0.2
0.2
0.3
0.3

0.8
0.6
0.1
0.0
0.2
0.2
0.4
0.3

0.9
0.8
0.1
0.0
0.2
0.2
0.3
0.3

-0.2 (0.22)
-0.2 (0.20)
-0.0 (0.05)
-0.0 (0.03)
-0.0 (0.12)
0.0 (0.09)
0.0 (0.10)
0.0 (0.11)

-0.1 (0.22)
-0.2 (0.20)
-0.0 (0.06)
-0.0 (0.04)
0.0 (0.12)
0.0 (0.09)
0.1 (0.11)
-0.0 (0.11)

2.6
1.7
0.1
0.1
0.9
0.7
1.3
1.1

2.8
1.6
0.1
0.1
1.0
0.8
1.5
1.3

2.5
1.7
0.1
0.1
0.8
0.5
1.1
1.0

0.1 (0.35)
0.0 (0.29)
0.0 (0.06)
-0.0 (0.04)
0.1 (0.20)
0.2 (0.16)
0.2 (0.21)
0.1 (0.21)

0.3 (0.36)
-0.0 (0.29)
0.0 (0.06)
-0.0 (0.04)
0.2 (0.23)
0.2 (0.18)
0.4 (0.23)
0.3 (0.22)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
a

Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type
psychotherapeutics. Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, or
prescription-type psychotherapeutics. Estimates for the QFT include relevant dependence or abuse data for methamphetamine.
5
Estimates for Prescription Drugs include misuse of pain relievers, tranquilizers, stimulants, or sedatives. Estimates for the QFT
do not include dependence or abuse data for methamphetamine.
6
Illicit Drugs Excluding Marijuana include dependence or abuse for cocaine, heroin, hallucinogens, inhalants, or prescriptiontype psychotherapeutics and require respondents not to have corresponding dependence or abuse for marijuana. Estimates for
the QFT include relevant dependence or abuse data for methamphetamine.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.
1

K-1

Table K-2 Substance Dependence or Abuse in the Past Year among Persons Aged 12 to 17, by
Survey Protocol: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011 QFT vs. 2012
Comparison, Comparison,
2012
2011
Difference
Comparison Comparison 2012 QFT Difference
(SE)
(SE)
(n = 22,419)1 (n = 10,465)1,2 (n = 541)1,3
2.5
1.9
0.1
0.1
0.6
0.5a

1.9
1.6
0.1
0.1
0.4
0.3a

0.1
N/A
0.9
0.6

0.1
N/A
0.5
0.4

2.1
1.7
0.2a
0.2
0.6a
0.5
0.9
0.8

2.0
1.7
0.2a
0.2
0.3a
0.2
0.6
0.5

4.7
3.6
0.3
0.3
1.2a
1.0a
1.7a
1.4

3.9
3.2
0.3
0.2
0.7
0.5
1.1
0.9

1.9
1.5
0.2
0.2
0.2
0.0*

0.6 (0.64)
0.4 (0.57)
-0.0 (0.16)
-0.1 (0.16)
0.4 (0.26)
0.5 (0.05)

0.0 (0.63)
0.0 (0.57)
-0.1 (0.16)
-0.1 (0.16)
0.2 (0.25)
0.3 (0.06)

N/A
N/A
0.5 (0.30)
0.2 (0.29)

N/A
N/A
0.1 (0.29)
-0.0 (0.29)

1.6
1.4
0.0*
0.4
0.0*
0.2
0.4
0.4

0.5 (0.65)
0.2 (0.61)
0.2 (0.04)
-0.2 (0.32)
0.6 (0.08)
0.2 (0.26)
0.4 (0.33)
0.4 (0.32)

0.3 (0.64)
0.2 (0.62)
0.2 (0.05)
-0.2 (0.32)
0.3 (0.07)
-0.0 (0.25)
0.1 (0.33)
0.1 (0.33)

3.5
3.0
0.2
0.6
0.2
0.2
0.8
0.8

1.1 (0.92)
0.6 (0.85)
0.2 (0.17)
-0.3 (0.35)
0.9 (0.28)
0.7 (0.27)
0.9 (0.45)
0.5 (0.43)

0.4 (0.90)
0.3 (0.84)
0.2 (0.17)
-0.3 (0.36)
0.4 (0.26)
0.3 (0.25)
0.3 (0.43)
0.1 (0.43)

N/A
0.2
0.4
0.4

K-2

Table K-3 Substance Dependence or Abuse in the Past Year among Persons Aged 18 to 25, by
Survey Protocol: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011 QFT vs. 2012
Comparison, Comparison,
2012
2011
Difference
Comparison Comparison 2012 QFT Difference
(SE)
(SE)
(n = 21,662)1 (n=10,336)1,2 (n = 504)1,3
5.4
3.8
0.2a
0.0
1.6
1.4

5.4
3.4
0.2a
0.0
1.9
1.5

0.1
N/A
2.1
1.6

0.1
N/A
2.5
2.0

2.2
2.0
0.3
0.1a
0.5
0.3
0.7
0.7

2.2
1.8
0.3
0.0
0.5
0.4
0.8
0.8

7.7
5.8
0.5
0.1a
2.1
1.8
2.8
2.3

7.6
5.2
0.4
0.1a
2.4
1.8
3.3
2.8

5.1
2.9
0.0*
0.0*
2.5
1.6

0.3 (1.05)
0.9 (0.86)
0.2 (0.05)
0.0 (0.01)
-0.9 (0.73)
-0.1 (0.59)

0.3 (1.08)
0.5 (0.87)
0.2 (0.04)
0.0 (0.02)
-0.7 (0.77)
-0.1 (0.60)

N/A
N/A
-0.9 (0.80)
-0.6 (0.73)

N/A
N/A
-0.5 (0.83)
-0.2 (0.74)

2.1
2.2
0.7
0.0*
0.5
0.4
0.8
0.9

0.1 (0.70)
-0.3 (0.76)
-0.4 (0.39)
0.1 (0.02)
-0.1 (0.30)
-0.1 (0.30)
-0.0 (0.38)
-0.2 (0.41)

0.1 (0.67)
-0.4 (0.74)
-0.4 (0.40)
0.0 (0.02)
0.0 (0.30)
-0.0 (0.31)
0.1 (0.40)
-0.1 (0.43)

7.2
5.1
0.7
0.0*
3.0
2.0
3.8
3.1

0.4 (1.26)
0.7 (1.12)
-0.2 (0.39)
0.1 (0.02)
-1.0 (0.81)
-0.2 (0.66)
-0.9 (0.93)
-0.8 (0.86)

0.4 (1.26)
0.1 (1.12)
-0.3 (0.40)
0.1 (0.03)
-0.7 (0.83)
-0.2 (0.66)
-0.5 (0.95)
-0.3 (0.87)

N/A
0.3
3.0
2.2

K-3

Table K-4 Substance Dependence or Abuse in the Past Year among Persons Aged 26 or Older, by
Survey Protocol: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

QFT vs. 2011 QFT vs. 2012
Comparison, Comparison,
2012
2011
Difference
Comparison Comparison 2012 QFT Difference
(SE)
(SE)
(n = 21,847)1 (n = 10,412)1,2 (n = 999)1,3
1.1
0.5
0.0a
0.0
0.5
0.4

1.3
0.5
0.0
0.0*
0.6a
0.5

0.9
0.4
0.0*
0.0*
0.2
0.2

0.2 (0.28)
0.0 (0.19)
0.0 (0.01)
0.0 (0.01)
0.2 (0.14)
0.2 (0.13)

0.5 (0.29)
0.1 (0.18)
0.0 (0.01)
0.0 (0.00)
0.4 (0.17)
0.3 (0.16)

0.1
N/A
0.7
0.6

0.0
N/A
0.9a
0.8

N/A
0.0*
0.4
0.4

N/A
N/A
0.3 (0.21)
0.2 (0.21)

N/A
N/A
0.5 (0.21)
0.4 (0.21)

0.3
0.2
0.0
0.0*
0.1
0.1
0.2
0.2

0.5
0.3
0.0a
0.0
0.2
0.1
0.3
0.2

0.6
0.4
0.0*
0.0*
0.2
0.1
0.2
0.2

-0.3 (0.24)
-0.2 (0.20)
0.0 (0.02)
0.0 (0.00)
-0.1 (0.14)
0.0 (0.09)
-0.0 (0.14)
-0.0 (0.14)

-0.2 (0.25)
-0.2 (0.20)
0.0 (0.02)
0.0 (0.02)
-0.0 (0.15)
0.0 (0.10)
0.1 (0.15)
0.0 (0.14)

1.4
0.7
0.1a
0.0
0.6
0.5
0.9
0.8

1.8
0.8
0.1a
0.0
0.8
0.6
1.2a
1.0

1.5
0.9
0.0*
0.0*
0.4
0.3
0.6
0.6

-0.0 (0.36)
-0.2 (0.28)
0.1 (0.02)
0.0 (0.01)
0.2 (0.20)
0.2 (0.16)
0.3 (0.23)
0.2 (0.22)

0.3 (0.38)
-0.1 (0.27)
0.1 (0.02)
0.0 (0.02)
0.3 (0.24)
0.3 (0.19)
0.5 (0.24)
0.4 (0.24)

K-4

Table K-5 Substance Use with a Needle in Lifetime, Past Year, and Past Month among Persons
Aged 12 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

2011
Substance Used with a
Comparison
Needle/Period of Use
(n = 65,928)1
USE OF HEROIN WITH A
NEEDLE
Lifetime
0.8
Past Year
0.1
Past Month
0.0
USE OF COCAINE WITH A
NEEDLE
Lifetime
0.8
Past Year
0.1a
Past Month
0.0a
USE OF METHAMPHETAMINE
WITH A NEEDLE
Lifetime
0.6
Past Year
0.1
Past Month
0.0
USE OF PRESCRIPTION
STIMULANTS WITH A
NEEDLE4
Past Year
0.1a
Past Month
0.0a
USE OF HEROIN, COCAINE,
METHAMPHETAMINE, OR
PRESCRIPTION STIMULANTS
WITH A NEEDLE4
Past Year
0.2
Past Month
0.1

2011
2012
Comparison Comparison
vs. QFT,
vs. QFT,
Difference Difference
(SE)
(SE)

2012
Comparison
(n = 31,213)1,2

2012 QFT
(n = 2,044)1,3

0.8
0.1
0.1

0.7
0.1
0.0

0.0 (0.27)
0.0 (0.04)
0.0 (0.02)

0.1 (0.27)
0.1 (0.05)
0.0 (0.03)

0.8
0.1a
0.0

1.0
0.0*
0.0*

-0.2 (0.33)
0.1 (0.02)
0.0 (0.01)

-0.3 (0.35)
0.1 (0.02)
0.0 (0.01)

0.7
0.1
0.0

0.8
0.2
0.2

-0.2 (0.27)
-0.1 (0.12)
-0.1 (0.12)

-0.1 (0.26)
-0.1 (0.12)
-0.1 (0.12)

0.1a
0.0

0.0*
0.0*

0.1 (0.01)
0.0 (0.01)

0.1 (0.02)
0.0 (0.01)

0.2
0.1

0.2
0.2

-0.1 (0.13)
-0.1 (0.12)

-0.0 (0.13)
-0.1 (0.12)

K-5

Table K-6 Demographic, Socioeconomic, and Household Characteristics among Persons Aged 12
or Older: Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

Characteristic
EDUCATION4
< High School
High School Graduate
Some College
College Graduate
OVERALL HEALTH5
Excellent
Very Good
Good
Fair/Poor
COVERED BY ANY HEALTH
INSURANCE
CURRENTLY EMPLOYED4
FAMILY INCOME
< $20,000
$20,000-$49,999
$50,000-$74,999
≥ $75,000
PARTICIPATED IN
GOVERNMENT PROGRAM6
RECEIVED INCOME
Social Security
Wages
Supplemental Security Income
Food Stamps
Welfare Payments
BETTER PROVIDER OF
INFORMATION5
USED PROXY

2012
2011
Comparison Comparison 2012 QFT
(n = 65,928)1 (n = 31,213)1,2 (n = 2,044)1,3

2011
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

2012
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

10.5
27.3
24.7
27.6

10.4
27.1
25.0
27.6

11.1
23.9
28.9
26.1

4.05, 0.004c

3.34, 0.0129c

24.2
38.2
25.7
11.8

23.4
38.0
26.2
12.5

22.3
40.4
26.2
11.2

1.19, 0.3185

1.04, 0.3772

86.3
63.8

87.0
65.2

85.7
66.2

0.33, 0.5665
1.61, 0.2073

1.89, 0.1724
0.29, 0.5936

18.2
31.0
17.5
33.3

18.5
31.7
16.8
33.0

19.4
33.3
16.3
31.0

1.01, 0.3905

0.50, 0.6854

19.1

20.5

24.7

12.96, 0.0005c

6.99, 0.0094 c

27.2
82.4
7.0
14.6
2.5

26.2
82.8
7.6
15.6
2.3

26.4
68.6
9.4
17.6
3.6

0.20, 0.6557
77.07, 0.0000c
7.66, 0.0067c
4.88, 0.0293c
4.70, 0.0324c

0.01, 0.9049
74.48, 0.0000c
3.50, 0.0641
1.98, 0.1628
7.46, 0.0074

19.0
13.7

20.1
13.9

22.3
15.7

7.82, 0.0062c
4.87, 0.0296c

3.48, 0.0650
4.03, 0.0473c

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Education and employment estimates are based only on respondents aged 18 or older. Sample sizes for respondents 18 or older are
n = 43,509 for 2011 comparison, n = 1,503 for QFT, and n = 20,748 for 2012 comparison.
5
Respondents with unknown data were excluded.
6
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI),
cash assistance (Temporary Assistance for Needy Families, TANF), noncash assistance, or food stamps.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-6

Table K-7 Demographic, Socioeconomic, and Household Characteristics among Persons Aged 12 to
17: Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

Characteristic
EDUCATION
< High School
High School Graduate
Some College
College Graduate
OVERALL HEALTH4
Excellent
Very Good
Good
Fair/Poor
COVERED BY ANY HEALTH
INSURANCE
CURRENTLY EMPLOYED
FAMILY INCOME
< $20,000
$20,000-$49,999
$50,000-$74,999
≥ $75,000
PARTICIPATED IN
GOVERNMENT PROGRAM5
RECEIVED INCOME
Social Security
Wages
Supplemental Security Income
Food Stamps
Welfare Payments
BETTER PROVIDER OF
INFORMATION4
USED PROXY

2012
2011
Comparison Comparison 2012 QFT
(n = 22,419)1 (n = 10,465)1,2 (n = 541)1,3

2011
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

2012
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

N/A
N/A
N/A
N/A

N/A

34.1
42.2
20.1
3.6

35.9
41.3
19.2
3.5

33.0
41.5
20.4
5.1

0.96, 0.4162

1.54, 0.2098

93.4
N/A

92.8
N/A

91.4
N/A

2.66, 0.1057
N/A

1.16, 0.2844
N/A

16.6
31.2
16.8
35.4

18.0
29.6
16.7
35.7

22.1
32.7
12.3
32.9

3.52, 0.0176c

2.65, 0.0530

25.4

26.4

32.2

7.66, 0.0067c

5.53, 0.0205c

12.2
89.4
7.6
20.9
4.2

11.1
89.6
7.8
21.4
4.0

12.7
65.6
9.9
27.7
5.6

0.08, 0.7725
140.89, 0.0000c
2.18, 0.1430
8.38, 0.0046c
1.72, 0.1927

0.80, 0.3728
148.82, 0.0000c
1.99, 0.1609
6.90, 0.0099c
2.60, 0.1098

88.2
83.8

89.2
84.5

90.4
83.8

1.36, 0.2465
0.00, 0.9779

0.39, 0.5322
0.09, 0.7711

* Low precision; estimate would be suppressed under NSDUH suppression rules.
N/A = not applicable; QFT = Questionnaire Field Test.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Respondents with unknown data were excluded.
5
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI),
cash assistance (Temporary Assistance for Needy Families, TANF), noncash assistance, or food stamps.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-7

Table K-8 Demographic, Socioeconomic, and Household Characteristics among Persons Aged 18 to
25: Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

2012
2011
Comparison Comparison 2012 QFT
(n = 21,662)1 (n = 10,336)1,2 (n = 504)1,3

2011
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

2012
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

15.6
34.0
35.7
14.7

12.0
35.7
36.4
15.9

13.8
34.9
37.6
13.7

0.36, 0.7811

0.57, 0.6356

30.4
42.3
22.1
5.2

29.9
41.9
22.7
5.5

33.0
38.8
23.1
5.1

0.67, 0.5718

0.67, 0.5706

75.9
63.8

78.6
66.5

75.6
69.9

0.02, 0.8850
6.35, 0.0133c

2.00, 0.1604
1.92, 0.1683

33.8
33.0
13.2
20.0

34.9
32.3
13.3
19.5

40.3
28.4
13.6
17.7

1.34, 0.2657

0.81, 0.4912

25.1

24.6

30.3

4.31, 0.0403c

5.21, 0.0245c

9.4
91.6
6.2
20.1
4.3

9.2
91.0
5.7
20.2
3.8

9.2
68.8
9.8
21.9
5.1

0.02, 0.8891
171.05, 0.0000c
6.55, 0.0119c
0.49, 0.4834
0.66, 0.4185

0.00, 0.9815
97.07, 0.0000c
8.35, 0.0047c
0.46, 0.5004
2.08, 0.1518

20.7
12.6

22.7
13.0

29.9
16.6

16.30, 0.0001c
5.14, 0.0255c

9.25, 0.0030c
4.27, 0.0412c

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Respondents with unknown data were excluded.
5
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI),
cash assistance (Temporary Assistance for Needy Families, TANF), noncash assistance, or food stamps.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-8

Table K-9 Demographic, Socioeconomic, and Household Characteristics among Persons Aged 26 or
Older: Percentages, Chi-Square Test Statistic, and P Value, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

2012
2011
Comparison Comparison 2012 QFT
(n = 21,847)1 (n = 10,412)1,2 (n = 999)1,3

2011
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

2012
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

10.9
29.7
26.0
33.4

11.4
29.1
26.2
33.3

12.1
25.1
31.1
31.7

4.99, 0.0028c

3.87, 0.0113c

21.9
37.0
27.1
14.1

20.6
36.8
27.7
14.9

19.0
40.5
27.4
13.1

1.71, 0.1687

1.35, 0.2609

87.2
63.8

87.8
65.0

86.8
65.6

0.14, 0.7125
0.64, 0.4241

0.76, 0.3858
0.08, 0.7800

15.6
30.7
18.3
35.4

15.7
31.8
17.5
35.1

15.3
34.3
17.3
33.1

1.21, 0.3111

0.45, 0.7197

17.3

19.0

22.7

10.39, 0.0017c

4.36, 0.0391c

32.3
79.8
7.0
12.7
2.0

31.2
80.4
8.0
14.0
1.8

31.3
69.0
9.3
15.5
3.1

0.23, 0.6293
32.13, 0.0000c
4.71, 0.0322c
3.80, 0.0538
4.36, 0.0393c

0.00, 0.9778
33.14, 0.0000c
1.39, 0.2404
1.00, 0.3191
5.90, 0.0168c

7.3
4.8

8.2
4.9

10.2
6.7

7.02, 0.0093c
5.74, 0.0183c

2.79, 0.0976
4.82, 0.0304c

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Respondents with unknown data were excluded.
5
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI),
cash assistance (Temporary Assistance for Needy Families, TANF), noncash assistance, or food stamps.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-9

Table K-10 Demographic and Geographic Characteristics among Persons Aged 12 or Older: Percentages, Chi-Square Test Statistic, and
P Value, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011 Comparison1

K-10

Characteristic
Education4
< High School
High School Graduate
Some College
College Graduate
Employment4
Full-Time
Part-Time
Unemployed
Other5
Region
Northeast
Midwest
South
West
County Type
Large Metro
Small Metro
Nonmetro

2012 Comparison1,2

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

5,922
14,119
13,434
10,034

13.6
32.5
30.9
23.1

11.6
30.3
27.4
30.6

2,483
6,859
6,466
4,940

12.0
33.1
31.2
23.8

11.5
30.1
27.7
30.7

20,420
8,615
3,899
10,575

46.9
19.8
9.0
24.3

49.7
14.1
5.8
30.4

10,345
3,934
1,701
4,768

49.9
19.0
8.2
23.0

12,701
19,008
22,158
12,061

19.3
28.8
33.6
18.3

18.6
22.6
37.4
21.4

6,480
9,099
9,724
5,910

28,475
23,627
13,826

43.2
35.8
21.0

52.6
31.3
16.1

13,865
10,789
6,559

2012 QFT1,3

Unwtd
n

QFT vs.
2011 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2011 ChiSquare
Statistic,
P Value
Unwtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Unwtd

Unwtd
Percent

Wtd
Percent

187
426
531
359

12.4
28.3
35.3
23.9

12.4
26.6
32.1
29.0

51.3
13.9
5.5
29.3

798
245
111
349

53.1
16.3
7.4
23.2

52.0
14.2
5.5
28.3

0.64, 0.5933 0.10, 0.9589 6.60, 0.0004c 2.80, 0.0437c

20.8
29.2
31.2
18.9

18.6
22.6
37.4
21.4

375
458
824
387

18.3
22.4
40.3
18.9

18.7
23.0
38.0
20.2

0.19, 0.9008 0.15, 0.9308 5.89, 0.0009c 11.07, 0.0000c

44.4
34.6
21.0

52.6
31.1
16.3

1,045
612
387

51.1
29.9
18.9

51.8
28.4
19.8

0.86, 0.4244 0.71, 0.4931 3.02, 0.0529 2.15, 0.1218

5.38, 0.0018c 4.45, 0.0055c 5.54, 0.0014c 6.27, 0.0006c

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test; unwtd = unweighted; wtd = weighted.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Education and employment estimates are based only on respondents aged 18 or older. Sample sizes for respondents 18 or older are n = 43,509 for 2011 comparison, n = 1,503 for
QFT, and n = 20,748 for 2012 comparison.
5
The Other Employment category includes student, persons keeping house or caring for children full time, retired or disabled person, or other persons not in the labor force.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table K-11 Geographic Characteristics among Persons Aged 12 to 17: Percentages, Chi-Square Test Statistic, and P Value, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011 Comparison1

Characteristic
Region
Northeast
Midwest
South
West
County Type
Large Metro
Small Metro
Nonmetro

2012 Comparison1,2

2012 QFT1,3

QFT vs.
2011 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2011 ChiSquare
Statistic,
P Value
Unwtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Unwtd

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

4,321
6,337
7,708
4,053

19.3
28.3
34.4
18.1

17.4
22.4
37.5
22.7

2,077
3,099
3,238
2,051

19.8
29.6
30.9
19.6

16.9
22.6
38.2
22.3

78
117
245
101

14.4
21.6
45.3
18.7

13.2
22.1
44.6
20.1

2.61, 0.0553 1.79, 0.1535 6.12, 0.0007c 9.02, 0.0000c

9,744
7,926
4,749

43.5
35.4
21.2

53.3
31.2
15.5

4,695
3,568
2,202

44.9
34.1
21.0

54.5
30.4
15.1

272
171
98

50.3
31.6
18.1

51.6
31.8
16.5

0.10, 0.9084 0.24, 0.7853 1.51, 0.2260 0.94, 0.3925

K-11

Table K-12 Demographic and Geographic Characteristics among Persons Aged 18 to 25: Percentages, Chi-Square Test Statistic, and
P Value, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011 Comparison1

K-12

Characteristic
Education
< High School
High School Graduate
Some College
College Graduate
Employment
Full-Time
Part-Time
Unemployed
Other4
Region
Northeast
Midwest
South
West
County Type
Large Metro
Small Metro
Nonmetro

2012 Comparison1,2

2012 QFT1,3

QFT vs.
2011 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2011 ChiSquare
Statistic,
P Value
Unwtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Unwtd

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

3,509
7,609
7,531
3,013

16.2
35.1
34.8
13.9

15.6
34.0
35.7
14.7

1,316
3,816
3,666
1,538

12.7
36.9
35.5
14.9

12.0
35.7
36.4
15.9

68
183
196
57

13.5
36.3
38.9
11.3

13.8
34.9
37.6
13.7

8,064
5,908
2,800
4,890

37.2
27.3
12.9
22.6

36.0
27.8
13.2
23.0

4,312
2,685
1,212
2,127

41.7
26.0
11.7
20.6

40.1
26.4
11.8
21.7

219
121
63
101

43.5
24.0
12.5
20.0

45.5
24.4
11.9
18.2

3.90, 0.0110c 1.35, 0.2637 1.95, 0.1255 0.30, 0.8266

4,148
6,236
7,253
4,025

19.1
28.8
33.5
18.6

18.2
22.0
37.1
22.7

2,203
2,909
3,340
1,884

21.3
28.1
32.3
18.2

18.8
20.7
38.7
21.8

100
118
193
93

19.8
23.4
38.3
18.5

20.8
22.7
37.5
19.0

0.41, 0.7453 0.34, 0.7955 1.39, 0.2512 1.83, 0.1459

9,409
7,989
4,264

43.4
36.9
19.7

53.5
32.4
14.0

4,640
3,672
2,024

44.9
35.5
19.6

54.8
31.5
13.7

259
150
95

51.4
29.8
18.8

54.2
28.3
17.5

0.84, 0.4362 0.82, 0.4421 2.05, 0.1335 1.37, 0.2583

0.36, 0.7811 0.57, 0.6356 1.79, 0.1544 1.64, 0.1843

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test; unwtd = unweighted; wtd = weighted.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The Other Employment category includes student, persons keeping house or caring for children full time, retired or disabled person, or other persons not in the labor force.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table K-13 Demographic and Geographic Characteristics among Persons Aged 26 or Older: Percentages, Chi-Square Test Statistic, and
P Value, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011 Comparison1

K-13

2012 Comparison1,2

2012 QFT1,3

QFT vs.
2011 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Wtd

QFT vs.
2011 ChiSquare
Statistic,
P Value
Unwtd

QFT vs.
2012 ChiSquare
Statistic,
P Value
Unwtd

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

Unwtd
Percent

Wtd
Percent

2,413
6,510
5,903
7,021

11.0
29.8
27.0
32.1

10.9
29.7
26.0
33.4

1,167
3,043
2,800
3,402

11.2
29.2
26.9
32.7

11.4
29.1
26.2
33.3

119
243
335
302

11.9
24.3
33.5
30.2

12.1
25.1
31.1
31.7

12,356
2,707
1,099
5,685

56.6
12.4
5.0
26.0

52.1
11.7
4.5
31.7

6,033
1,249
489
2,641

57.9
12.0
4.7
25.4

53.3
11.7
4.4
30.7

579
124
48
248

58.0
12.4
4.8
24.8

53.2
12.4
4.3
30.1

0.24, 0.8691 0.09, 0.9664 0.25, 0.8628 0.07, 0.9754

4,232
6,435
7,197
3,983

19.4
29.5
32.9
18.2

18.8
22.7
37.5
21.0

2,200
3,091
3,146
1,975

21.1
29.7
30.2
19.0

18.8
22.9
37.1
21.2

197
223
386
193

19.7
22.3
38.6
19.3

19.1
23.2
37.3
20.4

0.04, 0.9908 0.05, 0.9859 4.38, 0.0060c 7.07, 0.0002c

9,322
7,712
4,813

42.7
35.3
22.0

52.3
31.2
16.5

4,530
3,549
2,333

43.5
34.1
22.4

51.9
31.1
17.0

514
291
194

51.5
29.1
19.4

51.5
28.0
20.6

0.87, 0.4218 0.68, 0.5080 2.97, 0.0556 2.48, 0.0883

4.99, 0.0028c 3.87, 0.0113c 8.57, 0.0000c 9.06, 0.0000c

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test; unwtd = unweighted; wtd = weighted.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
The Other Employment category includes student, persons keeping house or caring for children full time, retired or disabled person, or other persons not in the labor force.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

Table K-14 Perceived Great Risk of Harm Associated with Substance Use among Persons Aged 12
or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Perception of Great Risk1
PERCEPTIONS OF GREAT RISK CIGARETTES
Smoke one or more packs per day
PERCEPTIONS OF GREAT RISK MARIJUANA
Smoke once a month
Smoke once or twice a week

2011
2012
Comparison Comparison
vs. QFT,
vs. QFT,
2011
2012
Comparison Comparison 2012 QFT Difference Difference
(SE)
(SE)
(n = 65,928)2 (n = 31,213)2,3 (n = 2,044)2,4

70.7

70.4

69.2

1.5 (1.48)

1.2 (1.49)

30.3
40.7

28.6
38.5

30.2
38.8

0.0 (1.56)
2.0 (1.63)

-1.6 (1.59)
-0.2 (1.70)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Response categories for the Perceptions of Risk questions include "No risk," "Slight risk," "Moderate risk," and "Great risk."
The estimates in this table correspond to persons reporting "Great risk." Respondents with unknown Perceptions of Risk data
were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-14

Table K-15 Number of Years Since Last Use for Selected Substances among Lifetime Users Aged
12 to 49: Averages, Differences, and Standard Error of Differences, 2011 Comparison,
2012 Comparison, and 2012 Questionnaire Field Test

Substance
Cigarettes
Alcohol
Marijuana
Cocaine
Hallucinogens
Inhalants

2011
2012
2012 QFT
Comparison Comparison
(n = 58,401)1 (n = 27,652)1,2 (n = 1,725)1,3
10.4
10.2
10.6
2.7
2.3
3.0
9.9
9.7
9.3
10.8
10.2
9.7
11.3a
10.9
9.6
13.4
13.5
13.3

2011
Comparison
vs. QFT,
Difference
(SE)
-0.2 (0.59)
-0.3 (0.36)
0.6 (0.58)
1.1 (0.75)
1.7 (0.72)
0.0 (0.91)

2012
Comparison
vs. QFT,
Difference
(SE)
-0.4 (0.60)
-0.7 (0.37)
0.4 (0.61)
0.5 (0.77)
1.2 (0.74)
0.2 (0.96)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
NOTE: If respondents reported last using a substance more than 30 days ago but within the past 12 months, the number of years
since last use was assumed to be zero, regardless of whether they reported last use more than a year ago based on the
age, year, or month when they last used. In addition, the number of years since last use was set to zero for past month
substance users, but they were not asked the questions pertaining to prior substance use.
NOTE: Within each set of data, sample sizes will vary by substance because nonusers of the substance were excluded from the
analysis.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and
2012.

K-15

Table K-16 Received Substance Use Treatment in Lifetime and Past Year and Types of Past Year
Substance Use Treatment among Persons Aged 12 or Older: Percentages, Differences,
and Standard Error of Differences, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Substance Use Treatment
LIFETIME TREATMENT
PAST YEAR TREATMENT
Alcohol use only
Drug use only
Both alcohol and drug use

2012
2011
2012
Comparison
Comparison vs. Comparison vs.
2011
(n =
Comparison
2012 QFT QFT, Difference QFT, Difference
(SE)
(SE)
(n = 65,928)1 31,213)1,2 (n = 2,044)1,3
5.9
6.2
6.6
-0.7 (0.78)
-0.4 (0.84)
1.4
1.4
1.5
-0.1 (0.32)
-0.0 (0.32)
0.6
0.6
0.5
0.1 (0.15)
0.1 (0.15)
0.4
0.5
0.4
-0.0 (0.15)
0.1 (0.15)
0.4
0.4
0.6
-0.2 (0.20)
-0.2 (0.21)

K-16

Table K-17 Adult Mental Health Treatment in the Past Year and Type of Facility Where Received
Treatment among Persons Aged 18 or Older: Percentages, Differences, and Standard
Error of Differences, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test

Past Year Mental Health
Treatment1
STAYED OVERNIGHT IN
HOSPITAL FOR MENTAL
HEALTH TREATMENT
FACILITY TYPE – OVERNIGHT
MENTAL HEALTH
TREATMENT5
Private or Public Psychiatric
Hospital
Psychiatric Unit – General
Hospital
Medical unit – General Hospital
Another Type of Hospital
Residential Treatment Center
Other Facility

2011
2012
Comparison Comparison
2012
vs. QFT,
vs. QFT,
Comparison
2011
Difference Difference
(n =
Comparison
2012 QFT
(SE)
(SE)
(n = 43,509)2 20,748)2,3 (n = 1,503)2,4

0.8

0.7

0.9

-0.1 (0.23)

-0.2 (0.23)

0.2

0.1

0.1 (0.10)

0.0 (0.11)

0.2
0.2
0.1a
0.1
0.1

0.2
0.2
0.1a
0.1
0.0

0.3
0.3
0.0*
0.1
0.1

-0.0 (0.12)
-0.1 (0.08)
0.1 (0.02)
-0.0 (0.08)
-0.1 (0.09)

-0.1 (0.12)
-0.1 (0.09)
0.1 (0.03)
-0.0 (0.09)
-0.1 (0.09)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Respondents with unknown mental health treatment information were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
QFT data collected from September 1 through November 3, 2012.
5
Respondents could indicate multiple locations for treatment; thus, these response categories are not mutually exclusive.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and
2012.

K-17

Table K-18 Youth Mental Health Treatment in the Past Year and Number of Nights Received
Treatment among Persons Aged 12 to 17: Percentages, Chi-Square Test Statistic, and
P Value, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Past Year Mental Health Treatment1
STAYED OVERNIGHT IN HOSPITAL
FOR MENTAL HEALTH
TREATMENT
Yes
No
NUMBER OF NIGHTS IN HOSPITAL
FOR MENTAL HEALTH
TREATMENT
1 Night

2012
Comparison
2011
(n =
Comparison
2012 QFT
(n = 22,419)2 10,465)2,3 (n = 541)2,4

1.8

2.0

2.3

98.2

98.0

97.7

2011
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

2012
Comparison
vs. QFT
Chi-Square
Statistic,
P Value

0.41, 0.5220

0.09, 0.7617

0.03, 0.9701

48.9

46.9

49.3

2 to 6 Nights

23.8

33.1

34.6*

0.31, 0.7322

7 or More Nights

27.3

20.0

16.1*

1.0

0.9

2.0

99.0

99.1

98.0

STAYED OVERNIGHT IN
RESIDENTIAL TREATMENT
CENTER FOR MENTAL HEALTH
TREATMENT
Yes
No
NUMBER OF NIGHTS IN
RESIDENTIAL TREATMENT
CENTER FOR MENTAL HEALTH
TREATMENT
1 Night

24.4

35.1

26.0

2 to 6 Nights

26.2

30.5

45.7*

7 or More Nights

38.8

43.4

29.9*

3.29, 0.0725

4.72, 0.0320c

0.60, 0.5481

0.33, 0.7180

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
c
Interaction between the characteristic and survey is significant at the 0.05 level.
1
Respondents with unknown mental health treatment information were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-18

Table K-19 Selected Mental Health Measures among Persons Aged 18 or Older: Percentages,
Differences, and Standard Error of Differences, 2011 Comparison, 2012 Comparison,
and 2012 Questionnaire Field Test

Mental Health Measure
Past Month SPD4
Past Year SPD4
Past Year Thoughts of Suicide5
Past Year Suicide Plans5
Past Year Attempted Suicide5
Several Days or Longer Felt Sad, Empty, or
Depressed6
Several Days When Most of the Day Felt Very
Discouraged6
Several Days or Longer Lost Interest in Things
Usually Enjoyable6
Average Past Month Total K6 Score7
Average Past Year Worst K6 Total Score7
Average WHODAS Score (0 to 24)

2011
Comparison
(n = 43,509)1
4.7a
10.4a
3.8
1.1
0.5

2012
2012 QFT
Comparison
(n = 20,748)1,2 (n = 1,503)1,3
5.3a
3.6
10.7a
8.5
3.9
3.0
1.0
1.2
0.5
0.6

2011
2012
Comparison Comparison
vs. QFT,
vs. QFT,
Difference Difference
(SE)
(SE)
1.1 (0.51)
1.9 (0.69)
0.8 (0.45)
-0.1 (0.31)
-0.1 (0.20)

1.6 (0.57)
2.1 (0.82)
0.9 (0.47)
-0.1 (0.31)
-0.1 (0.20)

31.2

31.1

28.7

2.6 (1.41)

2.4 (1.57)

12.5

12.0

11.3

1.2 (1.22)

0.7 (1.30)

4.2

4.3

4.7

-0.5 (1.07)

-0.5 (1.14)

3.5
4.6
3.3

0.2 (0.13)
0.3 (0.16)
0.3 (0.15)

0.3 (0.14)
0.3 (0.18)
0.4 (0.16)

3.8
4.9
3.5

3.9
5.0
3.7a

* Low precision; estimate would be suppressed under NSDUH suppression rules.
K6 = Kessler 6; QFT = Questionnaire Field Test; SPD = serious psychological distress; WHODAS = World Health Organization
Disability Assessment Schedule.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
SPD is defined as having a score of 13 or higher on the K6 scale.
5
Respondents with unknown suicide information were excluded.
6
Respondents with unknown depression information were excluded.
7
The K6 score is derived from 12 questions asking the frequency that a respondent experienced symptoms of psychological
distress. Six new questions were asked for the first time in 2008 to all respondents aged 18 or older about their past 30-day
symptoms. Responses to these six questions are combined to produce the past month score ranging from 0 to 24. The original
six questions are then only asked respondents who reported that there was a month in the past year when they felt more
symptoms than they felt in the past 30 days, and a score ranging from 0 to 24 is produced. The maximum of these two scores is
taken to create the past year K6 score.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-19

Table K-20 Adolescent Depression Characteristics among Persons Aged 12 to 17: Percentages,
Differences, and Standard Error of Differences, 2011 Comparison, 2012 Comparison,
and 2012 Questionnaire Field Test

Depression Characteristic1
Several Days or Longer Felt Sad, Empty or
Depressed
Several Days When Most of the Day Felt
Very Discouraged
Several Days or Longer Lost Interest in
Things Usually Enjoyable

2011
2012
2012
Comparison Comparison
Comparison
2011
vs. QFT,
vs. QFT,
(n =
Comparison
2012 QFT Difference Difference
(n = 22,419)2 10,465)2,3
(n = 541)2,4
(SE)
(SE)
43.2

43.0

43.4

-0.2 (2.38)

-0.4 (2.39)

8.4

8.0

7.7

0.7 (1.88)

0.2 (1.98)

14.6

15.0

14.3

0.3 (2.22)

0.7 (2.31)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Respondents with unknown depression information were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-20

Table K-21 Arrested and Booked in Lifetime and Past Year for Breaking the Law among Persons
Aged 12 or Older: Percentages, Differences, and Standard Error of Differences, 2011
Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

Arrested and Booked1
TIME PERIOD
Lifetime
Past Year

2011
2012
Comparison Comparison
2012 QFT
(n = 65,928)2 (n = 31,213)2,3 (n = 2,044)2,4
16.6
3.1

17.3
3.1

16.9
3.2

2011
Comparison
vs. QFT,
Difference
(SE)

2012
Comparison
vs. QFT,
Difference
(SE)

-0.3 (1.16)
-0.0 (0.43)

0.4 (1.22)
-0.1 (0.47)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
a
Difference between estimate and 2012 QFT estimate is statistically significant at the 0.05 level.
1
Respondents with unknown arrested and booked information were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2011 and 2012.

K-21

K-22

Appendix L: Detailed Tables for Prescription Drug Use and
Misuse in the 2012 Questionnaire Field Test and Data from
Sources Other than NSDUH

Table L-1 Comparison of Data for Pain Relievers from the 2012 NSDUH Questionnaire Field Test and the 2010 National Ambulatory
Medical Survey and 2010 National Hospital Ambulatory Medical Survey

L-1

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Any Prescription Pain Reliever5/Any Narcotic
Analgesic6
Vicodin®, Lortab®, Lorcet®, or Hydrocodone7
Vicodin®
Lortab®
Lorcet®
Hydrocodone7
OxyContin®, Percocet®, Percodan®, Tylox®,
or Oxycodone8,9
OxyContin®9
Percocet®
Percodan®
Tylox®
Oxycodone8
Darvocet®, Darvon®, or Propoxyphene7
Darvocet®
Darvon®
Propoxyphene7
Ultram®, Ultram® ER, Ultracet®, Ryzolt®, or
Tramadol7
Ultram®
Ultram® ER
Ultracet®
Ryzolt®
Tramadol7

NSDUH QFT,1
Percent (SE)
Any Past Year Use2

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3

NSDUH QFT,1
Percent (SE)
Past Year Misuse2

NAMCS, Number of
Mentions
in Thousands (SE)4
(6,493)
(3,520)
(1,650)
(1,996)
(941)
(1,393)

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4

38.9
25.4
12.9
5.5
1.1
14.4

(1.61)
(1.48)
(1.18)
(0.70)
(0.25)
(1.17)

32.9
21.5
10.5
4.5
0.8
12.4

(1.35)
(1.27)
(1.02)
(0.62)
(0.22)
(1.06)

6.0
3.8
2.4
1.0
0.3
1.9

(0.75)
(0.53)
(0.44)
(0.26)
(0.11)
(0.35)

77,194
35,868
15,684
9,671
1,529*
8,984

8,744 (1,161)
2,890
(378)
1,475
(259)
690
(160)
28*
(14)
697
(139)

12.6
2.4
6.5
0.4
0.3
6.8
2.1
1.6
0.5
0.2

(1.10)
(0.35)
(0.83)
(0.15)
(0.13)
(0.92)
(0.44)
(0.41)
(0.29)
(0.11)

10.5
1.6
5.4
0.2
0.3
5.6
2.0
1.5
0.5
0.2

(0.99)
(0.29)
(0.75)
(0.12)
(0.12)
(0.87)
(0.43)
(0.39)
(0.29)
(0.11)

2.1
0.8
1.0
0.2
0.0
1.2
0.1
0.1
0.0*
0.0*

(0.34)
(0.20)
(0.23)
(0.08)
(0.03)
(0.27)
(0.07)
(0.07)
(0.00)
(0.00)

13,517 (1,543)
1,708
(345)
7,125 (965)
51*
(51)
151* (101)
4,481
(630)
7,944 (1,158)
6,932
(996)
316* (203)
696* (219)

1,957
146
1,206
1*
18*
586
600
537
23*
40*

6.4
2.1
0.4
0.3
0.0
4.5

(0.78)
(0.55)
(0.23)
(0.15)
(0.02)
(0.56)

5.3
1.7
0.4
0.2
0.0
3.9

(0.68)
(0.42)
(0.23)
(0.12)
(0.02)
(0.54)

1.0
0.5
0.0*
0.1
0.0*
0.5

(0.26)
(0.18)
(0.00)
(0.10)
(0.00)
(0.16)

11,690 (1,563)
4,175
(877)
173* (103)
427* (181)
39*
(33)
6,876 (1,057)

1,548 (198)
456
(97)
0*
(0)
33* (21)
0*
(0)
1,059 (142)
(continued)

(284)
(37)
(196)
(1)
(18)
(105)
(142)
(132)
(13)
(22)

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates are
for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
7
For NAMCS/NHAMCS: generic or generic with acetaminophen.
8
For NAMCS/NHAMCS: generic, generic with acetaminophen, or generic with aspirin.
9
For NSDUH: The past year OxyContin® misuse estimate in these tables may differ from the estimate in the "Detailed Tables for Methamphetamine and Prescription Drug
Estimates" due to the availability of edited and imputed data.
(Source information is included on the last page of the table.)

Table L-1 Comparison of Data for Pain Relievers from the 2012 NSDUH Questionnaire Field Test and the 2010 National Ambulatory
Medical Survey and 2010 National Hospital Ambulatory Medical Survey (continued)

L-2

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Tylenol® with Codeine 3 or 4, or Codeine
Pills7
Tylenol® with Codeine 3 or 4
Codeine Pills7
Avinza®, Kadian®, MS Contin®, Oramorph®
SR, or Morphine
Avinza®
Kadian®
MS Contin®
Oramorph® SR
Morphine
Actiq®, Duragesic®, Fentora®, or Fentanyl
Actiq®
Duragesic®
Fentora®
Fentanyl
Suboxone®, Subutex®, or Buprenorphine
Suboxone®
Subutex®
Buprenorphine
Demerol®
Dilaudid®
Methadone
Opana® or Opana® ER
Opana®
Opana® ER

NSDUH QFT,1
Percent (SE)
Any Past Year
Use2

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3

NSDUH QFT,1
Percent (SE)
Past Year Misuse2

11.5 (0.99)
10.9 (0.98)
1.6 (0.30)

9.8 (0.93)
9.3 (0.93)
1.3 (0.28)

1.7 (0.29)
1.5 (0.27)
0.3 (0.11)

3,185 (476)
2,395 (391)
790* (262)

3.6
0.1
0.0
0.1
0.0*
3.3
0.8
0.1
0.1
0.0
0.6
0.6
0.5
0.2
0.0*
0.6
0.6
0.3
0.1
0.0
0.1

0.4
0.0*
0.0
0.0*
0.0*
0.4
0.1
0.0*
0.0*
0.0*
0.1
0.4
0.2
0.1
0.0
0.0
0.3
0.3
0.2
0.1
0.1

1,408
35*
124*
463*
26*
760
1,848
0*
572*
13*
1,263
1,535*
1,287*
8*
239*
310*
858
1,518
39*
19*
21*

4.0
0.1
0.1
0.1
0.0*
3.7
0.9
0.1
0.1
0.0
0.7
1.0
0.7
0.3
0.0
0.7
0.9
0.6
0.3
0.1
0.2

(0.59)
(0.11)
(0.05)
(0.06)
(0.00)
(0.54)
(0.27)
(0.11)
(0.05)
(0.04)
(0.23)
(0.25)
(0.23)
(0.11)
(0.04)
(0.15)
(0.23)
(0.17)
(0.09)
(0.06)
(0.08)

(0.57)
(0.11)
(0.04)
(0.06)
(0.00)
(0.52)
(0.27)
(0.11)
(0.05)
(0.04)
(0.24)
(0.22)
(0.21)
(0.08)
(0.00)
(0.15)
(0.21)
(0.13)
(0.05)
(0.04)
(0.06)

(0.15)
(0.00)
(0.03)
(0.00)
(0.00)
(0.14)
(0.05)
(0.00)
(0.00)
(0.00)
(0.05)
(0.13)
(0.10)
(0.07)
(0.04)
(0.04)
(0.08)
(0.11)
(0.07)
(0.05)
(0.05)

NAMCS, Number of
Mentions
in Thousands (SE)4

(272)
(26)
(82)
(156)
(26)
(155)
(325)
(0)
(174)
(13)
(280)
(650)
(471)
(8)
(211)
(154)
(218)
(341)
(25)
(14)
(21)

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
444 (86)
324 (67)
120* (37)
405
0*
55*
121*
0*
229
1,026*
4*
65*
0*
957*
88*
87*
1*
0*
343*
106*
146
5*
5*
0*

(120)
(0)
(42)
(50)
(0)
(65)
(372)
(4)
(30)
(0)
(369)
(32)
(32)
(1)
(0)
(251)
(36)
(38)
(4)
(4)
(0)
(continued)

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Talacen®, Talwin®, or Talwin® NX
Talacen®
Talwin®
Talwin® NX
Any Other Prescription Pain Reliever

NSDUH QFT,1
Percent (SE)
Any Past Year
Use2
0.1 (0.04)
0.0* (0.00)
0.0 (0.03)
0.0 (0.03)
8.7 (0.81)

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3
0.0 (0.03)
0.0* (0.00)
0.0* (0.00)
0.0 (0.03)
8.5 (0.80)

NSDUH QFT,1
Percent (SE)
Past Year Misuse2
0.0 (0.02)
0.0* (0.00)
0.0 (0.02)
0.0* (0.00)
0.2 (0.09)

NAMCS, Number of
Mentions
in Thousands (SE)4
117* (93)
91* (91)
27* (27)
0* (0)
N/A

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
0* (0)
0* (0)
0* (0)
0* (0)
N/A

L-3

*Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable (NSDUH) or not available (NAMCS/NHAMCS).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates are
for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Ambulatory Medical Care Survey (NAMCS), 2010, National Hospital Ambulatory Medical Care Survey
(NHAMCS), 2010.

L-4

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Any Tranquilizer5
Any Sedative5
Any Tranquilizer or Any Sedative6/Any
Anxiolytic, Sedative, Hypnotic, or Muscle
Relaxant7
Any Benzodiazepine
Xanax®, Xanax® XR, Alprazolam, or
Extended-Release Alprazolam7
Xanax®
Xanax® XR
Alprazolam
Extended-Release Alprazolam
Ativan® or Lorazepam8
Ativan®
Lorazepam
Klonopin® or Clonazepam8
Klonopin®
Clonazepam
Valium® or Diazepam8
Valium®
Diazepam
Librium®8
Tranxene®8
Oxazepam (also known as Serax®)8

NSDUH QFT,1
Percent (SE)
Any Past Year Use2
15.2 (1.23)
7.3 (0.78)

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3
12.9 (1.10)
6.5 (0.70)

NSDUH QFT,1
Percent (SE)
Past Year Misuse2
2.4 (0.38)
0.8 (0.22)

19.3 (1.32)
11.5 (1.12)

16.9 (1.16)
9.5 (1.00)

2.8 (0.41)
2.1 (0.37)

114,180 (8,913)
54,334 (4,534)

1.5
1.4
0.2
0.3
0.0
0.5
0.2
0.4
0.5
0.5
0.2
0.6
0.5
0.1
0.0
0.0*
0.0*

18,498 (1,808)
12,532 (1,300)
80* (61)
5,887 (935)
N/A
13,022 (1,447)
5,699
(884)
7,323 (1,050)
11,814 (1,578)
6,819 (1,228)
4,994
(658)
6,096
(841)
3,638
(520)
2,458
(555)
430* (212)
201*
(99)
164*
(61)

6.3
4.7
0.4
1.5
0.4
2.7
1.2
2.0
2.7
1.1
2.0
2.6
1.9
1.0
0.1
0.0
0.1

(0.81)
(0.67)
(0.15)
(0.34)
(0.24)
(0.41)
(0.31)
(0.32)
(0.47)
(0.26)
(0.40)
(0.50)
(0.41)
(0.27)
(0.07)
(0.03)
(0.05)

4.7
3.4
0.2
1.2
0.4
2.2
1.0
1.5
2.2
0.7
1.9
2.0
1.3
0.8
0.1
0.0
0.1

(0.70)
(0.58)
(0.10)
(0.32)
(0.24)
(0.36)
(0.30)
(0.28)
(0.41)
(0.19)
(0.39)
(0.44)
(0.36)
(0.25)
(0.06)
(0.03)
(0.05)

(0.28)
(0.27)
(0.11)
(0.11)
(0.02)
(0.15)
(0.07)
(0.14)
(0.18)
(0.16)
(0.07)
(0.17)
(0.16)
(0.07)
(0.02)
(0.00)
(0.00)

NAMCS, Number of
Mentions
in Thousands (SE)4
N/A
N/A

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
N/A
N/A
13,078 (1,745)
6,906 (1,139)
1,711 (289)
1,159 (223)
4*
(4)
548 (108)
N/A
1,716 (368)
881 (191)
835 (209)
1,455 (241)
720 (139)
735 (135)
461 (100)
239
(54)
222
(58)
18* (12)
5*
(5)
17* (17)
(continued)

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable (NSDUH) or not available (NAMCS/NHAMCS).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
Created from NSDUH QFT summary measures for any tranquilizer and any sedative use or misuse.
7
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
8
Benzodiazepine that is included in the NSDUH tranquilizers module.
(Source information is included on the last page of the table.)

Table L-2 Comparison of Data for Tranquilizers and Sedatives from the 2012 NSDUH Questionnaire Field Test and the 2010 National
Ambulatory Medical Survey and 2010 National Hospital Ambulatory Medical Survey (continued)
NSDUH QFT,1
Percent (SE)
Any Past Year Use2
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.4 (0.21)
0.2 (0.18)
0.2 (0.11)
0.7 (0.26)
0.1 (0.07)
0.6 (0.25)
5.4 (0.69)
4.2 (0.59)
1.4 (0.33)
0.4 (0.20)

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.4 (0.21)
0.2 (0.18)
0.2 (0.11)
0.6 (0.25)
0.0* (0.00)
0.6 (0.25)
4.7 (0.65)
3.8 (0.54)
1.0 (0.30)
0.4 (0.20)

NSDUH QFT,1
Percent (SE)
Past Year Misuse2
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.1 (0.07)
0.1 (0.07)
0.0* (0.00)
0.6 (0.16)
0.4 (0.13)
0.4 (0.11)
0.0 (0.02)

NAMCS, Number of
Mentions
in Thousands (SE)4
12* (12)
0*
(0)
12* (12)
97* (60)
44* (29)
53* (53)
2,333 (368)
1,298 (273)
1,035 (214)
11,442 (1,373)
8,438 (1,087)
3,004 (688)
2,330 (365)

NHAMCS Hospital
Outpatient, Number
of Mentions
in Thousands (SE)4
32* (26)
6* (6)
26* (25)
9* (5)
3* (1)
6* (5)
313* (97)
124* (48)
189* (58)
1,318 (188)
1,103 (164)
215* (68)
312 (64)

L-5

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Dalmane® or Flurazepam9
Dalmane
Flurazepam
Halcion® or Triazolam9
Halcion®
Triazolam
Restoril® or Temazepam9
Restoril®
Temazepam
Flexeril® or Soma®
Flexeril®
Soma®
Buspirone (also known as BuSpar®)
Hydroxyzine (also known as Atarax® or
Vistaril®)
0.6 (0.24)
0.6 (0.24)
0.0 (0.03)
3,649 (700)
676 (123)
Meprobamate (also known as Equanil® or
Miltown®)
0.0 (0.02)
0.0* (0.00)
0.0 (0.02)
114* (61)
0* (0)
Ambien®, Ambien® CR, Zolpidem, or
Extended-Release Zolpidem
5.8 (0.77)
5.1 (0.68)
0.7 (0.21)
17,051 (1,757)
1,312 (192)
Ambien®
4.5 (0.63)
4.1 (0.57)
0.4 (0.15)
11,870 (1,377)
1,090 (167)
Ambien® CR
0.7 (0.22)
0.6 (0.22)
0.0 (0.02)
462* (154)
72* (29)
Zolpidem
1.6 (0.46)
1.2 (0.40)
0.4 (0.18)
4,719 (738)
150 (40)
Extended-Release Zolpidem
0.1 (0.07)
0.1 (0.07)
0.0* (0.00)
N/A
N/A
®
Lunesta
1.1 (0.30)
0.9 (0.29)
0.1 (0.09)
2,365 (519)
119* (47)
Sonata® or Zaleplon
0.5 (0.24)
0.4 (0.24)
0.1 (0.06)
125* (53)
42* (20)
®
*
Sonata
0.5 (0.24)
0.4 (0.24)
0.1 (0.06)
125 (53)
22* (10)
Zaleplon
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0* (0)
21* (16)
(continued)
* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH
estimates are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or
older.
9
Benzodiazepine that is included in the NSDUH sedatives module.
(Source information is included on the last page of the table.)

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Butisol®, Seconal®, or Phenobarbital/
Barbiturates10
Butisol®
Seconal®
Phenobarbital
Any Other Prescription Tranquilizer
Any Other Prescription Sedative

NSDUH QFT,1
Percent (SE)
Any Past Year Use2
0.3
0.0
0.1
0.2
1.7
1.2

(0.17)
(0.03)
(0.07)
(0.15)
(0.35)
(0.27)

NSDUH QFT,1
Percent (SE)
Past Year Use But
Note Misuse3
0.2
0.0*
0.1
0.2
1.7
1.2

(0.16)
(0.00)
(0.07)
(0.15)
(0.35)
(0.27)

NSDUH QFT,1
Percent (SE)
Past Year Misuse2
0.0
0.0
0.0*
0.0
0.0*
0.0

(0.03)
(0.03)
(0.00)
(0.02)
(0.00)
(0.02)

NAMCS, Number of
Mentions
in Thousands (SE)4
673 (177)
0* (0)
N/A
527 (154)
N/A
N/A

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
72 (16)
0* (0)
N/A
64 (15)
N/A
N/A

L-6

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable (NSDUH) or not available (NAMCS/NHAMCS).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
10
NSDUH asks specifically about Butisol®, Seconal®, and phenobarbital. NAMCS and NHAMCS include a category for any barbiturates.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Ambulatory Medical Care Survey (NAMCS), 2010, National Hospital Ambulatory Medical Care Survey
(NHAMCS), 2010.

Table L-3 Comparison of Data for Stimulants from the 2012 NSDUH Questionnaire Field Test and the 2010 National Ambulatory
Medical Survey and 2010 National Hospital Ambulatory Medical Survey

L-7

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Any Prescription Stimulant5/Any Central
Nervous System Stimulant6
Adderall®, Adderall® XR, Dexedrine®,
Dextroamphetamine, or AmphetamineDextroamphetamine Combinations
Adderall®
Adderall® XR
Dexedrine®
Dextroamphetamine
Amphetamine-Dextroamphetamine
Combinations7
®
Ritalin , Ritalin® SR, Ritalin® LA, Concerta®,
Daytrana®, Metadate® CD, Metadate® ER,
Focalin®, Focalin® XR, Methylphenidate, or
Dexmethylphenidate
Ritalin®
Ritalin® SR or Ritalin® LA
Concerta®
Daytrana®
Metadate® CD
Metadate® ER
Focalin®
Focalin® XR
Methylphenidate
Dexmethylphenidate

NSDUH QFT,1
Percent (SE)
Any Past Year Use2

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3

NSDUH QFT,1
Percent (SE)
Past Year Misuse2

6.0 (0.64)

3.9 (0.48)

2.1 (0.39)

3.3
2.2
1.2
0.3
0.2

1.7
1.0
0.6
0.1
0.1

1.6
1.3
0.6
0.1
0.1

(0.49)
(0.37)
(0.23)
(0.11)
(0.10)

(0.32)
(0.21)
(0.16)
(0.08)
(0.05)

(0.32)
(0.28)
(0.15)
(0.08)
(0.09)

0.8 (0.27)

0.5 (0.22)

0.3 (0.12)

1.5
0.5
0.3
0.6
0.0
0.0
0.1
0.2
0.3
0.4
0.2

0.9
0.3
0.1
0.4
0.0*
0.0
0.1
0.1
0.2
0.3
0.1

0.6
0.2
0.2
0.2
0.0
0.0*
0.0*
0.1
0.1
0.1
0.1

(0.27)
(0.14)
(0.10)
(0.15)
(0.02)
(0.02)
(0.06)
(0.10)
(0.13)
(0.13)
(0.10)

(0.21)
(0.10)
(0.05)
(0.12)
(0.00)
(0.02)
(0.06)
(0.09)
(0.10)
(0.12)
(0.08)

(0.15)
(0.10)
(0.08)
(0.08)
(0.02)
(0.00)
(0.00)
(0.05)
(0.05)
(0.09)
(0.05)

NAMCS, Number of
Mentions
in Thousands (SE)4
17,054 (2,731)
4,860 (762)
3,464 (630)
1,153 (314)
193* (78)
13* (12)

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
1,437 (240)
351
241
101
2*
7*

(60)
(49)
(28)
(2)
(5)

38*

(28)

0* (0)

3,637
799
80*
1,470
112*
6*
114*
292*
294*
456*
14*

(664)
(209)
(75)
(327)
(85)
(6)
(94)
(124)
(123)
(153)
(11)

521 (120)
160 (46)
0* (0)
225 (57)
4* (3)
10* (9)
0* (0)
38* (17)
39* (37)
41* (16)
4* (3)
(continued)

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates
are for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
5
NSDUH QFT measure.
6
NAMCS/NHAMCS measure. NAMCS/NHAMCS mentions for specific drugs are limited to those that correspond to the drugs mentioned in the NSDUH screener questions.
7
For NAMCS/NHAMCS, mentions of the generic equivalent drug, excluding mentions of Adderall® or Adderall® XR.
(Source information is included on the last page of the table.)

Table L-3 Comparison of Data for Stimulants from the 2012 NSDUH Questionnaire Field Test and the 2010 National Ambulatory
Medical Survey and 2010 National Hospital Ambulatory Medical Survey (continued)

Reported Use (NSDUH) or Mention in
Ambulatory Medical Visits (NAMCS/NHAMCS)
Didrex® or Benzphetamine
Didrex®
Benzphetamine
Diethylpropion
Phendimetrazine
Phentermine
Provigil®
Tenuate®
Vyvanse®
Any Other Prescription Stimulant

NSDUH QFT,1
Percent (SE)
Any Past Year Use2
0.1 (0.04)
0.0 (0.03)
0.0 (0.03)
0.0 (0.02)
0.2 (0.15)
0.8 (0.23)
0.1 (0.06)
0.0* (0.00)
0.7 (0.23)
1.1 (0.25)

NSDUH QFT,1
Percent (SE)
Past Year Use But
Not Misuse3
0.1 (0.04)
0.0 (0.03)
0.0 (0.03)
0.0* (0.00)
0.2 (0.15)
0.7 (0.22)
0.1 (0.06)
0.0* (0.00)
0.5 (0.21)
1.0 (0.24)

NSDUH QFT,1
Percent (SE)
Past Year Misuse2
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
0.0 (0.02)
0.0* (0.00)
0.0 (0.03)
0.0* (0.00)
0.0* (0.00)
0.2 (0.09)
0.1 (0.07)

NAMCS, Number of
Mentions
in Thousands (SE)4
3* (3)
0* (0)
3* (3)
0* (0)
48* (48)
1,157* (515)
792 (209)
389* (279)
1,142 (279)
N/A

NHAMCS Hospital
Outpatient, Number of
Mentions
in Thousands (SE)4
6* (5)
6* (5)
0* (0)
0* (0)
6* (6)
111* (36)
73* (24)
19* (13)
130* (41)
N/A

L-8

* Low precision; estimate would be suppressed under NSDUH suppression rules or would not meet NAMCS and NHAMCS standards for reliability.
NAMCS = National Ambulatory Medical Survey; NHAMCS = National Hospital Ambulatory Medical Survey; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable (NSDUH) or not available (NAMCS/NHAMCS).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Data collected from September 1 through November 3, 2012. NSDUH estimates are
for the civilian, noninstitutionalized population aged 12 or older in the United States.
2
Persons with unknown data are excluded.
3
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with missing data for misuse are excluded.
4
Estimates are for the universe of annual outpatient office visits (NAMCS) or hospital outpatient department visits (NHAMCS) in the United States for persons aged 12 or older.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Ambulatory Medical Care Survey (NAMCS), 2010, National Hospital Ambulatory Medical Care Survey
(NHAMCS), 2010.

Table L-4 NSDUH Questionnaire Field Test and Monitoring the Future Comparisons for Past Year Misuse among Adolescents

Past Year Misuse1
Vicodin®, Lortab®, Lorcet®, or Hydrocodone
Vicodin®3
OxyContin®, Percocet®, Percodan®, Tylox®, or
Oxycodone
OxyContin®3
Prescription Tranquilizers
Prescription Stimulants4/Amphetamines5
Adderall®3
Ritalin®3

8th, 10th, 12th Graders
Aged 12 to 20 Years Old,
NSDUH QFT, Percent (SE)2
3.0 (1.20)
1.5 (0.93)
1.4 (0.69)
0.8 (0.54)
2.8 (1.12)
0.7 (0.55)
0.5* (0.51*)
0.0* (0.00*)

8th, 10th, 12th Graders,
2011 MTF,
Percent2
N/A
5.1
N/A
3.4
3.9
5.9
4.1
2.1

L-9

* NSDUH QFT low precision; estimate would be suppressed under NSDUH suppression rules.
MTF = Monitoring the Future; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable.
1
Defined in NSDUH as use "not directed for you by a doctor," including use without a prescription, in greater amounts, more often or longer than told to take a drug, or in some
other way not directed by a doctor. Defined in MTF as use "not under a doctor's orders."
2
NSDUH QFT data does not include Alaska or Hawaii and does not include Spanish-language interviews and were collected from September through November 3, 2012. MTF
data were collected in spring 2011. Published standard errors are not available for MTF data for combined 8th to 12th graders.
3
NSDUH QFT respondents in in grades 8, 10, or 12 and aged 12 to 20 with unknown data were excluded.
4
NSDUH question wording.
5
MTF question wording.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 - November 3, 2012;
University of Michigan, Monitoring the Future, 2011.

Table L-5 NSDUH Questionnaire Field Test and Monitoring the Future Comparisons for Past Year Misuse among Young Adults Aged
19 to 24

Past Year Misuse1
Prescription Pain Relievers3/Narcotics Other than
Heroin4
Vicodin®, Lortab®, Lorcet®, or Hydrocodone
Vicodin®5
OxyContin®, Percocet®, Percodan®, Tylox®, or
Oxycodone
OxyContin®5
Prescription Tranquilizers
Prescription Stimulants3/Amphetamines4
Adderall®5
Ritalin®5
Provigil®5
Prescription Sedatives3/Sedatives (Barbiturates)4

NSDUH
QFT, Aged
19 to 20,
Percent (SE)2

2011 MTF,
Aged 19
to 20,
Percent2

NSDUH
QFT, Aged
21 to 22,
Percent (SE)2

2011 MTF,
Aged 21
to 22,
Percent2

NSDUH
QFT, Aged
23 to 24,
Percent (SE)2

2011 MTF,
Aged 23
to 24,
Percent2

15.9 (3.51)
8.9 (2.91)
4.2 (2.18)

7.7
N/A
6.8

12.1 (3.03)
7.4 (2.30)
2.9 (1.51)

7.7
N/A
7.1

15.8* (4.63*)
11.6* (4.04*)
7.6* (3.95*)

7.8
N/A
7.7

8.2 (2.44)
3.6 (1.70)
6.6 (2.28)
8.1 (2.51)
5.1 (2.15)
0.0* (0.00*)
0.0* (0.00*)
0.7* 0.74*)

N/A
3.3
5.3
8.7
8.2
2.0
0.4
2.9

5.3 (2.02)
2.4 (1.41)
9.4 (2.75)
11.0 (3.05)
7.6 (2.50)
1.1 (0.85)
0.0* (0.00*)
0.7* 0.66*)

N/A
2.8
5.2
8.8
9.4
2.3
0.3
2.8

7.6 (2.55)
3.2* (2.06*)
9.7 (2.68)
6.0 (2.44)
4.6 (2.14)
1.0 (0.70)
0.0* (0.00*)
3.7 (2.12)

N/A
3.6
6.6
8.8
6.3
2.0
0.1
3.5

L-10

* NSDUH QFT low precision; estimate would be suppressed under NSDUH suppression rules.
MTF = Monitoring the Future; NSDUH QFT = NSDUH Questionnaire Field Test.
N/A: Not applicable.
1
Defined in NSDUH as use "not directed for you by a doctor," including use without a prescription, in greater amounts, more often or longer than told to take a drug, or in some
other way not directed by a doctor. Defined in MTF as use "not under a doctor's orders."
2
NSDUH QFT data does not include Alaska or Hawaii and does not include Spanish-language interviews and were collected September 1 through November 3, 2012. MTF
follow-up data were collected in spring 2011. Published standard errors are not available for MTF data for young adults.
3
NSDUH question wording.
4
MTF question wording.
5
NSDUH QFT young adults aged 19 to 24 with unknown misuse data were excluded.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 - November 3, 2012;
University of Michigan, Monitoring the Future, 2011.

Table L-6

Selected Characteristics among Persons Aged 12 or Older: Percentages and Standard
Errors, 2012 Questionnaire Field Test and 2011 National Health Interview Survey

Characteristic
HOUSEHOLD TELEPHONE SERVICE4,5
At least one telephone at address is not a cellular telephone
Anyone at address has a working cellular telephone
Cellular service only or no telephone service
Cellular telephone service only
No telephone service
NUMBER OF VISITS TO DOCTOR OR OTHER
HEALTH CARE PROFESSIONAL IN THE PAST
YEAR4,6
None
1
2 to 3
4 to 9
10 or more
HOSPITAL OVERNIGHT IN PAST YEAR4,5
EMERGENCY ROOM VISIT IN PAST YEAR4,6
CONDITIONS TOLD TO RESPONDENT BY DOCTOR
OR OTHER HEALTH CARE PROFESSIONAL
Any kind of heart condition or heart disease
Diabetes or sugar diabetes
Chronic bronchitis, emphysema, chronic obstructive
pulmonary disease, also called COPD
Cirrhosis of the liver
Hepatitis
Kidney disease, not including bladder infection or
incontinence
Asthma
Cancer or a malignancy of any kind
Hypertension, also called high blood pressure
See notes at end of table.

2012
Questionnaire Field
Test
(n = 2,044)1,2
Percent (SE)

NHIS, 2011
(n = 74,836)3
Percent (SE)

64.1
92.3
35.9
34.4
1.4

(1.68)
(0.82)
(1.68)
(1.63)
(0.33)

68.1
90.4
31.5
30.3
1.2

(.046)
(0.25)
(0.45)
(0.45)
(0.7)

15.5
21.0
30.2
22.7
10.6
9.7
26.5

(0.92)
(1.07)
(1.22)
(1.18)
(0.93)
(1.01)
(1.23)

17.2
18.0
27.4
24.3
13.1
8.3
20.3

(0.24)
(0.23)
(0.28)
(0.25)
(0.19)
(0.13)
(0.23)

10.4 (1.04)
9.0 (0.98)

10.8 (0.21)
8.1 (0.17)

3.3 (0.58)
0.2 (0.13)
2.1 (0.51)

5.7 (0.17)
1.3 (0.07)
3.0 (0.12)

1.3
11.1
6.1
17.8

(0.36)
(0.79)
(0.85)
(1.16)

1.8 (0.09)
13.6 (0.24)
8.6 (0.19)
30.3 (0.39)
(continued)

L-11

Table L-6 Selected Characteristics among Persons Aged 12 or Older: Percentages and Standard
Errors, 2012 Questionnaire Field Test and 2011 National Health Interview Survey
(continued)

Characteristic
DISABILITIES OR PHYSICAL LIMITATIONS
Deaf or serious hearing difficulty
Blind or serious difficulty seeing
Serious difficulty concentrating, remembering, or making
decisions
Serious difficulty walking or climbing stairs
Difficulty dressing or bathing
Difficulty doing errands alone, such as visiting a doctors'
office or shopping
FAMILY INCOME4,5
≤ $49,999
$50,000-$74,999
≥ $75,000
EDUCATION4,5,7
< High School
High School Graduate
Some College
College Graduate

2012
Questionnaire Field Test
(n = 2,044)1,2
Percent (SE)

NHIS, 2011
(n = 74,836)3
Percent (SE)

5.4 (0.61)
3.4 (0.58)

4.9 (0.21)
3.6 (0.18)

6.6 (0.68)
6.4 (0.89)
1.6 (0.36)

6.2 (0.25)
9.0 (0.28)
2.7 (0.15)

4.1 (0.68)

5.6 (0.21)

52.7 (2.05)
16.3 (1.22)
31.0 (1.97)

46.5 (0.54)
18.2 (0.33)
35.3 (0.55)

12.4
26.6
32.1
29.0

12.0
27.8
31.3
28.9

(1.26)
(1.92)
(1.42)
(2.48)

(0.20)
(0.29)
(0.26)
(0.38)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Sample includes Alaska and Hawaii and does not include Spanish-language interviews.
4
Respondents with unknown information were excluded.
5
NHIS weighted using person-level weights.
6
NHIS weighted using adult- and child-level weights, n = 33,961.
7
QFT and NHIS estimates are for persons aged 18 or older.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey, 2011.

L-12

Table L-7 2011 NHIS and 2009-2010 NHANES Height Statistics among Persons Aged 16 or Older
for Comparison with the 2012 Questionnaire Field Test
2012 QFT
Statistic

2009-2010 NHANES
1

Unbounded

NHIS Bounds

2011 NHIS

1,678

1,669

Mean

66.8

Standard Error

Self-Reported

Measured

31,999

5,261

5,845

66.4

66.8

67.1

66.5

0.27

0.21

0.03

0.06

0.07

Minimum

0.8

2.0

50.0

41.0

48.5

Maximum

158.0

76.0

80.0

79.8

66.2

66.5

66.4

Sample Size

Median

NHANES = National Health and Nutrition Examination Survey; NHIS = National Health Interview Survey; NSDUH QFT =
NSDUH Questionnaire Field Test.
1
Includes values up to 76 inches for men aged 18 or older and 70 inches for women aged 18 or older. For children, the weighted
1½ and 98½ percentiles for height were computed by age/gender. Respondents with values outside of these bounds were
excluded from the estimates.
2
For adults, these include values of 76 inches for men aged 18 or older and 70 inches for women aged 18 or older. For children,
the gender-specific height-for-age values of the highest 1½ percent of records and the lowest 1½ percent of records were
changed to "96" or "996" ("Not available"). In cases where extreme values were reported for either current height or current
weight, the data for both variables were changed to "96" or "996" ("Not available") on the public use data file.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
CDC, National Center for Health Statistics, National Health and Nutrition Examination Survey (NHANES), 20092010; National Health Interview Survey (NHIS), 2011.

L-13

Table L-8 2011 NHIS and 2009-2010 NHANES Weight Statistics among Persons Aged 16 or Older
for Comparison with the 2012 Questionnaire Field Test
2012 QFT1

2009-2010 NHANES
2

Self-Reported4

Measured

31,312

5,213

5,848

178.1

171.4

179.2

177.8

1.50

1.38

0.29

0.88

0.83

Minimum

100

76.0

55.3

Maximum

500

306

299

445.0

527.8

Median

172

167.4

174.0

171.0

Unbounded

NHIS Bounds

2011 NHIS

Sample Size

1,670

1,660

Mean

179.0

Statistic

Standard Error

NHANES = National Health and Nutrition Examination Survey; NHIS = National Health Interview Survey; NSDUH QFT =
NSDUH Questionnaire Field Test.
1
Pregnant women were asked to report their pre-pregnancy weight. Pregnancy status available for women aged 12 to 44.
2
For persons aged 18 or older, these include values between 126 and 299 pounds for men and 100 and 274 pounds for women.
For children, the weighted 1½ and 98½ percentiles for weight were computed by age. Respondents with values outside of these
bounds were excluded from the estimates.
3
For persons aged 18 or older, includes values between 126 and 299 pounds for men and 100 and 274 pounds for women. For
children, the gender-specific weight-for-age values of the highest 1½ percent of records and the lowest 1½ percent of records
were changed to "96" or "996" ("Not available"). In cases where extreme values were reported for either current height or
current weight, the data for both variables were changed to "96" or "996" ("Not available") on the public use data file.
4
Pregnant women were asked to report their pre-pregnancy weight. Pregnancy status available for women aged 20 to 44.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
CDC, National Center for Health Statistics, National Health and Nutrition Examination Survey (NHANES), 20092010; National Health Interview Survey (NHIS), 2011.

L-14

Table L-9 Received Income and Program Participation among Persons Aged 12 or Older: Percentages and Totals for 2011 Comparison, 2012
Comparison, 2012 Questionnaire Field Test, and Other Surveys

L-15

PERCENTAGES
TOTALS (in Thousands)
2011
2012
2011
2012
Comp.1
2011 ACS4 2011 NHIS5
2011 ACS4 2011 NHIS5
Comp.1,3
QFT1,2
Comp.1
Comp.1,3
QFT1,2
(SE)
(SE)
(SE)
(SE)
(SE)
Received Income
(SE)
(SE)
(SE)
(SE)
(SE)
Social Security
27.2
26.2
26.4
27.0
26.7
66,200
63,780
64,275
65,639
63,859
(0.42)
(0.53)
(1.70)
(0.05)
(0.35)
(1,316)
(1,727)
(5,216)
(123)
(994)
Wages
82.4
82.8
68.6
81.0
79.0
200,312
201,203
166,799
197,164
188,364
(0.38)
(0.48)
(1.77)
(0.04)
(0.32)
(2,158)
(3,028)
(8,293)
(111)
(2,197)
Supplemental
7.0
7.6
9.4
6.0
5.0
16,957
18,588
22,964
14,576
11,845
Security Income
(0.20)
(0.30)
(0.97)
(0.03)
(0.17)
(472)
(726)
(2,558)
(79)
(418)
Food Stamps
14.6
15.6
17.6
13.8
13.0
35,408
37,843
42,815
33,602
31,058
(0.32)
(0.46)
(1.49)
(0.05)
(0.32)
(755)
(1,141)
(3,786)
(110)
(824)
Welfare Payments
2.5
2.3
3.6
3.3
3.2
6,126
5,533
8,763
7,934
7,757
(0.11)
(0.16)
(0.56)
(0.03)
(0.14)
(278)
(373)
(1,434)
(65)
(338)
ACS = American Community Survey; Comp. = comparison; NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
NOTE: Unknown or invalid data were excluded from the analysis.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke English "not
well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table L-10 Received Income and Program Participation among Persons Aged 12 to 17: Percentages and Totals for 2011 Comparison, 2012
Comparison, 2012 Questionnaire Field Test, and Other Surveys

L-16

PERCENTAGES
TOTALS (in Thousands)
2011
2012
2011
2012
Comp.1
2011 ACS4 2011 NHIS5
2011 ACS4 2011 NHIS5
Comp.1,3
QFT1,2
Comp.1
Comp.1,3
QFT1,2
(SE)
(SE)
(SE)
(SE)
(SE)
Received Income
(SE)
(SE)
(SE)
(SE)
(SE)
Social Security
12.2
11.1
12.7
10.6
12.3
2,949
2,698
3,071
2,598
2,737
(0.39)
(0.42)
(1.74)
(0.10)
(0.66)
(96)
(112)
(501)
(25)
(158)
Wages
89.4
89.6
65.6
90.7
87.9
21,653
21,697
15,876
22,265
19,433
(0.36)
(0.41)
(2.67)
(0.11)
(0.64)
(297)
(435)
(1,178)
(46)
(451)
Supplemental
7.6
7.8
9.9
6.0
6.0
1,846
1,877
2,389
1,464
1,329
Security Income
(0.29)
(0.36)
(1.64)
(0.07)
(0.48)
(70)
(91)
(429)
(18)
(111)
Food Stamps
20.9
21.4
27.7
20.9
19.4
5,061
5,174
6,707
5,132
4,309
(0.44)
(0.64)
(2.54)
(0.13)
(0.85)
(126)
(178)
(729)
(33)
(213)
Welfare Payments
4.2
4.0
5.6
4.9
4.7
1,024
959
1,364
1,207
1,034
(0.23)
(0.31)
(1.15)
(0.07)
(0.47)
(59)
(77)
(296)
(17)
(106)
ACS = American Community Survey; Comp. = comparison; NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
NOTE: Unknown or invalid data were excluded from the analysis.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke English "not
well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table L-11 Received Income and Program Participation among Persons Aged 18 to 25: Percentages and Totals for 2011 Comparison, 2012
Comparison, 2012 Questionnaire Field Test, and Other Surveys

L-17

PERCENTAGES
TOTALS (in Thousands)
2011
2012
2011
2012
Comp.1
2011 ACS4 2011 NHIS5
2011 ACS4 2011 NHIS5
Comp.1,3
QFT1,2
Comp.1
Comp.1,3
QFT1,2
(SE)
(SE)
(SE)
(SE)
(SE)
Received Income
(SE)
(SE)
(SE)
(SE)
(SE)
Social Security
9.4
9.2
9.2
9.9
10.3
3,108
3,025
3,036
3,314
3,251
(0.29)
(0.41)
(1.44)
(0.10)
(0.82)
(104)
(127)
(496)
(31)
(268)
Wages
91.6
91.0
68.8
91.7
89.6
30,200
30,015
22,698
30,658
28,138
(0.31)
(0.74)
(2.55)
(0.08)
(0.70)
(513)
(65)
(2,067)
(54)
(795)
Supplemental
6.2
5.7
9.8
5.7
4.9
2,047
1,888
3,219
1,910
1,550
Security Income
(0.24)
(0.29)
(1.66)
(0.06)
(0.49)
(88)
(91)
(593)
(21)
(157)
Food Stamps
20.1
20.2
21.9
18.2
19.7
6,644
6,674
7,215
6,089
6,230
(0.46)
(0.64)
(2.47)
(0.09)
(0.86)
(160)
(215)
(881)
(31)
(305)
Welfare Payments
4.3
3.8
5.1
4.0
6.2
1,429
1,246
1,697
1,334
1,942
(0.20)
(0.27)
(1.04)
(0.06)
(0.54)
(70)
(91)
(343)
(20)
(180)
ACS = American Community Survey; Comp. = comparison; NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
NOTE: Unknown or invalid data were excluded from the analysis.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke English "not
well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table L-12 Received Income and Program Participation among Persons Aged 26 or Older: Percentages and Totals for 2011 Comparison, 2012
Comparison, 2012 Questionnaire Field Test, and Other Surveys

L-18

PERCENTAGES
TOTALS (in Thousands)
2011
2012
2011
2012
Comp.1
2011 ACS4 2011 NHIS5
2011 ACS4 2011 NHIS5
Comp.1,3
QFT1,2
Comp.1
Comp.1,3
QFT1,2
(SE)
(SE)
(SE)
(SE)
(SE)
Received Income
(SE)
(SE)
(SE)
(SE)
(SE)
Social Security
32.3
31.2
31.3
32.2
31.2
60,143
58,058
58,168
59,727
57,872
(0.53)
(0.65)
(2.10)
(0.04)
(0.39)
(1,285)
(1,689)
(5,116)
(93)
(928)
Wages
79.8
80.4
69.0
77.8
76.1
148,459
149,492
128,225
144,242
140,793
(0.48)
(0.59)
(2.10)
(0.04)
(0.35)
(1,967)
(2,594)
(7,326)
(97)
(1,642)
Supplemental
7.0
8.0
9.3
6.0
4.8
13,064
14,822
17,355
11,202
8,967
Security Income
(0.24)
(0.38)
(1.14)
(0.03)
(0.17)
(439)
(698)
(2,275)
(58)
(329)
Food Stamps
12.7
14.0
15.5
12.1
11.1
23,703
25,995
28,893
22,381
20,519
(0.37)
(0.51)
(1.56)
(0.04)
(0.28)
(679)
(992)
(2,959)
(75)
(539)
Welfare Payments
2.0
1.8
3.1
2.9
2.6
3,673
3,327
5,702
5,393
4,781
(0.13)
(0.17)
(0.61)
(0.02)
(0.12)
(250)
(315)
(1,157)
(44)
(217)
ACS = American Community Survey; Comp. = comparison; NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
NOTE: Unknown or invalid data were excluded from the analysis.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke English "not
well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table L-13 Health Insurance Coverage among Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item

2011 Comparison1
Percent (SE)

2012 Comparison1,3
Percent (SE)

2012 QFT1,2
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

Medicare (QHI01)

18.1 (0.38)

18.0 (0.53)

18.3 (1.58)

17.8 (0.02)

17.7 (0.25)

Medicaid (QHI02 and QHI02a)

11.6 (0.24)

11.5 (0.35)

13.4 (1.16)

12.9 (0.04)

10.6 (0.21)

4.7 (0.18)

4.6 (0.24)

5.0 (0.77)

4.8 (0.02)

3.5 (0.12)

62.1 (1.86)

67.5 (0.07)

68.7 (0.36)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military Health
Care (QHI03)
Private Health Insurance (QHI06)

67.1 (0.42)

67.5 (0.59)

L-19

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Department
of Veterans Affairs; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard error; TRICARE = Department of Defense heath care program with
three levels of coverage, prime, standard, and extra; VA = Department of Veterans Affairs.
NOTE: Unknown or invalid data were excluded from the analysis.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table L-14 Health Insurance Coverage among Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item
Medicare (QHI01)
Medicaid (QHI02 and QHI02a)
TRICARE, CHAMPUS,
CHAMPVA, VA, Military Health
Care (QHI03)
Private Health Insurance (QHI06)

2011 Comparison1
Percent (SE)

2012 Comparison1,3
Percent (SE)

2012 QFT1,2
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

0.4a (0.07)

0.4a (0.08)

1.8 (0.49)

0.6 (0.02)

0.2 (0.08)

31.8 (0.55)

32.8 (0.80)

36.2 (2.69)

30.7 (0.13)

27.9 (0.80)

3.1 (0.21)

2.9 (0.24)

2.6 (0.71)

2.3 (0.04)

2.3 (0.24)

60.6 (0.79)

54.9 (3.00)

62.0 (0.17)

67.9 (0.84)

61.3 (0.60)

L-20

Table L-15 Health Insurance Coverage among Persons Aged 18 to 25: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item
Medicare (QHI01)
Medicaid (QHI02 and QHI02a)
TRICARE, CHAMPUS,
CHAMPVA, VA, Military Health
Care (QHI03)
Private Health Insurance (QHI06)

2011 Comparison1
Percent (SE)

2012 Comparison1,3
Percent (SE)

2012 QFT1,2
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

0.6 (0.07)

0.8 (0.11)

1.6 (0.63)

0.7 (0.02)

0.5 (0.08)

15.7 (0.42)

15.5 (0.57)

15.9 (2.15)

13.7 (0.08)

14.3 (0.52)

2.6 (0.17)

2.7 (0.24)

2.9 (1.01)

2.4 (0.04)

2.1 (0.19)

56.5 (0.56)

58.7 (0.78)

52.3 (3.31)

61.0 (0.12)

62.3 (0.79)

L-21

Table L-16 Health Insurance Coverage among Persons Aged 26 or Older: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item

2011 Comparison1
Percent (SE)

2012 Comparison1,3
Percent (SE)

2012 QFT1,2
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

Medicare (QHI01)

23.5 (0.49)

23.3 (0.67)

23.4 (1.94)

23.2 (0.02)

22.7 (0.30)

Medicaid (QHI02 and QHI02a)

8.3 (0.25)

8.1 (0.38)

10.0 (1.21)

10.4 (0.04)

7.9 (0.17)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military Health
Care (QHI03)

5.3 (0.23)

5.2 (0.30)

5.6 (0.92)

5.6 (0.02)

3.9 (0.13)

64.8 (2.16)

69.3 (0.07)

69.9 (0.35)

Private Health Insurance (QHI06)

69.8 (0.50)

69.9 (0.68)

L-22

Table L-17 Income among Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS
Income Level
<$49,999
$50,000 - $74,999
$75,000 or More

2011 Comparison1
Percent (SE)
49.2 (0.49)
17.5 (0.28)
33.3 (0.53)

2012 Comparison1,2
Percent (SE)
50.2 (0.63)
16.8 (0.42)
33.0 (0.63)

2012 QFT1,3
Percent (SE)
52.7 (2.05)
16.3 (1.22)
31.0 (1.97)

2011 NHIS4
Percent (SE)
46.5 (0.54)
18.2 (0.33)
35.3 (0.55)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

L-23

Table L-18 Income among Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS
Income Level
<$49,999
$50,000 - $74,999
$75,000 or More

2011 Comparison1
Percent (SE)
47.8a (0.63)
16.8a (0.38)
35.4 (0.57)

2012 Comparison1,2
Percent (SE)
47.6a (0.98)
16.7a (0.52)
35.7 (0.82)

2012 QFT1,3
Percent (SE)
54.9 (3.15)
12.3 (1.60)
32.9 (3.01)

2011 NHIS4
Percent (SE)
41.1 (1.11)
17.2 (0.91)
41.7 (1.10)

L-24

Table L-19 Income among Persons Aged 18 to 25: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS Data
Income Level
<$49,999
$50,000 - $74,999
$75,000 or More

2011 Comparison1
Percent (SE)
66.8 (0.65)
13.2 (0.39)
20.0 (0.52)

2012 Comparison1,2
Percent (SE)
67.2 (0.98)
13.3 (0.59)
19.5 (0.64)

2012 QFT1,3
Percent (SE)
68.7 (3.01)
13.6 (2.19)
17.7 (2.18)

2011 NHIS4
Percent (SE)
61.2 (1.31)
15.8 (0.85)
23.0 (1.16)

L-25

Table L-20 Income among Persons Aged 26 or Older: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and NHIS Data
Income Level
<$49,999
$50,000 - $74,999
$75,000 or More

2011 Comparison1
Percent (SE)
46.3 (0.57)
18.3 (0.36)
35.4 (0.60)

2012 Comparison1,2
Percent (SE)
47.5 (0.72)
17.5 (0.55)
35.1 (0.74)

2012 QFT1,3
Percent (SE)
49.6 (2.36)
17.3 (1.46)
33.1 (2.42)

NHIS4
Percent (SE)
44.6 (0.52)
18.7 (0.33)
36.7 (0.54)

L-26

Table L-21 Levels of Current Employment among Persons Aged 18 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data

Current Employment
Full-Time
Part-Time
Unemployed
Other5

2011 Comparison1
Percent (SE)
49.7 (0.49)
14.1 (0.26)
5.8 (0.14)
30.4 (0.43)

2012
Comparison1,2
Percent (SE)
51.3 (0.63)
13.9 (0.39)
5.5 (0.20)
29.3 (0.65)

2012 QFT1,3
Percent (SE)
52.0 (1.65)
14.2 (1.15)
5.5 (0.65)
28.3 (1.70)

CPS Q3 & Q44
Percent (SE)
49.2 (0.07)
11.2 (0.05)
4.9 (0.03)
34.7 (0.07)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison
compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
5
The Other Employment category includes students, person keeping house or caring for children full time, retired or disabled
persons, or other persons not in the labor force.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

L-27

Table L-22 Levels of Current Employment among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data

Current Employment
Full-Time
Part-Time
Unemployed
Other5

2011 Comparison1
Percent (SE)
36.0a (0.56)
27.8 (0.42)
13.2 (0.33)
23.0a (0.43)

2012
Comparison1,2
Percent (SE)
40.1 (0.86)
26.4 (0.67)
11.8 (0.41)
21.7 (0.91)

2012 QFT1,3
Percent (SE)
45.5 (2.98)
24.4 (2.29)
11.9 (1.58)
18.2 (1.83)

CPS Q3 & Q44
Percent (SE)
35.0 (0.19)
22.4 (0.17)
9.4 (0.12)
33.2 (0.19)

L-28

Table L-23 Levels of Current Employment among Persons Aged 26 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data

Current Employment
Full-Time
Part-Time
Unemployed
Other5

2011 Comparison1
Percent (SE)
52.1 (0.55)
11.7 (0.30)
4.5 (0.16)
31.7 (0.51)

2012
Comparison1,2
Percent (SE)
53.3 (0.72)
11.7 (0.43)
4.4 (0.23)
30.7 (0.75)

2012 QFT1,3
Percent (SE)
53.2 (1.90)
12.4 (1.34)
4.3 (0.70)
30.1 (2.01)

CPS Q3 & Q44
Percent (SE)
51.5 (0.08)
9.3 (0.04)
4.2 (0.03)
35.0 (0.08)

L-29

Table L-24 Unemployment Rates among Persons Aged 18 or Older, by Age Group: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field
Test, and CPS Data

Age/Unemployment Rate
18 or Older
Unemployment Rate
18 to 25
Unemployment Rate
26 or Older
Unemployment Rate

2011 Comparison1
Percent (SE)

2012
Comparison1,2
Percent (SE)

8.4 (0.21)

7.8 (0.29)

17.2 (0.21)

15.0 (0.48)

6.6 (0.23)

6.3 (0.34)

2012 QFT1,3
Percent (SE)
7.6 (0.91)
14.6 (1.93)
6.2 (1.00)

CPS Q3 & Q44
Percent (SE)
7.6 (0.05)
14.0 (0.18)
6.5 (0.05)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison
compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

L-30

Table L-25 Levels of Education among Persons Aged 18 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011
NHIS
Level of Education
< High School
High School Graduate
Some College
College Graduate

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)
11.6 (0.24)
11.5 (0.35)
30.3 (0.38)
30.1 (0.61)
27.4a (0.37)
27.7a (0.48)
30.6 (0.41)
30.7 (0.67)

2012 QFT1,3
Percent (SE)
12.4 (1.26)
26.6 (1.92)
32.1 (1.42)
29.0 (2.48)

2011 NHIS4
Percent (SE)
12.0 (0.20)
27.8 (0.29)
31.3 (0.26)
28.9 (0.38)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison
compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012; CDC, National Center for Health Statistics, National Health Interview Survey (NHIS),
2011.

L-31

Table L-26 Levels of Education among Persons Aged 18 to 25: Percentages and Standard Errors,
2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011 NHIS
Level of Education
< High School
High School Graduate
Some College
College Graduate

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)
15.6 (0.40)
12.0 (0.42)
34.0 (0.55)
35.7 (1.04)
35.7 (0.59)
36.4 (0.90)
14.7 (0.46)
15.9 (0.60)

2012 QFT1,3
Percent (SE)
13.8 (1.92)
34.9 (2.56)
37.6 (3.40)
13.7 (2.30)

2011 NHIS4
Percent (SE)
14.0 (0.49)
29.6 (0.65)
43.0 (0.83)
13.5 (0.54)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison
compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

L-32

Table L-27 Levels of Education among Persons Aged 26 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011
NHIS
Level of Education
< High School
High School Graduate
Some College
College Graduate

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)
10.9 (0.28)
11.4 (0.41)
29.7a (0.43)
29.1 (0.69)
26.0a (0.41)
26.2a (0.57)
33.4 (0.47)
33.3 (0.77)

2012 QFT1,3
Percent (SE)
12.1 (1.39)
25.1 (2.16)
31.1 (1.76)
31.7 (2.77)

2011 NHIS4
Percent (SE)
11.6 (0.21)
27.5 (0.31)
29.3 (0.25)
31.6 (0.40)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison
compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1
through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

L-33

L-34

Appendix M: Estimates for New Items in the 2012
Questionnaire Field Test That Were Included in the 2013
NSDUH Main Study Questionnaire

Table M-1 Estimates and Standard Errors for New Items in the 2012 Questionnaire Field Test
That Were Included in the 2013 NSDUH Main Study Questionnaire among Persons
Aged 12 or Older

Instrument Item
Race3,4 (QD05)
White (QD051)
Black or African American (QD052)
American Indian or Alaska Native (American Indian includes North
American, Central American, and South American Indians)
(QD053)
Native Hawaiian (QD054)
Guamanian or Chamorro (QD055)
Samoan (QD056)
Other Pacific Islander (QD057)
Asian (Including: Asian, Indian, Chinese, Filipino, Japanese, Korean,
and Vietnamese (QD058)
Other (Specify) (QD059)
Member of a Reserve Component Currently Serving Full-Time in an
Active-Duty Status (V2a)
Serving Full-Time in a Reserve Component (V2b)
Ever Served on Active Duty in the United States Armed Forces or
Reserve Components (QD10a)
Time Served4,5 (QD10b)
September 2001 or Later (QD10b11)
August 1990 to August 2001 (Including Persian Gulf War)
(QD10b12)
May 1975 to July 1990 (QD10b13)
Vietnam Era (August 1964 to April 1975) (QD10b14)
February 1955 to July 1964 (QD10b15)
Korean War (July 1950 to January 1955) (QD10b16)
January 1947 to June 1950 (QD10b17)
World War II (December 1941 to December 1946) (QD10b18)
November 1941 or Earlier (QD10b19)
Drew Imminent Danger Pay or Hostile Fire Pay5 (QD10c)
Any Marijuana Use in the Past 12 Months Recommended by Doctor
(MJMM)
All Marijuana Use in the Past 12 Months Recommended by Doctor6
(MJMM01)
See notes at end of table.

2012 QFT
Estimate
(n =
2,044)1,2

Standard
Error

Unweighted
Total

78.0
13.5

(1.93)
(1.63)

1,479
353

1.8
0.1
0.0*
0.1
0.3

(0.42)
(0.06)
(0.00)
(0.09)
(0.11)

82
3
0
2
19

5.3
2.7

(0.89)
(0.49)

107
81

0.0*
0.0*

(0.00)
(0.00)

0
0

7.5

(0.86)

10.8*

(2.88)

18.1*
20.9*
45.4*
8.9*
8.4*
0.9*
5.4*
0.0*
36.8*

(4.77)
(5.32)
(5.96)
(3.28)
(3.21)
(0.94)
(2.71)
(0.00)
(6.71)

15
17
30
7
6
1
4
0
38

0.5

(0.16)

41.5*

(15.49)

5
(continued)

M-1

Table M-1 Estimates and Standard Errors for New Items in the 2012 Questionnaire Field Test
That Were Also Included in the 2013 NSDUH Main Study Questionnaire among
Persons Aged 12 or Older (continued)

Instrument Item
Average Weight3,8 (HLTH10-14)
Average Number of Times Treated in an Emergency Room3 (HLTH16)
Stayed Overnight or Longer as an Inpatient in a Hospital3 (HLTH17)
Average Number of Nights Inpatient in a Hospital3,9 (HLTH18)
Average Number Times Visited a Doctor about Own Health at a
Doctor's Office3 (HLTH19)
Doctor Asked, Either in Person or on a Form, about Use3,10
(HLTH20)
Smoke Cigarettes or Use Any Other Tobacco Products (HLTH20a)
Drink Alcohol (HLTH20b)
Use Illegal Drugs (HLTH20c)
TRICARE, or CHAMPUS, CHAMPVA, the VA, or Military Health
Care3 (QHI03)
Social Security or Railroad Retirement Payment3 (QI01n)

2012 QFT
Estimate
(n =
2,044)1,2
176.0
0.5
9.7
4.6
3.9

Standard
Error
(1.44)
(0.04)
(1.01)
(0.75)
(0.18)

Unweighted
Total
N/A
N/A
173
N/A
N/A

71.2
67.9
51.0
5.0

(1.37)
(1.50)
(1.55)
(0.77)

1,137
1,067
865
77

26.5

(1.69)

351

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test.
NOTE: All estimates are based on the raw data, with no edits applied.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Respondents with unknown or missing data were excluded from the analysis.
4
Respondents could report multiple responses to these items.
5
Estimates are among only respondents who reported serving on active duty in the United States Armed Forces or Reserve
components.
6
Estimates are among only respondents who reported some of their marijuana use in the past year was recommended by a doctor.
7
Average is reported in inches.
8
Average is reported in pounds and includes pre-pregnancy weight of pregnant females as reported in HLTH13 and HLTH14.
9
Estimates are among only respondents who reported staying overnight or longer in a hospital in the past 12 months.
10
Estimates are among only respondents who reported being treated at an emergency room at least once, stayed overnight or
longer in a hospital, or visited a doctor, nurse, physician assistant or nurse practitioner about your own health at a doctor's
office, a clinic, or some other place in the past 12 months.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

M-2

Appendix N: Moved Demographic and Household Items in
the 2012 Questionnaire Field Test: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

49.8 (0.46)

49.7 (0.59)

51.0 (2.03)

5.6 (0.21)

5.5 (0.30)

4.9 (0.81)

Divorced or Separated

13.7 (0.28)

14.1 (0.44)

13.8 (1.19)

Have Never Married

30.9 (0.36)

30.6 (0.48)

30.2 (1.54)

How many times have you been
married? (QD08) 4,5,6

1.4 (0.01)

1.3 (0.01)

1.4 (0.03)

How many times in the past 12 months
have you moved? (QD13)6,7

0.3 (0.01)

0.4 (0.03)

Were you born in the United States?
(QD14)4

88.8 (0.30)

88.9 (0.39)

87.9 (1.29)

How many years have you lived in the
United States? (QD16b) 5,6

22.5 (0.40)

22.3 (0.59)

23.7 (1.56)

Are you now attending or are you
currently enrolled in school? (QD17)4,5

21.1 (0.26)

20.7 (0.32)

18.9 (1.07)

Instrument Item
Are you now married, widowed,
divorced or separated, or have you
never married? (QD07)4,5
Married
Widowed

What grade or year of school are you
now attending? (QD18) 4,5
1st Grade
2nd Grade

0.0* (0.00* )

0.0 (0.00 )

0.2 (0.15)

0.0 (0.00 )

0.3 (0.23)

3rd Grade

0.0 (0.00 )

0.0 (0.01)

0.0* (0.00* )

4th Grade

0.0a (0.00)

0.0* (0.00* )

5th Grade

0.2a (0.02)

0.1a (0.02)

0.0* (0.00* )

6th Grade

2.7 (0.11)

1.3 (0.09)

1.2 (0.43)

7th Grade

7.1 (0.18)

7.4 (0.23)

7.7 (0.92)

8th Grade

7.9 (0.18)

8.0 (0.25)

9.8 (1.17)

9th Grade

7.9 (0.16)

8.3 (0.26)

9.7 (1.19)

10th Grade

8.5 (0.21)

8.4 (0.24)

8.3 (0.91)

11th Grade

8.1 (0.20)

8.3 (0.28)

8.2 (0.98)

12th Grade

8.8 (0.24)

8.9 (0.31)

9.1 (0.99)

College or University/1st Year

10.7 (0.34)

12.1 (0.76)

12.2 (1.54)

College or University/2nd Year

11.0 (0.38)

10.0 (0.43)

8.8 (1.34)

College or University/3rd Year

9.7 (0.37)

9.8 (0.47)

8.5 (1.44)

College or University/4th Year
See notes at end of table.

6.2 (0.30)

6.1 (0.38)

5.1 (1.24)
(continued)

N-1

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

Full-Time

81.8 (0.53)

83.1 (0.65)

80.7 (2.14)

Part-Time

18.2 (0.53)

16.9 (0.65)

19.3 (2.14)

During the past 30 days how many
whole days of school did you miss
because you were sick or injured?
(QD20)5,6,7

0.8 (0.02)

0.7 (0.03)

0.8 (0.16)

During the past 30 days how many
whole days of school did you miss
because you skipped or “cut” or just
didn’t want to be there? (QD21)5,6,7

0.4 (0.01)

0.3 (0.02)

0.4 (0.07)

Did you work at a job or business at
any time last week? (QD26)4,5

57.4 (0.42)

57.7 (0.62)

60.0 (1.72)

Even though you did not work at any
time last week, did you have a job or
business? (QD27)4,5

10.5 (0.32)

13.7 (0.63)

12.1 (1.68)

How many hours did you work last
week at all jobs or businesses?
(QD28)5,6,7

38.6 (0.14)

39.0 (0.22)

38.5 (0.51)

Do you usually work 35 hours or more
per week at all jobs or businesses?
(QD29)4,5

76.5 (0.41)

77.2 (0.54)

77.0 (1.53)

Vacation/Sick/Furlough/Strike/
Other Temporary Absence/
Maternity Leave

54.6a (1.71)

55.9a (2.47)

33.0* (5.79* )

Layoff, Not Looking for Work

3.1 (0.44)

2.9 (0.52)

3.6* (2.19* )

Layoff, Looking for Work

4.6 (0.58)

3.2 (0.56)

9.8* (4.37* )

Waiting to Report to New Job

5.3 (0.62)

6.0 (1.02)

4.3 (1.88)

Self-Employed, No Business
Last Week

14.5 (1.33)

13.2 (1.65)

15.4* (5.46* )

7.2 (0.48)

6.1 (0.58)

11.7 (3.42)

10.8 (1.21)

12.9 (1.80)

22.1* (5.73* )

15.7 (0.33)

15.6 (0.55)

16.3 (1.90)

1.7 (0.15)

1.5 (0.19)

Instrument Item
Are you a full-time student or a parttime student? (QD19)4,5

Which one of these reasons best
describes why you did not work last
week? (QD30)4,5

Going to School/Training
Some Other Reason
Which one of these reasons best
describes why you did not have a job or
business last week? (QD31)4,5
Looking for Work
On Layoff, Not Looking for Work
See notes at end of table.

N-2

1.5 (0.46)
(continued)

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

10.1 (0.33)

11.0 (0.56)

11.8 (1.89)

Going to School/Training

13.9 (0.31)

13.0 (0.50)

9.9 (1.08)

Retired

39.3 (0.73)

38.0 (0.97)

38.0 (2.90)

Disabled

13.8 (0.47)

15.4 (0.78)

14.7 (1.99)

Didn't Want A Job
Some Other Reason

3.9 (0.20)
a

2.3 (0.55)

4.2 (0.28)

1.7 (0.15)

1.3 (0.17)

5.5 (0.98)

During the past 30 days, did you make
specific efforts to find work? (QD32)4,5

87.7 (0.79)

88.6 (0.97)

82.1 (3.68)

Did you work at a job or business at
any time during the past 12 months?
(QD33)4,5

19.8 (0.44)

19.8 (0.66)

18.9 (2.04)

How many different employers have
you had in the past 12 months? (QD35
and QD36)5,6

1.3 (0.01)

1.4 (0.05)

During the past 12 months, was there
ever a time when you did not have at
least one job or business? (QD37)4,5

12.4a (0.30)

12.3a (0.33)

15.6 (1.35)

In how many weeks during the past 12
months did you not have at least one
job or business? (QD38)5,6

17.1a (0.29)

17.9a (0.44)

13.8 (0.99)

During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)5,6,7

0.6 (0.02)

0.7 (0.04)

0.7 (0.12)

During the past 30 days, how many
whole days of work did you miss
because you just didn’t want to be
there? (QD41)5,6,7

0.2 (0.01)

0.2 (0.02)

0.2 (0.03)

How many people work for your
employer out of this office, store, etc.?
(QD42)4,5
Fewer Than 10 People

29.3 (0.45)

28.3 (0.55)

30.3 (1.93)

10 to 24 People

16.7 (0.32)

18.2 (0.53)

18.3 (1.36)

25 to 99 People

22.3a (0.38)

21.4a (0.41)

18.6 (1.28)

100 to 499 People

17.8 (0.41)

18.2 (0.48)

18.4 (1.59)

14.0 (0.35)

13.9 (0.52)

14.4 (1.66)
(continued)

500 People or More
See notes at end of table.

N-3

Table N-1 Moved Demographic and Household Items in the 2012 Questionnaire Field Test:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)
Instrument Item
At your workplace, is there a written
policy about employee use of alcohol or
drugs? (QD43)4,5

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

77.1 (0.41)

77.9 (0.49)

80.1 (1.63)

0.7 (0.09)

0.6 (0.08)

1.1 (0.49)

Does this policy cover only alcohol,
only drugs, or both alcohol and drugs?
(QD44)4,5
Only Alcohol
Only Drugs
Both Alcohol and Drugs

3.0 (0.18)

3.5 (0.21)

2.3 (0.52)

96.3 (0.20)

95.9 (0.22)

96.5 (0.73)

Through your workplace, is there
access to any type of employee
assistance program or other type of
counseling program for employees who
have alcohol or drug-related problems?
(QD46)4,5

53.6 (0.56)

53.6 (0.68)

53.5 (1.98)

Does your workplace ever test its
employees for alcohol use? (QD47)4,5

33.2 (0.51)

33.3 (0.62)

31.5 (1.71)

Does your workplace ever test its
employees for drug use? (QD48)4,5

48.9 (0.52)

50.4 (0.71)

48.1 (2.05)

Does your workplace test its employees
for drug or alcohol use as part of the
hiring process? (QD49)4,5

86.7 (0.45)

87.5 (0.63)

87.6 (1.71)

Does your workplace test its employees
for drug or alcohol use on a random
basis? (QD50)4,5

62.2 (0.64)

62.4 (0.92)

59.8 (3.18)

Handled on Individual
Basis/Policy Does Not Specify
What Happens

20.9 (0.64)

18.6a (0.74)

24.3 (2.51)

Employee Is Fired

50.3 (0.75)

52.1 (1.12)

47.1 (2.65)

Employee Referred for
Treatment/Counseling

26.2 (0.74)

26.2 (0.70)

23.6 (2.17)

0.2 (0.04)

0.4 (0.11)

1.6 (0.85)

2.3 (0.19)

2.7 (0.29)

3.4 (1.00)
(continued)

According to the policy at your
workplace, what happens to an
employee the first time he or she tests
positive for illicit drugs? (QD51)4,5

Nothing Happens
Something Else Happens
See notes at end of table.

N-4

Table N-1 Moved Demographic and Household Items in the 2012 Questionnaire Field Test:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)
Instrument Item
Would you be more or less likely to
want to work for an employer that tests
its employees for drug use as part of the
hiring process? (QD52)4,5
More Likely
Less Likely
Would Make No Difference

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

44.0a (0.44)

44.4 (0.64)

48.3 (1.85)

4.2 (0.23)
a

7.2 (0.82)

4.3 (0.25)

51.8 (0.46)

51.3 (0.63)

44.6 (1.57)

36.6a (0.47)

37.1a (0.59)

43.1 (1.77)

Would you be more or less likely to
want to work for an employer that tests
its employees for drug or alcohol use on
a random basis? (QD53)4,5
More Likely
Less Likely

8.5 (0.30)
a

11.5 (1.24)

8.3 (0.32)

Would Make No Difference

54.9 (0.48)

54.6 (0.60)

45.4 (1.66)

Was [SAMPLE MEMBER] private
health insurance obtained through
work? (QHI07)4,5

87.0 (0.37)

87.2 (0.51)

88.6 (1.47)

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)4,5

83.7a (0.45)

84.0a (0.67)

74.2 (1.99)

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for drug abuse? (QHI09)4,5

82.9a (0.44)

83.3a (0.68)

73.2 (2.04)

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for mental or emotional
problems? (QHI10)4,5

91.1a (0.28)

91.7a (0.45)

85.0 (1.62)

[SAMPLE MEMBER A] currently
covered by any kind of health insurance
including Indian Health Insurance?
(QHI11)4

10.3a (0.42)

12.7a (0.75)

21.9 (2.71)

In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)8

27.2 (0.42)

26.2 (0.53)

26.4 (1.70)

8.1a (0.14)

8.4a (0.18)

For how many months in [YEAR] did
you or your [RELATIONSHIP] receive
any type of welfare or public assistance,
not including food stamps? (QI12AN
and QI12BN)6,8
See notes at end of table.

N-5

6.0 (0.51)
(continued)

Table N-1 Moved Demographic and Household Items in the 2012 Questionnaire Field Test:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)
Instrument Item
Of these income groups, which
category best represents [SAMPLE
MEMBER] total personal income
during [YEAR]?(QI21B)5
Less Than $1,000
$1,000-$1,999

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

16.6 (0.22)

16.7a (0.35)

14.9 (0.84)

2.2 (0.08)

2.5 (0.13)

2.9 (0.38)

$2,000-$2,999

1.8 (0.09)

1.6 (0.09)

1.2 (0.23)

$3,000-$3,999

1.5 (0.07)

1.7 (0.12)

1.4 (0.30)

$4,000-$4,999

1.3 (0.06)

1.2 (0.09)

1.1 (0.27)

$5,000-$5,999

1.6a (0.08)

1.4 (0.10)

0.9 (0.23)

1.5 (0.12)

0.9 (0.27)

$6,000-$6,999

1.5 (0.09)
a

$7,000-$7,999

1.7 (0.09)

1.7 (0.14)

0.4 (0.19)

$8,000-$8,999

1.9 (0.10)

2.0 (0.15)

1.3 (0.32)

$9,000-$9,999

1.9 (0.09)

1.9 (0.14)

2.6 (0.51)

$10,000-$10,999

2.1 (0.10)

2.2 (0.14)

2.3 (0.44)

$11,000-$11,999

1.5 (0.07)

1.7 (0.13)

1.4 (0.36)

$12,000-$12,999

2.2 (0.12)

2.5 (0.20)

1.4 (0.35)

$13,000-$13,999

1.6 (0.10)

1.3 (0.11)

1.3 (0.37)

$14,000-$14,999

1.5 (0.09)

1.5 (0.12)

1.3 (0.31)

$15,000-$15,999

1.8 (0.09)

1.5 (0.10)

1.8 (0.39)

$16,000-$16,999

1.2 (0.08)

1.3 (0.11)

1.5 (0.32)

$17,000-$17,999

1.4 (0.07)

1.1 (0.09)

1.8 (0.41)

$18,000-$18,999

1.7 (0.10)

1.5 (0.12)

1.7 (0.38)

$19,000-$19,999

1.8 (0.11)

1.6 (0.15)

1.8 (0.38)

$20,000-$24,999

6.4 (0.20)

6.3 (0.27)

8.7 (0.85)

$25,000-$29,999

6.1 (0.23)

5.7 (0.25)

5.5 (0.68)

$30,000-$34,999

5.3 (0.19)

5.4 (0.22)

4.8 (0.72)

$35,000-$39,999

4.4 (0.17)

4.4 (0.24)

5.6 (0.78)

$40,000-$44,999

4.0 (0.16)

4.2 (0.23)

4.8 (0.79)

$45,000-$49,999

3.7 (0.14)

4.2 (0.23)

4.9 (0.77)

$50,000-$74,999

10.4 (0.25)

10.5 (0.37)

10.8 (1.08)

$75,000-$99,999

4.8 (0.18)

4.9 (0.28)

4.4 (0.74)

6.1 (0.26)

6.0 (0.37)

6.6 (1.21)
(continued)

$100,000 or More
See notes at end of table.

N-6

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

$20,000 or More

82.0 (0.34)

81.6 (0.51)

79.5 (1.53)

Less Than $20,000

18.0 (0.34)

18.4 (0.51)

20.5 (1.53)

Instrument Item
Before taxes and other deductions, was
the total combined family income
during [YEAR] more or less than
20,000 dollars? (QI22) 8

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test; SE = standard error.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being intervieweradministered to self- administered.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or
missing data.
5
Estimate is based on an edited version of the variable.
6
Estimate is an average based on valid responses to the relevant question(s). Respondents with unknown or missing
data were excluded.
7
The estimated mean includes zeroes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012.

N-7

N-8

Appendix O: Estimates and Standard Errors for New,
Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older
Type of
Change1

QFT Instrument Item
Race (QD05)

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Description of Change Estimate2,3 Error
Added response categories
for Guamanian or
Chamorro and Samoan.

White (QD051)

78.0

(1.93)

1,479

2,040

Black or African American (QD052)
American Indian or Alaska Native
(American Indian Includes North
American, Central American, and
South American Indians)
(QD053)

13.5

(1.63)

353

2,040

1.8

(0.42)

2,040

Native Hawaiian (QD054)

0.1

(0.06)

2,040

Guamanian or Chamorro (QD055)

0.0*

(0.00)

2,040

Samoan (QD056)

0.1

(0.09)

2,040

Other Pacific Islander (QD057)
Asian (Including: Asian, Indian,
Chinese, Filipino, Japanese,
Korean, and Vietnamese (QD058)

0.3

(0.11)

2,040

5.3

(0.89)

107

2,040

2.7

(0.49)

2,040

0.0*

(0.00)

2,044

7.5

(0.86)

2,044

10.8*

(2.88)

18.1*

(4.77)

20.9

(5.32)

45.4*

(5.96)

8.9*

(3.28)

8.4*

(3.21)

0.9*

(0.94)

5.4*

(2.71)

0.0*

(0.00)

Other (Specify) (QD059)
Are you currently serving full-time in
a Reserve component? (V2b)
Have you ever served on active duty
in the United States Armed Forces
or Reserve components? (QD10a)
When did you serve on active duty in
the United States Armed Forces or
Reserve components? (QD10b1)4
September 2001 or Later
(QD10b11)
August 1990 to August 2001
(Including Persian Gulf War)
(QD10b12)

N
N
N

Added two questions about
serving in reserve
components.
Added three questions
about active-duty U.S.
military service.
Added three questions
about active-duty U.S.
military service.

May 1975 to July 1990 (QD10b13)
Vietnam Era (August 1964 to
April 1975) (QD10b14)
February 1955 to July 1964
(QD10b15)
Korean War (July 1950 to
January 1955) (QD10b16)
January 1947 to June 1950
(QD10b17)
World War II (December 1941 to
December 1946) (QD10b18)
November 1941 or Earlier
(QD10b19)
See notes at end of table.

83
(continued)

O-1

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Did you ever serve on active duty in
the U.S. Armed Forces or Reserve
components in a military combat
zone or an area where you drew
imminent danger pay or hostile fire
pay? (QD10c) 4
What is the highest grade or year of
school you have completed?
(QD11)

Type of
Change1

Description of Change

Added three questions
about active-duty U.S.
military service.

Changed response
categories.

No Schooling

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3 Error

36.8*

(6.71)

0.1

(0.04)

2,044

1st Grade

0.0

(0.00)

2,044

2nd Grade

0.0

(0.03)

2,044

3rd Grade

0.0

(0.03)

2,044

4th Grade

0.4

(0.23)

2,044

5th Grade

0.4

(0.16)

2,044

6th Grade

1.9

(0.28)

2,044

7th Grade

2.9

(0.41)

113

2,044

8th Grade

3.4

(0.43)

113

2,044

9th Grade

2.9

(0.38)

105

2,044

10th Grade

3.3

(0.42)

119

2,044

11th Grade

3.9

(0.49)

132

2,044

20.0

(1.53)

351

2,044

12th Grade, No Diploma

1.9

(0.42)

2,044

GED Certificate

4.0

(0.58)

2,044

Regular High School Diploma

Some College, No Degree

19.5

(1.18)

382

2,044

Associate's Degree

9.4

(0.86)

149

2,044

Bachelor's Degree

16.5

(1.61)

235

2,044

Master's Degree

7.1

(0.87)

2,044

Doctorate Degree (e.g., PhD)
Professional Degree Beyond
Bachelor's Degree (e.g., MD)

1.1

(0.32)

2,044

1.4

(0.36)

2,044

Previously served as a proxy for
another respondent? (PREVCOM)

Added two questions to
determine if R had
previously served as a
proxy.

Yes

10.5

(1.69)

766

57.5

(1.87)

1,276

1,969

0.1

(0.09)

695

0.0*

(0.00)

I am not sure
Previously completed any part of this
interview yourself, including
answering questions on behalf of a
member of your household?
(PREVCOM2)4

Added two questions to
determine if R had
previously served as a
proxy.

See notes at end of table.

2
(continued)

O-2

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Use of "smokeless" tobacco such as
snuff, dip, chewing tobacco, or
"snus." (CG25)
How old were you the first time you
used "smokeless" tobacco? (CG26)5
How long has it been since you last
used, have you used "smokeless"
tobacco? (CG27and CG28)

Type of
Change1

Description of Change

Edited to include all forms
of smokeless tobacco.
Edited to include all forms
of smokeless tobacco.

Edited to include all forms
of smokeless tobacco.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3 Error
17.4

(1.06)

332

2,043

18.3

(0.68)

N/A

332

Within the past 30 days
More than 30 days ago but within
the past 12 months

5.2

(0.57)

2,042

1.6

(0.31)

2,042

More than 12 months ago

1.5

(0.28)

2,042

9.1

(0.82)

146

2,042

24.0

(1.19)

503

2,024

1.4

(0.30)

2,042

0.6

(0.18)

2,041

2.4

(0.45)

2,041

0.0

(0.04)

2,041

0.3

(0.14)

2,041

1.0

(0.25)

2,041

More than 3 years ago
During the past 30 days, did you have
[Insert #] or more drinks on the
same occasion? (AL08)6

Ever used Ketamine (LS01i)

Ever used DMT, AMT, or Foxy
(LS01j)

Ever used Salvia divinorum (LS01k)

How long has it been since you last
used Ketamine? (LS33)

Changed question wording
for women to "4 or more
drinks."
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added 3 questions to
measure Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum use.
Added these items to
measure time since last use
of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.

Within the past 30 days
More than 30 days ago but within
the past 12 months
More than 12 months ago

How long has it been since you last
used DMT, AMT, or Foxy? (LS34)

Added these items to
measure time since last use
of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.

Within the past 30 days
More than 30 days ago but within
the past 12 months

0.1

(0.04)

2,040

0.2

(0.10)

2,040

More than 12 months ago

0.3

(0.14)

See notes at end of table.

2,040
(continued)

O-3

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item

How long has it been since you last
used Salvia divinorum? (LS35)

Type of
Change1

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Description of Change Estimate2,3 Error
Added these items to
measure time since last use
of Ketamine,
DMT/AMT/Foxy, and
Salvia divinorum.

Within the past 30 days
More than 30 days ago but within
the past 12 months
More than 12 months ago
Have you ever, inhaled felt-tip pens,
felt-tip markers, or magic markers for
kicks or to get high? (IN01h1)
Have you ever inhaled computer
keyboard cleaner, also known as air
duster, for kicks or to get high?
(IN01ii)
Have you ever used methamphetamine?
(ME01)
How old were you the first time you
used methamphetamine? (ME02)5
How long has it been since you last used
methamphetamine? (MELAST3)

N
N
N
N

Added question to measure
use of felt-tip pens, felt-tip
markers, or magic markers.
Added question to measure
use computer keyboard
cleaner, also known as air
duster.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.

0.1

(0.08)

2,041

0.3

(0.12)

2,041

2.1

(0.36)

2,041

3.3

(0.35)

105

2,041

1.2

(0.25)

2,042

6.5

(0.83)

112

2,043

20.7

(0.63)

N/A

112

Within the past 30 days
More than 30 days ago but within
the past 12 months

0.4

(0.16)

2,043

0.1

(0.07)

2,043

More than 12 months ago
How many days you’ve used
methamphetamine during the past 12
months. (MEFRAME3, MEYRAVE,
MEMONAVE, MEWKAVE)5
During the past 30 days, on how many
days did you use methamphetamine?
(ME06)5
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR01)

6.0

(0.79)

100

2,043

161.2

(45.87)

N/A

(4.51)

N/A

Vicodin

N
N
N

Added to measure use of
methamphetamine.
Added to measure use of
methamphetamine.
Added questions to
indicate use of prescription
pain relievers.

17.7*

12.9

(1.18)

242

2,029

Lortab®

5.5

(0.70)

103

2,029

Lorcet®

1.1

(0.25)

2,029

14.4

(1.17)

264

Hydrocodone
See notes at end of table.

2,029
(continued)

O-4

Type of
Change1
N

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Description of Change Estimate2,3 Error
Added questions to indicate
use of prescription pain
relievers.

OxyContin

2.4

(0.35)

2,026

6.5

(0.83)

128

2,026

0.4

(0.15)

2,026

0.3

(0.13)

2,026

6.8

(0.92)

128

2,026

1.6

(0.41)

2,027

0.5

(0.29)

2,027

0.2

(0.11)

2,027

2.1

(0.55)

2,028

0.4

(0.23)

2,028

Ultracet

0.3

(0.15)

2,028

Ryzolt®

0.0

(0.02)

2,028

4.5

(0.56)

2,028

10.9

(0.98)

233

2,025

1.6

(0.30)

2,025

Avinza

0.1

(0.11)

2,030

Kadian®

0.1

(0.05)

2,030

(0.06)

2,030

(0.00)

2,030

(0.54)

Percocet

Percodan

Tylox

Oxycodone
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR03)

Added questions to indicate
use of prescription pain
relievers.

Darvocet®
Darvon

Propoxyphene
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR04)

Added questions to indicate
use of prescription pain
relievers.

Ultram

Ultram ER
®

Tramadol
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR05)

Added questions to indicate
use of prescription pain
relievers.

Tylenol with Codeine 3 or 4
Codeine Pills
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR06)

Added questions to indicate
use of prescription pain
relievers.

MS Contin

0.1

Oramorph SR

0.0

Morphine

3.7

See notes at end of table.

2,030
(continued)

O-5

Type of
Change1
N

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Description of Change Estimate2,3 Error
Added questions to indicate
use of prescription pain
relievers.

Actiq

0.1

(0.11)

2,029

0.1

(0.05)

2,029

0.0

(0.04)

2,029

0.7

(0.23)

2,029

Suboxone

0.7

(0.23)

2,029

Subutex®

0.3

(0.11)

2,029

0.0

(0.04)

2,029

0.7

(0.15)

2,028

0.9

(0.23)

2,028

0.6

(0.17)

2,028

0.1

(0.06)

2,028

0.2

(0.08)

2,028

0.0*

(0.00)

2,028

0.0

(0.03)

2,028

0.0

(0.03)

2,028

8.7

(0.81)

178

2,027

66.8

(1.61)

1,158

2,017

4.7

(0.67)

100

2,037

Xanax XR

0.4

(0.15)

2,037

Alprazolam

1.5

(0.34)

2,037

Extended-Release Alprazolam

0.4

(0.24)

Duragesic
®

Fentora

Fentanyl
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR08)

Added questions to indicate
use of prescription pain
relievers.

Buprenorphine
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR09)

Added questions to indicate
use of prescription pain
relievers.

Demerol

Dilaudid

Methadone
®

Opana

Opana® ER
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR10)

Added questions to indicate
use of prescription pain
relievers.

Talacen

Talwin

Talwin NX
In the past 12 months, have you used
any other prescription pain
reliever? (PR11)
Have you ever used any prescription
pain reliever? (PR12)
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR01)
Xanax

N
N
N

Added questions to indicate
use of prescription pain
relievers.
Added questions to indicate
use of prescription pain
relievers.
Added questions to indicate
use of prescription
tranquilizers.

See notes at end of table.

2,037
(continued)

O-6

Type of
Change1
N

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Description of Change Estimate2,3 Error
Added questions to indicate
use of prescription
tranquilizers.

1.2

(0.31)

2,037

Klonopin

1.1

(0.26)

2,037

Lorazepam

2.0

(0.32)

2,037

2.0

(0.40)

2,037

Valium

1.9

(0.41)

2,037

Diazepam

1.0

(0.27)

2,037

0.1

(0.07)

2,037

0.0

(0.03)

2,037

0.1

(0.05)

2,037

4.2

(0.59)

2,037

1.4

(0.33)

2,037

0.4

(0.20)

2,037

0.6

(0.24)

2,037

0.0

(0.02)

2,037

1.7

(0.35)

2,037

25.7

(1.54)

413

Clonazepam
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR03)

Added questions to indicate
use of prescription
tranquilizers.

Librium

Tranxene
Oxazepam (also known as
Serax®)
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR04)

Added questions to indicate
use of prescription
tranquilizers.

®`

Flexeril
®`

Soma
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR05)
Buspirone (also known as
BuSpar®)
Hydroxyzine (also known as
Atarax® or Vistaril®)
Meprobamate
In the past 12 months, have you used
any other prescription tranquilizer?
(TR06)
Have you ever, even once, used any
prescription tranquilizer?
(TR07)

N
N

Added questions to indicate
use of prescription
tranquilizers.

Added questions to indicate
use of prescription
tranquilizers.
Added questions to indicate
use of prescription
tranquilizers.

See notes at end of table.

2,033
(continued)

O-7

Type of
Change1 Description of Change
Added questions to
indicate use of
N
prescription stimulants.

Adderall®
®

Adderall XR
Dexedrine

Dextroamphetamine
AmphetamineDextroamphetamine
Combinations
In the past 12 months, which, if any,
of these stimulants have you used?
(ST02)

2.2

(0.37)

2,038

1.2

(0.23)

2,038

0.3

(0.11)

2,038

0.2

(0.10)

2,038

0.8

(0.27)

2,038

0.5

(0.14)

2,038

Added questions to
indicate use of
prescription stimulants.

Ritalin®
®

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Ritalin SR or Ritalin LA

0.3

(0.10)

2,038

0.6

(0.15)

2,038

0.0

(0.02)

2,038

0.4

(0.13)

2,038

Metadate CD

0.0

(0.02)

2,038

Metadate® ER

0.1

(0.06)

2,038

0.2

(0.10)

2,038

0.3

(0.13)

2,038

0.2

(0.10)

2,038

Benzphetamine

0.0

(0.03)

2,038

0.0

(0.03)

2,038

Diethylpropion

0.0

(0.02)

2,038

Phendimetrazine

0.2

(0.15)

2,038

0.8

(0.23)

2,038

0.1

(0.06)

2,038

Concerta

Daytrana

Methylphenidate
In the past 12 months, which, if any,
of these stimulants have you used?
(ST03)

Added questions to
indicate use of
prescription stimulants.

Focalin

Focalin XR
Dexmethylphenidate
In the past 12 months, which, if any,
of these stimulants have you used?
(ST04)
Didrex

Phentermine
In the past 12 months, which, if any,
of these stimulants have you used?
(ST05)

Added questions to
indicate use of
prescription stimulants.

Provigil

Tenuate

0.0

Vyvanse®
In the past 12 months, have you used
any other prescription stimulant?
(ST06)

Have you ever, even once, used any
prescription stimulant? (ST07)

Added questions to
indicate use of
prescription stimulants.
Added questions to
indicate use of
prescription stimulants.

See notes at end of table.

(0.00)

2,038

0.7

(0.23)

2,038

1.1

(0.25)

2,037

11.5

(0.95)

249

2,035
(continued)

O-8

Type of
Change1 Description of Change
Added questions to
indicate use of
N
prescription sedatives.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Ambien®

4.5

(0.63)

2,037

Ambien CR

0.7

(0.22)

2,037

Zolpidem

1.6

(0.46)

2,037

0.1

(0.07)

2,037

Lunesta

1.1

(0.30)

2,038

Sonata®

0.5

(0.24)

2,038

(0.00)

2,038

0.0*

(0.00)

2,038

0.2

(0.18)

2,038

0.0

(0.00)

2,038

0.2

(0.11)

2,038

0.1

(0.07)

2,038

0.6

(0.25)

2,038

0.0

(0.03)

2,038

0.1

(0.07)

2,038

0.2

(0.15)

2,038

1.2

(0.27)

2,038

16.2

(1.30)

240

2,033

11.8

(0.94)

259

2,013

2.4

(0.44)

2,034

23.9

(2.11)

N/A

Extended-Release Zolpidem
In the past 12 months, which, if any,
of these sedatives have you used?
(SV02)

Added questions to
indicate use of
prescription sedatives.

Zaleplon
In the past 12 months, which, if any,
of these sedatives have you used?
(SV03)

0.0
N

Added questions to
indicate use of
prescription sedatives.

Dalmane
®

Halcion

Flurazepam
Triazolam
In the past 12 months, which, if any,
of these sedatives have you used?
(SV04)

Added questions to
indicate use of
prescription sedatives.

Restoril

Temazepam
In the past 12 months, which, if any,
of these sedatives have you used?
(SV05)

Added questions to
indicate use of
prescription sedatives.

Butisol

Seconal

Phenobarbital
In the past 12 months, have you used
any other prescription sedative?
(SV06)
Have you ever used any prescription
sedative? (SV07)
Have you ever, even once, used any
prescription pain reliever in any
way a doctor did not direct you to
use it? (PRL01 and PRL02)
In the past 12 months, did you use
Vicodin in any way a doctor did
not direct you to use it? (PRY01)
How old were you when you first
used Vicodin in a way a doctor did
not direct you to use it? (PRY01a)5

N
N

N
N
N

Added questions to
indicate use of
prescription sedatives.
Added questions to
indicate use of
prescription sedatives.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

See notes at end of table.

58
(continued)

O-9

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
Lortab in a way a doctor did not
direct you to use it? (PRY02)
How old were you when you first
used Lortab in a way a doctor did
not direct you to use it?
(PRY02a)5
In the past 12 months, did you use
Lorcet in any way a doctor did not
direct you to use it? (PRY03)
How old were you when you first
used Lorcet in a way a doctor did
not direct you to use it?
(PRY03a)5
In the past 12 months, did you use
hydrocodone in any way a doctor
did not direct you to use it?
(PRY04)
How old were you when you first
used hydrocodone in a way a
doctor did not direct you to use it?
(PRY04a)5
In the past 12 months, did you use
OxyContin in any way a doctor
did not direct you to use it?
(PRY05)
How old were you when you first
used OxyContin in a way a doctor
did not direct you to use it?
(PRY05a)5
In the past 12 months, did you use
Percocet in any way a doctor did
not direct you to use it? (PRY06)
How old were you when you first
used Percocet in a way a doctor
did not direct you to use
it? (PRY06a)5
In the past 12 months, did you use
Percodan in any way a doctor did
not direct you to use it? (PRY07)
How old were you when you first
used Percodan in a way a doctor
did not direct you to use it?
(PRY07a)5
In the past 12 months, did you use
Tylox in any way a doctor did not
direct you to use it? (PRY08)

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

Added questions to
indicate misuse of
prescription pain relievers.

16.6*

(2.06)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

1.9

(0.35)

2,033

Added questions to
indicate misuse of
prescription pain relievers.

25.1

(2.48)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.8

(0.20)

2,033

20.8

(1.98)

N/A

1.0

(0.23)

2,032

23.2

(2.27)

N/A

0.2

(0.08)

2,033

19.6*

(2.46)

N/A

0.0

(0.03)

N
N

See notes at end of table.

23.3

(2.53)

N/A

0.3

(0.11)

2,034

2,033
(continued)

O-10

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used Tylox in a way a doctor did
not direct you to use it?
(PRY08a)5
In the past 12 months, did you use
oxycodone in any way a doctor
did not direct you to use it?
(PRY09)
How old were you when you first
used oxycodone in a way a doctor
did not direct you to use it?
(PRY09a)5
In the past 12 months, did you use
Darvocet in a way a doctor did not
direct you to use it? (PRY10)5
How old were you when you first
used Darvocet in a way a doctor
did not direct you to use it?
(PRY10a)5
In the past 12 months, did you use
Darvon in any way a doctor did
not direct you to use it? (PRY11)
In the past 12 months, did you use
propoxyphene in any way a doctor
did not direct you to use it?
(PRY12)
In the past 12 months, did you use
Ultram in any way a doctor did
not direct you to use it? (PRY13)
How old were you when you first
used Ultram in a way a doctor did
not direct you to use it?
(PRY13a)5
In the past 12 months, did you use
Ultram ER in any way a doctor
did not direct you to use it?
(PRY14)
In the past 12 months, did you use
Ultracet in any way a doctor did
not direct you to use it? (PRY15)
How old were you when you first
used Ultracet in a way a doctor did
not direct you to use it?
(PRY15a)5
In the past 12 months, did you use
Ryzolt in any way a doctor did not
direct you to use it? (PRY16)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription pain relievers.

15.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

1.2

(0.27)

2,032

23.4

(1.73)

N/A

0.1

(0.07)

2,034

16.2*

(0.67)

N/A

0.0*

(0.00)

2,034

0.0*

(0.00)

2,034

0.5

(0.18)

2,033

33.3*

(5.80)

N/A

0.0*

(0.00)

2,034

0.1

(0.10)

2,034

33.6*

(11.61)

N/A

0.0*

(0.00)

N
N

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

Added questions to
indicate misuse of
prescription pain relievers.

N
N

See notes at end of table.

2,034
(continued)

O-11

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
tramadol in any way a doctor did
not direct you to use it? (PRY17)
How old were you when you first
used tramadol in a way a doctor
did not direct you to use it?
(PRY17a)5
In the past 12 months, did you use
Tylenol with codeine 3 or 4 in any
way a doctor did not direct you to
use it? (PRY18)
How old were you when you first
used Tylenol with codeine 3 or 4
in a way a doctor did not direct
you to use it? (PRY18a)5
In the past 12 months, did you use
codeine pills in any way a doctor
did not direct you to use them?
(PRY19)
How old were you when you first
used codeine pills in a way a
doctor did not direct you to use
them? (PRY19a)5
In the past 12 months, did you use
Avinza in any way a doctor did
not direct you to use it? (PRY20)
In the past 12 months, did you use
Kadian in any way a doctor did
not direct you to use it? (PRY21)
How old were you when you first
used Kadian in a way a doctor did
not direct you to use it?
(PRY21a)5
In the past 12 months, did you use
MS Contin in any way a doctor
did not direct you to use it?
(PRY22)
In the past 12 months, did you use
morphine in any way a doctor did
not direct you to use it? (PRY24)
How old were you when you first
used morphine in a way a doctor
did not direct you to use it?
(PRY24a)5
In the past 12 months, did you use
Actiq in any way a doctor did not
direct you to use it? (PRY25)

Added questions to
indicate misuse of
prescription pain relievers.

26.4

(3.15)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

1.5

(0.27)

2,030

Added questions to
indicate misuse of
prescription pain relievers.

26.0

(4.59)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.3

(0.11)

2,031

17.2

(0.71)

N/A

0.0*

(0.00)

2,034

0.0

(0.03)

2,034

17.0*

(0.00)

N/A

0.0*

(0.00)

2,034

0.4

(0.14)

2,034

17.5

(1.49)

N/A

(0.00)

Added questions to
indicate misuse of
prescription pain relievers.

N
N

See notes at end of table.

0.0*

2,034
(continued)

O-12

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
Duragesic in any way a doctor did
not direct you to use it? (PRY26)
In the past 12 months, did you use
Fentora in any way a doctor did
not direct you to use it? (PRY27)
In the past 12 months, did you use
fentanyl in any way a doctor did
not direct you to use it? (PRY28)
How old were you when you first
used fentanyl in a way a doctor
did not direct you to use it?
(PRY28a)5
In the past 12 months, did you use
Suboxone in any way a doctor did
not direct you to use it? (PRY29)
How old were you when you first
used Suboxone in a way a doctor
did not direct you to use it?
(PRY29a)5
In the past 12 months, did you use
Subutex in any way a doctor did
not direct you to use it? (PRY30)
How old were you when you first
used Subutex in a way a doctor
did not direct you to use it?
(PRY30a)5
In the past 12 months, did you use
buprenorphine in any way a doctor
did not direct you to use it?
(PRY31)
How old were you when you first
used buprenorphine in a way a
doctor did not direct you to use it?
(PRY31a)5
In the past 12 months, did you use
Demerol in any way a doctor did
not direct you to use it? (PRY32)
How old were you when you first
used Demerol in a way a doctor
did not direct you to use it?
(PRY32a)5
In the past 12 months, did you use
Dilaudid in any way a doctor did
not direct you to use it? (PRY33)

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
Added questions to
indicate misuse of
N
prescription pain relievers.
0.0*
(0.00)
0
2,034
Added questions to
indicate misuse of
N
prescription pain relievers.
0.0*
(0.00)
0
2,034
Added questions to
indicate misuse of
N
prescription pain relievers.
0.1
(0.05)
2
2,034

N
N

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

22.1*

(2.83)

N/A

0.2

(0.10)

2,034

24.2*

(2.03)

N/A

0.1

(0.07)

2,034

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

Added questions to
indicate misuse of
prescription pain relievers.

23.5*

(0.65)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.0

(0.04)

2,034

17.0*

(0.00)

N/A

0.0

(0.04)

2,034

18.6*

(0.61)

N/A

0.3

(0.08)

N
N

See notes at end of table.

2,034
(continued)

O-13

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used Dilaudid in a way a doctor
did not direct you to use it?
(PRY33a)5
In the past 12 months, did you use
methadone in any way a doctor
did not direct you to use it?
(PRY34)
How old were you when you first
used methadone in a way a doctor
did not direct you to use it?
(PRY34a)5
In the past 12 months, did you use
Opana in any way a doctor did not
direct you to use it? (PRY35)
How old were you when you first
used Opana in a way a doctor did
not direct you to use it?
(PRY35a)5
In the past 12 months, did you use
Opana ER in any way a doctor did
not direct you to use it? (PRY36)
How old were you when you first
used Opana ER in a way a doctor
did not direct you to use it?
(PRY36a)5
In the past 12 months, did you use
Talwin in any way a doctor did
not direct you to use it? (PRY38)
How old were you when you first
used Talwin in a way a doctor did
not direct you to use it?
(PRY38a)5
In the past 12 months, did you use
Talwin NX in any way a doctor
Did not direct you to use it?
(PRY39)
In the past 12 months, did you use
any other prescription pain
reliever in a way a doctor did not
direct you to use it? (PRY40)
How old were you when you first
used any other prescription pain
reliever in a way a doctor did not
direct you to use it? (PRY40a)5
In the past 30 days, did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it?
(PRM01)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription pain relievers.

21.5*

(2.42)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.3

(0.11)

2,034

20.9*

(2.42)

N/A

0.1

(0.05)

2,034

16.2*

(1.16)

N/A

0.1

(0.05)

2,034

17.7*

(0.24)

N/A

0.0

(0.02)

2,034

N
N

Added questions to
indicate misuse of
prescription pain relievers.
Added questions to
indicate misuse of
prescription pain relievers.

Added questions to
indicate misuse of
prescription pain relievers.

13.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.0*

(0.00)

2,034

Added questions to
indicate misuse of
prescription pain relievers.

0.2

(0.09)

2,030

Added questions to
indicate misuse of
prescription pain relievers.

20.6*

(2.46)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

2.0

(0.36)

See notes at end of table.

2,025
(continued)

O-14

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 30 days, on how
many days did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it?
(PRM02)5
During the past 30 days, did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (PRM03)
Which of these statements describe
your use of [PRNAMEFILL] at
any time in the past 12 months?
(PRY41)4
I used [PRNAMEFILL] without
a prescription of my own.
I used [PRNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [PRNAMEFILL] more
often than it was/they were
prescribed.
I used [PRNAMEFILL] for
longer than it was/they were
prescribed.
I used [PRNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
What were the reasons you used
[PRLASTFILL2] that time?
(PRYMOTIV)4

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription pain relievers.

8.2

(1.35)

N/A

Added questions to
indicate misuse of
prescription pain relievers.

0.7

(0.21)

2,025

Added questions to
indicate misuse of
prescription pain relievers.
67.4

(4.48)

149

23.1

(4.43)

149

20.2

(4.31)

149

12.5

(3.27)

149

23.0

(4.18)

149

To relieve physical pain

70.2

(4.36)

144

To relax or relieve tension
To experiment or to see what it's/
they're like

26.1

(4.52)

144

8.1

(3.08)

144

To feel good or get high

22.3

(4.19)

144

To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason

14.5

(2.98)

144

9.3

(3.24)

144

2.0

(1.29)

144

1.6

(1.11)

144

2.1*

(1.54)

Added questions to
indicate misuse of
prescription pain relievers.

See notes at end of table.

144
(continued)

O-15

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
Added questions to
indicate misuse of
N
prescription pain relievers.

To relieve physical pain
To relax or relieve tension
To experiment or to see what it's/
they're like

(8.73)

(7.47)

0.0*

(0.00)

(6.90)

(6.67)

8.3*

(4.85)

0.0*

(0.00)

4.8*

(3.88)

0.0

(0.00)

27.1

(4.59)

149

2.0*

(1.72)

149

0.2

(0.24)

149

45.5

(4.66)

149

11.3

(2.77)

149

4.0

(1.65)

149

5.5

(1.49)

149

4.3*

(2.59)

20.2

To feel good or get high

17.6

To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
Now think about the last time you
used [PRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[PRLASTFILL]? (PRY42B)4
I got a prescription for the
[PRLASTFILL] from just one
doctor
I got prescriptions for the
[PRLASTFILL] from more
than one doctor
I stole the [PRLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [PRLASTFILL] from a
friend or relative for free
I bought the [PRLASTFILL]
from a friend or relative
I took the [PRLASTFILL] from a
friend or relative without
asking
I bought the [PRLASTFILL]
from a drug dealer or other
stranger
I got the [PRLASTFILL] in some
other way

31.3*

17.8

Added "fill" and moved
from the noncore prior
substance use module.

See notes at end of table.

149
(continued)

O-16

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How did your friend or relative get
the [PRLASTFILL]? (PRY42C)4
He or she got a prescription for
the [PRLASTFILL] from just
one doctor
He or she got prescriptions for
the [PRLASTFILL] from more
than one doctor
He or she stole the
[PRLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[PRLASTFILL] from a friend
or relative for free
He or she bought the
[PRLASTFILL] from a friend
or relative
He or she took the
[PRLASTFILL] from a friend
or relative without asking
He or she bought the
[PRLASTFILL] from a drug
dealer or other stranger
He or she got the
[PRLASTFILL] in some other
way
Have you ever, even once, used any
prescription tranquilizer in any
way a doctor did not direct you to
use it? (TRL01 and TRL02)
In the past 12 months, did you use
Xanax in any way a doctor did not
direct you to use it? (TRY01)
How old were you when you first
used Xanax in a way a doctor did
not direct you to use it?
(TRY01a)5
In the past 12 months, did you use
Xanax XR in a way a doctor did
not direct you to use it? (TRY02)
How old were you when you first
used Xanax XR in a way a doctor
did not direct you to use it?
(TRY02a)5
In the past 12 months, did you use
alprazolam in any way a doctor
did not direct you to use it?
(TRY03)

Type of
Change1 Description of Change
Added "fill" and moved
from the noncore prior
R
substance use module.

N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

90.1*

(4.61)

0.0*

(0.00)

0.0*

(0.00)

2.4*

(1.76)

0.0*

(0.00)

1.1*

(1.08)

1.4*

(1.36)

5.1*

(3.99)

5.6

(0.77)

112

2,033

1.4

(0.27)

2,038

20.8

(1.47)

N/A

0.2

(0.11)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

24.9*

(6.18)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.3

(0.11)

See notes at end of table.

2,038
(continued)

O-17

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used alprazolam in a way a doctor
did not direct you to use it?
(TRY03a)5
In the past 12 months, did you use
extended-release alprazolam in
any way a doctor did not direct
you to use it? (TRY04)
How old were you when you first
used extended-release alprazolam
in a way a doctor did not direct
you to use it? (TRY04a)5
In the past 12 months, did you use
Ativan in any way a doctor did not
direct you to use it? (TRY05)
How old were you when you first
used Ativan in a way a doctor did
not direct you to use it?
(TRY05a)5
In the past 12 months, did you use
Klonopin in any way a doctor did
not direct you to use it? (TRY06)
How old were you when you first
used Klonopin in a way a doctor
did not direct you to use it?
(TRY06a)5
In the past 12 months, did you use
lorazepam in any way a doctor did
not direct you to use it? (TRY07)
How old were you when you first
used lorazepam in a way a doctor
did not direct you to use it?
(TRY07a)5
In the past 12 months, did you use
clonazepam in any way a doctor
did not direct you to use it?
(TRY08)
How old were you when you first
used clonazepam in a way a doctor
did not direct you to use it?
(TRY08a)5
In the past 12 months, did you use
Valium in any way a doctor did
not direct you to use it? (TRY09)
How old were you when you first
used Valium in a way a doctor did
not direct you to use it?
(TRY09a)5

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription tranquilizers.

20.9

(3.54)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.0

(0.02)

2,038

13.0*

(0.00)

N/A

0.2

(0.07)

2,038

24.8*

(4.08)

N/A

0.5

(0.16)

2,038

18.7

(0.84)

N/A

0.4

(0.14)

2,038

N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

26.2

(4.11)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.2

(0.07)

2,038

17.6*

(1.69)

N/A

0.5

(0.16)

2,038

20.4

(2.44)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

See notes at end of table.

15
(continued)

O-18

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
Librium in any way a doctor did
not direct you to use it? (TRY10)
How old were you when you first
used Librium in a way a doctor
did not direct you to use it?
(TRY10a)5
In the past 12 months, did you use
Tranxene in any way a doctor did
not direct you to use it? (TRY11)
In the past 12 months, did you use
diazepam in any way a doctor did
not direct you to use it? (TRY12)
How old were you when you first
used diazepam in a way a doctor
did not direct you to use it?
(TRY12a)5
In the past 12 months, did you use
oxazepam, also known as Serax,
in any way a doctor did not direct
you to use it? (TRY13)
In the past 12 months, did you use
Flexeril in any way a doctor did
not direct you to use it? (TRY14)
How old were you when you first
used Flexeril in a way a doctor did
not direct you to use it?
(TRY14a)5
In the past 12 months, did you use
Soma in any way a doctor did not
direct you to use it? (TRY15)
How old were you when you first
used Soma in a way a doctor did
not direct you to use it?
(TRY15a)5
In the past 12 months, did you use
buspirone, also known as BuSpar,
in any way a doctor did not direct
you to use it? (TRY16)
How old were you when you first
used buspirone, also known as
BuSpar, in a way a doctor did not
direct you to use it? (TRY16a)5
In the past 12 months, did you use
hydroxyzine, also known as
Atarax or Vistaril, in any way a
doctor did not direct you to use it?
(TRY17)

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
Added questions to
indicate misuse of
N
prescription tranquilizers.
0.0
(0.02)
1
2,038

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

N
N

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

17.0*

(0.00)

N/A

0.0*

(0.00)

2,038

0.1

(0.07)

2,038

20.9*

(2.58)

N/A

0.0*

(0.00)

2,038

0.4

(0.13)

2,038

29.6

(4.17)

N/A

0.4

(0.11)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.
Added questions to
indicate misuse of
prescription tranquilizers.

Added questions to
indicate misuse of
prescription tranquilizers.

19.6

(1.11)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.0

(0.02)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.

13.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.0

(0.03)

See notes at end of table.

2,038
(continued)

O-19

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used hydroxyzine, also known as
Atarax or Vistaril, in a way a
doctor did not direct you to use it?
(TRY17a)5
In the past 12 months, did you use
meprobamate, also known as
Equanil or Miltown, in any way a
doctor did not direct you to use it?
(TRY18)
How old were you when you first
used meprobamate, also known as
Equanil or Miltown, in a way a
doctor did not direct you to use it?
(TRY18a)5
In the past 12 months, did you use
any other prescription tranquilizer
in a way a doctor did not direct
you to use it? (TRY19)
In the past 30 days, did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it?
(TRM01)
During the past 30 days, on how
many days did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it?
(TRM02)5
During the past 30 days, did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (TRM03)
Which of these statements describe
your use of [TRNAMEFILL] at
any time in the past 12 months?
(TRY20)4
I used [TRNAMEFILL] without
a prescription of my own.
I used [TRNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [TRNAMEFILL] more
often than it was/they were
prescribed.
I used [TRNAMEFILL] for
longer than it was/they were
prescribed.
I used [TRNAMEFILL] in some
other way a doctor did not
direct me to use it/them.

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription tranquilizers.

16.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.0

(0.02)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.

13.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.0*

(0.00)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.

0.9

(0.23)

2,038

Added questions to
indicate misuse of
prescription tranquilizers.

5.8

(1.49)

N/A

Added questions to
indicate misuse of
prescription tranquilizers.

0.3

(0.14)

2,037

Added questions to
indicate misuse of
prescription tranquilizers.

See notes at end of table.

78.7*

(5.47)

18.7*

(5.18)

6.9*

(2.97)

2.7*

(1.99)

9.9*

(3.22)

69
(continued)

O-20

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
What were the reasons you used
[TRLASTFILL2] that time?
(TRYMOTIV)4

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
Added questions to
indicate misuse of
N
prescription tranquilizers.

To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
Which was the main reason you
used [TRLASTFILL2] that time?
(TRYMOT1)4

(6.54)

11.1*

(4.00)

(5.63)

28.5

(7.38)

21.4*

(5.50)

9.5*

(4.49)

0.0*

(0.00)

2.1*

(2.11)

49.5*

(10.81)

5.5*

(5.28)

22.5

Added questions to
indicate misuse of
prescription tranquilizers.

To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high

8.5

To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
Now think about the last time you
used [TRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[TRLASTFILL]? (TRY21B)4
I got a prescription for the
[TRLASTFILL] from just one
doctor
I got prescriptions for the
[TRLASTFILL] from more
than one doctor
I stole the [TRLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [TRLASTFILL] from a
friend or relative for free
I bought the [TRLASTFILL]
from a friend or relative

65.7*

(4.87)

17.1*

(11.10)

13.1*

(6.59)

6.4*

(5.19)

0.0*

(0.00)

0.0

(0.00)

16.5*

(6.70)

0.0*

(0.00)

0.0*

(0.00)

53.7*

(6.74)

9.9*

(3.66)

Added "fill" and moved
from the noncore prior
substance use module.

See notes at end of table.

68
(continued)

O-21

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
I took the [TRLASTFILL] from
a friend or relative without
asking
I bought the [TRLASTFILL]
from a drug dealer or other
stranger
I got the [TRLASTFILL] in
some other way
How did your friend or relative get
the [TRLASTFILL]? (TRY21C)4
He or she got a prescription for
the [TRLASTFILL] from just
one doctor
He or she got prescriptions for
the[TRLASTFILL] from more
than one doctor
He or she stole the
[TRLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[TRLASTFILL] from a friend
or relative for free
He or she bought the
[TRLASTFILL] from a friend
or relative
He or she took the
[TRLASTFILL] from a friend
or relative without asking
He or she bought the
[TRLASTFILL] from a drug
dealer or other stranger
He or she got the
[TRLASTFILL] in some other
way
Have you ever, even once, used any
prescription stimulant in any way
a doctor did not direct you to use
it? (STL01 and STL02)
In the past 12 months, did you use
Adderall in any way a doctor did
not direct you to use it? (STY01)
How old were you when you first
used Adderall in a way a doctor
did not direct you to use it?
(STY01a)5
In the past 12 months, did you use
Adderall XR in any way a doctor
did not direct you to use it?
(STY02)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
12.5*

(5.42)

5.7*

(3.19)

1.9*

(1.94)

90.0*

(4.99)

2.7*

(2.72)

0.0*

(0.00)

2.1*

(2.06)

5.2*

(3.72)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

3.9

(0.58)

2,034

1.3

(0.28)

2,038

Added "fill" and moved
from the noncore prior
substance use module.

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

19.1

(0.57)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.6

(0.15)

See notes at end of table.

2,037
(continued)

O-22

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used Adderall XR in a way a
doctor did not direct you to use it?
(STY02a)5
In the past 12 months, did you use
Dexedrine in any way a doctor did
not direct you to use it? (STY03)
How old were you when you first
used Dexedrine in a way a doctor
did not direct you to use it?
(STY03a)5
In the past 12 months, did you use
dextroamphetamine in any way a
doctor did not direct you to use it?
(STY04)
How old were you when you first
used dextroamphetamine in a way
a doctor did not direct you to use
it? (STY04a)5
In the past 12 months, did you use
mixed amphetamine
dextroamphetamine pills other
than Adderall in any way a doctor
did not direct you to use them?
(STY05)
How old were you when you first
used mixed amphetamine
dextroamphetamine pills other
than Adderall in a way a doctor
did not direct you to use them?
(STY05a)5
In the past 12 months, did you use
Ritalin in any way a doctor did not
direct you to use it? (STY06)
How old were you when you first
used Ritalin in a way a doctor did
not direct you to use it? (STY06a)5
In the past 12 months, did you use
Ritalin SR or Ritalin LA in any
way a doctor did not direct you to
use it? (STY07)
How old were you when you first
used Ritalin SR or Ritalin LA in a
way a doctor did not direct you to
use it? (STY07a)5
In the past 12 months, did you use
Concerta in any way a doctor did
not direct you to use it? (STY08)
How old were you when you first
used Concerta in a way a doctor
did not direct you to use it?
(STY08a)5

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

17.6*

(0.44)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.1

(0.09)

2,038

Added questions to
indicate misuse of
prescription stimulants.

18.3*

(0.26)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.3

(0.12)

2,038

20.2*

(1.26)

N/A

0.2

(0.10)

2,038

26.3*

(6.68)

N/A

0.2

(0.08)

2,038

18.2*

(0.63)

N/A

0.2

(0.08)

2,038

17.5*

(0.79)

N/A

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

18.6

(0.79)

N/A

0.1

(0.08)

2,038

9
(continued)

O-23

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
Daytrana in any way a doctor did
not direct you to use it? (STY09)
How old were you when you first
used Daytrana in a way a doctor
did not direct you to use it?
(STY09a)5
In the past 12 months, did you use
methylphenidate in any way a
doctor did not direct you to use it?
(STY10)
How old were you when you first
used methylphenidate in a way a
doctor did not direct you to use it?
(STY10a)5
In the past 12 months, did you use
Metadate CD in any way a doctor
did not direct you to use it?
(STY11)
In the past 12 months, did you use
Metadate ER in any way a doctor
did not direct you to use it?
(STY12)
In the past 12 months, did you use
Focalin in any way a doctor did
not direct you to use it? (STY13)
How old were you when you first
used Focalin in a way a doctor did
not direct you to use it? (STY13a)5
In the past 12 months, did you use
Focalin XR in any way a doctor
did not direct you to use it?
(STY14)
How old were you when you first
used Focalin XR in a way a doctor
did not direct you to use it?
(STY14a)5
In the past 12 months, did you use
dexmethylphenidate in any way a
doctor did not direct you to use it?
(STY15)
How old were you when you first
used dexmethylphenidate in a way
a doctor did not direct you to use
it? (STY15a)5
In the past 12 months, did you use
benzphetamine in any way a
doctor did not direct you to use it?
(STY16)
In the past 12 months, did you use
Didrex in any way a doctor did not
direct you to use it? (STY17)

Type of
Change1 Description of Change
Added questions to
indicate misuse of
N
prescription stimulants.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
0.0

(0.02)

2,038

Added questions to
indicate misuse of
prescription stimulants.

19.6*

(2.47)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.1

(0.09)

2,038

Added questions to
indicate misuse of
prescription stimulants.

30.1*

(11.21)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.0*

(0.00)

2,038

0.0*

(0.00)

2,038

0.1

(0.05)

2,038

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

17.7*

(1.05)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.1

(0.05)

2,038

Added questions to
indicate misuse of
prescription stimulants.

17.3*

(0.45)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.1

(0.05)

2,038

Added questions to
indicate misuse of
prescription stimulants.

17.4*

(0.92)

N/A

0.0*

(0.00)

2,038

0.0*

(0.00)

N
N

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

See notes at end of table.

2,038
(continued)

O-24

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
diethylpropion in any way a
doctor did not direct you to use it?
(STY18)
How old were you when you first
used diethylpropion in a way a
doctor did not direct you to use it?
(STY18a)5
In the past 12 months, did you use
phendimetrazine in any way a
doctor did not direct you to use it?
(STY19)
In the past 12 months, did you use
phentermine in any way a doctor
did not direct you to use it?
(STY20)
How old were you when you first
used phentermine in a way a
doctor did not direct you to use it?
(STY20a)5
In the past 12 months, did you use
Provigil in any way a doctor did
not direct you to use it? (STY21)
In the past 12 months, did you use
Tenuate in any way a doctor did
not direct you to use it? (STY22)
In the past 12 months, did you use
Vyvanse in any way a doctor did
not direct you to use it? (STY23)
How old were you when you first
used Vyvanse in a way a doctor
did not direct you to use it?
(STY23a)5
In the past 12 months, did you use
any other prescription stimulant in
a way a doctor did not direct you
to use it? (STY24)
How old were you when you first
used any other prescription
stimulant in a way a doctor did not
direct you to use it? (STY24a)5
In the past 30 days, did you use
[STNAMEFILL] in any way a
doctor did not direct you to use it?
(STM01)
During the past 30 days, on how
many days did you use
[STNAMEFILL’] in any way a
doctor did not direct you to use it?
(STM02)5

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription stimulants.

0.0

(0.02)

2,038

Added questions to
indicate misuse of
prescription stimulants.

12.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.0*

(0.00)

2,038

Added questions to
indicate misuse of
prescription stimulants.

0.0

(0.03)

2,038

21.4*

(1.06)

N/A

0.0*

(0.00)

2,038

0.0*

(0.00)

2,038

0.2

(0.09)

2,037

Added questions to
indicate misuse of
prescription stimulants.

17.9*

(0.64)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.1

(0.07)

2,038

Added questions to
indicate misuse of
prescription stimulants.

20.8*

(1.17)

N/A

Added questions to
indicate misuse of
prescription stimulants.

0.5

(0.13)

2,037

10.1

(3.53)

N/A

N
N
N

See notes at end of table.

16
(continued)

O-25

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 30 days, did you use
[STNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (STM03)
Which of these statements describe
your use of [STNAMEFILL] at
any time in the past 12 months?
(STY25)4
I used [STNAMEFILL] without
a prescription of my own.
I used [STNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [STNAMEFILL] more
often than it was/they were
prescribed.
I used [STNAMEFILL] for
longer than it was/they were
prescribed.
I used [STNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
At any time in the past 12 months,
did you ever use a needle to inject
[STNAMEFILL]? (STY25a)
How long has it been since you last
used a needle to inject
[STNAMEFILL]? (STY25b)
Within the past 30 days
More than 30 days ago but
within the past 12 months
What were the reasons you used
[STLASTFILL2] that time?
(STYMOTIV)4

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

0.2

N
N

2,037

Added questions to
indicate misuse of
prescription stimulants.

Added questions to
indicate misuse of
prescription stimulants.
Added questions to
indicate misuse of
prescription stimulants.

81.2*

(5.72)

22.1*

(6.70)

12.0*

(5.23)

9.6*

(5.40)

14.0*

(4.52)

0.0*

(0.00)

2,037

(0.00)

2,037

0.0*

(0.00)

2,037

8.1*

(3.68)

46.8*

(8.71)

52.1*

(6.20)

39.0

(9.40)

13.0*

(4.25)

(6.19)

0.0*

(0.00)

0.0*

(0.00)

5.1*

(3.02)

0.0

(0.09)

Added questions to
indicate misuse of
prescription stimulants.

To help me lose weight
To help me concentrate
To help me be alert or stay
awake
To help me study
To experiment or to see what it's
like
To feel good or get high
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason

19.5

See notes at end of table.

56
(continued)

O-26

Type of
Change1 Description of Change
Added questions to
indicate misuse of
N
prescription stimulants.

6.4*

To help me lose weight

(4.60)

24.1

(11.13)

14.2*

(8.29)

(14.35)

2.4*

(2.40)

(5.57)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

8.4*

(3.83)

3.3*

(3.17)

0.0*

(0.00)

60.1*

(7.16)

14.1*

(4.70)

2.9*

(2.04)

5.9*

(3.92)

5.2*

(4.14)

79.9*

(7.41)

0.0*

(0.00)

To help me concentrate
To help me be alert or stay
awake

To help me study
To experiment or to see what it's
like

45.8

To feel good or get high
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
How did you get the
[STLASTFILL]? (STY26b)4
I got a prescription for the
[STLASTFILL] from just one
doctor
I got prescriptions for the
[STLASTFILL] from more
than one doctor
I stole the [STLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [STLASTFILL] from a
friend or relative for free
I bought the [STLASTFILL]
from a friend or relative
I took the [STLASTFILL] from a
friend or relative without
asking
I bought the [STLASTFILL]
from a drug dealer or other
stranger
I got the [STLASTFILL] in some
other way
How did your friend or relative get
the [STLASTFILL]? (STY26c)4
He or she got a prescription for
the [STLASTFILL] from just
one doctor
He or she got prescriptions for
the [STLASTFILL] from more
than one doctor

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

7.2

Added "fill" and moved
from the noncore prior
substance use module.

See notes at end of table.

28
(continued)

O-27

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
He or she stole the
[STLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[STLASTFILL] from another
friend or relative for free
He or she bought the
[STLASTFILL] from another
friend or relative
He or she took the
[STLASTFILL] from another
friend or relative without
asking
He or she bought the
[STLASTFILL] from a drug
dealer or other stranger
He or she got the
[STLASTFILL] in some other
Way
Have you ever, even once, used any
prescription sedative in any way a
doctor did not direct you to use it?
(SVL01 and SVL02)
In the past 12 months, did you use
Ambien in any way a doctor did
not direct you to use it? (SVY01)
How old were you when you first
used Ambien in a way a doctor
did not direct you to use it?
(SVY01a)5
In the past 12 months, did you use
Ambien CR in a way a doctor did
not direct you to use it? (SVY02)
How old were you when you first
used Ambien CR in a way a
doctor did not direct you to use it?
(SVY02a)5
In the past 12 months, did you use
zolpidem in any way a doctor did
not direct you to use it? (SVY03)
How old were you when you first
used zolpidem in a way a doctor
did not direct you to use it?
(SVY03a)5
In the past 12 months, did you use
extended-release zolpidem in any
way a doctor did not direct you to
use it? (SVY04)
In the past 12 months, did you use
Lunesta in any way a doctor did
not direct you to use it? (SVY05)

Type of
Change1 Description of Change

N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

See notes at end of table.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

0.0*

(0.00)

0.0*

(0.00)

6.0*

(3.45)

2.7*

(2.76)

6.5*

(4.89)

4.8*

(4.60)

3.4

(0.56)

2,033

0.4

(0.15)

2,039

24.8

(2.55)

N/A

0.0

(0.02)

2,039

18.9*

(2.12)

N/A

0.4

(0.18)

2,039

45.4*

(7.55)

N/A

0.0*

(0.00)

2,039

0.1

(0.09)

2,039
(continued)

O-28

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when you first
used Lunesta in a way a doctor did
not direct you to use it?
(SVY05a)5
In the past 12 months, did you use
Sonata in any way a doctor did not
direct you to use it? (SVY06)
How old were you when you first
used Sonata in a way a doctor did
not direct you to use it?
(SVY06a)5
In the past 12 months, did you use
zaleplon in any way a doctor did
not direct you to use it? (SVY07)
In the past 12 months, did you use
Dalmane in any way a doctor did
Not direct you to use it? (SVY08)
In the past 12 months, did you use
Halcion in any way a doctor did
not direct you to use it? (SVY09)
In the past 12 months, did you use
triazolam in any way a doctor did
not direct you to use it? (SVY11)
In the past 12 months, did you use
Restoril in any way a doctor did
not direct you to use it? (SVY12)
How old were you when you first
used Restoril in a way a doctor
did not direct you to use it?
(SVY12a)5
In the past 12 months, did you use
temazepam in any way a doctor
did not direct you to use it?
(SVY13)
In the past 12 months, did you use
Butisol in any way a doctor did
not direct you to use it? (SVY14)
How old were you when you first
used Butisol in a way a doctor did
not direct you to use it?
(SVY14a)5
In the past 12 months, did you use
Seconal in any way a doctor did
Not direct you to use it? (SVY15)
In the past 12 months, did you use
phenobarbital in any way a doctor
did not direct you to use it?
(SVY16)
How old were you when you first
used phenobarbital in a way a
doctor did not direct you to use it?
(SVY16a)5

Type of
Change1 Description of Change

N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

N
N
N
N
N

N
N

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

57.0*

(12.65)

N/A

0.1

(0.06)

2,039

16.0*

(0.00)

N/A

0.0*

(0.00)

2,039

0.0*

(0.00)

2,039

0.0*

(0.00)

2,039

0.0*

(0.00)

2,039

0.1

(0.07)

2,039

16.2*

(0.22)

N/A

0.0*

(0.00)

2,039

0.0

(0.03)

2,039

17.0*

(0.00)

N/A

0.0*

(0.00)

2,039

Added questions to
indicate misuse of
prescription sedatives.
Added questions to
indicate misuse of
prescription sedatives.

Added questions to
indicate misuse of
prescription sedatives.

0.0

(0.02)

Added questions to
indicate misuse of
prescription sedatives.

20.0*

(0.00)

N/A

See notes at end of table.

1
(continued)

O-29

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
In the past 12 months, did you use
any other prescription sedative in
a way a doctor did not direct you
to use it? (SVY17)
How old were you when you first
used any other prescription
sedative in a way a doctor did not
direct you to use it? (SVY17a)5
In the past 30 days, did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it?
(SVM01)
During the past 30 days, on how
Many days did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it?
(SVM02)5
During the past 30 days, did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (SVM03)
Which of these statements describe
your use of [SVNAMEFILL] at
any time in the past 12 months?
(SVY18)4
I used [SVNAMEFILL] without
a prescription of my own.
I used [SVNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [SVNAMEFILL] more
often than it was/they were
prescribed
I used [SVNAMEFILL] for
longer than it was/they were
prescribed.
I used [SVNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
What were the reasons you used
[SVLASTFILL2] that time?
(SVYMOTIV)4

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added questions to
indicate misuse of
prescription sedatives.

0.0

(0.02)

2,038

Added questions to
indicate misuse of
prescription sedatives.

16.0*

(0.00)

N/A

Added questions to
indicate misuse of
prescription sedatives.

0.3

(0.15)

2,038

Added questions to
indicate misuse of
prescription sedatives.

11.2*

(5.80)

N/A

Added questions to
indicate misuse of
prescription sedatives.

0.1

(0.10)

2,038

Added questions to
indicate misuse of
prescription sedatives.

53.6*

(14.03)

22.7*

(12.04)

16.4*

(11.68)

0.0*

(0.00)

24.2*

(13.23)

Added questions to
indicate misuse of
prescription sedatives.

To relax or relieve tension
To experiment or to see what it's/
they're like

29.0

To feel good or get high

(13.13)

5.6*

(4.08)

(4.82)

(10.38)

2.0*

(1.88)

3.8*

(2.64)

9.3

To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug

75.0

See notes at end of table.

17
(continued)

O-30

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Because I am "hooked" or I have
to have it/them
The other reason I reported
Which was the main reason you
used [SVLASTFILL] that time?
(SVYMOT1)4

Type of
Change1 Description of Change

To feel good or get high

(0.00)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

(22.23)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

0.0

(0.00)

45.2*

(14.38)

0.0*

(0.00)

0.0*

(0.00)

38.8*

(13.62)

5.5*

(4.03)

0.0*

(0.00)

8.5*

(8.13)

1.9*

(1.88)

79.6*

(13.03)

5.0*

(5.18)

23.8

To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them

76.2

The other reason I reported

How did your friend or relative get
the [SVLASTFILL]? (SVY19C)4
He or she got a prescription for
the [SVLASTFILL] from just
one doctor
He or she got prescriptions for
the [SVLASTFILL] from more
than one doctor

0.0*
Added questions to
indicate misuse of
prescription sedatives.

To relax or relieve tension
To experiment or to see what it's/
they're like

How did you get the
[SVLASTFILL]? (SVY19B)4
I got a prescription for the
[SVLASTFILL] from just one
doctor
I got prescriptions for the
[SVLASTFILL] from more
than one doctor
I stole the [SVLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [SVLASTFILL] from a
friend or relative for free
I bought the [SVLASTFILL]
from a friend or relative
I took the [SVLASTFILL] from
a friend or relative without
asking
I bought the [SVLASTFILL]
from a drug dealer or other
stranger
I got the [SVLASTFILL] in
some other way

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Added "fill" and moved
from the noncore prior
substance use module.

See notes at end of table.

7
(continued)

O-31

Have you ever, even once, used a
needle to inject any drug that was
not prescribed for you? (SD15)
Was any of your marijuana use in
the past 12 months recommended
by a doctor? (MJMM)
Was all of your marijuana use in the
past 12 months recommended by a
doctor? (MJMM01)4
During the past 12 months, was
there a month or more when you
spent a lot of your time getting or
using methamphetamine?
(DRME01)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting
over the effects of the
methamphetamine you used?
(DRME02)
During the past 12 months, did you
try to set limits on how often or
how much methamphetamine you
would use? (DRME04)
Were you able to keep to the limits
you set, or did you often use
methamphetamine more than you
intended to? (DRME05)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

0.0*

(0.00)

15.4*

(11.58)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

0.0*

(0.00)

0.8

(0.26)

2,044

0.5

(0.16)

2,044

QFT SD15 is similar to
2012 SD10c, with edits to
the wording to ask about
any other drug and to
remove "only for the
experience or feeling that
it caused."
New medical marijuana
questions in blunts
module
New medical marijuana
questions in blunts
module

41.5*

(15.49)

New questions about
dependence and abuse of
methamphetamine

0.1

(0.07)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0*

(0.00)

2,043

New questions about
dependence and abuse of
methamphetamine

0.1

(0.04)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.02)

M
N

See notes at end of table.

2,043
(continued)

O-32

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 12 months, did you
need to use more
methamphetamine than you used
in order to get the effect you
wanted? (DRME06)
During the past 12 months, did you
notice that using the same amount
of methamphetamine had less
effect on you than it used to?
(DRME07)
During the past 12 months, did you
want to or try to cut down or stop
using methamphetamine?
(DRME08)
During the past 12 months, were
you able to cut down or stop
using methamphetamine every
time you wanted to or tried to?
(DRME09)
During the past 12 months, have you
felt kind of blue or down when
you cut down or stopped using
methamphetamine? (DRME10)
During the past 12 months, have you
felt kind of blue or down when
you cut down or stopped using
methamphetamine? (DRME10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped
using methamphetamine?
(DRME11)
During the past 12 months, did you
have 2 or more of these symptoms
at the same time that lasted for
longer than a day after you cut
back or stopped using
methamphetamine? (DRME12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of
methamphetamine? (DRME13)
Did you continue to use
methamphetamine even though
you thought it was causing you to
have problems with your
emotions, nerves, or mental
health? (DRME14)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.1

(0.06)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.1

(0.05)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.12)

2,043

New questions about
dependence and abuse of
methamphetamine

0.2

(0.11)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.03)

See notes at end of table.

2,043
(continued)

O-33

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of
methamphetamine? (DRME15)
Did you continue to use
methamphetamine even though
this was causing you to have
physical problems? (DRME16)
During the past 12 months, did
using methamphetamine cause
you to give up or spend less time
doing these types of important
activities? (DRME17)
During the past 12 months, did
using methamphetamine cause
you to have serious problems
either at home, work, or school?
(DRME18)
During the past 12 months, did you
regularly use methamphetamine
and then do something where
using methamphetamine might
have put you in physical danger?
(DRME19)
During the past 12 months, did
using methamphetamine cause
you to do things that repeatedly
got you in trouble with the law?
(DRME20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused
by your use of methamphetamine?
(DRME21)
Did you continue to use
methamphetamine even though
you thought it caused problems
with family or friends? (DRME22)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting or
using prescription stimulants?
(DRST01)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting
over the effects of the prescription
stimulants you used? (DRST02)
During the past 12 months, did you
try to set limits on how often or
how much prescription stimulants
you would use? (DRST04)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

New questions about
dependence and abuse of
methamphetamine

0.0*

(0.00)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0*

(0.00)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.02)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.02)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.03)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.02)

2,043

New questions about
dependence and abuse of
methamphetamine

0.1

(0.06)

2,043

New questions about
dependence and abuse of
methamphetamine

0.0

(0.02)

2,043

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.1

(0.06)

2,034

0.0*

(0.00)

2,034

0.5

(0.15)

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

See notes at end of table.

2,034
(continued)

O-34

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Were you able to keep to the limits
you set, or did you often use
prescription stimulants more than
you intended to? (DRST05)
During the past 12 months, did you
need to use more prescription
stimulants than you used to in
order to get the effect you wanted?
(DRST06)
During the past 12 months, did you
notice that using the same amount
of prescription stimulants had less
effect on you than it used to?
(DRST07)
During the past 12 months, did you
want to or try to cut down or stop
using prescription stimulants?
(DRST08)
During the past 12 months, were
you able to cut down or stop
Using prescription stimulants
every time you wanted to or tried
to? (DRST09)
During the past 12 months, did you
cut down or stop using
Prescription stimulants at least one
time? (DRST10)
During the past 12 months, have
you felt kind of blue or down
when you cut down or stopped
using prescription stimulants?
(DRST10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped
using prescription stimulants?
(DRST11)
During the past 12 months, did you
have 2 or more of these symptoms
at the same time that lasted for
longer than a day after you cut
back or stopped using prescription
stimulants? (DRST12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of prescription
stimulants? (DRST13)

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
Question text the same.
Universe edited to remove
meth users from these
R
stimulant questions.
0.4
(0.15)
14
2,034

0.3

(0.12)

2,034

0.1

(0.07)

2,034

0.5

(0.16)

2,034

0.4

(0.15)

2,034

0.3

(0.09)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.3

(0.11)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.3

(0.11)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.2

(0.08)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.2

(0.09)

See notes at end of table.

2,034
(continued)

O-35

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Did you continue to use prescription
stimulants even though you
thought this was causing you to
have problems with your
emotions, nerves, or mental
health? (DRST14)
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of prescription
stimulants? (DRST15)
Did you continue to use prescription
stimulants even though this was
causing you to have physical
problems? (DRST16)
During the past 12 months, did
using prescription stimulants
cause you to give up or spend less
time doing these types of
important activities? (DRST17)
During the past 12 months, did
using prescription stimulants
cause you to have serious
problems either at home, work, or
school? (DRST18)
During the past 12 months, did you
regularly use prescription
stimulants and then do something
where using prescription
stimulants might have put you in
physical danger? (DRST19)
During the past 12 months, did
using prescription stimulants
cause you to do things that
repeatedly got you in trouble with
the law? (DRST20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused
by your use of prescription
stimulants? (DRST21)
Did you continue to use prescription
stimulants even though you
thought this caused problems with
family or friends? (DRST22)

Type of
Change1 Description of Change

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

0.1

(0.08)

2,034

0.0

(0.04)

2,034

0.0

(0.04)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

(0.00)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0

(0.02)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

(0.00)

2,034

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

0.0*

(0.00)

2,034

0.0*

(0.00)

2,034

0.0*

(0.00)

Question text the same.
Universe edited to remove
meth users from these
stimulant questions.
Question text the same.
Universe edited to remove
meth users from these
stimulant questions.

See notes at end of table.

2,034
(continued)

O-36

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item

How old were you the last time you
used any methamphetamine for
kicks or to get high? (LU17)5
Height in inches (HLTH05HLTH08)5
Weight in pounds (HLTH10-14)5,7
During the past 12 months, how
many times have you visited a
doctor, nurse, physician assistant
or nurse practitioner about your
own health at a doctor’s office, a
clinic, or some other place?
(HLTH19)5,8
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you smoke
cigarettes or use any other tobacco
products? (HLTH20a)4
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you drink alcohol?
(HLTH20b)4
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you use illegal
drugs? (HLTH20c)4
During the past 12 months, did any
doctor or other health care
professional advise you to quit
smoking cigarettes or quit using
any other tobacco products?
(HLTH21)4
Choose the statement or statements
below that describe any
discussions you may have had in
person with a doctor or other
health professional about your
alcohol use. (HLTH22)4
The doctor asked how much I
drink.
The doctor asked how often I
drink.
The doctor asked if I have any
problems because of my
drinking.

Type of
2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Change1 Description of Change Estimate2,3
Error
In the 2012 interview, this
was about pain relievers.
In the QFT, it is about
meth. The prescription
drug questions were
R
deleted from this module.
24.5
(0.81)
N/A
101
New questions about
N
height and weight.
66.6
(0.26)
N/A
2,007
New questions about
N
height and weight.
176.0
(1.44)
N/A
2,001

New questions about
health.

3.9

(0.18)

N/A

1,971

New questions about
health.

71.2

(1.37)

1,137

1,677

New questions about
health.

67.9

(1.50)

1,067

1,675

New questions about
health.

51.0

(1.55)

865

1,675

New questions about
health.

28.8

(2.01)

310

994

New questions about
health.
33.5

(1.97)

329

1,031

32.8

(1.97)

325

1,031

5.9

(0.89)

See notes at end of table.

1,031
(continued)

O-37

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
The doctor advised me to cut
down on my drinking.
The doctor offered to give me
more information about alcohol
use and treatment for problems
with alcohol use.
The doctor didn't discuss my
alcohol use with me in the past
12 months.
During the past 12 months, did any
doctor or other health care
professional talk to you about your
use of marijuana, cocaine, crack,
heroin, inhalants, hallucinogens,
or methamphetamine? (HLTH23)4
During the past 12 months, did you
have a sexually transmitted
disease such as chlamydia,
gonorrhea, herpes or syphilis?
(HLTH24)
Conditions that a doctor or other
health care professional has ever
told you that you had (HLTH25)
Any kind of heart condition or
heart disease

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
2.3

(0.55)

1,031

0.9

(0.27)

1,031

54.0

(1.95)

561

1,031

New questions about
health.

17.2

(2.74)

297

New questions about
health.

1.6

(0.30)

2,038

New questions about
health.

10.4

(1.04)

124

2,027

Diabetes or sugar diabetes
Chronic bronchitis, emphysema,
chronic obstructive pulmonary
disease, also called COPD

9.0

(0.98)

109

2,027

3.3

(0.58)

2,027

Cirrhosis of the liver

0.2

(0.13)

2,027

Hepatitis
Kidney disease, not including
bladder infection or
incontinence

2.1

(0.51)

2,027

1.3

(0.36)

2,027

(0.79)

256

2,027

(0.00)

2,027

6.1

(0.85)

2,027

17.8

(1.16)

199

2,027

57.3

(1.62)

1,381

2,027

Asthma
HIV or AIDS
Cancer or a malignancy of any
Kind
Hypertension, also called high
blood pressure
None of the above – I have never
had any of these conditions
What kind of cancer was it?
(HLTH26)4

11.1
0.0

New questions about
health.
0.0*

(0.00)

(1.67)

(0.27)

Brain

1.9

(1.86)

Breast

24.8*

(6.34)

Bladder
Blood

2.0

Bone

0.3

See notes at end of table.

(continued)

O-38

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Estimate2,3
Error
Total
Sample Size
13.7*

(4.47)

5.2

(2.40)

Esophagus

3.5

(2.23)

Gallbladder

0.0*

(0.00)

3.0

(2.08)

0.0

(0.00)

2.3

(1.69)

(0.00)

(2.35)

Lymphoma

9.2

(4.70)

Melanoma

11.2*

(4.86)

(0.00)

(1.85)

(3.46)

(3.10)

0.0

(0.00)

16.9*

(5.22)

(4.25)

(0.00)

0.0

(0.00)

Thyroid

2.7

(2.03)

Uterus (Females Only)

3.5*

(3.41)

(2.35)

Cervix (Females Only)
Colon

Kidney
Larynx/Windpipe
Leukemia
Liver

0.0

Lung

3.2

Mouth/Tongue/Lip

0.0

Ovary (Females Only)

2.0

Pancreas

3.5

Prostate (Males Only)

5.4

Rectum
Skin (Not Melanoma)
Skin (Don't Know Which Kind)

4.5

Soft Tissue (Muscle or Fat)

0.0

Stomach

0.0

Testis (Males Only)

0.0

Throat/Pharynx

Other
How old were you when your blood
cancer was first diagnosed?
(HLTH28a)5
How old were you when your bone
cancer was first diagnosed?
(HLTH28b)5
How old were you when your brain
cancer was first diagnosed?
(HLTH28c)5
How old were you when your breast
cancer was first diagnosed?
(HLTH28d)5
How old were you when your
cervical cancer was first
diagnosed? (HLTH28e)5
How old were you when your colon
cancer was first diagnosed?
(HLTH28f)5

3.4
N

New questions about
health.

4.0*

(0.00)

N/A

New questions about
health.

5.0*

(0.00)

N/A

New questions about
health.

50.0*

(0.00)

N/A

New questions about
health.

50.8

(3.16)

N/A

New questions about
health.

34.5

(3.97)

N/A

New questions about
health.

51.1*

(5.49)

N/A

See notes at end of table.

5
(continued)

O-39

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when your
esophageal cancer was first
diagnosed? (HLTH28g)5
How old were you when your
kidney cancer was first
diagnosed? (HLTH28i)5
How old were you when your
leukemia was first diagnosed?
(HLTH28k)5
How old were you when your lung
cancer was first diagnosed?
(HLTH28m)5
How old were you when your
lymphoma was first diagnosed?
(HLTH28n)5
How old were you when your
melanoma was first diagnosed?
(HLTH28o)5
How old were you when your
ovarian cancer was first
diagnosed? (HLTH28q)5
How old were you when your
pancreatic cancer was first
diagnosed? (HLTH28r)5
How old were you when your
prostate cancer was first
diagnosed? (HLTH28s)5
How old were you when your skin
[not melanoma] cancer was first
diagnosed? (HLTH28u)5
How old were you when your skin
cancer was first diagnosed?
(HLTH28v)5
How old were you when your
thyroid cancer was first
diagnosed? (HLTH28aa)5
How old were you when your
uterine cancer was first
diagnosed? (HLTH28bb)5
How old were you when the type of
cancer listed below was first
diagnosed? (HLTH28cc)5
Did you have cancer during the past
12 months? (HLTH29)4
How old were you when your heart
condition or heart disease was first
diagnosed? (HLTH30)5,8
Did you have any kind of heart
condition or heart disease in the
past 12 months? (HLTH31)4
How old were you when your
diabetes or sugar diabetes was first
diagnosed? (HLTH32)5,8

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

New questions about
health.

63.4*

(9.11)

N/A

New questions about
health.

44.8*

(6.58)

N/A

New questions about
health.

26.5*

(7.52)

N/A

New questions about
health.

58.7*

(10.48)

N/A

New questions about
health.

56.0*

(5.42)

N/A

New questions about
health.

37.8*

(3.81)

N/A

New questions about
health.

56.7*

(2.94)

N/A

New questions about
health.

64.0*

(0.00)

N/A

New questions about
health.

66.0*

(1.42)

N/A

New questions about
health.

54.5*

(2.99)

N/A

New questions about
health.

46.0*

(0.00)

N/A

New questions about
health.

35.6*

(2.48)

N/A

New questions about
health.

40.0*

(0.00)

N/A

47.7*

(10.47)

N/A

New questions about
health.
New questions about
health.

34.9*

(7.47)

New questions about
health.

43.4

(1.94)

N/A

122

New questions about
health.

42.5

(5.70)

116

New questions about
health.

43.2

(1.60)

N/A

See notes at end of table.

107
(continued)

O-40

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
How old were you when your
chronic bronchitis, emphysema, or
chronic obstructive pulmonary
disease, also called COPD were
first diagnosed? (HLTH33)5
How old were you when your
cirrhosis of the liver was first
diagnosed? (HLTH34)5
How old were you when your
hepatitis was first diagnosed?
(HLTH35)5
How old were you when your
kidney disease was first
diagnosed? (HLTH36)5
How old were you when your
asthma was first diagnosed?
(HLTH37)5
Do you still have asthma?
(HLTH38)4
Are you currently taking
prescription medicine for your
high blood pressure? (HLTH40)4
How old were you when your high
blood pressure was first
diagnosed? (HLTH41)5
How many times in the past 12
months have you moved?
(QD13)5,8
Were you born in the United States?
(QD14)
Have you lived in the United States
for at least one year? (QD16a)4
For how many years have you lived
in the United States? (QD16b)5
For how many months have you
lived in the United States?
(QD16c)5
Are you now attending or are you
currently enrolled in school?
(QD17)
What grade or year of school are
you now attending? (QD18)4

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Error
Total
Sample Size
Estimate2,3

New questions about
health.

35.0

(3.27)

N/A

New questions about
health.

47.6*

(4.41)

N/A

New questions about
health.

27.0

(3.96)

N/A

New questions about
health.

41.0

(4.47)

N/A

18.5

(1.77)

N/A

232

64.3

(4.06)

169

249

New questions about
health.
New questions about
health.

New questions about
health.

86.7

(2.35)

153

199

New questions about
health.

45.1

(1.04)

N/A

147

0.4

(0.03)

N/A

2,014

87.9

(1.29)

1,803

2,042

95.9

(1.52)

227

238

23.7

(1.56)

N/A

227

(2.28)

N/A

18.9

(1.07)

804

2,040

1st Grade

0.3

(0.23)

802

2nd Grade

0.2

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

M
M
M

M
M

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

3rd Grade
4th Grade

6.7*

(0.15)

802

0.0

(0.00)

802

0.0

(0.00)

802

5th Grade

0.0

(0.00)

802

6th Grade

1.2

(0.43)

802

7th Grade

7.7

(0.92)

802

8th Grade

9.8

(1.17)

See notes at end of table.

802
(continued)

O-41

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

9th Grade

9.7

(1.19)

802

10th Grade

8.3

(0.91)

802

11th Grade

8.2

(0.98)

802

9.1

(0.99)

802

College or University/1st Year

12th Grade

12.2

(1.54)

802

College or University/2nd Year

8.8

(1.34)

802

College or University/3rd Year

8.5

(1.44)

802

College or University/4th Year
College or University/5th Year or
Higher
Are you a full-time student or a part
time student? (QD19)4

5.1

(1.24)

802

10.9

(2.09)

802

80.7

(2.14)

690

792

19.3

(2.14)

102

792

Administered in ACASI
instead of CAPI.

Full-Time
Part-Time
During the past 30 days, how many
whole days of school did you miss
because you were sick or injured?
(QD20)5,8
During the past 30 days, how many
whole days of school did you miss
because you skipped or “cut” or
just didn’t want to be there? (QD21)5,8
Are you now married, widowed,
divorced or separated, or have you
never married? (QD07)4

Administered in ACASI
instead of CAPI.

0.8

(0.16)

N/A

584

Administered in ACASI
instead of CAPI.

0.4

(0.07)

N/A

587

Administered in ACASI
instead of CAPI.
51.0

(2.03)

639

1,771

4.9

(0.81)

1,771

Married
Widowed
Divorced or Separated

13.8

(1.19)

174

1,771

30.2

(1.54)

912

1,771

1.4

(0.03)

N/A

857

6.2

(0.70)

143

2,021

My spouse

7.6

(3.20)

123

Unmarried partner

3.4

(1.74)

123

Have Never Married
How many times have you been
married? (QD08)5
Is anyone in your immediate family
currently serving in the United
States military? (QD10d)
Which member or members of your
immediate family are currently in
the United States military?
(QD10e)4

Administered in ACASI
instead of CAPI.
New question on
immediate family serving
in the military.

New question on
immediate family serving
in the military.

My mother

1.5

(0.75)

123

My father

5.1

(1.55)

123

33.4*

(6.40)

123

(2.66)

123

(6.19)

123

(0.61)

My son or sons
My daughter or daughters

3.6

My brother or brothers

47.2

My sister or sisters

1.2

See notes at end of table.

123
(continued)

O-42

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Did you work at a job or business at
any time last week? (QD26)4
Even though you did not work at
any time last week, did you have
a job or business? (QD27)4
How many hours did you work last
week at all jobs or businesses?
(QD28)5
Do you usually work 35 hours or
more per week at all jobs or
businesses? (QD29)4
Which one of these reasons best
describes why you did not work
last week? (QD30)4
Vacation/Sick/Furlough/Strike/
Other Temporary
Absence/Maternity Leave

Type of
Change1 Description of Change
Administered in ACASI
M
instead of CAPI.

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
60.0

(1.72)

1,025

1,773

Administered in ACASI
instead of CAPI.

12.1

(1.68)

104

744

Administered in ACASI
instead of CAPI.

38.5

(0.51)

N/A

1,020

Administered in ACASI
instead of CAPI.

77.0

(1.53)

812

1,126

Administered in ACASI
instead of CAPI.
33.0*

(5.79)

104

(2.19)

104

(4.37)

104

Layoff, Not Looking for Work

3.6

Layoff, Looking for Work

9.8

Waiting to Report to New Job
Self-Employed, No Business
Last Week

4.3

(1.88)

104

15.4*

(5.46)

104

Going to School/Training

11.7

(3.42)

104

22.1

(5.73)

104

16.3

(1.90)

156

636

1.5

(0.46)

636

11.8

(1.89)

636

9.9

(1.08)

151

636

Retired

38.0

(2.90)

104

636

Disabled

14.7

(1.99)

636

2.3

(0.55)

636

5.5

(0.98)

636

82.1

(3.68)

119

156
642

Some Other Reason
Which one of these reasons best
describes why you did not have a
job or business last week?
(QD31)4

Administered in ACASI
instead of CAPI.

Looking for Work
On Layoff, Not Looking for
Work
Keeping House/Caring for
Children Full Time
Going to School/Training

Didn't Want a Job
Some Other Reason
During the past 30 days, did you
make specific efforts to find
work? (QD32)4
Did you work at a job or business at
any time during the past 12
months? (QD33)4
How many different employers have
you had in the past 12 months?
(QD35 and QD36)5

Administered in ACASI
instead of CAPI.

18.9

(2.04)

158

Administered in ACASI
instead of CAPI.

1.4

(0.05)

N/A

See notes at end of table.

1,272
(continued)

O-43

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 12 months, was
there ever a time when you did not
have at least one job or business?
(QD37)4
In how many weeks during the past
12 months did you not have at
least one job or business?
(QD38)5
During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)5,8
During the past 30 days, how many
whole days of work did you miss
because you just didn’t want to be
there? (QD41)5,8
Thinking about the location where
you work, how many people work
for your employer out of this
office, store, etc.? (QD42)4

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Administered in ACASI
instead of CAPI.

15.6

(1.35)

249

1,126

Administered in ACASI
instead of CAPI.

13.8

(0.99)

N/A

234

Administered in ACASI
instead of CAPI.

0.7

(0.12)

N/A

1,116

Administered in ACASI
instead of CAPI.

0.2

(0.03)

N/A

1,116

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

30.3

(1.93)

326

1,110

10 to 24 People

18.3

(1.36)

229

1,110

25 to 99 People

18.6

(1.28)

230

1,110

100 to 499 People

18.4

(1.59)

190

1,110

14.4

(1.66)

135

1,110

80.1

(1.63)

858

1,092

1.1

(0.49)

853

Less Than 10 People

500 People or More
At your workplace, is there a written
policy about employee use of
alcohol or drugs? (QD43)4
Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)4
Only Alcohol

M
M

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Only Drugs

2.3

(0.52)

853

96.5

(0.73)

819

853

Yes

33.2

(2.03)

343

1,121

49.0

(2.11)

568

1,121

17.9

(1.43)

210

1,121

53.5

(1.98)

488

Both Alcohol and Drugs
At your workplace, have you ever
been given any educational
information regarding the use of
alcohol or drugs? (QD45)4

Don't Remember
Through your workplace, is there
access to any type of employee
assistance program or other type
of counseling program for
employees who have alcohol or
drug-related problems? (QD46)4

Administered in ACASI
instead of CAPI.

See notes at end of table.

1,040
(continued)

O-44

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Does your workplace ever test its
employees for alcohol use?
(QD47)4
Does your workplace ever test its
employees for drug use? (QD48)4
Does your workplace test its
employees for drug or alcohol use
as part of the hiring process?
(QD49)4
Does your workplace test its
employees for drug or alcohol use
on a random basis? (QD50)4
According to the policy at your
workplace, what happens to an
employee the first time he or she
tests positive for illicit drugs?
(QD51)4
Handled on Individual
Basis/Policy Does Not Specify
What Happens

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
31.5

(1.71)

337

1,083

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

48.1

(2.05)

524

1,094

Administered in ACASI
instead of CAPI.

87.6

(1.71)

450

525

Administered in ACASI
instead of CAPI.

59.8

(3.18)

315

511

Administered in ACASI
instead of CAPI.
24.3

(2.51)

122

472

47.1

(2.65)

238

472

23.6

(2.17)

472

1.6

(0.85)

472

3.4

(1.00)

472

More Likely

48.3

(1.85)

516

1,121

Less Likely

7.2

(0.82)

1,121

44.6

(1.57)

509

1,121

More Likely

43.1

(1.77)

458

1,122

Less Likely

11.5

(1.24)

146

1,122

45.4

(1.66)

518

1,122

Employee Is Fired
Employee Referred for
Treatment/Counseling
Nothing Happens
Something Else Happens
Would you be more or less likely to
want to work for an employer that
tests its employees for drug use as
part of the hiring process?
(QD52)4

Would Make No Difference
Would you be more or less likely to
want to work for an employer that
tests its employees for drug or
alcohol use on a random basis?
(QD53)4

Would Make No Difference
How well do you speak English?
(QD55)

Administered in ACASI
instead of CAPI.

New questions.
90.9

(0.92)

1,874

2,042

Well

Very well

8.6

(0.92)

151

2,042

Not well

0.5

(0.14)

2,042

Not at all

0.0

(0.03)

See notes at end of table.

2,042
(continued)

O-45

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Are you deaf or do you have serious
difficulty hearing? (QD56)
Are you blind or do you have
serious difficulty seeing, even
when wearing glasses? (QD57)
Because of a physical, mental or
emotional condition, do you have
serious difficulty concentrating,
remembering, or making
decisions? (QD58)
Do you have serious difficulty
walking or climbing stairs?
(QD59)
Do you have difficulty dressing or
bathing? (QD60)
Because of a physical, mental or
emotional condition, do you have
difficulty doing errands alone such
as visiting a doctors’ office or
shopping? (QD61)4
Covered by Medicare? (QHI01)
Covered by Medicaid/[CHIPFILL]
(QHI02 and QHI02a)
Covered by TRICARE, CHAMPUS
CHAMPVA, VA, Military Health
Care (QHI03)
Covered by Private Health
Insurance (QHI06)
Was [MEMBER] private health
insurance obtained through work,
such as through an employer,
union, or professional association?
(QHI07)4
Does [MEMBER] private health
insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)4
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4
Does [MEMBER] private health
insurance include coverage for
treatment for mental or emotional
problems? (QHI10)4
[MEMBER] currently covered by
any kind of health insurance,
including Indian Health
Insurance? (QHI11)4
Any Health Insurance Coverage
(Recode)

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

New questions.

5.4

(0.61)

2,040

New questions.

3.4

(0.58)

2,038

New questions.

6.6

(0.68)

161

2,036

New questions.

6.4

(0.89)

2,040

New questions.

1.6

(0.36)

2,042

New questions.
Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

4.1

(0.68)

1,773

18.3

(1.58)

181

2,025

13.4

(1.16)

390

2,015

M
M

5.0

(0.77)

2,027

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

62.1

(1.86)

1,148

2,012

Administered in ACASI
instead of CAPI.

88.6

(1.47)

1,053

1,144

Administered in ACASI
instead of CAPI.

74.2

(1.99)

594

826

Administered in ACASI
instead of CAPI.

73.2

(2.04)

582

818

Administered in ACASI
instead of CAPI.

85.0

(1.62)

795

939

Administered in ACASI
instead of CAPI.

21.9

(2.71)

412

86.1

(1.03)

1,685

See notes at end of table.

2,010
(continued)

O-46

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
During the past 12 months, was
there any time when [MEMBER]
did not have any kind of health
insurance or coverage? (QHI13)4
During the past 12 months, about
How many months without any
kind of health insurance or
coverage? (QHI14)5
About how long has it been since
[MEMBER] last had any kind of
health care coverage? (QHI15)4
Within the Past 6 Months
More Than 6 Months Ago but
Within the Past Year
More Than 1 Year Ago but
Within the Past 3 Years
More Than 3 Years Ago
Never Had Coverage
Which of these reasons is the main
reason why [MEMBER] stopped
being covered by health
insurance? (QHI17)4
Person in Family with Health
Insurance Lost Job/Changed
Employer
Lost Medicaid Coverage Because
of New Job/Increase in Income
Lost Medicaid Coverage for
Some Other Reason
Cost Is Too High/Can't Afford
Premiums
Became Ineligible Because of
Age/Leaving School
Employer Does Not Offer
Coverage or Not Eligible for
Coverage
Divorced/Separated from Person
With Insurance

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Administered in ACASI
instead of CAPI.

7.3

(0.75)

155

1,677

Administered in ACASI
instead of CAPI.

4.2

(0.41)

N/A

153

Administered in ACASI
instead of CAPI.

15.6

(2.42)

319

7.8

(1.62)

319

21.9

(3.14)

319

35.6

(3.18)

103

319

19.0

(2.63)

319

28.4

(4.19)

250

7.1

(1.49)

250

4.6

(1.38)

250

26.7

(3.74)

250

9.9

(2.09)

250

3.8

(1.13)

250

1.2

(0.69)

250

0.2

(0.21)

250

1.1*

(0.92)

250

3.5

(1.53)

250

2.8

(1.01)

250

10.8

(2.38)

Administered in ACASI
instead of CAPI.

Death of Spouse/Parent
Insurance Company Refused
Coverage
Don't Need It
Received Medicaid/Insurance
Only While Pregnant
Some Other Reason
See notes at end of table.

250
(continued)

O-47

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
Which of these reasons describe
why [SAMPLE MEMBER] never
had health insurance coverage?
(QHI18)4
Cost Too High/ Can't Afford
Premiums
Employer Does Not Offer
Coverage or Not Eligible for
Coverage
Insurance Company Refused
Coverage

Type of
Change1 Description of Change

Administered in ACASI
instead of CAPI.

Don't Need It
Some Other Reason
In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)
In [YEAR], did you receive
Supplemental Security Income or
SSI? (QI03N)
In [YEAR], did you receive income
from wages or pay earned while
working at a job or business? (QI05N)
In [YEAR], did you receive food
stamps? (QI07N)
At any time during [YEAR], did
you receive any cash assistance
from a state or county welfare
program such as [TANFFILL]?
(QI08N)
In [YEAR], because of low income,
did you receive any other kind of
non-monetary welfare or public
assistance? (QI10N)
For how many months in [YEAR]
did you or your
[RELATIONSHIP] receive any
type of welfare or public
assistance, not including food
stamps? (QI12AN and QI12BN)5
Before taxes and other deductions,
was your total personal income
from all sources during [YEAR]
more or less than 20,000 dollars?
(QI20N)

44.0*

(6.55)

5.1*

(2.63)

1.0*

(0.96)

11.8*

(4.11)

38.1*

(8.53)

Administered in ACASI
instead of CAPI.

26.5

(1.69)

351

2,011

Administered in ACASI
instead of CAPI.

9.5

(0.98)

177

1,990

68.6

(1.78)

1,379

2,006

17.6

(1.49)

454

2,020

Administered in ACASI
instead of CAPI.
Administered in ACASI
instead of CAPI.

Administered in ACASI
instead of CAPI.

3.4

(0.54)

2,007

Administered in ACASI
instead of CAPI.

3.4

(0.52)

2,016

Administered in ACASI
instead of CAPI.

6.1

(0.55)

N/A

147

Administered in ACASI
instead of CAPI.
55.7

(1.60)

769

1,970

44.3

(1.60)

1,201

1,970

$20,000 or More
Less Than $20,000
Of these income groups, which
category best represents
[MEMBER] total personal income
during [YEAR]? (QI21A and QI21B)

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

Administered in ACASI
instead of CAPI.

Less Than $1,000

14.9

See notes at end of table.

(0.84)

555

1,895
(continued)

O-48

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
$6,000-$6,999
$7,000-$7,999
$8,000-$8,999
$9,000-$9,999
$10,000-$10,999
$11,000-$11,999
$12,000-$12,999
$13,000-$13,999
$14,000-$14,999
$15,000-$15,999
$16,000-$16,999
$17,000-$17,999
$18,000-$18,999
$19,000-$19,999
$20,000-$24,999
$25,000-$29,999
$30,000-$34,999
$35,000-$39,999
$40,000-$44,999
$45,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000-$149,999
$150,000 or More
Before taxes and other deductions,
was the total combined family
income during [YEAR] more or
less than 20,000 dollars? (QI22)
$20,000 or More
Less Than $20,000
Of these income groups, which
category best represents your total
combined family income during
[YEAR]. (QI23A and QI23B)
Less Than $1,000
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error
2.9
(0.38)
84
1,895
1.2
(0.23)
41
1,895
1.4
(0.30)
34
1,895
1.1
(0.27)
27
1,895
0.9
(0.23)
24
1,895
0.9
(0.27)
20
1,895
0.4
(0.19)
9
1,895
1.3
(0.32)
25
1,895
2.6
(0.51)
47
1,895
2.3
(0.44)
43
1,895
1.4
(0.36)
22
1,895
1.4
(0.35)
24
1,895
1.3
(0.37)
21
1,895
1.3
(0.31)
21
1,895
1.8
(0.39)
35
1,895
1.5
(0.32)
27
1,895
1.8
(0.41)
28
1,895
1.7
(0.38)
29
1,895
1.8
(0.38)
34
1,895
8.7
(0.85)
146
1,895
5.5
(0.68)
88
1,895
4.8
(0.72)
78
1,895
5.6
(0.78)
65
1,895
4.8
(0.79)
63
1,895
4.9
(0.77)
54
1,895
10.8
(1.08)
128
1,895
4.4
(0.74)
56
1,895
3.9
(0.85)
47
1,895
2.7
(0.88)
20
1,895

Administered in ACASI
instead of CAPI.
79.7
20.3

(1.55)
(1.55)

1,449
500

1,949
1,949

2.3
1.0
0.6
0.9
0.4
0.4

(0.42)
(0.30)
(0.18)
(0.25)
(0.16)
(0.17)

71
25
21
20
13
11

1,797
1,797
1,797
1,797
1,797
1,797

Administered in ACASI
instead of CAPI.

See notes at end of table.

(continued)

O-49

Table O-1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field
Test among Persons Aged 12 or Older (continued)
QFT Instrument Item

Type of
Change1 Description of Change

2012 QFT Standard Unweighted Unweighted
Total
Sample Size
Estimate2,3
Error

$6,000-$6,999

0.5

(0.19)

1,797

$7,000-$7,999

0.2

$8,000-$8,999

0.6

(0.10)

1,797

(0.25)

1,797

$9,000-$9,999

0.8

(0.19)

1,797

$10,000-$10,999

1.2

(0.29)

1,797

$11,000-$11,999

0.6

(0.20)

1,797

$12,000-$12,999

0.8

(0.18)

1,797

$13,000-$13,999

1.1

(0.40)

1,797

$14,000-$14,999

1.2

(0.30)

1,797

$15,000-$15,999

0.9

(0.24)

1,797

$16,000-$16,999

0.7

(0.19)

1,797

$17,000-$17,999

1.6

(0.40)

1,797

$18,000-$18,999

0.9

(0.25)

1,797

$19,000-$19,999

2.0

(0.47)

1,797

$20,000-$24,999

7.7

(0.93)

138

1,797

$25,000-$29,999

4.2

(0.51)

1,797

$30,000-$34,999

5.2

(0.69)

101

1,797

$35,000-$39,999

5.2

(0.77)

1,797

$40,000-$44,999

6.3

(1.11)

102

1,797

$45,000-$49,999

5.0

(0.64)

1,797

$50,000-$74,999

15.9

(1.25)

249

1,797

$75,000-$99,999

11.6

(0.98)

195

1,797

$100,000-$149,999

12.1

(1.41)

194

1,797

7.8

(1.17)

114

1,797

$150,000 or More
Is there at least one telephone at this
address that is not a cell phone?
(CELL1)
Do you or anyone at this address
have a working cell phone?
(CELL2)

New item.

64.1

(1.68)

1,143

2,032

New item.

92.3

(0.82)

1,913

2,037

*Low precision; estimate would be suppressed due to not meeting the NSDUH suppression rule.
ACASI = audio computer-assisted self-interviewing; CAPI = computer-assisted personal interviewing; N/A = not applicable; QFT =
Questionnaire Field Test; R = respondent.
1
Changes to questionnaire items fall under three categories: N = new item, R= revised item, and M= no changes to item but moved to another
place in the questionnaire or moved from being interviewer-administered to self- administered.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. QFT data were collected from September 1
through November 3, 2012.
3
Estimates are percentages of all persons aged 12 or older, except where noted.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing data.
5
Estimate is an average based on valid responses to the relevant question(s). Respondents with unknown or missing data were excluded.
6
Data in the source question are continuous. The estimate is expressed as a percentage for persons reporting valid nonzero values.
7
Includes pre-pregnancy weight of pregnant females as reported in HLTH13 and HLTH14.
8
The estimated mean includes zeroes.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health, 2012.

O-50

Appendix P: Proxy Reports from the QFT and the
Comparison Samples

Table P-1 Distribution of Respondent Relationship with Proxy among Persons Aged 12 or Older Who Obtained a Proxy, by Age Group:
Percentages, and Standard Errors, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test

P-1

Proxy Relationship
Father
Mother
Son / Daughter
Brother / Sister
Husband / Wife
Live-in Boyfriend / Girlfriend
Son-in-law / Daughter-in-law
Grandson / Granddaughter
Father-in-law / Mother-in-law
Grandfather / Grandmother
Other Adult Relative

2011
Comparison1
12-17,
Percent (SE)
23.7 (0.42)
69.7 (0.45)
0.0* (0.00)
1.7 (0.15)
0.0* (0.00)
0.0 (0.01)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
3.0 (0.17)
1.9 (0.15)

2012
Comparison1,2
12-17,
Percent (SE)
23.7 (0.63)
69.3 (0.70)
0.0 (0.02)
1.8 (0.17)
0.0* (0.00)
0.0 (0.02)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
3.2 (0.24)
2.0 (0.22)

2012 QFT1,3
12-17,
Percent (SE)
25.1 (2.62)
67.8 (2.76)
0.2 (0.16)
1.9 (0.72)
0.0* (0.00)
0.2 (0.19)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
2.3 (0.62)
2.6 (0.98)

2011
Comparison1 2012 Comparison1,2
18 or Older,
18 or Older,
Percent (SE)
Percent (SE)
6.2 (0.44)
6.4 (0.60)
22.6 (0.86)
22.9 (1.28)
6.1a (1.09)
5.1a (1.22)
1.1 (0.25)
1.1 (0.34)
58.2 (1.18)
57.4 (1.85)
2.8 (0.47)
4.0 (0.77)
0.0* (0.00)
0.4 (0.38)
0.3 (0.19)
0.3 (0.30)
0.4 (0.22)
0.5 (0.36)
0.9 (0.17)
0.9 (0.18)
a
1.5 (0.37)
1.0 (0.38)

2012 QFT1,3
18 or Older,
Percent (SE)
4.6 (1.49)
23.2 (3.39)
0.0* (0.00)
2.2 (1.31)
62.0 (4.04)
6.7 (2.60)
0.0* (0.00)
0.0* (0.00)
0.0* (0.00)
1.1 (0.62)
0.2 (0.23)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not answer HASJOIN, he or she is defined
as not having used a proxy. Respondents who were legitimately skipped from answering question QP01 were excluded from this analysis. Edited
variables PRXYANS2 for HASJOIN and PRXRELAT for QP02 were used in this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison proxy compared with 2012 QFT
proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

Table P-2 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 12 or
Older: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Instrument Item
Covered by Private Health
Insurance? (QHI06)4,5

2012
2011
2012
2011
Comparison1 Comparison1,3 2012 QFT1,2 Comparison1 Comparison1,3 2012 QFT1,2
Proxy Percent
Proxy
Proxy
No Proxy
No Proxy
No Proxy
(SE)
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)
64.6 (0.79)

65.3 (0.96)

59.5 (3.04)

69.6a (0.49)

69.4 (0.67)

64.9 (2.19)

Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5

84.7a (0.88)

85.1a (1.05)

73.7 (5.07)

84.9a (0.52)

84.7a (0.82)

76.8 (2.13)

Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5

84.7a (0.89)

84.6a (1.04)

76.3 (3.65)

84.0a (0.53)

84.3a (0.85)

74.8 (2.26)

Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems?
(QHI10)4,5

91.7a (0.54)

91.3a (0.74)

83.3 (3.24)

91.9a (0.32)

92.4a (0.55)

85.7 (1.80)

In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5

21.1 (0.73)

19.7 (1.18)

22.2 (2.86)

27.6 (0.53)

26.3 (0.60)

26.4 (2.06)

In [YEAR], did [FILL] receive
supplemental Security Income
or SSI? (QI03N)4,5

8.6 (0.44)

8.8 (0.53)

10.0 (1.84)

6.5a (0.23)

7.6 (0.39)

9.4 (1.18)

In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job
or business? (QI05N)4,5

84.9a (0.60)

86.3a (0.79)

63.8 (2.66)

87.2a (0.42)

87.5a (0.50)

71.6 (1.90)

In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5

18.2a (0.62)

18.0a (0.74)

23.9 (2.50)

13.3 (0.36)

14.6 (0.47)

15.2 (1.67)

At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or
county welfare program such as
[TANFFILL]? (QI08N)4,5

3.4 (0.24)

3.1 (0.26)

3.9 (0.92)

2.3 (0.13)

2.0 (0.16)

2.7 (0.59)

In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5

3.9 (0.25)

4.2 (0.34)

4.9 (1.21)

3.0 (0.15)

2.7 (0.16)

2.9 (0.58)

Before taxes and other
deductions, was [MEMBER]
total personal income from all
sources during [YEAR] more or
less than 20,000 dollars?
(QI20N)4,5
$20,000 or More
Less Than $20,000

14.1 (0.80)
85.9 (0.80)

15.0 (0.99)
85.0 (0.99)

19.2 (2.64)
80.8 (2.64)

58.4a (0.46)
41.6a (0.46)

58.4a (0.62)
41.6a (0.62)

64.9 (1.74)
35.1 (1.74)

See notes at end of table.

(continued)

P-2

Table P-2 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 12 or
Older: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B)4,5
Less Than $1,000
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
$6,000-$6,999
$7,000-$7,999
$8,000-$8,999
$9,000-$9,999
$10,000-$10,999
$11,000-$11,999
$12,000-$12,999
$13,000-$13,999
$14,000-$14,999
$15,000-$15,999
$16,000-$16,999
$17,000-$17,999
$18,000-$18,999
$19,000-$19,999
$20,000-$24,999
$25,000-$29,999
$30,000-$34,999
$35,000-$39,999
$40,000-$44,999
$45,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000 or More
$100,000-$149,999
$150,000 or More

2011
2012
2012
2011
Comparison1 Comparison1,3 2012 QFT1,2 Comparison1 Comparison1,3 2012 QFT1,2
Proxy Percent
Proxy
Proxy
No Proxy
No Proxy
No Proxy
(SE)
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)

60.2a (0.84)
4.1 (0.17)
3.0 (0.22)
1.9 (0.16)
1.4 (0.12)
2.0a (0.26)
1.9 (0.37)
1.4 (0.16)
1.2 (0.14)
1.6 (0.27)
1.2 (0.18)
0.7 (0.13)
1.0 (0.24)
0.8a (0.20)
0.6 (0.16)
0.5 (0.10)
0.2 (0.09)
0.8 (0.29)
0.9a (0.21)
0.8 (0.17)
2.4 (0.32)
2.3 (0.35)
1.7 (0.32)
1.2 (0.22)
1.3 (0.24)
1.1 (0.22)
2.4 (0.31)
0.8 (0.19)
0.4 (0.13)
-- (--)
-- (--)

60.1a (1.10)
4.3 (0.31)
2.7 (0.24)
2.1 (0.24)
1.4 (0.15)
1.2 (0.21)
1.1 (0.14)
1.1 (0.18)
1.5 (0.26)
1.7 (0.47)
1.4 (0.22)
1.0 (0.20)
1.4 (0.34)
1.0a (0.27)
0.5 (0.14)
0.6 (0.17)
0.4 (0.17)
0.2 (0.08)
0.8 (0.21)
0.7 (0.25)
2.6 (0.42)
1.7 (0.32)
1.8 (0.36)
1.4 (0.40)
1.7 (0.50)
1.3 (0.29)
2.4 (0.37)
0.6 (0.17)
1.2 (0.36)
-- (--)
-- (--)

53.7 (2.84)
4.5 (0.86)
1.9 (0.87)
2.1 (0.65)
2.9 (1.25)
0.9 (0.37)
0.9 (0.40)
0.5 (0.43)
1.1 (0.50)
2.1 (1.21)
3.1 (1.30)
0.5 (0.33)
0.7 (0.58)
0.2 (0.19)
0.9 (0.65)
0.3 (0.25)
1.4 (0.95)
1.3 (0.95)
0.3 (0.22)
1.5 (0.84)
4.1 (1.28)
2.7 (1.19)
2.4 (1.25)
1.0 (0.71)
1.2 (0.77)
2.3 (1.19)
2.7 (1.26)
1.9 (1.10)
1.0 (0.62)
1.0 (0.62)
0.0* (0.00* )

10.5a (0.23)
1.9 (0.10)
1.6a (0.09)
1.4 (0.09)
1.3a (0.08)
1.6a (0.10)
1.4 (0.11)
1.6a (0.11)
1.8 (0.11)
1.8 (0.11)
2.2 (0.15)
1.5 (0.10)
2.2a (0.13)
1.5 (0.11)
1.5a (0.11)
1.8 (0.11)
1.2 (0.10)
1.4 (0.09)
1.8 (0.11)
1.8 (0.12)
6.8 (0.24)
6.6 (0.31)
5.9 (0.26)
5.0 (0.23)
4.4 (0.20)
4.2 (0.18)
12.0 (0.34)
5.7 (0.23)
7.8 (0.35)
-- (--)
-- (--)

10.4a (0.34)
2.0 (0.14)
1.4 (0.11)
1.5 (0.15)
1.1 (0.11)
1.4 (0.11)
1.6 (0.17)
1.6a (0.18)
1.8 (0.17)
1.8 (0.16)
2.1 (0.17)
1.8 (0.18)
2.6a (0.24)
1.3 (0.12)
1.7a (0.15)
1.6 (0.14)
1.3 (0.12)
1.2 (0.12)
1.7 (0.16)
1.7 (0.16)
6.8 (0.33)
6.2 (0.32)
5.7 (0.26)
5.0 (0.33)
4.4 (0.27)
4.8 (0.29)
12.2 (0.45)
5.5 (0.36)
7.5 (0.49)
-- (--)
-- (--)

7.6
2.4
1.0
1.1
0.7
0.9
1.0
0.4
1.3
2.7
2.2
1.7
1.3
1.2
0.9
2.1
1.6
1.2
1.9
2.0
8.5
6.2
5.3
7.0
5.3
6.0
12.2
5.7
8.9
5.1
3.8

(0.80)
(0.42)
(0.22)
(0.31)
(0.20)
(0.30)
(0.34)
(0.25)
(0.40)
(0.66)
(0.53)
(0.50)
(0.38)
(0.35)
(0.30)
(0.50)
(0.39)
(0.40)
(0.49)
(0.50)
(1.06)
(0.92)
(0.93)
(1.08)
(0.90)
(1.04)
(1.47)
(1.00)
(1.64)
(1.15)
(1.26)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
-- Not available.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to selfadministered.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not answer HASJOIN,
he or she is defined as not having used a proxy. Respondents who were legitimately skipped from answering question QP01 were
excluded from this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (i.e., 2011 comparison proxy
compared with 2012 QFT proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing data.
5
Estimate is based on an edited version of the variable.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

P-3

Table P-3 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 12 to 17:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire
Field Test

Instrument Item
Covered by Private Health
Insurance? (QHI06) 4,5

2011
2012
2012
2011
Comparison1 Comparison1,3 2012 QFT1,2 Comparison1 Comparison1,3 2012 QFT1,2
Proxy
Proxy
Proxy
No Proxy
No Proxy
No Proxy
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)
63.0 (0.58)

62.5 (0.78)

58.9 (3.06)

51.7a (1.37)

49.2a (2.04)

31.5* (5.84*)

Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5

86.8a (0.54)

87.6a (0.78)

78.0 (3.52)

64.6 (2.29)

60.4 (3.50)

43.3* (16.72*)

Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5

86.7a (0.56)

86.8a (0.81)

78.1 (3.16)

64.6 (2.34)

59.3 (3.52)

44.6* (17.16*)

Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems? (QHI10)4,5

92.9 (0.36)

92.8 (0.45)

88.6 (2.69)

82.7 (1.57)

81.1 (2.74)

57.9* (16.19*)

In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5

11.9 (0.41)

10.7 (0.43)

12.1 (1.88)

14.3 (0.97)

13.4 (1.12)

16.4* (4.18*)

In [YEAR], did [FILL] receive
supplemental Security Income or
SSI? (QI03N)4,5

7.5 (0.31)

8.0 (0.39)

9.4 (1.81)

8.2 (0.73)

6.2 (0.81)

14.5* (5.42*)

In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job or
business? (QI05N)4,5

89.4a (0.36)

89.4a (0.47)

64.0 (2.73)

91.8a (0.73)

92.5a (0.91)

74.8* (7.17*)

In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5

20.2a (0.45)

20.4a (0.65)

26.7 (2.64)

25.0 (1.15)

26.9 (1.56)

37.9* (7.59*)

At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or county
welfare program such as
[TANFFILL]? (QI08N)4,5

4.1 (0.23)

3.9 (0.33)

5.5 (1.20)

5.1 (0.63)

4.3 (0.62)

5.7* (3.25*)

In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5

4.2 (0.21)

4.2 (0.29)

6.3 (1.33)

5.9a (0.60)

5.5a (0.80)

0.0* (0.00*)

Before taxes and other deductions,
was [MEMBER] total personal
income from all sources during
[YEAR] more or less than
20,000 dollars? (QI20N)4,5
$20,000 or More
Less Than $20,000

0.4a (0.07)
99.6a (0.07)

0.4a (0.10)
99.6a (0.10)

6.5 (1.42)
93.5 (1.42)

0.5a (0.13)
99.5a (0.13)

0.9 (0.30)
99.1 (0.30)

10.1* (4.73*)
89.9* (4.73*)

See notes at end of table.

(continued)

P-4

Table P-3 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 12 to 17:
Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and Questionnaire Field
Test Data (continued)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B) 4,5
Less Than $1,000
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
$6,000-$6,999
$7,000-$7,999
$8,000-$8,999
$9,000-$9,999
$10,000-$10,999
$11,000-$11,999
$12,000-$12,999
$13,000-$13,999
$14,000-$14,999
$15,000-$15,999
$16,000-$16,999
$17,000-$17,999
$18,000-$18,999
$19,000-$19,999
$20,000-$24,999
$25,000-$29,999
$30,000-$34,999
$35,000-$39,999
$40,000-$44,999
$45,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000 or More
$100,000-$149,999
$150,000 or More

85.3 (0.35)
4.4 (0.16)
2.4a (0.17)
1.6 (0.13)
1.2 (0.10)
0.9 (0.09)
0.8 (0.09)
0.7a (0.08)
0.6 (0.10)
0.4a (0.07)
0.3 (0.05)
0.2 (0.04)
0.3 (0.09)
0.1 (0.04)
0.1a (0.04)
0.1 (0.04)
0.0 (0.02)
0.0a (0.01)
0.1 (0.03)
0.1 (0.04)
0.1a (0.02)
0.1 (0.03)
0.0 (0.02)
0.0 (0.01)
0.0* (0.00*)
0.0* (0.00*)
0.1 (0.03)
0.0 (0.02)
0.0a (0.02)
-- (--)
-- (--)

85.8 (0.46)
4.3 (0.29)
2.2a (0.19)
1.6 (0.16)
1.1 (0.13)
0.6 (0.10)
0.6 (0.09)
0.8a (0.10)
0.7 (0.10)
0.4a (0.09)
0.5 (0.08)
0.2 (0.06)
0.3 (0.07)
0.1 (0.04)
0.1a (0.05)
0.1 (0.05)
0.1 (0.04)
0.1 (0.03)
0.1 (0.04)
0.1 (0.04)
0.2a (0.05)
0.1 (0.05)
0.1 (0.03)
0.0* (0.00* )
0.0 (0.02)
0.0* (0.00* )
0.0* (0.00* )
0.0* (0.00* )
0.1a (0.04)
-- (--)
-- (--)

82.2 (2.18)
4.1 (1.14)
0.8 (0.48)
1.4 (0.65)
1.0 (0.50)
0.4 (0.30)
0.8 (0.50)
0.2 (0.18)
0.4 (0.30)
0.0* (0.00*)
0.3 (0.27)
0.2 (0.23)
0.2 (0.20)
0.1 (0.10)
0.0* (0.00*)
0.5* (0.48*)
0.3 (0.24)
0.0* (0.00*)
0.1 (0.09)
0.5 (0.39)
4.2 (1.06)
0.8 (0.45)
0.4* (0.44*)
0.0* (0.00*)
0.0* (0.00*)
0.2 (0.23)
0.4 (0.26)
0.2 (0.24)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)

78.6a (0.98)
7.5 (0.64)
4.2 (0.44)
2.5 (0.35)
1.4 (0.26)
1.2a (0.28)
1.1 (0.27)
0.3a (0.10)
0.4a (0.12)
0.3a (0.11)
0.7 (0.16)
0.1 (0.08)
0.1 (0.06)
0.1 (0.05)
0.1 (0.09)
0.5 (0.17)
0.0 (0.03)
0.4a (0.17)
0.0 (0.04)
0.0 (0.03)
0.1 (0.06)
0.0 (0.02)
0.0* (0.00*)
0.0 (0.03)
0.0* (0.00*)
0.0* (0.00*)
0.0 (0.03)
0.0* (0.00*)
0.0 (0.03)
-- (--)
-- (--)

78.8a (1.30)
9.3 (0.95)
3.5 (0.54)
2.5 (0.48)
1.1 (0.25)
0.6a (0.19)
0.9 (0.33)
0.7a (0.22)
0.4a (0.17)
0.0 (0.05)
0.6 (0.27)
0.3 (0.17)
0.1 (0.06)
0.1 (0.12)
0.0 (0.02)
0.1 (0.05)
0.0 (0.04)
0.0* (0.00*)
0.1 (0.15)
0.0* (0.00*)
0.3 (0.22)
0.0* (0.00*)
0.3 (0.17)
0.1 (0.07)
0.0* (0.00*)
0.1 (0.07)
0.0* (0.00*)
0.0* (0.00*)
0.2 (0.10)
-- (--)
-- (--)

63.6* (7.10*)
11.7* (4.46*)
2.7* (2.73*)
2.3* (2.25*)
1.3* (1.29*)
0.0* (0.00*)
1.7* (1.73*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
1.3* (1.36*)
0.0* (0.00*)
2.0* (1.97*)
1.5* (1.46*)
0.0* (0.00*)
1.4* (1.42*)
1.5* (1.53*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
2.4* (2.20*)
0.0* (0.00*)
4.3* (3.07*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
2.1* (1.93*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
-- Not available.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to selfadministered.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not answer
HASJOINhe or she is defined as not having used a proxy. Respondents who were legitimately skipped from answering question
QP01 were excluded from this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (i.e., 2011 comparison proxy
compared with 2012 QFT proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing data.
5
Estimate is based on an edited version of the variable.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

P-5

Table P-4 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 18 or
Older: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and 2012
Questionnaire Field Test

Instrument Item
Covered by Private Health
Insurance? (QHI06) 4,5

69.6 (1.84)

60.1 (5.55)

70.0 (0.50)

69.8 (0.67)

65.5 (2.24)

Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5

81.7 (1.82)

81.5 (2.27)

69.2* (8.71*)

85.1a (0.53)

85.0a (0.82)

77.0 (2.14)

Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5

81.8 (1.88)

81.3 (2.28)

74.4* (6.19*)

84.2a (0.54)

84.6a (0.85)

75.0 (2.26)

Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems? (QHI10)4,5

89.8a (1.28)

89.2 (1.68)

77.6* (5.92*)

92.0a (0.33)

92.5a (0.55)

85.9 (1.78)

In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5

35.4 (1.61)

33.3 (2.60)

33.7 (5.20)

27.9 (0.54)

26.6 (0.61)

26.6 (2.09)

In [YEAR], did [FILL] receive
supplemental Security Income
or SSI? (QI03N)4,5

10.2 (0.97)

10.0 (1.12)

10.7 (3.20)

6.5a (0.23)

7.6 (0.40)

9.3 (1.18)

In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job or
business? (QI05N) 4,5

78.0a (1.38)

81.4a (1.78)

63.5 (4.30)

87.0a (0.43)

87.4a (0.51)

71.5 (1.93)

In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5

15.2 (1.25)

14.4 (1.31)

20.7 (3.99)

13.0 (0.36)

14.3 (0.47)

14.8 (1.66)

At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or county
welfare program such as
[TANFFILL]? (QI08N)4,5

2.3 (0.38)

2.0 (0.41)

2.1 (1.30)

2.2 (0.13)

2.0 (0.16)

2.6 (0.60)

In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5

3.5 (0.52)

4.1 (0.70)

3.3 (1.77)

3.0 (0.15)

2.6 (0.16)

2.9 (0.59)

Before taxes and other
deductions, was [MEMBER]
total personal income from all
sources during [YEAR] more or
less than 20,000 dollars?
(QI20N)4,5
$20,000 or More
Less Than $20,000

35.5 (1.81)
64.5 (1.81)

37.6 (2.01)
62.4 (2.01)

33.7 (5.05)
66.3 (5.05)

59.8a (0.46)
40.2a (0.46)

59.7a (0.62)
40.3a (0.62)

65.8 (1.76)
34.2 (1.76)

See notes at end of table.

(continued)

P-6

Table P-4 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons Aged 18 or
Older, Percentages and Standard Errors, 2011 Comparison, 2012 Comparison, and
Questionnaire Field Test Data (continued)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B) 4,5
Less Than $1,000
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
$6,000-$6,999
$7,000-$7,999
$8,000-$8,999
$9,000-$9,999
$10,000-$10,999
$11,000-$11,999
$12,000-$12,999
$13,000-$13,999
$14,000-$14,999
$15,000-$15,999
$16,000-$16,999
$17,000-$17,999
$18,000-$18,999
$19,000-$19,999
$20,000-$24,999
$25,000-$29,999
$30,000-$34,999
$35,000-$39,999
$40,000-$44,999
$45,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000 or More
$100,000-$149,999
$150,000 or More

20.4 (1.24)
3.6 (0.39)
3.8 (0.50)
2.4 (0.37)
1.9 (0.27)
3.7a (0.64)
3.7a (0.91)
2.6 (0.39)
2.0 (0.30)
3.5 (0.67)
2.7 (0.46)
1.5 (0.34)
2.2 (0.61)
1.8a (0.50)
1.5 (0.42)
1.2a (0.25)
0.6 (0.23)
1.9 (0.76)
2.2a (0.54)
2.0 (0.44)
6.1 (0.80)
5.9 (0.89)
4.3 (0.83)
3.0 (0.56)
3.4 (0.63)
2.9 (0.56)
6.1 (0.77)
2.2 (0.50)
1.1 (0.33)
-- (--)
-- (--)

19.3 (1.14)
4.3 (0.62)
3.4 (0.55)
2.8 (0.54)
1.8 (0.34)
2.1 (0.52)
1.8 (0.37)
1.7 (0.43)
2.7 (0.66)
3.8 (1.18)
3.0 (0.58)
2.1 (0.50)
3.3 (0.87)
2.4a (0.70)
1.0 (0.37)
1.4a (0.42)
1.0 (0.42)
0.5 (0.21)
1.9a (0.54)
1.7 (0.64)
6.6 (1.06)
4.3 (0.81)
4.6 (0.94)
3.7 (1.01)
4.4 (1.25)
3.4 (0.76)
6.3 (0.96)
1.5 (0.46)
3.1 (0.92)
-- (--)
-- (--)

21.6 (4.06)
4.9 (1.27)
3.1 (1.83)
2.8 (1.14)
4.9* (2.75*)
1.4 (0.71)
1.1 (0.65)
0.9* (0.89*)
1.9 (1.03)
4.4* (2.58*)
6.3 (2.58)
0.9 (0.65)
1.2* (1.22*)
0.4* (0.40*)
1.9* (1.37*)
0.0* (0.00*)
2.7* (1.96*)
2.7* (1.99*)
0.5* (0.46*)
2.5* (1.72*)
4.0* (2.42*)
4.8 (2.50)
4.5* (2.56*)
2.2* (1.50*)
2.6 (1.61)
4.7* (2.52*)
5.2 (2.64)
3.8* (2.30*)
2.2 (1.33)
2.2 (1.33)
0.0* (0.00*)

8.9a (0.22)
1.7 (0.10)
1.5a (0.09)
1.3 (0.09)
1.3a (0.08)
1.6a (0.10)
1.4 (0.11)
1.6a (0.11)
1.8 (0.11)
1.8 (0.11)
2.3 (0.15)
1.6 (0.10)
2.2a (0.13)
1.6 (0.12)
1.6a (0.11)
1.8 (0.11)
1.3 (0.10)
1.4 (0.09)
1.8 (0.11)
1.8 (0.12)
6.9 (0.24)
6.8 (0.32)
6.1 (0.27)
5.1 (0.23)
4.5 (0.21)
4.3 (0.19)
12.3 (0.35)
5.8 (0.24)
8.0 (0.36)
-- (--)
-- (--)

8.8a (0.34)
1.8 (0.15)
1.4 (0.12)
1.5 (0.15)
1.1 (0.12)
1.4 (0.11)
1.7 (0.17)
1.6a (0.18)
1.8 (0.18)
1.8 (0.16)
2.2 (0.17)
1.8 (0.18)
2.7a (0.25)
1.3 (0.13)
1.8a (0.16)
1.7 (0.14)
1.3 (0.12)
1.2 (0.12)
1.7 (0.17)
1.8 (0.17)
6.9 (0.34)
6.4 (0.33)
5.9 (0.27)
5.2 (0.33)
4.5 (0.28)
4.9 (0.30)
12.5 (0.46)
5.7 (0.37)
7.7 (0.51)
-- (--)
-- (--)

6.7
2.3
1.0
1.1
0.6
0.9
0.9
0.4
1.3
2.8
2.2
1.7
1.2
1.1
0.9
2.1
1.6
1.2
1.9
2.0
8.6
6.3
5.3
7.1
5.4
6.1
12.4
5.8
9.0
5.2
3.8

(0.81)
(0.42)
(0.22)
(0.32)
(0.20)
(0.30)
(0.34)
(0.25)
(0.41)
(0.67)
(0.54)
(0.51)
(0.38)
(0.35)
(0.30)
(0.50)
(0.40)
(0.40)
(0.50)
(0.51)
(1.08)
(0.94)
(0.94)
(1.09)
(0.91)
(1.06)
(1.49)
(1.02)
(1.67)
(1.17)
(1.28)

P-7

P-8

Appendix Q: Protocol Changes Considered for the Dress
Rehearsal and Whether the Changes Will Be Implemented
for the Dress Rehearsal

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Screening

N/A

Program a Spanish-language version LeBaron
of the screening program for the DR.

Change for DR.

Yes

Screening

N/A

In the screening, if a R indicates
"Other" or "Don't Know/Refused"
on the Race or Hispanic questions,
remove the "Other" and
"Unspecified" designation that FIs
read to the R when verifying the
roster information. There will be no
automatic fill for the race or
ethnicity of the roster member in
cases where the response is "Don't
Know," "Refused," or "Other." Fills
will only be provided for items
where the R has chosen one of the
offered response categories.

LeBaron

Change for DR.

Yes

Changes mirror updates made
to the screening program for
the 2013 NSDUH.

Yes

Screening

N/A

Make edits to the screening program LeBaron
to exit when the SR is younger than
17.

Change for DR.

Yes

Changes mirror updates made
to the screening program for
the 2013 NSDUH.

Yes

Debriefing
questions

LeBaron
Section 5.3 For QFTDBF17a, "Which of the
following describes the problems
with the proxy's use of ACASI in
answering the income and health
insurance questions?" 72% answered
"Other." Consider adding an
"OTHER, Specify" question.

Change for DR.

Yes

There was no follow-up
question in the QFT to clarify
the "other" category.
SAMHSA approved the
addition of this item.

Yes

Debriefing
questions

Section
5.5.4.2

Change for DR.

Yes

Main study debriefing does
have an open-ended question
for comments. SAMHSA
approved the addition of this
item.

Yes

Q-1

During focus groups, FIs suggested
adding a field to the debriefing
questions to record comments about
the case.

LeBaron

Yes

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Q-2

Debriefing
questions

N/A

Edit debriefing items to reflect
analytic goals of the DR and to
measure functionality of items that
have the potential to change.

LeBaron

Change for DR.

Yes

SAMHSA approved the FI
Debriefing items on 4/17/13.

Yes

Screening

N/A

Delete the physical characteristics
screen of the screener, as it is not
used in analysis.

LeBaron

Change for DR.

Yes

RTI and SAMHSA confirmed
the deletion of this screen on
4/23/13.

Yes

Screening

N/A

Delete the controlled access screen
of the screener, as it is not used in
analysis.

LeBaron

Change for DR.

Yes

RTI and SAMHSA confirmed
the deletion of this screen on
4/23/13.

Yes

Screening

N/A

Correct bug in the screening
program that causes the instrument
to freeze.

LeBaron

Change for DR.

Yes

This bug was corrected so that
the DR instrument performed
as intended, and was not a
change from the QFT per se.

Yes

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; FI = field interviewer; NSDUH = National Survey on Drug Use and Health; QFT = Questionnaire
Field Test; R = respondent; RTI = Research Triangle Institute; SR = screening respondent; SAMHSA = Substance Abuse and Mental Health Services Administration.

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Develop Spanish-language version
of questionnaire for DR.

LeBaron

Change for DR.

Yes

CAI

N/A

Investigate the limits for the hard
error after QD11 using QFT data.

LeBaron

No change for
DR.

Yes

Limits were investigated and a
decision was reached with
SAMHSA not to add a hard
error.

N/A

CAI

N/A

Add PENTER1 before
ENDAUDIO to lock the ACASI
portion of the interview.

LeBaron

Change for DR.

Yes

Edit should match change in
2013 questionnaire.

Yes

CAI

N/A

LeBaron
Add adult family members to the
list of available proxies (QP02)
when the adult family members
ages=DK or REF. Add language in
the specifications to note that this
edit was made.

Change for DR.

Yes

Edit matches change in 2013
questionnaire; added a note in
the specs to make clear that this
change was made.

Yes

CAI

N/A

Change logic in MJMM so that
anyone reporting past year blunt
use in BL02 is routed to MJMM.

LeBaron

Change for DR.

Yes

Edit should match change in
2013 questionnaire.

Yes

CAI

N/A

Remove PREVCOM when R is 12
to 17 because R could not have
been a proxy on a previous
interview.

LeBaron

Change for DR.

Yes

Approved for revision during
QFT training, but reserved for
DR update.

Yes

CAI

N/A

Change the data structure on TX10 LeBaron
to allow R to choose all 12 possible
options.

Change for DR.

Yes

CAI

N/A

Remove "…including Indian
Health Insurance" from QHI11.

Change for DR.

Yes

Q-3

Item
No.

LeBaron

DR Action

RTI Comments

Revision
Complete

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Q-4

Item
No.

Activity

CAI

N/A

Fix skip pattern for "source of
prescription drug" questions so
they no longer skip 12 to 17 year
olds per Larry Kroutil's email on
9/27/12 (PRY42C, TRY21C,
STY26C, SVY19C).

LeBaron

Change for DR.

Yes

CAI

N/A

Add "headphones" back to
IntroACASI1 "…you will do an
important part of this interview on
your own, using the computer and
headphones."

LeBaron

Change for DR.

Yes

CAI

N/A

In ANYQUES, add "please" back
to the question and re-record.

LeBaron

Change for DR.

Yes

CAI

N/A

On CG39, RCG39, and RRCG39,
Macanudo should be singular.

LeBaron

Change for DR.

Yes

CAI

N/A

For PRINTROYR2 and similar
LeBaron
questions, add "and" before the last
drug in the list.

Change for DR.

Yes

CAI

N/A

For PRYMOTIV, the upward
inflection after "...that time?"
sounds strange and should be rerecorded.

LeBaron

Change for DR.

Yes

CAI

N/A

Bold "feet," "inches," "meters,"
"centimeters," "pounds," and
"kilograms" in the specifications
(HLTH05-HLTH14). No update
needed for the DR instrument
because the QFT instrument
included this bolding.

LeBaron

Change for DR
(specs only).

Yes

DR Action

RTI Comments

Revision
Complete
Yes

Gather feedback from DR FIs.

The instrument was correct;
only the specs need to be
updated.

Yes

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Change logic on HLTH29 so that if LeBaron
a respondent reports age at first
cancer diagnosis as current age,
HLTH29 is skipped (per an email
sent to SAMHSA on 10/2/12).

Change for DR.

Yes

CAI

N/A

Reword BACKUP/BACKUPB to LeBaron
be less confusing. The revised
question will read: "If you want to
change or see your answer to a
previous question, you can back up
using the [F9] key. Each time you
press the [F9] key, the computer
will go back one question. You can
tell the computer to repeat a
question by pressing [F10]. Try
this now. When you are finished,
press [ENTER] to continue."

Change for DR.

Yes

CAI

N/A

Remove F7 functionality (mute)
from the entire interview. Remove
the introduction to this
functionality from IntroACASI1
and IntrAcasi1b.

LeBaron

Change for DR.

Yes

CAI

N/A

Change TOALLR3 to "As you can
see, this is kept separate from the
answers that were entered, so they
will still be completely private."

LeBaron

Change for DR.

Yes

CAI

N/A

Revise DR with 2013 Medicaid
and CHIP program names in
MEDIFILL, CHIPFILL, and
TANFFILL.

LeBaron

Change for DR.

Yes

Q-5

Item
No.

DR Action

RTI Comments
Correct in QFT specs; Blaise
changes only for DR

Wording change only.

Revision
Complete
Yes

Yes

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

LeBaron
Some fills sound awkward due to
inconsistent inflection. Vicodin and
Provigil are two examples. Need to
determine if fills should be rerecorded. Assess quality of existing
wav files and reach determination
about re-recording.

No change for
DR.

Yes

CAI

N/A

Added questions about sexual
orientation using NCHS as a
model.

LeBaron

Change for DR.

Yes

CAI

N/A

There is a concern that the ACASI
voice does not pronounce the drug
names until the response options
are read. Respondents often do not
wait to hear all response options
before entering their answer. Once
a response is entered, the audio
pauses.

LeBaron

No change for
DR.

Yes

CAI

N/A

Change INTROINC to make audio LeBaron
transitions less choppy, and use
passive voice to list the family
members. For example, "…kinds
and amounts of income received by
your son and his family, that is,
your son, you, his father and sister
living here."

Change for DR.

Yes

CAI

N/A

Revise the response for reporting
no use of prescription drugs in the
prescription drug screeners (PR01,
etc.), perhaps by changing it from
95 to 0.

No change for
DR.

Yes

N/A

Q-6

Item
No.

LeBaron

DR Action

RTI Comments
Defer assessment of quality due
to TTS investigation.

Revision
Complete
N/A

Yes

Investigate rates for these first
drugs after the first few weeks,
and again at the end of data
collection.

N/A

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Q-7

Item
No.

Activity

CAI

N/A

Skip the lead xxM01 question for
LeBaron
prescription drugs (e.g., PRM01) if
the respondent is a past month
initiate (e.g., PR30ANYINIT=1).

Change for DR.

Yes

CAI

N/A

Delete QD42 from the instrument.

LeBaron

Change for DR.

Yes

CAI

N/A

Add "or other health professional"
to the medical marijuana (MJMM,
MJMM01) questions.

LeBaron

Change for DR.

Yes

CAI

N/A

Change the allowable range of the
30-day frequency questions for
prescription drugs (e.g., PRM02)
from 0 to 30 to 1 to 30.

LeBaron

Change for DR.

Yes

CAI

N/A

LeBaron
Add language that references
reports of methamphetamine use in
the special drug module (SD14)
into logic for creating
MET12MON in the substance
dependence and abuse module.

Change for DR.

Yes

CAI

N/A

Add a question to the prescription LeBaron
drug modules that measures
initiation of misuse of prescription
drugs. This issue was first
communicated to SAMHSA on
10/31/12 and 11/1/12. On
11/16/12, Jonaki Bose sent a
proposed follow-up question if Rs
report only past year initiation.
A proposed revision to the question
was sent to SAMHSA on 11/27/12.
The question is XXL03.

Change for DR.

Yes

DR Action

RTI Comments
Correct in QFT specs; Blaise
changes only for DR.

Revision
Complete
Yes

Yes
Edit should match change in
2013 questionnaire.

DR testing will focus on this
question to ensure that the specs
are working correctly.

Yes

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Q-8

Item
No.

Activity

CAI

N/A

Edit the logic for the motivation
questions (XXYMOTIV) so that it
no longer skips Rs out of these
questions if the only drug they
misused in the past year is "any
other drug" in the category. This
issue was noted to the Instrument
Development team on 11/29/12.

LeBaron

Change for DR.

Yes

CAI

N/A

Edit QD10 to match the war era
LeBaron
categories to those of the VA.
Vietnam era should start 3/1961 for
those who served in Vietnam in
that period.

Change for DR.

Yes

CAI

Chapter 5

LeBaron
Based on results in debriefing
question QFTDBF12, edit the
wording to PLAYINFO so as to
explain the steps the R must take
more clearly. In some cases, it was
not clear what to do after entering
F2, with some respondents perhaps
not realizing that they must enter a
response after seeing the pop-up
instruction box.

Change for DR.

Yes

Also add reminder to training to
tell FIs what to do if a R asks
about F2.

Yes

CAI

Section
5.5.4.3

In a focus group, an FI suggested a LeBaron
darker color to highlight dates
because the current colors are
difficult to see in sunlight.

No change for
DR.

Yes

There are no plans to change the
color for CAI dates. Gather
feedback from DR FIs on
visibility of new laptop screen in
different environments.

N/A

DR Action

RTI Comments

Revision
Complete

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

Section
5.5.4.3

In the focus groups, FIs suggested LeBaron
the tutorial be clearly labeled as a
practice session, or the introduction
be emphasized. They reported that
Rs struggled with providing
accurate answers to questions and
were confused by the lack of
concordance with the question
topics and the NSDUH study
description. To address this issue,
label the tutorial items as Practice
Question #1, Practice Question #2,
etc.

Change for DR.

Yes

CAI

Section
5.5.4.6

In the focus groups, FIs provided
Zelko/
LeBaron
general feedback that they would
like to do away with the showcards
and move the demographic
questions to be self-administered.

No change for
DR.

Yes

CAI

Add an "OTHER, Specify"
question to the prescription drug
reasons for misuse decomposition
question.

LeBaron

No change for
DR.

Yes

N/A

CAI

Chapter 9

Add "OTHER, Specify" questions
for the prescription drug screeners.

LeBaron

No change for
DR.

Yes

N/A

CAI

N/A

Due to respondent complaints and
confusion that the type of music
they listen to is not listed on
ALLAPPLY in the tutorial
questions, delete "9 Techno" and
replace it with "9 Something Else"
to limit respondent issues.

LeBaron

Change for DR.

Yes

Q-9

Item
No.

DR Action

RTI Comments

Revision
Complete
Yes

SAMHSA reviewed electronic
showcards and the text was too
small on the screen. There are
no plans to move demographics
to ACASI.

N/A

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Change the wording of TOALLR3I LeBaron
to remind FIs that, in an interview
with a minor R, a parent or
guardian should sign the QC form
if possible. Specifically, change the
first interviewer note to "[GIVE
QUALITY CONTROL FORM
AND ENVELOPE TO
RESPONDENT (OR
PARENT/GUARDIAN OF
YOUTH RESPONDENT, IF
AVAILABLE]."

Change for DR.

Yes

CAI

N/A

Edit the ranges to the height
questions (HLTH05 - HLTH08).
This change was also made to the
2013 (Q2-Q4) and 2014
instrument.

LeBaron

Change for DR.

Yes

CAI

N/A

Edit the language to the military
family questions (QD10d-QD10f).
QD10f will only be included for
the DR and will be deleted for the
2015 instrument.

LeBaron

Change for DR.

Yes

CAI

N/A

Edited response options for
QD10b1 to reflect correct eras for
military service.

LeBaron

Change for DR.

Yes

CAI

N/A

Edited INTRO2 to instruct R that
LeBaron
he or she can turn down the volume
of the voice.

Change for DR.

Yes

Q-10

Item
No.

DR Action

RTI Comments

Revision
Complete
Yes

Edit should match change in
2013 questionnaire.

Yes

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Added logic to define new
variables PRYRDKRE1,
TRYDKRE1, STYDKRE1, and
SVYDKRE1. These variables will
be used in routing Rs with
unknown recent initiation to the
new questions, PRL03, TRL03,
STL03, and SVL03.

LeBaron

Change for DR.

Yes

CAI

N/A

Added logic to define Rs with
unknown recent initiation of
prescription drug use in all
prescription drug modules.

LeBaron

Change for DR.

Yes

CAI

N/A

Corrected question wording of
PRY02 to be consistent with the
wording of other questions in the
module.

LeBaron

Change for DR.

Yes

CAI

N/A

LeBaron
Added new questions, PRL03,
TRL03, STL03, and SVL03 (and
appropriate routing), which ask
about initiation of misuse of
prescription drugs more than 12
months ago if the only definite
reports of initiation occurred in the
past 12 months, or all initiation
data were missing. These questions
were added to produce accurate
estimates of recent initiation.

Change for DR.

Yes

CAI

N/A

Removed unnecessary routing
logic from PRYMOTIV,
TRYMOTIV, STYMOTIV, and
SVYMOTIV for accuracy.

Change for DR.

Yes

Q-11

Item
No.

LeBaron

DR Action

RTI Comments

Revision
Complete

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Q-12

Item
No.

Activity

CAI

N/A

Edited PRYMOT1 response
options for clarity. Reversed the
order of response options 5 and 6
to match the order of response
options in the tranquilizers and
sedatives main modules. Revised
the wording of the new response
option 5 so that it is parallel to
similar response options.

LeBaron

Change for DR.

Yes

CAI

N/A

Deleted extraneous routing of
LeBaron
PRY42BSP, PRY42C,
TRY21BSP, TRY21C,
STY26BSP, STY26C, SVY19BSP,
and SVY19C for accuracy.

Change for DR.

Yes

CAI

N/A

Renumbered TRY21B to be
consecutive.

LeBaron

Change for DR.

Yes

CAI

N/A

Edited routing of MJMM01 to
LeBaron
include Rs who used blunts in the
past year but didn't report past year
marijuana use in the core module.

Change for DR.

Yes

CAI

N/A

Renamed medical marijuana
questions to MJMM01 and
MJMM02 for consistency with the
2013 NSDUH questions.

LeBaron

Change for DR.

Yes

CAI

N/A

Added "B or C" to "Hepatitis" in
HLTH25 for more precise
description of condition.

LeBaron

Change for DR.

Yes

DR Action

RTI Comments

Revision
Complete

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Switched the order of response
options 12 and 13 for QD18CC04
to match QD11. Corrected the
response option numbers for
doctorate and professional degrees
to be consecutive.

LeBaron

Change for DR.

Yes

CAI

N/A

Added a new variable,
PENTER1B, which instructs
respondents to lock the ACASI
portion of the instrument before
returning the computer to the
interviewer.

LeBaron

Change for DR.

Yes

CAI

N/A

Discussed adding DAUTYPE and
SONTYPE back into instrument
from 2013 main study, but with
modified logic. This decision was
ultimately reversed, and the
variable will not be added.

LeBaron

No change for
DR.

Yes

N/A

CAI

N/A

Edited the ranges for the weight
items (HTH10-HLTH14) to be
more inclusive of extreme values

LeBaron

Change for DR.

Yes

CAI

N/A

Skip the xxM03 30-day
prescription drug with alcohol
questions (e.g., PRM03) if
ALCUSE30 NE 1.

LeBaron

Change for DR.

Yes

CAI

N/A

Edited routing of MJMM01 to
include Rs who used blunts in the
past 30 days but didn't report past
year marijuana use in the core
module.

LeBaron

Change for DR.

Yes

Q-13

Item
No.

DR Action

RTI Comments

This was specified correctly, but
was not programmed correctly
in the QFT instrument.

Revision
Complete

Yes

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Q-14

Item
No.

Activity

CAI

N/A

PR07: Can we make the Duragesic LeBaron
picture large enough to read the
largest type?

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

New audio needs to be recorded for LeBaron
QHI07, QHI08, QHI13, and
PRY01 and parallel questions.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Audio edited in LS01i and
HALINTRO to fix tone and
pronunciation issues.

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Change to MJMM01 logic to
include BL04 = 2

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

LeBaron
Edit the specs to base logic in
QP02 on the presence of an "Adult
Family Member," as opposed to an
"Other Person" in the household.
No changes to the CAI are
required.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

In IntrAcasi1b, an optional
transition will be added to this
interviewer-administered question.
This intro will say, "Your
[daughter, etc.] has said you are
better able to answer the questions
about [her] health insurance and
the family income."

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

In Anyques, add the word "Please" LeBaron
to the screen. It was missing during
testing.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

DR Action

RTI Comments

Revision
Complete

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

In calendr3, add a statement that
says, "Press F1 again to close the
calendar."

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Delete reminders about the F2
LeBaron
function in the prescription drug
main modules in all questions other
than the Age at First Use questions.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Edit TRY13a, TRY16a, TRY17a, LeBaron
and TRY18a to remove the "also
known as" phrase. This phrase will
also be dropped from month and
year of last use questions,
consistency check questions, and
the TRFILL2 and TRNAMEFILL
fills.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Edit QD26 and QD27 to change
the text about the F2 note. The
instruction should say, "Press F2
for information about unpaid
work."

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Re-record audio files for zolpidem
and meprobamate to correct
pronunciation.

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

Q-15

Item
No.

DR Action

RTI Comments

Revision
Complete

Item
No.

Activity

CAI

QFT Report
Section
QFT Issue/Potential DR Change
N/A

Edit QHI06. The new text should
read, "Private health insurance can
be obtained through work, such as
through an employer, union, or
professional association, or by
paying premiums directly to a
health insurance company. It
includes coverage by a health
maintenance organization (HMO),
fee for service plans, and single
service plans.

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

LeBaron

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

[Are you/Is SP] covered by private
health insurance?"

Q-16

CAI

N/A

LeBaron
Edit QI03N. The new wording is,
"Supplemental Security Income or
SSI is a program administered by a
government agency that makes
assistance payments to low income,
aged, blind, and disabled persons.
This is not the same as Social
Security. In [CURRENT YEAR 1], did you receive Supplemental
Security Income or SSI?"

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

Delete QI05N, the question about
receiving wages from a job or
business.

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

LeBaron

QFT Report
Section
QFT Issue/Potential DR Change

Responsible
Person(s)

SAMHSA
Approved DR
Action

Activity

CAI

N/A

Edit the list of income sources used LeBaron
in INTRTINN, as well as the
introductory text. The new wording
is, "Below is a list of some possible
sources of income. When you
answer the next questions, please
consider these income sources as
well as those asked about in earlier
questions."
Income earned at a job or business
Retirement , disability, or survivor
pension
Unemployment or worker's
compensation
Veteran's Administration payments
Child support
Alimony
Interest income
Dividends from stocks or mutual
funds
Income from rental properties,
royalties, estates or trusts

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

CAI

N/A

LeBaron
Edit the wording to QI07N. The
new wording will be, "The
Supplemental Nutrition Assistance
Program (SNAP), formerly known
as food stamps, provides assistance
for buying food. A special card is
issued which can be used to buy
food in grocery stores. In [year],
did [you/family member fill]
receive food stamp benefits?"

Change for DR.

Yes

Requested by Peggy Barker
during testing.

Yes

Q-17

Item
No.

DR Action

RTI Comments

Revision
Complete

Item
No.

QFT Report
Section
QFT Issue/Potential DR Change

Activity

CAI

N/A

Create a fill for QI07N to
customize State-specific names of
food stamp programs.

CAI

N/A

CAI

Responsible
Person(s)
LeBaron

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Yes.

LeBaron
Added language to the specs
describing the hard error in
HLTH27 through HLTH28cc that
is triggered if an age at first
diagnosis is older than current age.
This change was made to the specs
only because the hard error was
already present in the program.

Change for DR
(specs only).

Yes

N/A

Edit the specs and program so that LeBaron
IntrAcasi1B, IntrAcasi3b, and
IntrAcasi4b in the back-end proxy
tutorial are "Press 1 and Enter to
continue," as opposed to just
requiring that "Enter" is pressed.
This will allow bilingual
interviewers to toggle between
languages in the event that a proxy
wishes to complete the back-end
ACASI in a different language than
the respondent.

New for DR,
given inclusion of
Spanish.

Yes

CAI

N/A

Correct bug in one of the testing
LeBaron
versions, where audio was dropped
for four tranquilizers in TRY21a in
the main module.

Update DR test
program.

Yes

This edit was made so that the
instrument performed as
intended and was not a change
from the QFT per se.

Yes

CAI

N/A

Edit INTRTNN, because the word
"earned" was spelled wrong.

Update DR test
program.

Yes

This edit refined the change
requested in item 78.

Yes

Q-18

No change for
DR.

LeBaron

Due to time constraints, reserve
this item for 2015 specifications.

Revision
Complete
N/A

Item
No.

Activity

CAI

QFT Report
Section
QFT Issue/Potential DR Change
N/A

Update a few items in the Spanish
instrument to reflect current
wording and translations.
Corresponding edits were not
needed in the English instrument.

Responsible
Person(s)
LeBaron

DR Action
Edit Spanishlanguage DR
specs and
program.

SAMHSA
Approved DR
Action
Yes

RTI Comments

Revision
Complete
Yes

ACASI = audio computer-assisted self-interviewing; CAI = computer-assisted interviewing; CHIP = Children's Health Insurance Program; DR = Dress Rehearsal; FI = field
interviewer; NSDUH = National Survey on Drug Use and Health; N/A = not applicable; NCHS = National Center for Health Statistics; QC = quality control; QFT = Questionnaire
Field Test; R = respondent; RTI = Research Triangle Institute; SAMHSA = Substance Abuse and Mental Health Services Administration; specs = specifications; TTS = text to
speech; VA = Department of Veterans Affairs.

Q-19

Item
No.

Activity

QFT Report
Section

QFT Issue/Potential DR
Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Materials

N/A

Add the OMB number to the
study description.

McKamey

Change for DR.

Yes

Spanish will be on the
reverse side of the SD.

Yes

Materials

N/A

Add the burden statement to the
study description.

McKamey

Change for DR.

Yes

Spanish will be on the
reverse side of the SD.

Yes

Training/
handbook

Section 5.2

During field observations, two FIs McKamey
had issues troubleshooting
unexpected events with the tablet,
such as an alarm going off during
a screening. These
troubleshooting issues will be
handled for the DR by addressing
these specific items during
training and adding
documentation to the FI handbook
on how to resolve these
occurrences.

Change for DR.

Yes

This topic was included in
the DR FI training agenda
approved by SAMHSA on
3/6/12.

Yes

Materials

Section
5.5.4.1

During focus groups, when
discussing the lead letter, some
FIs mentioned that they
appreciated that the letter was
addressed to "[NAME
County/Parish/District] Resident
at:" and did not just say
"Resident." During training, one
New York City FI indicated that
listing county/parish/district
would not resonate with Rs in his
region.

No change for
DR.

Yes

Because the New York City
FI comment was made at
training before the FI entered
the field and no similar
comments were made after
data collection, no changes
are recommended for DR.

N/A

Q-20

McKamey

Item
No.

Activity

QFT Report
Section

QFT Issue/Potential DR
Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Q-21

FI handbook

Section 5.4.4 In the QFT equipment survey, FIs McKamey
mentioned that they did not
recognize the view letters
function on the tablet. This
feature is available and will be
clarified in the DR FI handbook
and training sessions.

Change for DR.

Yes

FIs can view letters once the
FS has sent them.

Yes

Materials

Section
5.5.4.1

In the focus groups, one FI noted McKamey
that a respondent is pictured using
a paper reference date calendar in
a graphic in the redesigned Q&A
brochure.

Change for DR.

Yes

Picture has been removed
and replaced on the brochure.

Yes

FI portfolio

Section
5.5.4.2

Cohen/Payne
In the focus groups, FIs pointed
out pros and cons of the new
portfolio that was provided at
training. Some said they disliked
the portfolio enough to revert to
using the old one, which is sturdy
and professional. The new one is
slippery and hard to hold. The
tablet, when placed on it, falls off
and materials fall out of it. The
closure is flimsy. FIs would have
preferred a zip closure similar to
the main study portfolio. It also is
difficult to write on top of it, such
as when filling out the quality
control letters. FIs do, however,
like the number of slots in the
portfolio and the clear pockets for
easier access to materials. For the
DR, investigate other portfolio
options and associated costs.

Change for DR.

Yes

Two local FIs reviewed the
selected portfolio options.
RTI sent the FI feedback and
the RTI-recommended
portfolio to SAMHSA on
6/4/13 for review and
approval. Received
SAMHSA approval of
recommended portfolio on
6/10/13 and placed the final
portfolio order on 6/13/13.

Yes

Item
No.

Activity

QFT Report
Section

QFT Issue/Potential DR
Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Q-22

Training

Section
5.5.4.2

McKamey
In the focus groups, some FIs
reported getting into programs or
onto screens early in their
fieldwork that they had not seen
in training and did not know how
to return to the screening
program. Although they felt
comfortable conducting the
screening with the tablet, they
would have preferred more handson training on how to deal with
these unexpected navigational
errors. Additional training on
correcting navigational issues and
potential errors will be
incorporated into the DR training.

Change for DR.

Yes

It is not possible to remove
the multiple home screens
and unused features of the
tablet, so more practice on
how to move off these
screens will be provided in
training. This topic was
included in the DR FI
training agenda approved by
SAMHSA on 3/6/13.

Yes

Training

Section
5.5.4.5

McKamey
In focus groups, FIs mentioned
challenges associated with
making sure that the parent does
not leave the household or
become unavailable before the
child reaches the back end of the
instrument. DR training will be
amended to remind FIs to do their
best to confirm the parent will be
in the house for the entirety of the
interview.

Change for DR.

Yes

This will be addressed in DR
training, but further
discussion with SAMHSA is
needed to determine if this
should be done in a more
formal, standardized manner
in the future. This topic was
included in the DR FI
training agenda approved by
SAMHSA on 3/6/13.

Yes

Item
No.
10

Activity
Materials

QFT Report
Section
Section
5.5.4.5

QFT Issue/Potential DR
Change

Responsible
Person(s)

Q-23

Payne/Zelko
In focus groups, the moderator
asked FIs how they would feel
about having an additional tool
available to help with doorstep
screenings. This tool would
consist of a 20- to 30-second
video clip of the NSDUH press
conference, would be available on
the tablet, and could help with
gaining cooperation. FIs were
enthusiastic about this idea, if the
video was optional and not a
required part of the screening.
One FI suggested having multiple
videos designed to address
common respondent concerns,
such as confidentiality, or
targeted to specific populations,
such as parents or elderly persons.
They said respondents would
think that if it is on television, it is
true. It would also help with
legitimacy and would be short
enough to use at the doorstep.

DR Action
No change for
DR.

SAMHSA
Approved DR
Action
Yes

RTI Comments
Good idea, but consider for
use in the 2015 NSDUH due
to OMB schedules.
Functionality issues within
the tablet also need to be
investigated.

Revision
Complete
N/A

Item
No.

Activity

QFT Report
Section

QFT Issue/Potential DR
Change

Responsible
Person(s)

Materials

N/A

LeBaron/
Make a change to the Intro and
McKamey
Informed Consent for 12 to 17
year olds. Consider removing the
option to skip giving the
respondent a study description at
this point if they have already
received one. Youths are not to
serve as screening respondents, so
would not have had the
opportunity to receive the SD.
The only time youths may have
already received it would be for
youths living independently
without a parent/guardian in the
home if no residents 18 or older
who was SR and then selected.
In that rare case, the youth would
receive two study descriptions.

Materials

N/A

Change the intro and informed
consent text for both youths and
adults to match the wording used
at the end of the interview during
the QC process. Change
"mailing" address to "current"
address.

Materials

N/A

Q-24

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

No change for
DR

Yes

N/A

LeBaron/
McKamey

Change for DR.

Yes

Change wording of Showcard 4 to LeBaron/
match QD10 (Vietnam era should McKamey
start 3/1961 for those who served
in Vietnam in that period.)

Change for DR.

Yes

Item
No.

Activity

QFT Report
Section

QFT Issue/Potential DR
Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Materials

N/A

Add the words "Open/Close" to
the F1 keyboard label that says
"Calendar."

McKamey

Change for DR.

Yes

This phrase will be added on
the label for F1.

Yes

Materials

N/A

Minor updates to the DR
summary of the questionnaire,
including revisions to make all
text in the third person voice.

McKamey

Change for DR.

Yes

Received SAMHSA approval
of revised DR summary on
6/19/13.

Yes

DR = Dress Rehearsal; FI = field interviewer; FS = field supervisor; NSDUH = National Survey on Drug Use and Health; N/A = not applicable; OMB = Office of Management
and Budget; Q&A = question and answer; QC = quality control; QFT = Questionnaire Field Test; R = respondent; RTI = Research Triangle Institute; SAMHSA = Substance
Abuse and Mental Health Services Administration; SD = study description.

Q-25

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Laptop

N/A

New laptop (and case) for
interviewing. Incorporate into
handbook and training for DR.

Meyer/
McKamey

Change for DR.

Yes

SAMHSA selected the
Samsung Ultrabook, and 200
units have arrived at RTI. Two
local FIs reviewed the two
laptop bag options. After
reviewing the FI feedback and
the RTI-recommended laptop
bag, SAMHSA approved the
bag for purchase 6/10/13.
Computer bags have been
ordered.

Yes

Laptop

N/A

The laptops that will be purchased Meyer/
include Ethernet adaptors for FIs McKamey
who do not have wi-fi. However,
there are some areas of the
country where FIs can only
transmit via dial-up when on
travel status. RTI would like to
purchase a small supply (10) of
USB modems for FIs in remote
areas who cannot transmit via the
Internet. In these rare situations,
tech support will FedEx the USB
modem to the FI and provide
instructions for transmission over
the phone.

Change for DR.

Yes

RTI received SAMHSA
approval to purchase the 10
USB modems for the DR
laptops on 5/6/13. The USB
modems were ordered on
6/3/13.

Yes

N/A

Provide a two-way RTI email
account for FIs to use on the
Samsung tablet. Add training on
new tablet email function to
handbook and training.

Change for DR.

Yes

Q-26

Meyer/
McKamey

Yes

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Q-27

Change for DR.

Yes

This option will supplement
the traditional tethered
tablet/laptop transmission
method that they currently and
can continue to use.

Yes

Zelko
In the tablet equipment survey,
two FIs suggested that finalized
cases should be removed from the
select case screen. The view/sort
function on the tablet already
allows FIs to select whether they
want to view pending or final
cases on the select case screen.

No change for
DR.

Yes

Modifying the tablet to hide
finalized cases automatically
could introduce errors.

N/A

Zelko/
In the tablet equipment survey,
McKamey
two FIs noted it would be useful
to have the call distribution
feature available on the tablet so
that they could review the
different days and times they had
visited households. This feature
will be implemented as part of the
DR version of the tablet and
included in training.

Change for DR.

Yes

Because of time constraints in
the development of the QFT
screening program, the call
distribution feature that is
currently on the iPAQ was not
implemented.

Yes

Transmission

N/A

Provide optional wireless tablet
transmission capability that will
allow FIs to transmit the from the
tablet data wirelessly and
independently of laptop.

Tablet view

Section
5.4.4

Tablet features

Section
5.4.4

Meyer/
McKamey

QFT
Report
Section

Q-28

Item
No.

Activity

Tablet features

Section
5.5.4

In focus groups, FIs mentioned
that they would like to have a
larger calendar for appointments,
which is not possible. The
calendar is a default app on the
tablet that cannot be modified or
reformatted (to be "larger").
However, since the QFT was
fielded, a mechanism has been
built into the screening program
for FIs to schedule appointments
for specific cases, integrated with
the default calendar app. DR
training will cover using this tool
with the FIs.

Tablet
accessories

Section
5.4.4

Zelko
In the tablet equipment survey,
several FIs indicated that the
carrying case could be improved
by adding a pen holder in addition
to the stylus holder so that they
could have easy access to a pen
for writing on appointment cards.
Although a couple of FIs
indicated that the neck strap was
too wide on the case and that the
snap was hard to use, a number of
FIs commented that they were
happy the Velcro closure had
been removed.

QFT Issue/Potential DR Change

Responsible
Person(s)
Zelko/
McKamey

DR Action

SAMHSA
Approved DR
Action

Change for DR.

Yes

No change for
DR.

Yes

RTI Comments

Revision
Complete
Yes

Design changes for carrying
case will be considered prior
to the 2015 redesign.

N/A

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Q-29

Tablet
accessories

Section
5.5.4.2

In focus groups, FIs liked the case Zelko
that was designed for the tablet. It
was easy to flip the cover open to
charge. Many FIs reported
disliking the strap for the tablet,
felt it was too bulky and thick,
and indicated that it interfered
with badges and necklaces. Some
reported they would like a pen
holder on the side of the case
opposite the stylus. Several FIs
preferred the magnetic snap
closure to the Velcro closure on
the current iPAQ case.

No change for
DR.

Yes

Design changes for carrying
case will be considered prior
to the 2015 redesign.

N/A

Tablet
functions

Section
5.5.4.2

In focus groups, FIs reported they Zelko
could delete a code, but did not
have the capability to change it.
The difference in the QFT from
the main study was that the FIs
could not "Edit" the numeric code
in the ROC from the dropdown
list (but they can do that on the
iPAQ before the case is
transmitted). This was essentially
a bug in the program, and thus it
should be fixed.

Change for DR.

Yes

The DR screening program
has been modified so that FIs
have the ability to "Edit" a
ROC code (not just
comments) in the same way as
in the iPAQ. Note that after
ROCS are transmitted, they
are frozen, and no edits to the
codes or comments can be
made.

Yes

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)
McKamey

Q-30

Tablet
functions

Section
5.5.4.2

In focus groups, FIs stated that it
was tricky to navigate back to the
verification screen for the
"vacants," but it is possible.
Additional training will be given
to DR FIs on tablet navigation.

Tablet
functions

Section
5.5.4.2

In a focus group, it was reported Zelko
that reentering cases in the tablet
created a time discrepancy in the
case. One FI reported that
pressing "Commit" and pressing
"Done" created two different time
stamps.

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Change for DR.

Yes

To view verification
information on a case coded
10 for vacant , the FI simply
taps and holds the case on
select case screen and selects
"View Verification
Information." The FI is then
taken directly to the
verification screen where he
or she can see information that
has been entered and edit if
needed.

Yes

No change for
DR.

Yes

During the QFT, there was a
data processing issue with the
ROC time discrepancy report
that was incorrectly showing
the modify times (every time
the FI made an edit to
comments) rather than the
create date times, which
caused some confusion for the
field and led to some FIs
showing up on that report who
should not have been. During
the QFT, the data processing
error was fixed so the ROC
time discrepancy report was
showing the correct
information.

N/A

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

RTI Comments

Revision
Complete

Tablet
functions

Section
5.5.4.2

Zelko/
In focus groups, FIs provided
McKamey
feedback on the keyboard. FIs
would like to have apostrophe and
quotation marks available and be
able to remove unnecessary
symbols from the keyboard. They
also indicated that the question
mark was hard to find and
requested that the period be
placed on the same keyboard as
the letters and be available if a
user inserts two spaces after a
sentence. Based on tablet
keyboard evaluation, make
Samsung and hacker keyboards
available to FIs for DR and
provide training on both versions.

Change for DR.

Yes

The layout/design of the
Samsung keyboard cannot be
altered, but the hacker
keyboard will be available for
the DR. Gather FI feedback
after data collection.
Regarding apostrophes and
quotations, those are not
allowed because they could
cause problems with the
coding and data transmission.

Yes

Tablet
functions

Section
5.5.4.2

Zelko
A mixed stylus review was
received from the focus groups;
some FIs did not use the stylus,
saying it was slippery and hard to
insert into the holder on the case,
which caused the holder on the
case to tear.

No change for
DR.

Yes

Investigate stylus options for
the 2015 redesign.

N/A

Q-31

Item
No.

Activity

QFT
Report
Section

QFT Issue/Potential DR Change

Responsible
Person(s)

DR Action

SAMHSA
Approved DR
Action

Q-32

Tablet
accessories

Section
5.5.4.2

Zelko
In focus groups, several FIs
mentioned that a car charger
would be appreciated because the
battery did not last all day. A
travel kit with a car charger is
provided for the iPAQ on the
main study. The iPAQ car charger
can be used to charge the tablet.
However, if a tablet charger is
used on the iPAQ, it could
damage the iPAQ.

Change for DR.

Yes

Tablet
functions

Section
5.5.4.2

In focus groups, FIs reported that Zelko
they would like several of the
iPAQ features to be transferred to
the tablet, specifically for the
CaseID to remain at the top of the
screen on the selections and ROC
screen and a selected line remain
highlighted on the select case
screen. Although it is not possible
to have a selected case remain
highlighted, the highlighting will
remain for a longer time for the
DR. RTI will display the entire
Case ID rather than the last 3
digits on the selections and ROC
screens as it is on the iPAQ.

Change for DR.

Yes

RTI Comments
RTI received SAMHSA
approval to purchase tablet car
chargers on 5/6/13. The car
chargers were ordered on
6/3/13.

Revision
Complete
Yes

Yes

DR = Dress Rehearsal; FI = field interviewer; NSDUH = National Survey on Drug Use and Health; N/A = not applicable; QFT = Questionnaire Field Test; ROC = record of call;
RTI = Research Triangle Institute; SAMHSA = Substance Abuse and Mental Health Services Administration; USB = universal serial bus; wi-fi = wireless connection.

Appendix R: 2012 Questionnaire Field Test—Investigation
of Data Quality Issues for Items Moved from CAPI to
ACASI

R.1

Background and Introduction

R.1.1 Background on the 2012 QFT and Items Moved to ACASI
R.1.1.1 Overview of the 2012 QFT Data Collection Protocol and Outcomes

This appendix describes data collection results and analysis conducted for questionnaire
items moved from computer-assisted personal interviewing (CAPI) to audio computer-assisted
self-interviewing (ACASI) administration in the 2012 Questionnaire Field Test (QFT) instrument
for the National Survey on Drug Use and Health (NSDUH). The findings for these questionnaire
items include comparisons with current and comparable NSDUH main study data and other
comparable sources of survey data. Sponsored by the Substance Abuse and Mental Health
Services Administration (SAMHSA), NSDUH is a national survey of the U.S. civilian,
noninstitutionalized population aged 12 or older. The annual conduct of NSDUH is paramount in
meeting a critical objective of SAMHSA's mission to maintain current data on the prevalence of
substance use in the United States. In order to continue producing data that accurately reflect
current conditions, SAMHSA's Center for Behavioral Health Statistics and Quality (CBHSQ)
must update NSDUH periodically to reflect changing substance use and mental health issues.
The NSDUH questionnaire used in the 2012 QFT was revised to improve some of the
questions that cause known or suspected problems with data from the current questionnaire.
New content that addresses current data needs was also added. Revisions designed to reduce
errors associated with usability problems in the design and layout of the computer-assisted
interviewing (CAI) instrument were added. These changes included revising the prescription
drug modules, the front-end demographics, the binge drinking definition for women, the special
drugs module, and the back-end demographics section, as well as including a new
methamphetamine module.
Similar to the NSDUH main study, the respondent universe for the QFT was the civilian,
noninstitutionalized population aged 12 or older. In order to control costs, persons residing in
Alaska and Hawaii, as well as persons who were not able to complete the interview in English,
were excluded from the QFT sample. Therefore, the sample was representative of members of
the noninstitutionalized population aged 12 or older in the contiguous United States who are able
to complete the interview in English. NSDUH main study comparison data from 2011 and 2102
quarters 3 and 4, as well as other survey data used for comparison with the QFT, were adjusted
to account for the lack of Alaska and Hawaii residents and those who did not complete the
interview in English.
To make the QFT sample representative of the target population, a probability
proportional to size (PPS) sample of 213 State sampling (SS) regions was selected from all
876 SS regions. From these 213 SS regions, 5,358 dwelling units were sampled, 3,837 dwelling
units were screened as eligible, and 2,823 people were selected from within these eligible
dwelling units. Among persons selected for the QFT interview, a total of 2,044 completed
interviews were yielded during the field period of September 1, 2012, through November 3,
2012. The weighted overall response rate (combining the screening and interview response rates)
for the 2012 QFT sample was 57.71 percent compared with 61.30 percent for the 2011 main
study comparison sample and 60.98 percent for the 2012 quarters 3 and 4 main study comparison

R-1

sample. The overall lower response rate for the QFT could have introduced some unique
nonresponse bias for specific QFT estimates most likely to be affected by this difference in
response rate levels. Direct analysis was not undertaken of the impact of the approximately
4 percent lower response rate in potentially adding nonresponse bias for specific QFT estimates.
The focus of this appendix is examining multiple data quality indicators for items moved from
CAPI to ACASI administration in the 2012 QFT instrument.
R.1.1.2 Items Moved from CAPI to ACASI Administration

In the 2012 QFT questionnaire, the following back-end demographics items were moved
from the CAPI administration part of the NSDUH interview to the ACASI administration part of
the interview:
•

marital status and number of times married;46

•

moves in the past year and State of residence 1 year ago;

•

born in the United States or, if not, length of time residing in the United States;

•

education, including current enrollment in school, grade in school, and fullor part-time student status for postsecondary students, and related items;47

•

employment, including current job or business, hours worked at current job or
business, number of employers in the past year, employee assistance programs,
employer alcohol and drug use policies, and related items;

•

health insurance, including type or source of health insurance coverage, lack of health
insurance coverage, and whether health insurance covers substance abuse or mental
health problems; and

•

income, including receipt of five types of income from the government or
participation in government assistance programs and overall income level for the
prior calendar year.

As in the main study, the QFT protocol allowed the primary respondent to identify a proxy to
answer the questions in the last two sections (i.e., health insurance and income). (See
Section R.3.4 for comparisons of the distribution of relationships of proxy reporters to the
primary respondent and comparisons of estimates based on proxy report status.) All other items
were answered by the primary respondent, when logically applicable to the respondent, based on
responses to prior questions, the respondent's age, and other logical criteria.
To accommodate the transition from an interviewer-administered CAPI mode to ACASI
mode, the text and format of some of these questions required revisions. For example, questions
46

The items on current marital status and number of times married were actually moved from the front-end
demographic section of the CAPI-administered part of the interview to the back-end demographic section in ACASI
in the QFT instrument.
47
New questions on respondent disability, ability to speak English, whether any family members
were currently serving in the military, and cellular phone and land line telephone service in the household were
added to the ACASI portion of the QFT interview protocol in these sections. Because these items were new to the
NSDUH instrument, data quality indicators for these items could not be compared with the 2011 and 2012 quarters 3
and 4 data.

R-2

throughout the health insurance and income modules had contained notes for field interviewer
(FI) use in CAPI. These FI notes provided additional information about terms or constructs in the
questions. FIs are trained to read these notes to respondents when they feel that this additional
information would help the respondent to provide an accurate answer. Respondents who exhibit
confusion, ask for clarification, or hesitate to provide a response are likely to hear the
information contained within the interviewer note.
During instrumentation development for the QFT, this information was either moved to
the question text itself, deleted, or added as a note that respondents could view using the F2
function key on the laptop. In this way, F2 notes functioned similarly to the interviewer notes in
CAPI mode. QFT respondents were instructed to press F2 for more information about terms in
the question. In ACASI mode, the burden was on the QFT respondents to access this
information, as opposed to FIs in CAPI mode determining when to provide the information.
Relevant research shows that respondents using self-administered modes are less likely to
consult definitions when they have to request them, as opposed to when they appear on the
screen along with the question (Peytchev, Conrad, Couper, & Tourangeau, 2010). As a result,
providing notes via the F2 function key may have inadvertently created a barrier to QFT
respondents accessing this information in ACASI.
Despite these changes to QFT items moved to ACASI administration, data quality
indicators for these items could still be directly compared with the parallel items administered
via CAPI in the current NSDUH main study interviews.
R.1.2 Indicators Used to Evaluate the Effect of ACASI on Data Quality
As part of the QFT analysis and reporting, the following three data quality indicators
were used to examine the potential impact of moving items from CAPI to ACASI in the NSDUH
questionnaire:
1. comparing item missingness rates for the QFT items with item missingness rates for
the same items in the 2011 and 2012 quarters 3 and 4 main study comparison
datasets;
2. comparing QFT estimates for items moved to ACASI with (1) estimates for the same
items in the 2011 and 2012 quarters 3 and 4 main study comparison datasets and
(2) other national survey estimates with the same target population and comparable
survey items; and
3. for health insurance and income items, comparing QFT estimates with 2011 and 2012
quarters 3 and 4 main study comparison estimates for proxy versus self-reported data.
In addition to examining these three indicators of data quality for items moved to ACASI
administration, a literature review, communications with other Federal agencies, input from RTI
methodologists, and other steps were taken to understand the implications of the QFT results, as
described in Section R.3.1. For moved items with observed data quality issues, Section R.3.2
provides a summary of item missingness rates, Section R.3.3 summarizes benchmarking of
estimates to other surveys, and Section R.3.4 summarizes the impact of proxy reporting on
estimates for health insurance and income items.

R-3

R.1.3 Items Examined and Indication of Data Quality Issues
Table R-1 lists the items moved from CAPI to ACASI in the QFT instruments that were
examined for this appendix and indicates the nature of the data quality issues for those items.
Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly Higher
than Comparison Data3,4

Estimate Was
Significantly Different
from Comparison
Data5,6

Are you now married, widowed, divorced, or
separated, or have you never married? (QD07)

Yes

How many times have you been married? (QD08)

How many times in the past 12 months have you
moved? (QD13)

Yes

In what State did you live one year ago today?
(QD13a)

Yes

N/A

How many years have you lived in the United States?
(QD16b)

Are you now attending or are you currently enrolled
in school? (QD17)

What grade or year of school are you now attending?
(QD18)

Yes

Are you a full-time student or a part-time student?
(QD19)

Yes

During the past 30 days, how many whole days of
school did you miss because you were sick or
injured? (QD20)

Yes

During the past 30 days, how many whole days of
school did you miss because you skipped or "cut" or
just didn't want to be there? (QD21)

Yes

Did you work at a job or business at any time last
week? (QD26)

Yes

Even though you did not work at any time last week,
did you have a job or business? (QD27)

How many hours did you work last week at all jobs or
businesses? (QD28)

See notes at end of table.

(continued)

R-4

Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed (continued)

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Estimate Was
Significantly Different
from Comparison
Data5,6

Do you usually work 35 hours or more per week at all
jobs or businesses? (QD29)

Which one of these reasons best describes why you
did not work last week? (QD30)

Yes

Which one of these reasons best describes why you
did not have a job or business last week? (QD31)

Yes

During the past 30 days, did you make specific efforts
to find work? (QD32)

Did you work at a job or business at any time during
the past 12 months? (QD33)

Yes

How many different employers have you had in the
past 12 months? (QD36)

Yes

During the past 12 months, was there ever a
time when you did not have at least one job
or business? (QD37)

Yes

In how many weeks during the past 12 months did
you not have at least one job or business? (QD38)

Yes

In what year did you last work at a job or business?
(QD39a)

Yes

N/A

In what month in did you last work at a job
or business? (QD39b)

N/A

During the past 30 days, how many whole days of
work did you miss because you were sick or injured?
(QD40)

Yes

During the past 30 days, how many whole days of
work did you miss because you just didn't want to be
there? (QD41)

Yes

How many people work for your employer out of this
office, store, etc.? (QD42)

Yes

At your workplace, is there a written policy about
employee use of alcohol or drugs? (QD43)

See notes at end of table.

(continued)

R-5

Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed (continued)

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Estimate Was
Significantly Different
from Comparison
Data5,6

Does this policy cover only alcohol, only drugs, or
both alcohol and drugs? (QD44)

At your workplace, have you ever been given any
educational information regarding the use of alcohol
or drugs? (QD45)

Through your workplace, is there access to any type
of employee assistance program or other type of
counseling program for employees who have alcohol
or drug-related problems? (QD46)

Does your workplace ever test its employees
for alcohol use? (QD47)

Does your workplace ever test its employees for drug
use? (QD48)

Does your workplace test its employees for drug or
alcohol use as part of the hiring process? (QD49)

Does your workplace test its employees for drug or
alcohol use on a random basis? (QD50)

According to the policy at your workplace,
what happens to an employee the first time he or
she tests positive for illicit drugs? (QD51)

Yes

Would you be more or less likely to want to work for
an employer that tests its employees for drug use as
part of the hiring process? (QD52)

yes

Would you be more or less likely to want to work for
an employer that tests its employees for drug or
alcohol use on a random basis? (QD53)

yes

[SAMPLE MEMBER A] covered by
Medicare? (QHI01)

Yes

You have indicated that [SAMPLE MEMBER B]
covered by Medicare. Is this correct? (QHI01v)

Yes

[SAMPLE MEMBER A] covered by
Medicaid? (QHI02)

See notes at end of table.

(continued)

R-6

Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed (continued)

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Estimate Was
Significantly Different
from Comparison
Data5,6

You have indicated that [SAMPLE MEMBER B]
covered by Medicaid. Is this correct? (QHI02v)

[SAMPLE MEMBER A] currently covered
by [CHIPFILL]? (QHI02A)

[SAMPLE MEMBER A] currently covered
by TRICARE, or CHAMPUS, CHAMPVA, the VA,
or military health care? (QHI03)

[SAMPLE MEMBER A] currently covered by private
health insurance? (QHI06)

Yes

Was [SAMPLE MEMBER] private health insurance
obtained through work? (QHI07)

Does [SAMPLE MEMBER] private health insurance
include coverage for treatment for alcohol abuse or
alcoholism? (QHI08)

Yes

Does [SAMPLE MEMBER] private health insurance
include coverage for treatment for drug abuse?
(QHI09)

yes

Does [SAMPLE MEMBER] private health insurance
include coverage for treatment for mental or
emotional problems? (QHI10)

Yes

[SAMPLE MEMBER A] currently covered by any
kind of health insurance including Indian Health
Insurance? (QHI11)

Yes

In [YEAR], did you receive Social Security or
Railroad Retirement payments? (QI01N)

In [YEAR], did you receive Supplemental Security
Income or SSI? (QI03N)

Yes

In [YEAR], did you receive income from wages or
pay earned while working at a job or business?
(QI05N)

Yes

In [YEAR], did you receive food stamps? (QI07N)

Yes

At any time during [YEAR], even for 1 month, did
you receive any cash assistance from a State or county
welfare program such as [TANFFILL]? (QI08N)

Yes

See notes at end of table.

(continued)

R-7

Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed (continued)

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Estimate Was
Significantly Different
from Comparison
Data5,6

In [YEAR], because of low income, did you receive
any other kind of nonmonetary welfare or public
assistance? (QI10N)

Yes

For how many months in [YEAR] did you or your
[RELATIONSHIP] receive any type of welfare or
public assistance? (QI12AN)

Yes

For how many months in [YEAR] did you or your
[RELATIONSHIP] receive any type of welfare or
public assistance, not including food stamps?
(QI12BN)

Yes

Before taxes and other deductions, was your total
personal income from all sources during [YEAR]
more or less than $20,000? (QI20N)7

Yes

Of these income groups, which category best
represents [SAMPLE MEMBER] total personal
income during [YEAR]? (QI21A)

Yes

Of these income groups, which category
best represents [SAMPLE MEMBER] total personal
income during [YEAR]?(QI21B)

Yes

Before taxes and other deductions, was the
total combined family income during [YEAR] more
or less than 20,000 dollars? (QI22) 7

Of these income groups, which category
best represents your total combined family
income during [YEAR]? (QI23A)

Yes

See notes at end of table.

(continued)

R-8

Table R-1

Items Moved from CAPI to ACASI in the QFT Instruments and Data Quality Issues
Observed (continued)

QFT Questionnaire Item1,2

Item Missingness Rate
Was Significantly
Higher than
Comparison Data3,4

Of these income groups, which category
best represents your total combined family
income during [YEAR]? (QI23B)

Estimate Was
Significantly Different
from Comparison
Data5,6
Yes

CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and
Medical Program of the Department of Veteran's Affairs; DR = Dress Rehearsal; N/A = not applicable; Q =
question; QFT = Questionnaire Field Test; VA = Department of Veteran's Affairs.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
QFT data collected from September 1 through November 3, 2012.
3
Missing data include selection of responses of either "don't know" or "refused" for the question.
4
Item missingness rates for QFT questionnaire items were compared only with the 2011 main study data and the
2012 quarters 3 and 4 main study comparison data.
5
QFT estimates were compared with estimates from other survey data sources based on the comparability of the
survey design and questions. As detailed in Section R.3, the other data sources used for comparing estimates
included the 2011 National Survey on Drug Use and Health (NSDUH) main study, the 2012 quarters 3 and 4
NSDUH main study, the 2011 National Health Interview Survey (NHIS), the 2009-2010 National Health and
Nutrition Examination Survey (NHANES), the 2011 American Community Survey (ACS), and the Current
Population Survey (CPS).
6
Items marked N/A in this column indicate those for which the estimate from the item was not compared with any
of the other data sources listed in footnote 5. Given the units of analysis reported for these items, indicators were
not developed to compare QFT estimates with any of these other data sources.
7
Analysis variables for items QI20N and QI22 were edited to include the results of edited nonresponse follow-up
questions for respondents who initially entered a "refused" response to these questions. Both missingness rates and
estimates for these two items incorporated any further responses to the nonresponse follow-up-items.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health,
2012.

R-9

R.2

Items with No Observed Data Quality Issues

Missingness rates for many of the items moved to (ACASI in the QFT instrument were
similar to the missingness rates for these items when they were administered by CAPI in the
2011 and 2012 quarters 3 and 4 comparison interviews. However, some moved items had lower
missingness rates in the QFT data, and several items had higher missingness rates in the QFT
data. This section provides details for selected moved items that did not have any observed data
quality issues, especially those that had significantly lower missingness rates than either the 2011
or 2012 quarters 3 and 4 comparison data. Section R.3 presents and discusses moved items that
did have observed data quality issues, including having higher missingness rates and producing
significantly different estimates from National Survey on Drug Use and Health comparison data
and comparison data from other surveys.
Table R-2 provides two sets of items administered in ACASI for the QFT that had
significantly lower missingness rates than in the 2011 and 2012 quarters 3 and 4 comparison
data, including the following:
•

•

For the other items in Table R-2, no significant differences in missingness rates were found
between the QFT data and the 2011 and 2012 quarters 3 and 4 comparison datasets. As denoted
by an asterisk in Table R-2, estimates of missingness rates for the QFT data, the 2011
comparison data, or the 2012 quarters 3 and 4 comparison data had low precision. As with the
items where no differences in missingness rates were observed between the QFT data and the
2011 and 2012 quarters 3 and 4 comparison datasets, items with low precision rates were treated
as those with no observed data quality issues even when missingness rates appeared to differ
between the datasets. In addition, some QFT missingness rates in Table R-2 differed
significantly from either the 2011 comparison data or the 2012 quarters 3 and 4 comparison data,
but not both. Because these QFT items had relatively low missingness rates, these items were
also treated as those with no observed data quality issues. Items in Section R.3 treated as items
with observed data quality issues include those with significantly higher missingness rates and/or
significantly different estimates from multiple sources of comparison data.

R-10

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

R-11

How many times have you been
married? (QD08)

20,247

0.0

9,659

0.0

Were you born in the United States?
(QD14)

65,914

0.0

31,212

Have you lived in the United States for
at least one year? (QD16a)

5,101

0.0*

2,437

How many years have you lived in the
United States? (QD16b)

4,872

0.1a

228

0.0*

Are you now attending or are you
currently enrolled in school? (QD17)

65,914

What grade or year of school are you
now attending? (QD18)

34,297

Even though you did not work at any
time last week, did you have a job or
business? (QD27)
How many hours did you work last
week at all jobs or businesses?
(QD28)

How many months have you lived in the
United States? (QD16c)

See notes at end of table.

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

859

0.2

0.0*

2,043

0.0

0.0*

239

0.3

2,337

0.1

227

0.0*

100

0.0*

19.7*

0.0

31,212

0.0*

2,043

0.1

0.0

15,915

0.2

804

0.5

25,795

0.0

11,746

0.0

747

0.5

29,144

0.1

14,288

0.1

1,025

0.3
(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-12

Do you usually work 35 hours or more
per week at all jobs or businesses?
(QD29)

32,036

0.0

15,921

0.1

1,129

0.2

Which one of these reasons best
describes why you did not work last
week? (QD30)

2,892

0.0

1,633

0.1

104

0.0*

Which one of these reasons best
describes why you did not have a job
or business last week? (QD31)

22,903

0.1

10,113

0.0a

643

0.8

During the past 30 days, did you make
specific efforts to find work? (QD32)

5,851

0.1

2,607

0.0*

156

0.0*

During the past 12 months, was there
ever a time when you did not have at
least one job or business? (QD37)

32,036

0.0

15,921

0.0

1,129

0.3

In what month in did you last work at a
job or business? (QD39b)

7,413

0.4

3,335

0.5

175

0.7*

At your workplace, is there a written
policy about employee use of alcohol
or drugs? (QD43)

32,036

1,656

4.4a

15,921

872

4.7a

1,129

3.0

Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)

23,221

404

2.0a

11,463

198

1.8a

858

0.4

See notes at end of table.

(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-13

At your workplace, have you ever been
given any educational information
regarding the use of alcohol or drugs?
(QD45)

32,036

190

0.7

15,921

107

0.7

1,129

0.4

Through your workplace, is there access
to any type of employee assistance
program or other type of counseling
program for employees who have
alcohol or drug-related problems?
(QD46)

32,036

4,428

11.8a

15,921

2,231

11.9a

1,129

7.7

Does your workplace ever test its
employees for alcohol use? (QD47)

32,036

1,805

5.4a

15,921

907

5.3a

1,129

3.2

Does your workplace ever test its
employees for drug use? (QD48)

32,036

1,441

4.3

15,921

741

4.4a

1,129

3.0

Does your workplace test its employees
for drug or alcohol use as part of the
hiring process? (QD49)

14,351

230

2.0

7,214

112

1.8

530

1.2

Does your workplace test its employees
for drug or alcohol use on a random
basis? (QD50)

14,351

806

5.5

7,214

418

5.3

530

3.7

According to the policy at your
workplace, what happens to an
employee the first time he or she tests
positive for illicit drugs? (QD51)

14,351

1,865

14.0

7,214

937

13.0

530

11.3

See notes at end of table.

(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-14

Would you be more or less likely to
want to work for an employer that
tests its employees for drug use as
part of the hiring process? (QD52)

32,036

0.2

15,921

0.2

1,129

0.5

Would you be more or less likely to
want to work for an employer that
tests its employees for drug or alcohol
use on a random basis? (QD53)

32,036

0.2

15,921

0.2

1,129

0.3

[SAMPLE MEMBER A] covered by
Medicaid? (QHI02)

65,914

360

0.3

31,211

235

0.4

2,042

0.8

You have indicated that [SAMPLE
MEMBER B] covered by Medicaid.
Is this correct? (QHI02v)

220

102

0.0*

[SAMPLE MEMBER A] currently
covered by [CHIPFILL]? (QHI02A)

28,126

567

1.9

13,131

312

2.5

663

3.8

[SAMPLE MEMBER A] currently
covered by TRICARE, or
CHAMPUS, CHAMPVA, the VA, or
military health care? (QHI03)

65,914

194

0.2

31,211

142

0.2

2,042

0.6

Was [SAMPLE MEMBER] private
health insurance obtained through
work? (QHI07)

40,366

149

0.2

19,247

0.2

1,148

0.1

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)

40,366

18,327

43.8a

19,247

8,785

44.5a

1,148

322

See notes at end of table.

0.4*

0.0*

26.4
(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-15

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for drug abuse? (QHI09)

40,366

18,195

43.8a

19,247

8,748

44.8a

1,148

330

27.6

Does [SAMPLE MEMBER] private
health insurance include coverage for
treatment for mental or emotional
problems? (QHI10)

40,366

10,900

26.9a

19,247

5,187

26.4a

1,148

209

18.2

[SAMPLE MEMBER A] currently
covered by any kind of health
insurance including Indian Health
Insurance? (QHI11)

10,940

0.2a

5,061

0.3

412

0.0*

During the past 12 months, was there
any time when [SAMPLE MEMBER]
did not have any kind of health
insurance or coverage? (QHI13)

55,956

143

0.2

26,605

0.1

1,685

0.2

During the past 12 months, about how
many months without any kind of
health insurance or coverage?
(QHI14)

4,873

0.6

2,046

0.4

155

1.1

About how long has it been since
[SAMPLE MEMBER] last had any
kind of health care coverage?
(QHI15)

9,498

0.5

4,297

0.2

325

See notes at end of table.

0.8
(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item
Which of these reasons is the main
reason why [SAMPLE MEMBER]
stopped being covered by health
insurance? (QHI17)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

8,524

0.4

3,857

0.4

258

1.6

974

0.6

440

0.7

0.6*

65,913

616

0.6

31,211

341

0.6

2,042

1.0

For how many months in [YEAR] did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance? (QI12AN)

1,181

3.0

492

5.3

3.6*

For how many months in [YEAR] did
you or your [RELATIONSHIP]
receive any type of welfare or public
assistance, not including food stamps?
(QI12BN)

3,583

123

3.0

1,645

5.0

114

5.1*

Which of these reasons describe
why [SAMPLE MEMBER] never had
health insurance coverage? (QHI187)

R-16

In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)

See notes at end of table.

(continued)

Table R-2

Item Missingness Rates for Moved Items with No Observed Data Quality Issues in the 2012 Questionnaire Field Test and
Item Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item
Before taxes and other deductions, was
the total combined family income
during [YEAR] more or less than
20,000 dollars? (QI22)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

43,440

2,582

2012 Comparison Data1,2

Missing
Data4
(weighted)

7.8

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

20,458

1,293

QFT1,3

Missing
Data4
(weighted)

8.1

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

1,131

Missing
Data4
(weighted)

9.5

R-17

* Low precision.
CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Veterans Administration; DMT =
dimethyltryptamine; QFT = Questionnaire Field Test, VA = Department of Veterans Affairs.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to self-administered.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Missing data include selection of responses of either "don't' know" or "refused" for the question. "Missing Data (weighted)" denotes the weighted percentage of missing data.
Denominators for these percentages were based on the total number of cases (i.e., respondents) who were asked the question.
5
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not asked the question if all response options were
coded as "blank" (e.g., 98 for 2-digit variables).
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

R.3

Items with Observed Data Quality Issues

R.3.1 Review of the Literature and Other Steps Taken to Understand Findings for Items
with Observed Data Quality Issues
R.3.1.1 Summary of Relevant Literature

In an effort to shed light on observed differences in missingness rates and estimates for
Questionnaire Field Test (QFT) items with observed data quality issues, an extensive literature
review was conducted. The literature search was based mainly on publication databases, such as
the Web of Science (http://thomsonreuters.com/web-of-science/), to find relevant published
journal articles and was complemented by a Web search using Google Scholar
(http://scholar.google.com/). The search was supplemented further by reviewing the proceedings
of the Survey Research Methods Section of the American Statistical Association (ASA) and
research presented at recent conferences of the American Association for Public Opinion
Research (AAPOR).
The first step of the online literature search was to enter all of the combinations of the
following key words:
•

data quality,

•

ACASI (i.e., audio computer-assisted self-interviewing ),

•

CAPI (i.e., computer-assisted personal interviewing),

•

item nonresponse,

•

income, and

•

health insurance.

When no literature was found that met all of these specific criteria, the number of key words
used in the search was limited to fewer words. Despite this expanded effort, the search results
produced research that was only partially related to the topic. For this reason, the final phase of
the search went beyond the original key words and touched on all research related to ACASI and
CAPI data quality, regardless of the topic.
The literature review was not able to identify research studies that specifically compared
missingness rates for items such as income, employment, or health insurance coverage between
ACASI and CAPI. However, several articles were found that more generally compared data
quality between self-administered and interviewer-administered surveys for other types of survey
questions. For example, van den Brakel, Vis-Visschers, and Schmeets (2013) reported an
increased rate of "don't know" responses in the data collected via computer-assisted selfinterviewing (CASI) than CAPI for 14 attitudinal questions.
Another set of research findings compiled by Langhaug, Sherr, and Cowan (2010)
examined the effect of questionnaire delivery modes on item nonresponse rates. By searching
Medline, Embase, PyschINFO, and International Society for Sexually Transmitted Diseases

R-18

Research (ISSTDR) conference proceedings, these authors identified surveys using different
questionnaire delivery modes to collect data about sexual behavior in developing countries.
Overall, the existing research found lower item nonresponse rates in interviewer-administered
interviews than in self-administered interviews. Some of these findings for questions on sexual
behavior included the following:
•

Langhaug et al. (2007) reported the highest item nonresponse rates in selfadministered questionnaires using paper-and-pencil (SAQ) and audio-SAQ48 than in
interviewer-administered surveys.

•

Jaspan et al. (2007) reported approximately 7 times more item nonresponse in
computer self-administered interview than interviewer-administered personal digital
assistant (PDA) interviews.

•

Plummer et al. (2004a, 2004b) reported a higher proportion of "don't know" responses
in a derivative of self-completion questionnaires where the questions were read aloud
in a group setting than face-to-face interviewing.

•

Lara, Strickler, Olavarrieta, and Ellertson (2004) reported that paper-and-pencil SAQ
produced the highest level of item nonresponse compared with face-to-face
interviewing, ACASI, and the random response technique.

Although the authors could not fully explain these findings, the primary explanation offered is
that interviewer presence makes it more difficult for respondents to avoid providing a response to
questions. Given that interviewer training typically instructs interviewers to probe further when a
respondent fails to respond or provides a "don't know" response, respondents likely feel pressure
to provide an answer rather than skip the question. In self-administration, this pressure from
interviewers is absent and, therefore, can make it easier for respondents to feel comfortable when
declining to answer questions. The findings on SAQs might not apply directly to the comparison
of ACASI with CAPI missingness rates. Only the Lara et al. (2004) study directly compared
paper-and-pencil SAQ and ACASI, with the item nonresponse rate being higher for paper-andpencil SAQ. This finding could have resulted from greater difficulty of SAQ respondents
following the protocol than ACASI respondents.
Even if the assumption is correct that higher missingness rates in ACASI compared with
CAPI result from the lack of interviewer presence, the finding of higher missingness rates does
not necessarily indicate lower overall data quality in ACASI reports. Item nonresponse is only
one indicator of data quality. For other aspects of data quality, reports in self-administered
surveys, such as ACASI or CASI, may be superior to interviewer-administered surveys. For
example, Chang and Krosnick (2010) reported on the results of a laboratory study in which
respondents were randomly assigned to answer questions on a computer or by an interviewer
over an intercom. For a number of attitudinal questions on political candidates, issues, and
ideology, respondents in the self-administered (computer) mode provided responses with higher
concurrent validity, less survey "satisficing" (i.e., putting forth minimal cognitive effort to
answer questions, as explained by Krosnick [1991]), and less socially desirable reporting than

SAQ = self-administered questionnaire, where questions, instructions, and responses are heard through
headphones.

R-19

those in the interviewer-administered mode. The differences were more pronounced among those
with more limited cognitive skills.
For questions where respondents might view their responses as sensitive, there is
considerable research that focuses on higher levels of reporting of such items in selfadministered versus interviewer-administered modes. Beginning with the Tourangeau and Smith
(1996) study on sexual behaviors, ACASI has become known as a valuable method for collecting
accurate responses on sensitive questions, such as sexual behavior or substance use (de Leeuw,
Hox, & Kef, 2003; Tourangeau & Yan, 2007; Turner et al., 1998).
Income could be considered a sensitive question, and item nonresponse rates for these
questions tend to be high for any survey mode. It is feasible that the ACASI responses to the
income level questions provided were generally more accurate than those provided in CAPI,
which could counter reductions in data quality because of the higher missingness rates.
Determining the full impact of higher missingness rates on the quality of income estimates
requires comparing the QFT results with the results of other surveys that can be considered
highly accurate.
Questions on health insurance coverage would not seem to fall clearly under the category
of sensitive reporting in surveys. One possible explanation for the higher missingness rates for
these items could be respondent confusion about the various types of health insurance coverage,
which could not be resolved via self-administration with ACASI as it could with interviewer
administration with CAPI. Potdar and Koenig (2005) argued that respondents' unfamiliarity with
certain terms, which could be easily clarified by interviewers, explained inconsistencies observed
between ACASI and face-to-face interviews. These authors concluded that respondents were
more likely to encounter difficulty in comprehending questions in ACASI, leading to "don't
know" or "refuse" responses. These findings suggest that the absence of interviewer assistance in
ACASI could be one possible explanation for the increased missingness rates for the health
insurance items, especially for the "private health insurance" question.
R.3.1.2 Communications with the Survey Research Community and Other Federal Agencies

To solicit input from the community of survey researchers and those working on other
Federal agency surveys on possible explanations for the higher QFT missingness rates and
differences in estimates for several ACASI items, the following outreach efforts were undertaken
in June 2013.
R.3.1.2.1 SRMSnet and AAPORnet Email Inquiries

A request for input was submitted to the Survey Research Methods Section (SRMS) of
the ASA and the AAPOR email lists (or "listservs"). The message provided a summary of
missingness rates and differences in estimates for several QFT items moved to ACASI and asked
whether recipients were aware of any research looking at the impact of moving from CAPI to
ACASI on data quality for these specific kinds of questions. This request also asked for
recommendations on sources of data for benchmarking estimates of participation in food stamp
programs at the family level.

R-20

A total of nine email responses were received in response to the SRMS message.
Although well-intentioned, respondents were unable to provide responses focused on the kinds
of demographic and household items that exhibited high missingness rates in the QFT.
The recent research identified focused mainly on "sensitive items," such as sexual orientation,
sexual behavior, and substance use. A few emails identified data sources for benchmarking
estimates of food stamp program participation at the family level, but these sources were either
already identified or incompatible with the QFT data.
R.3.1.2.2 Communication with Staff Working on the NHIS, NHANES, and NSFG

SAMHSA and RTI also reached out to researchers working on three other Federal
surveys that could have data to inform the QFT results on demographic and household items
moved to ACASI. These surveys included the National Health Interview Survey (NHIS), the
National Health and Nutrition Examination Survey (NHANES), and the National Survey of
Family Growth (NSFG). Like the responses to the SRMSnet and AAPORnet email inquiries, the
primary use of ACASI for the NHIS and NSFG was for asking questions on sensitive topics,
such as sexual orientation (NHIS) and sexual behavior and substance use (NSFG). None of the
three surveys had tested and compiled results from asking the same demographic and household
items in ACASI compared with results from CAPI.
R.3.1.3 Input from a Discussion with RTI Survey Methodologists

On June 12, 2013, RTI held a meeting with a panel of survey methodologist to solicit
their input on possible explanations for the higher QFT missingness rates and differences in
estimates for several ACASI items. The panel of RTI survey methodologists consisted of Paul P.
Biemer, Rachel A. Caspar, Joseph J. Murphy, and Andy Peytchev. Several members of RTI's
National Survey on Drug Use and Health (NSDUH) management team and QFT report team also
participated in this hour-long discussion. In advance of this meeting, the RTI NSDUH team
provided participants with an overview of the QFT design features and key outcomes, such as
response rates from the draft QFT report. For efficiency, the QFT results presented to
participants focused on the following three items: (1) current coverage by private health
insurance (QHI06), (2) receipt of income from wages or pay earned while working at a job or
business in the prior year (QI05N), and (3) receipt of food stamps in the past year (QI07N).
The participants offered several comments and thoughts on the nature of the higher missingness
rates and differences in estimates for these three QFT items and, possibly, other items, as
summarized below:
•

The magnitude of some differences was surprising, especially for items that would
not seem to elicit strong socially desirable reporting, such as income from wages.
The recent status of the economy could have increased the sensitivity of this item to
QFT respondents, although a similar impact would be expected in the main study
data.

•

Additional subgroup analysis or predictive validity with correlates could be useful for
estimating measurement error for each of the affected items. Subgroup analysis could
focus on which sets of respondents are reporting differently for each item. Such an
analysis could be informed by consulting with experts in these areas for
characteristics of respondents that may be related to differences in reporting.
R-21

•

For some items, it is possible that NSDUH CAPI estimates are underestimates.
Benchmarking NSDUH CAPI estimates to other sources of CAPI survey data should
answer this question.

•

Interviewer variance would be higher for CAPI mode, but CAPI administration could
also include standard probes for clarification of questions. In ACASI mode,
interviewing notes were available via the F2 key.

•

Further debriefing with main study and QFT field interviewers (FIs) could provide
some insights on any observed differences in how respondents reacted to these
questions in ACASI mode in the QFT versus CAPI mode in the main study.

•

If appropriate data are available, behavior coding could also help understand
differences in the ACASI versus CAPI experiences of QFT versus main study
respondents.

•

Similar health insurance questions created a lot of confusion on at least one recent
RTI survey. Improvements to these items might be needed for ACASI administration.

•

Overall, it is difficult to determine which ACASI estimates might have higher or
lower data quality than comparison estimates, given the multiple sources of error that
cannot be fully assessed. Some of these items might be better in one mode versus the
other.

•

Given that sources of differences between the QFT results and comparison results
cannot be definitively tested, the default position could be to keep the affected items
in CAPI.

•

One further step is to complete an analysis of the distribution of demographic and
geographic characteristics of the QFT and NSDUH comparison samples in order to
ensure that these results are not the result of some anomalous distribution of the QFT
sample.

Overall, the RTI panel was similarly uncertain about the likely explanations for the higher
missingness rates and differences in estimates for these QFT items. As noted in multiple
comments, panel members acknowledged that the explanations could differ for specific items.
R.3.2 Item Missingness Rates for Items with Observed Data Quality Issues
As shown in Table R-3, several types of items that were moved to ACASI for the QFT
had significantly higher missingness rates than the CAPI items from the 2011 and 2012 quarters
3 and 4 comparison samples:
•

R-22

Table R-3

Item Missingness Rates for Moved Items with Observed Data Quality Issues in the 2012 Questionnaire Field Test and Item
Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-23

Are you now married, widowed,
divorced, or separated, or have you
never married? (QD07)

54,954

0.0a

26,036

0.0a*

1,778

0.4

How many times in the past 12 months
have you moved? (QD13)

65,914

0.1a

31,212

0.0a

2,043

0.8

In what State did you live in one year
ago today? (QD13a)

20,017

0.0a

9,585

0.0a

618

0.7

Are you a full-time student or a parttime student? (QD19)

34,297

0.0a

15,915

0.0a

804

During the past 30 days, how many
whole days of school did you miss
because you were sick or injured?
(QD20)

1.4
31,249

During the past 30 days, how many
whole days of school did you miss
because you skipped or "cut" or just
didn't want to be there? (QD21)

26,816

Did you work at a job or business at any
time last week? (QD26)
Did you work at a job or business at any
time during the past 12 months?
(QD33)
See notes at end of table.

1.0

0.3

14,472

0.2

690

13
1.5

0.1

10,528

0.1

54,944

0.0a

26,035

0.0a*

22,908

0.1a

10,114

0.0a

597

1,778

649

0.2

0.6
(continued)

Table R-3

Item Missingness Rates for Moved Items with Observed Data Quality Issues in the 2012 Questionnaire Field Test and Item
Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item
How many different employers have
you had in the past 12 months?
(QD36)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-24

32,855

0.0a

15,906

0.1a

1,066

0.8

In how many weeks during the past 12
months did you not have at least one
job or business? (QD38)

7,023

0.7a

3,615

0.9a

249

4.3

In what month in did you last work at a
job or business? (QD39b)

7,413

0.4

3,335

0.5

175

0.7*

During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)

32,036

0.0a

15,921

0.1a

1,129

0.6

At your workplace, is there a written
policy about employee use of alcohol
or drugs? (QD43)

32,036

1,656

4.4a

15,921

872

4.7a

1,129

3.0

Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)

23,221

404

2.0a

11,463

198

1.8a

858

0.4

[SAMPLE MEMBER A] covered by
Medicare? (QHI01)

65,914

193

0.2

31,211

130

0.3

2,042

0.6

You have indicated that [SAMPLE
MEMBER B] covered by Medicare.
Is this correct? (QHI01v)

1,208

0.0

620

0.1

1.1*

[SAMPLE MEMBER A] currently
covered by private health insurance?
(QHI06)

65,914

382

0.3a

31,211

261

0.4

2,042

See notes at end of table.

0.7
(continued)

Table R-3

Item Missingness Rates for Moved Items with Observed Data Quality Issues in the 2012 Questionnaire Field Test and Item
Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

R-25

In [YEAR], did you receive
Supplemental Security Income or
SSI? (QI03N)

65,913

883

0.8a

31,211

459

0.8a

2,042

1.5

In [YEAR], did you receive income
from wages or pay earned while
working at a job or business? (QI05N)

65,913

162

0.2a

31,211

103

0.3a

2,042

1.1

In [YEAR], did you receive food
stamps? (QI07N)

65,912

236

0.3

31,211

165

0.3

2,042

0.5

At any time during [YEAR], even for
one month, did you receive any cash
assistance from a State or county
welfare program such as
[TANFFILL]? (QI08N)

65,912

462

0.4a

31,211

239

0.4a

2,042

1.0

In [YEAR], because of low income,
did you receive any other kind of nonmonetary welfare or public
assistance? (QI10N)

65,912

349

0.3a

31,211

191

0.3a

2,042

0.6

Before taxes and other deductions, was
your total personal income from all
sources during [YEAR] more or less
than 20,000 dollars? (QI20N)

65,912

785

1.9a

31,211

393

1.9a

2,042

3.7

See notes at end of table.

(continued)

Table R-3

Item Missingness Rates for Moved Items with Observed Data Quality Issues in the 2012 Questionnaire Field Test and Item
Missingness Rates for these Items in the 2011 Comparison Sample and the 2012 Quarters 3 and 4 Comparison Sample
(continued)
2011 Comparison Data1

Instrument Item

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

2012 Comparison Data1,2

Missing
Data4
(weighted)

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

QFT1,3

Missing
Data4
(weighted)

R-26

Of these income groups, which category
best represents [SAMPLE
MEMBER] total personal income
during [YEAR]?(QI21A)

47,732

581

2.2a

22,448

258

2.2a

Of these income groups, which category
best represents your total combined
family income during [YEAR]?
(QI23A)

9,445

605

6.1

4,572

298

Of these income groups, which category
best represents your total combined
family income during [YEAR]?
(QI23B)

44,537

2,810

6.4

20,887

1,314

Number of
Number of
Cases with
Cases Asked
Missing
the Question
Data4
(unweighted) (unweighted)

Missing
Data4
(weighted)

1,196

4.6

6.9

365

9.7

6.3

1,328

6.1

•

Most of the items asking about receipt of various sources of income or participation
in government assistance programs—QI03N for receipt of Supplemental Security
Income (SSI), QI05N for wages or pay from a job or business, QI07N for receipt of
food stamps, QI08N for receipt of State or county welfare programs, and QI10N for
receipt of any other kind of nonmonetary welfare or public assistance—had
significantly higher item missingness rates in the QFT data than in the 2011 or 2012
quarters 3 and 4 comparison data. Missingness rates for all of these items were quite
similar in the 2011 and 2012 quarters 3 and 4 comparison data, but proportionately
higher in the QFT data.

•

R-27

R.3.3 Distribution of "Don't Know" and "Refused" Item Response Rates for Items with
Observed Data Quality Issues
Table R-4 presents the distribution of "don't know" and "refused" responses for the
22 items moved to ACASI for the QFT that had significantly higher missingness rates than the
CAPI items from the 2011 and 2012 quarters 3 and 4 comparison samples. The distribution of
"don't know" and "refused" responses varied, with some items having rather similar proportions
and others having markedly different proportions. QD07 on marital status, QD13 on moving
home in the past year, QD26 about work at a job or business at any time in the past week, QD33
on working at a job or business in the past year, QD36 on the number different employers in the
past year, and QD40 on workdays missed due to sickness or injury appeared to have no
meaningful differences in the proportions of "don't know" and "refused" responses.
For items where the proportions of "don't know" and "refused" responses appeared to
differ meaningfully, the most common pattern among these items was a higher proportion of
"don't know" responses. A total of 15 items followed this pattern of higher proportions of "don't
know" than "refused" responses, including the following:
•

QD13a on State of residence 1 year ago;

•

QD19 on full-time or a part-time student status;

•

QD20 on school days missed due to sickness or injury;

•

QD21 on school days missed due to "skipping," "cutting," or not wanting to be there;

•

QD38 on the number of weeks during the past 12 months without at least one job or
business;

•

QD39b on month of last work at a job or business;

•

QD43 on whether workplace has a written policy about employee use of alcohol or
drugs;

•

QD44 on whether workplace policy covers only alcohol, only drugs, or both alcohol
and drugs;

•

QHI06 on private health insurance coverage;

•

QI03N on receipt of SSI;

•

QI05N on wages or pay from a job or business;

•

QI07N on receipt of food stamps;

•

QI08N on receipt of State or county welfare programs;

•

QI010N on receipt of any other kind of nonmonetary welfare or public assistance;
and

•

QI21A on personal income level.

Only 1 of the 22 items—QI20N on personal income level—had a higher proportion of "refused"
than "don't know" responses. These results suggest that QFT respondents answering these
questions in ACASI were unsure of the most appropriate answers to provide.
R-28

Table R-4

Distribution of "Don't Know" and "Refused" Item Response Rates for Moved Items with Observed Data Quality Issues in
the 2012 Questionnaire Field Test and Item Missingness Rates for These Items in the 2011 Comparison Sample and the 2012
Quarters 3 and 4 Comparison Sample

R-29

Instrument Item
Are you now married,
widowed, divorced,
or separated, or have
you never
married? (QD07)
How many times in the
past 12 months have
you moved? (QD13)
In what State did you
live in one year ago
today? (QD13a)
Are you a full-time
student or a parttime student? (QD19)
During the past 30 days,
how many whole
days of school did you
miss because you
were sick or injured?
(QD20)
During the past 30 days,
how many whole
days of school did you
miss because you
skipped or "cut" or
just didn't want to be
there? (QD21)
Did you work at a job or
business at any
time last week?
(QD26)
Did you work at a job or
business at any
time during the past
12 months? (QD33)
See notes at end of table.

2012 Comparison Data1,2
QFT1,3
2011 Comparison Data1
Number of
Number of
Number of
Number of
Number of
Number of
Don't
Don't
Don't
Cases Asked
Cases Asked
Cases Asked
Cases with
Cases with
Cases with
Refused6 the Question Missing Data4 Know5
Refused6 the Question Missing Data4 Know5
Refused6
the Question Missing Data4 Know5
(unweighted) (unweighted) (weighted) (weighted) (unweighted) (unweighted) (weighted) (weighted) (unweighted) (unweighted) (weighted) (weighted)

54,954

0.0*

0.0

26,036

0.0*

1,778

0.2

65,914

0.0a

31,212

0.0a

2,043

0.5

0.4

20,017

0.0*

0.0

9,585

0.0

618

0.5

0.2

34,297

0.0a

0.0*

15,915

0.0a

0.0

804

1.0

0.0*

31,249

0.2a

0.0

14,472

0.2a

0.0*

690

1.3

0.1

26,816

0.1a

0.1

10,528

0.1a

0.0*

597

1.3

0.2

54,944

0.0*

0.0

26,035

0.0*

1,778

0.1

22,908

0.0*

0.1

10,114

0.0

649

0.3

0.4
(continued)

Table R-4

Distribution of "Don't Know" and "Refused" Item Response Rates for Moved Items with Observed Data Quality Issues in
the 2012 Questionnaire Field Test and Item Missingness Rates for These Items in the 2011 Comparison Sample and the 2012
Quarters 3 and 4 Comparison Sample (continued)
2011 Comparison Data1

Instrument Item

2012 Comparison Data1,2

QFT1,3

Number of
Number of
Number of
Number of
Don't
Don't
Cases Asked
Cases Asked
Cases with
Cases with
Refused6 the Question Missing Data4 Know5
Refused6
the Question Missing Data4 Know5
(unweighted) (unweighted) (weighted) (weighted) (unweighted) (unweighted) (weighted) (weighted)

Number of
Number of
Don't
Cases Asked
Cases with
Refused6
the Question Missing Data4 Know5
(unweighted) (unweighted) (weighted) (weighted)

R-30

How many different
employers have you had
in the past 12 months?
(QD36)

32,855

0.0a

0.0

15,906

0.0

1,066

0.3

0.4

In how many weeks during
the past 12 months did
you not have at least one
job or business? (QD38)

7,023

0.7a

0.0*

3,615

0.9a

0.0

249

3.4

0.9

In what month in did you
last work at a job
or business? (QD39b)

7,413

0.4

0.0

3,335

0.5

0.0*

175

0.7*

0.0*

During the past 30 days,
how many whole days of
work did you miss
because you were sick
or injured? (QD40)

32,036

0.0a

15,921

0.1

0.0a

1,129

0.3

At your workplace, is there
a written policy about
employee use of alcohol
or drugs? (QD43)

32,036

1,656

4.4a

0.0*

15,921

872

4.7a

0.0

1,129

2.9

0.1

Does this policy cover only
alcohol, only drugs, or
both alcohol and drugs?
(QD44)

23,221

404

2.0a

0.0

11,463

198

1.8a

0.0*

858

0.4

0.0*

[SAMPLE MEMBER A]
currently covered
by private health
insurance? (QHI06)

65,914

382

0.2a

0.0

31,211

261

0.4

0.1

2,042

0.6

See notes at end of table.

0.1
(continued)

Table R-4

Distribution of "Don't Know" and "Refused" Item Response Rates for Moved Items with Observed Data Quality Issues in
the 2012 Questionnaire Field Test and Item Missingness Rates for These Items in the 2011 Comparison Sample and the 2012
Quarters 3 and 4 Comparison Sample (continued)
2011 Comparison Data1

Instrument Item

2012 Comparison Data1,2

QFT1,3

Number of
Number of
Don't
Cases Asked
Cases with
Refused6
the Question Missing Data4 Know5
(unweighted) (unweighted) (weighted) (weighted)

R-31

In [YEAR], did you receive
Supplemental Security
Income or SSI? (QI03N)

65,913

883

0.7

0.1

31,211

459

0.6a

0.1

2,042

1.1

0.5

In [YEAR], did you receive
income from wages or
pay earned while
working at a job or
business? (QI05N)

65,913

162

0.1a

0.1

31,211

103

0.2a

0.1

2,042

0.9

0.3

In [YEAR], did you receive
food stamps? (QI07N)

65,912

236

0.1a

0.1

31,211

165

0.2

0.1

2,042

0.4

0.1

At any time during
[YEAR], even for
one month, did you
receive any cash
assistance from a State or
county welfare program
such as [TANFFILL]?
(QI08N)

65,912

462

0.3a

0.1

31,211

239

0.3a

0.1

2,042

0.9

0.1

In [YEAR], because of low
income, did you receive
any other kind of nonmonetary welfare or
public assistance?
(QI10N)

65,912

349

0.2a

0.1

31,211

191

0.2a

0.1

2,042

0.5

0.1

Before taxes and other
deductions, was
your total personal
income from all sources
during [YEAR] more or
less than 20,000
dollars? (QI20N)

65,912

785

0.5a

1.4

31,211

393

0.5a

1.4

2,042

1.3

See notes at end of table.

2.4
(continued)

Table R-4

Distribution of "Don't Know" and "Refused" Item Response Rates for Moved Items with Observed Data Quality Issues in
the 2012 Questionnaire Field Test and Item Missingness Rates for These Items in the 2011 Comparison Sample and the 2012
Quarters 3 and 4 Comparison Sample (continued)
2011 Comparison Data1

Instrument Item
Of these income groups,
which category
best represents
[SAMPLE MEMBER]
total personal income
during [YEAR]?(QI21A)

2012 Comparison Data1,2

QFT1,3

47,732

581

1.5a

0.7

22,448

258

1.4a

0.7

Number of
Number of
Don't
Cases Asked
Cases with
Refused6
the Question Missing Data4 Know5
(unweighted) (unweighted) (weighted) (weighted)

1,196

3.3

1.3

R-32

a
1

Difference between estimate and QFT estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Missing data include selection of responses of either "don't know" or "refused" for the question.
5
"Don't Know (weighted)" denotes the weighted percentage of responses of "don't know" for the question. Denominators for these percentages were based on the total number of cases (i.e., respondents)
who were asked the question.
6
"Refused (weighted)" denotes the weighted percentage of responses of "refused" for the question. Denominators for these percentages were based on the total number of cases (i.e., respondents) who
were asked the question.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

R.3.4 Benchmarking of Estimates to Other Surveys for Items with Observed Data Quality
Issues
Estimates for most demographic and household items from the QFT data were similar to
the 2011 and 2012 quarters 3 and 4 comparison estimates. The majority of differences observed
indicated that the QFT sample members were associated with lower socioeconomic status.
For example, the QFT estimates for participating in government programs, such as food stamps,
were significantly higher than those for the 2011 and 2012 quarters 3 and 4 comparison data.
Differences in missingness rates and estimates for items that were most highly correlated with
socioeconomic status could have been affected by these observed differences in socioeconomic
status between the QFT sample and the 2011 and 2012 quarters 3 and 4 comparison samples.
Because the noncore demographic and household questions were administered via ACASI for
QFT respondents and via CAPI for 2011 and 2012 quarters 3 and 4 respondents, the effects of
this mode difference cannot be disentangled from the effects of differences in socioeconomic
status.
As shown in Table R-3 earlier, missingness rates for several QFT ACASI items were
significantly higher than the missingness rates in the 2011 and 2012 quarters 3 and 4 comparison
data for the parallel CAPI items. Although missingness rates for the first six items in
Table R-3—QD07 on marital status, QD13 on moving home in the past year, QD13a on State of
residence 1 year ago, QD19 on full-time or a part-time student status, QD20 on school days
missed due to sickness or injury, and QD21 on school days missed due to "skipping" or
"cutting"—were generally higher than the missingness rates in the 2011 and 2012 quarters 3 and
4 comparison datasets, concern about the data quality of these items was limited. The same
conclusion was reached for several other items asking about employment history and workplace
policies—QD33 on working at a job or business in the past year, QD36 on the number of
different employers in the past year, QD38 on the number of weeks during the past 12 months
without at least one job or business, QD39b on the month of last work at a job or business, QD40
on workdays missed due to sickness or injury, QD43 on whether workplace has a written policy
about employee use of alcohol or drugs, and QD44 on whether workplace policy covers only
alcohol, only drugs, or both alcohol and drugs. For these two sets of items, no benchmarking
analyses were conducted to understand the implications for overall data quality for these items.
For items where the findings on item missingness rates raised significant concerns about
data quality, benchmarking comparisons to both the 2011 and 2012 quarters 3 and 4 datasets and
to other national surveys was undertaken. This benchmarking was intended to determine whether
and how the QFT estimates differed from other national survey estimates with the same target
population and comparable survey items. The following sets of QFT items shown in Table R-3
were benchmarked to other survey data:
•

received income and participation in government assistance programs,

•

health insurance coverage,

•

income,

•

employment status and unemployment rates, and

•

education.

R-33

The following five sections present and discuss the results of benchmarking these sets of items to
other survey data sources. In addition, given that health insurance and income items allow for
proxy reports, Section R.3.4 presents and discusses the potential impact of proxy reports on the
missingness rates and estimates for these two sets of items.
R.3.4.1 Received Income and Participation in Government Assistance Programs

In Tables R-5 through R-8,49 QFT estimates for five types of received income or
participation in government assistance programs for all persons aged 12 or older and three
separate age groups are presented with parallel estimates from the 2011 comparison sample, the
2012 quarters 3 and 4 comparison sample, the 2011 American Community Survey (ACS), and
the 2011 National Health Interview Survey (NHIS). The three separate age groups are persons
aged 12 to 17, 18 to 25, and 26 or older. Estimates for all data sources are provided in both
percentages and thousands of persons, with standard errors in parentheses. Several notable
comparisons can be observed from these tables:
•

For all persons aged 12 or older (Table R-5), estimates for receipt of social security
were very similar across all five survey data sources at about 27 percent. Estimates
for social security were also similar across these data sources for the three separate
age groups (Tables R-6 through R-8).

•

The QFT estimate for receipt of wages for all persons aged 12 or older (68.6 percent)
was significantly lower than the estimates from the four other data sources, which
were all close to 80 percent. This pattern held for receipt of wages across all three
separate age groups.

•

For SSI, the QFT estimate for all persons aged 12 or older (9.4 percent) was generally
higher than the estimates from most of the other data sources. Estimates for SSI from
the other surveys ranged from 5.0 percent in the 2011 NHIS to 7.6 percent in the 2012
quarters 3 and 4 comparison sample. This pattern for receipt of SSI was very similar
across the three separate age groups.

•

The QFT estimate for participation in food stamp50 programs for all persons aged 12
or older (17.6 percent) was also generally higher than the estimates from the four
other data sources. Estimates for food stamp receipt from the other surveys ranged
from 13.0 percent in the 2011 NHIS to 15.6 percent in the 2012 quarters 3 and 4
comparison sample. This pattern for receipt of food stamps was very similar across
the three separate age groups.

•

For receipt of welfare payments, such as those from Temporary Assistance for Needy
Families (TANF), the QFT estimate for all persons aged 12 or older (3.6 percent) was
higher than the estimates from the 2011 comparison sample (2.5 percent) and the
2012 quarters 3 and 4 comparison sample (2.3 percent), but it was similar to the 2011
ACS estimate (3.3 percent) and the 2011 NHIS estimate (3.2 percent). The pattern for

To aid in their readability, Table R-5 through Table R-23 appear together at the end of their discussion in
this Section R.3.4.
50
Food stamp programs are now more commonly known as the Supplemental Nutrition Assistance
Program (SNAP).

R-34

receipt of welfare payments generally held across the three separate age groups, with
the QFT estimates being somewhat higher than the 2011 and 2012 quarters 3 and 4
comparison estimates, but similar to the 2011 ACS and 2011 NHIS estimates.
Benchmarking QFT estimates for five types of received income or participation in government
assistance programs to both recent NSDUH data and other national survey data revealed mixed
results. Estimates for receipt of social security payments were quite similar across all five
surveys. The QFT estimate for receipt of wages was substantially lower than the estimates from
the other four survey sources. For receipt of welfare payments, QFT estimates were generally
similar to the 2011 ACS and 2011 NHIS estimates, but higher than the 2011 and 2012 quarters 3
and 4 comparison estimates. Estimates of participation in two programs—SSI and food stamps—
appeared to be clearly greater for the QFT sample than in the other four surveys. These findings
suggest that QFT respondents had a somewhat lower socioeconomic status than the 2011 and
2012 quarters 3 and 4 comparisons samples. This difference could have accounted for some of
the observed differences between the QFT estimates and the 2011 and 2012 quarters 3 and 4
comparison estimates for those items that were the most highly correlated with socioeconomic
status (SES).
In principle, the weighting adjustments for nonresponse and undercoverage applied to the
QFT data would have eliminated differences in SES to the extent that the measures used in the
weighting adjustments were themselves correlated with SES. However, the correlations between
the variables used in weighting adjustments, such as combined median rent and housing value, at
the segment-level and individual-level SES have not been examined. In addition, it is unknown
whether the same correlations in the main survey samples would be similar to those in the QFT
sample. Given these considerations, weighting more explicitly by SES might not eliminate
differences in estimates, such as program participation between the QFT and main survey
comparison samples.
R.3.4.2 Health Insurance Coverage

In Tables R-9 through R-12, QFT estimates for four types of health insurance coverage
for all persons aged 12 or older and three separate age groups are presented with parallel
estimates from the 2011 comparison sample, the 2012 quarters 3 and 4 comparison sample, the
2011 ACS, and the 2011 NHIS. The three separate age groups are persons aged 12 to 17, 18 to
25, and 26 or older. A few notable comparisons can be observed from these tables:
•

For all persons aged 12 or older (Table R-9), estimates for the first three types of
health insurance coverage—Medicare, Medicaid, and TRICARE, CHAMPUS, or
other military health care sources—were generally similar across all five survey data
sources. This pattern generally held for these three types of health insurance coverage
across the three separate age groups (Tables R-10 through R-12).

•

17 (Table R-10), where the QFT estimate was at least 5 percent higher than the
estimates from the other four data sources.
•

In addition, the 2011 NHIS estimate for health insurance coverage via TRICARE,
CHAMPUS, or other military health care sources for all persons aged 12 or older
(3.5 percent) was lower than the estimates from the other four data sources, which
were all close to 5 percent. This difference appeared to be driven mostly by the
estimate for persons aged 12 to 17 (Table R-10), where the 2011 NHIS estimate of
3.9 percent was higher than the estimates from the other four data sources, which
ranged from 5.2 to 5.6 percent.

•

In Tables R-13 through R-16, QFT estimates for three income categories for all persons
aged 12 or older and three separate age groups are presented with parallel estimates from the
2011 comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2011 NHIS.
The three separate age groups are persons aged 12 to 17, 18 to 25, and 26 or older. Two notable
comparisons can be observed from these tables:
•

For all persons aged 12 or older (Table R-13), the QFT estimate for family income of
$49,999 or less (52.1 percent) was only slightly higher than the 2011 and 2012
quarters 3 and 4 comparison estimate, but it was significantly higher than the 2011
NHIS estimate (46.5 percent). Correspondingly, the QFT estimates for a family
income of $50,000 to $74,999 and a family income of $75,000 or greater were lower
than estimates for the 2011 comparison sample, the 2012 quarters 3 and 4 comparison
sample, and the 2011 NHIS. QFT estimates for these two income categories were
somewhat closer to the 2011 and 2012 quarters 3 and 4 comparison estimates than to
the 2011 NHIS estimates.

•

This pattern generally held for the three separate age groups (Tables R-14 through
R-16), although the differences between the QFT estimates and the other three
sources were most pronounced for persons aged 12 to 17 (Table R-14). This finding
suggests that proxy and self-reports of income from QFT respondents aged 12 to 17
contributed the most to the observed differences in estimates for all persons compared
with the other three surveys.
R-36

Overall, the QFT estimates resulted in higher proportions of persons at lower income levels and
lower proportions at higher income levels compared with three other sources of survey data. This
difference could have accounted for some of the observed differences between QFT estimates
and the 2011 and 2012 quarters 3 and 4 comparison estimates for those items that were the most
highly correlated with income level.
R.3.4.4 Employment Status and Unemployment Rates

In Tables R-17 through R-19, QFT estimates for four employment categories for all
persons aged 18 or older and two separate age groups are presented with parallel estimates from
the 2011 comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2012
quarters 3 and 4 Current Population Survey (CPS). The two separate age groups are persons aged
18 to 25 and those aged 26 or older. A few notable comparisons can be observed from these
tables:
•

For all persons aged 18 or older (Table R-17), the QFT estimate of persons employed
full time (52.0 percent) was slightly higher than the 2011 comparison estimate
(49.7 percent) and the 2012 quarters 3 and 4 CPS estimate (49.2 percent), but it was
similar to the 2012 quarters 3 and 4 comparison estimate (51.3 percent). A similar
pattern was observed for adults aged 26 or older (Table R-19), but the differences
between the QFT and three other survey estimates of full-time employment were
more pronounced for adults aged 18 to 25 (Table R-18). This finding suggest that
reports of full-time employment from QFT respondents aged 18 to 25 contributed the
most to the observed differences in estimates for all persons compared with the other
three surveys.

•

For all persons aged 18 or older, the QFT estimate of persons with an employment
status of other (28.3 percent), such as being retired or otherwise not in the labor force,
was lower than the 2012 quarters 3 and 4 CPS estimate (34.7 percent), but it was
similar to the 2011 comparison estimate (30.4 percent) and the 2012 quarters 3 and 4
comparison estimate (29.3 percent). A similar pattern was observed for adults aged 26
or older, but the differences between the QFT and three other survey estimates for
persons with an employment status of other were more pronounced for adults aged 18
to 25. This finding suggest that reports of an employment status of "other" from QFT

R-37

respondents aged 18 to 25 contributed the most to the observed differences in
estimates for all persons compared with the other three surveys.
In addition, Table R-20 provides calculated unemployment rate estimates among persons
aged 18 or older for three age groups for the QFT, the 2011 comparison sample, the 2012
quarters 3 and 4 comparison sample, and the 2012 quarters 3 and 4 CPS. QFT unemployment
rate estimates were similar to the 2012 quarters 3 and 4 comparison sample and the 2012 quarters
3 and 4 CPS for all persons aged 18 or older and for persons aged 18 to 25. Unemployment rate
estimates for the 2011 comparison sample were higher than the other three surveys for all
persons aged 18 or older and for persons aged 18 to 25. These differences in estimates from the
lone 2011 source and the three 2012 sources could simply reflect a trend of declining
unemployment rates for adults aged 18 to 25. For adults aged 26 or older, unemployment rate
estimates were similar across all four surveys.
Overall, comparisons between the QFT and three other sources of survey data on
employment status and unemployment rates showed significant differences mostly for adults
aged 18 to 25. Observed differences for all adults and adults aged 26 or older were relatively
small. These results could be attributable to either differences in reporting employment status
among respondents aged 18 to 25 in the QFT sample or the impact of actual trends in
employment for adults aged 18 to 25 from 2011 to 2012.
R.3.4.5 Education

In Tables R-21 through R-23, QFT estimates for four education categories for all persons
aged 18 or older and two separate age groups are presented with parallel estimates from the 2011
comparison sample, the 2012 quarters 3 and 4 comparison sample, and the 2011 NHIS. The two
separate age groups are persons aged 18 to 25 and those aged 26 or older. A few notable
comparisons can be observed from these tables:
•

For all persons aged 18 or older (Table R-21), estimates for less than a high school
education and having a college degree were similar across the four surveys.

•

Differences in estimates between the QFT sample and the 2011 and 2012 quarters 3
and 4 comparison samples for the high school graduate and some college categories
were more pronounced among adults aged 26 or older (Table R-23). Among adults
aged 18 to 25, QFT estimates for the high school graduate and some college
categories were actually very similar to the 2011 and 2012 quarters 3 and 4
comparison estimates.

R-38

•

In contrast, differences in estimates between the QFT sample and the 2011 NHIS for
the high school graduate and some college categories were more pronounced among
adults aged 18 to 25 (Table R-22). Among adults aged 26 or older, QFT estimates for
the high school graduate and some college categories were similar the 2011 NHIS
estimates.

R-39

Table R-5

Received Income and Program Participation among Persons Aged 12 or Older: Percentages and Totals for 2011 Comparison,
2012 Comparison, 2012 Questionnaire Field Test, and Other Surveys
PERCENTAGES

Received Income
Social Security
Wages

Supplemental
Security Income
Food Stamps

R-40

Welfare Payments

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

TOTALS (in Thousands)
2011 ACS4
(SE)

2011 NHIS5
(SE)

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

2011 ACS4
(SE)

2011 NHIS5
(SE)

27.2

26.2

26.4

27.0

26.7

66,200

63,780

64,275

65,639

63,859

(0.42)

(0.53)

(1.70)

(0.05)

(0.35)

(1,316)

(1,727)

(5,216)

(123)

(994)

82.4

82.8

68.6

81.0

79.0

200,312

201,203

166,799

197,164

188,364

(0.38)

(0.48)

(1.77)

(0.04)

(0.32)

(2,158)

(3,028)

(8,293)

(111)

(2,197)

7.0

7.6

9.4

6.0

5.0

16,957

18,588

22,964

14,576

11,845

(0.20)

(0.30)

(0.97)

(0.03)

(0.17)

(472)

(726)

(2,558)

(79)

(418)

14.6

15.6

17.6

13.8

13.0

35,408

37,843

42,815

33,602

31,058

(0.32)

(0.46)

(1.49)

(0.05)

(0.32)

(755)

(1,141)

(3,786)

(110)

(824)

2.5

2.3

3.6

3.3

3.2

6,126

5,533

8,763

7,934

7,757

(0.11)

(0.16)

(0.56)

(0.03)

(0.14)

(278)

(373)

(1,434)

(65)

(338)

ACS = American Community Survey; Comp. = comparison; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard error.
NOTE: Unknown or invalid data were excluded from the analysis.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke
English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table R-6

Received Income and Program Participation among Persons Aged 12 to 17: Percentages and Totals for 2011 Comparison,
2012 Comparison, 2012 Questionnaire Field Test, and Other Surveys
PERCENTAGES

Received Income
Social Security
Wages

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

12.2

11.1

12.7

(0.39)

(0.42)

(1.74)

89.4

89.6

65.6

(0.36)

(0.41)

(2.67)

R-41

Supplemental
Security Income

7.6

7.8

9.9

(0.29)

(0.36)

(1.64)

Food Stamps

20.9

21.4

27.7

(0.44)

(0.64)

(2.54)

4.2

4.0

5.6

(0.23)

(0.31)

(1.15)

Welfare Payments

TOTALS (in Thousands)
2011 ACS4
(SE)

2011 NHIS5
(SE)

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

2011 ACS4
(SE)

2011 NHIS5
(SE)

10.6

12.3

2,949

2,698

3,071

2,598

2,737

(0.10)

(0.66)

(96)

(112)

(501)

(25)

(158)

90.7

87.9

21,653

21,697

15,876

22,265

19,433

(0.11)

(0.64)

(297)

(435)

(1,178)

(46)

(451)

6.0

1,846

1,877

2,389

1,464

1,329

(0.07)

(0.48)

(70)

(91)

(429)

(18)

(111)

20.9

19.4

5,061

5,174

6,707

5,132

4,309

(0.13)

(0.85)

(126)

(178)

(729)

(33)

(213)

4.9

4.7

1,024

959

1,364

1,207

1,034

(0.07)

(0.47)

(59)

(77)

(296)

(17)

(106)

Table R-7

Received Income and Program Participation among Persons Aged 18 to 25: Percentages and Totals for 2011 Comparison,
2012 Comparison, 2012 Questionnaire Field Test, and Other Surveys
PERCENTAGES

Received Income
Social Security
Wages

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

9.4

9.2

(0.29)

(0.41)

(1.44)

91.6

91.0

68.8

(0.31)

(0.74)

(2.55)

R-42

Supplemental
Security Income

6.2

5.7

9.8

(0.24)

(0.29)

(1.66)

Food Stamps

20.1

20.2

21.9

(0.46)

(0.64)

(2.47)

4.3

3.8

5.1

(0.20)

(0.27)

(1.04)

Welfare Payments

TOTALS (in Thousands)
2011 ACS4
(SE)

2011 NHIS5
(SE)

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

2011 ACS4
(SE)

2011 NHIS5
(SE)

9.9

10.3

3,108

3,025

3,036

3,314

3,251

(0.10)

(0.82)

(104)

(127)

(496)

(31)

(268)

91.7

89.6

30,200

30,015

22,698

30,658

28,138

(0.08)

(0.70)

(513)

(65)

(2,067)

(54)

(795)

5.7

4.9

2,047

1,888

3,219

1,910

1,550

(0.06)

(0.49)

(88)

(91)

(593)

(21)

(157)

18.2

19.7

6,644

6,674

7,215

6,089

6,230

(0.09)

(0.86)

(160)

(215)

(881)

(31)

(305)

4.0

6.2

1,429

1,246

1,697

1,334

1,942

(0.06)

(0.54)

(70)

(91)

(343)

(20)

(180)

Table R-8

Received Income and Program Participation among Persons Aged 26 or Older: Percentages and Totals for 2011 Comparison,
2012 Comparison, 2012 Questionnaire Field Test, and Other Surveys
PERCENTAGES

Received Income
Social Security
Wages

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

32.3

31.2

31.3

(0.53)

(0.65)

(2.10)

79.8

80.4

69.0

(0.48)

(0.59)

(2.10)

R-43

Supplemental
Security Income

7.0

8.0

9.3

(0.24)

(0.38)

(1.14)

Food Stamps

12.7

14.0

15.5

(0.37)

(0.51)

(1.56)

2.0

1.8

3.1

(0.13)

(0.17)

(0.61)

Welfare Payments

TOTALS (in Thousands)
2011 ACS4
(SE)

2011 NHIS5
(SE)

2011
Comp.1
(SE)

2012
Comp.1,2
(SE)

QFT1,3
(SE)

2011 ACS4
(SE)

2011 NHIS5
(SE)

32.2

31.2

60,143

58,058

58,168

59,727

57,872

(0.04)

(0.39)

(1,285)

(1,689)

(5,116)

(93)

(928)

77.8

76.1

148,459

149,492

128,225

144,242

140,793

(0.04)

(0.35)

(1,967)

(2,594)

(7,326)

(97)

(1,642)

6.0

4.8

13,064

14,822

17,355

11,202

8,967

(0.03)

(0.17)

(439)

(698)

(2,275)

(58)

(329)

12.1

11.1

23,703

25,995

28,893

22,381

20,519

(0.04)

(0.28)

(679)

(992)

(2,959)

(75)

(539)

2.9

2.6

3,673

3,327

5,702

5,393

4,781

(0.02)

(0.12)

(250)

(315)

(1,157)

(44)

(217)

Table R-9

Health Insurance Coverage among Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data

Instrument Item

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 ACS4
Percent (SE)

Medicare (QHI01)

18.1 (0.38)

18.0 (0.53)

18.3 (1.58)

17.8 (0.02)

17.7 (0.25)

Medicaid (QHI02 and QHI02a)

11.6 (0.24)

11.5 (0.35)

13.4 (1.16)

12.9 (0.04)

10.6 (0.21)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military
Health Care (QHI03)

4.7 (0.18)

4.6 (0.24)

5.0 (0.77)

4.8 (0.02)

3.5 (0.12)

62.1 (1.86)

67.5 (0.07)

68.7 (0.36)

Private Health Insurance (QHI06)

67.1 (0.42)

67.5 (0.59)

2011 NHIS5
Percent (SE)

R-44

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and
Medical Program of the Department of Veterans Affairs; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard
error; TRICARE = Department of Defense heath care program with three levels of coverage, prime, standard, and extra; VA = Department of Veterans Affairs.
NOTE: Unknown or invalid data were excluded from the analysis.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke
English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table R-10 Health Insurance Coverage among Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

0.4a (0.07)

0.4a (0.08)

1.8 (0.49)

0.6 (0.02)

0.2 (0.08)

Medicaid (QHI02 and QHI02a)

31.8 (0.55)

32.8 (0.80)

36.2 (2.69)

30.7 (0.13)

27.9 (0.80)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military
Health Care (QHI03)

3.1 (0.21)

2.9 (0.24)

2.6 (0.71)

2.3 (0.04)

2.3 (0.24)

60.6 (0.79)

54.9 (3.00)

62.0 (0.17)

67.9 (0.84)

Medicare (QHI01)

Private Health Insurance (QHI06)

61.3 (0.60)

R-45

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and
Medical Program of the Department of Veterans Affairs; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard
error; TRICARE = Department of Defense heath care program with three levels of coverage, prime, standard, and extra; VA = Department of Veterans Affairs.
NOTE: Unknown or invalid data were excluded from the analysis.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke
English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table R-11 Health Insurance Coverage among Persons Aged 18 to 25: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item
Medicare (QHI01)

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

0.6 (0.07)

0.8 (0.11)

1.6 (0.63)

0.7 (0.02)

0.5 (0.08)

Medicaid (QHI02 and QHI02a)

15.7 (0.42)

15.5 (0.57)

15.9 (2.15)

13.7 (0.08)

14.3 (0.52)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military
Health Care (QHI03)

2.6 (0.17)

2.7 (0.24)

2.9 (1.01)

2.4 (0.04)

2.1 (0.19)

56.5 (0.56)

58.7 (0.78)

52.3 (3.31)

61.0 (0.12)

62.3 (0.79)

Private Health Insurance (QHI06)

R-46

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and
Medical Program of the Department of Veterans Affairs; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard
error; TRICARE = Department of Defense heath care program with three levels of coverage, prime, standard, and extra; VA = Department of Veterans Affairs.
NOTE: Unknown or invalid data were excluded from the analysis.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke
English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table R-12 Health Insurance Coverage among Persons Aged 26 or Older: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, Questionnaire Field Test, 2011 ACS, and 2011 NHIS Data
Instrument Item
Medicare (QHI01)

2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 ACS4
Percent (SE)

2011 NHIS5
Percent (SE)

23.5 (0.49)

23.3 (0.67)

23.4 (1.94)

23.2 (0.02)

22.7 (0.30)

Medicaid (QHI02 and QHI02a)

8.3 (0.25)

8.1 (0.38)

10.0 (1.21)

10.4 (0.04)

7.9 (0.17)

TRICARE, CHAMPUS,
CHAMPVA, VA, Military
Health Care (QHI03)

5.3 (0.23)

5.2 (0.30)

5.6 (0.92)

5.6 (0.02)

3.9 (0.13)

64.8 (2.16)

69.3 (0.07)

69.9 (0.35)

Private Health Insurance (QHI06)

69.8 (0.50)

69.9 (0.68)

R-47

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and
Medical Program of the Department of Veterans Affairs; QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard
error; TRICARE = Department of Defense heath care program with three levels of coverage, prime, standard, and extra; VA = Department of Veterans Affairs.
NOTE: Unknown or invalid data were excluded from the analysis.
a
Difference between estimate and QFT estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, persons in institutional group quarters, and those who spoke
English "not well" or "not at all."
5
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011;
U.S. Census Bureau, American Community Survey (ACS), 2011.

Table R-13 Income among Persons Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS
2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 NHIS4
Percent (SE)

<$49,999

49.2 (0.49)

50.2 (0.63)

52.7 (2.05)

46.5 (0.54)

$50,000 - $74,999

17.5 (0.28)

16.8 (0.42)

16.3 (1.22)

18.2 (0.33)

$75,000 or More

33.3 (0.53)

33.0 (0.63)

31.0 (1.97)

35.3 (0.55)

Income Level

R-48

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test; NHIS = National Health Interview Survey; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011 comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

Table R-14 Income among Persons Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS
2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 NHIS4
Percent (SE)

<$49,999

47.8a (0.63)

47.6a (0.98)

54.9 (3.15)

41.1 (1.11)

$50,000 - $74,999

16.8a (0.38)

16.7a (0.52)

12.3 (1.60)

17.2 (0.91)

$75,000 or More

35.4 (0.57)

35.7 (0.82)

32.9 (3.01)

41.7 (1.10)

Income Level

R-49

Table R-15 Income among Persons Aged 18 to 25: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and 2011 NHIS Data
2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 NHIS4
Percent (SE)

<$49,999

66.8 (0.65)

67.2 (0.98)

68.7 (3.01)

61.2 (1.31)

$50,000 - $74,999

13.2 (0.39)

13.3 (0.59)

13.6 (2.19)

15.8 (0.85)

$75,000 or More

20.0 (0.52)

19.5 (0.64)

17.7 (2.18)

23.0 (1.16)

Income Level

R-50

Table R-16 Income among Persons Aged 26 or Older: Percentages and Standard Errors, 2011 Comparison Data, 2012 Comparison Data,
2012 Questionnaire Field Test, and NHIS Data
2011 Comparison1
Percent (SE)

2012 Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

NHIS4
Percent (SE)

<$49,999

46.3 (0.57)

47.5 (0.72)

49.6 (2.36)

44.6 (0.52)

$50,000 - $74,999

18.3 (0.36)

17.5 (0.55)

17.3 (1.46)

18.7 (0.33)

$75,000 or More

35.4 (0.60)

35.1 (0.74)

33.1 (2.42)

36.7 (0.54)

Income Level

R-51

Table R-17 Levels of Current Employment among Persons Aged 18 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data
2011 Comparison1
Percent (SE)

2012
Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

CPS Q3 & Q44
Percent (SE)

Full-Time

49.7 (0.49)

51.3 (0.63)

52.0 (1.65)

49.2 (0.07)

Part-Time

14.1 (0.26)

13.9 (0.39)

14.2 (1.15)

11.2 (0.05)

5.8 (0.14)

5.5 (0.20)

5.5 (0.65)

4.9 (0.03)

30.4 (0.43)

29.3 (0.65)

28.3 (1.70)

34.7 (0.07)

Current Employment

Unemployed
Other

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
5
The Other Employment category includes students, person keeping house or caring for children full time, retired or
disabled persons, or other persons not in the labor force.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

R-52

Table R-18 Levels of Current Employment among Persons Aged 18 to 25: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data
2011 Comparison1
Percent (SE)

2012
Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

CPS Q3 & Q44
Percent (SE)

Full-Time

36.0a (0.56)

40.1 (0.86)

45.5 (2.98)

35.0 (0.19)

Part-Time

27.8 (0.42)

26.4 (0.67)

24.4 (2.29)

22.4 (0.17)

Unemployed

13.2 (0.33)

11.8 (0.41)

11.9 (1.58)

9.4 (0.12)

21.7 (0.91)

18.2 (1.83)

33.2 (0.19)

Current Employment

Other

23.0 (0.43)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
5
The Other Employment category includes students, person keeping house or caring for children full time, retired or
disabled persons, or other persons not in the labor force.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

R-53

Table R-19 Levels of Current Employment among Persons Aged 26 or Older: Percentages and
Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test,
and CPS Data
2011 Comparison1
Percent (SE)

2012
Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

CPS Q3 & Q44
Percent (SE)

Full-Time

52.1 (0.55)

53.3 (0.72)

53.2 (1.90)

51.5 (0.08)

Part-Time

11.7 (0.30)

11.7 (0.43)

12.4 (1.34)

9.3 (0.04)

4.5 (0.16)

4.4 (0.23)

4.3 (0.70)

4.2 (0.03)

31.7 (0.51)

30.7 (0.75)

30.1 (2.01)

35.0 (0.08)

Current Employment

Unemployed
Other

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
5
The Other Employment category includes students, person keeping house or caring for children full time, retired or
disabled persons, or other persons not in the labor force.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

R-54

Table R-20 Unemployment Rates among Persons Aged 18 or Older, by Age Group: Percentages
and Standard Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field
Test, and CPS Data

Age/Unemployment Rate

2012
Comparison1,2
Percent (SE)

2012 QFT1,3
Percent (SE)

8.4 (0.21)

7.8 (0.29)

7.6 (0.91)

7.6 (0.05)

17.2 (0.21)

15.0 (0.48)

14.6 (1.93)

14.0 (0.18)

6.6 (0.23)

6.3 (0.34)

6.2 (1.00)

6.5 (0.05)

2011 Comparison1
Percent (SE)

CPS Q3 & Q44
Percent (SE)

18 or Older
Unemployment Rate
18 to 25
Unemployment Rate
26 or Older
Unemployment Rate

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; Q = quarter; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample does not include Alaska or Hawaii.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
U.S. Census Bureau and U.S. Bureau of Labor Statistics (BLS), Current Population Survey (CPS).

R-55

Table R-21 Levels of Education among Persons Aged 18 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011
NHIS
Level of Education

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 NHIS4
Percent (SE)

< High School

11.6 (0.24)

11.5 (0.35)

12.4 (1.26)

12.0 (0.20)

High School Graduate

30.3 (0.38)

30.1 (0.61)

26.6 (1.92)

27.8 (0.29)

Some College

27.4a (0.37)

27.7a (0.48)

32.1 (1.42)

31.3 (0.26)

College Graduate

30.6 (0.41)

30.7 (0.67)

29.0 (2.48)

28.9 (0.38)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012; CDC, National Center for Health Statistics, National Health
Interview Survey (NHIS), 2011.

R-56

Table R-22 Levels of Education among Persons Aged 18 to 25: Percentages and Standard Errors,
2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011 NHIS
Level of Education

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)

2012 QFT1,3
Percent (SE)

2011 NHIS4
Percent (SE)

< High School

15.6 (0.40)

12.0 (0.42)

13.8 (1.92)

14.0 (0.49)

High School Graduate

34.0 (0.55)

35.7 (1.04)

34.9 (2.56)

29.6 (0.65)

Some College

35.7 (0.59)

36.4 (0.90)

37.6 (3.40)

43.0 (0.83)

College Graduate

14.7 (0.46)

15.9 (0.60)

13.7 (2.30)

13.5 (0.54)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

R-57

Table R-23 Levels of Education among Persons Aged 26 or Older: Percentages and Standard
Errors, 2011 Comparison, 2012 Comparison, 2012 Questionnaire Field Test, and 2011
NHIS
Level of Education
< High School
High School Graduate

2011 Comparison1 2012 Comparison1,2
Percent (SE)
Percent (SE)
10.9 (0.28)
11.4 (0.41)
29.7a (0.43)
a

29.1 (0.69)
a

2012 QFT1,3
Percent (SE)
12.1 (1.39)

2011 NHIS4
Percent (SE)
11.6 (0.21)

25.1 (2.16)

27.5 (0.31)

Some College

26.0 (0.41)

26.2 (0.57)

31.1 (1.76)

29.3 (0.25)

College Graduate

33.4 (0.47)

33.3 (0.77)

31.7 (2.77)

31.6 (0.40)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
NHIS = National Health Interview Survey; QFT = NSDUH Questionnaire Field Test; SE = standard error.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (e.g., 2011
comparison compared with 2012 QFT).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Sample only includes interviews done in English.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health,
September 1 through November 3, 2012;
CDC, National Center for Health Statistics, National Health Interview Survey (NHIS), 2011.

R-58

R.3.5 Potential Impact of Proxy Reporting for Items with Observed Data Quality Issues
Two sets of questionnaire items that were moved from CAPI to ACASI administration in
the QFT questionnaire—health insurance and income—allowed for a proxy respondent to
answer these questions in lieu of the primary respondent. For example, about 75 percent of youth
respondents aged 12 to 17 nominate a parent or other adult in their household to answer these
questions instead of them. QFT respondents were significantly more likely to use a proxy
reporter for these questions than 2011 and 2012 quarters 3 and 4 comparison respondents. One
further difference for all persons aged 12 or older was that QFT respondents were more likely
than 2011 and 2012 quarters 3 and 4 respondents to use a proxy reporter for the health insurance
and income items. Among QFT respondents, 15.7 percent reported using a proxy compared with
13.7 percent among 2011 comparison sample respondents and 13.9 percent among 2012 quarters
3 and 4 comparison sample respondents.
Given this difference, reporting patterns among proxies could be one possible source of
observed differences between QFT estimates and 2011 and 2012 quarters 3 and 4 comparison
estimates for these items. This section presents and discusses two types of data on proxy reports
in the QFT data compared with the 2011 and 2012 quarters 3 and 4 comparison data:
•

the distribution of proxy relationships to the primary respondent and

•

estimates for proxy reports versus respondent reports for these items.

These analyses will provide some insight on whether the greater use of proxy reporters in the
QFT appeared to have any impact on differences observed between the QFT estimates and the
2011 and 2012 quarters 3 and 4 comparison estimates for these items.
Table R-24 shows the distribution of respondents' relationships with their proxy reporters
for youths aged 12 to 17 and adults aged 18 or older for the QFT sample, the 2011 comparison
sample, and the 2012 quarters 3 and 4 comparison sample.51 Overall, the distributions of proxy
relationships across 11 types of relationships were very similar across all three datasets for both
youths and adults. For youths aged 12 to 17 in all three samples, a little over two thirds of
proxies were mothers of the primary respondents, and about one quarter were fathers. For adults
aged 18 or older in all three samples, about 60 percent of proxies were spouses, and about
23 percent were mothers. Proportions for other relationship categories for both youths and adults
were relatively small. Only one difference among all relationship categories was statistically
significant. For adult respondents, the QFT sample proportion (0.2 percent) for using another
adult relative as a proxy was significantly lower than the 2011 comparison sample proportion
(1.5 percent). This proportion was 1.0 percent for the 2012 quarters 3 and 4 comparison sample,
but the difference between the QFT and the 2012 quarters 3 and 4 proportions was not
statistically significant. The lack of significant differences in the distribution of respondents'
relationships with their proxy reporters across the three datasets indicates that proxy relationships
to those respondents who used proxies were not a factor in explaining differences in estimates
between the samples for items where proxy reporting was allowed.

To aid in their readability, Table R-24 through Table R-27 appear together at the end of their discussion
in this Section R.3.5.

R-59

Although the relationship of proxy reporters to primary respondents was not a factor in
observed differences in relevant estimates among the three datasets, the higher overall use of
proxy reporters could have been a contributor to these observed differences. To explore this
possibility, Tables R-25 through R-27 compare estimates from proxy reports versus primary
respondent reports for three age group categories: (1) all respondents aged 12 or older, (2) youth
respondents aged 12 to 17, and (3) adult respondents aged 18 or older. If the greater use of proxy
reporters in the QFT was at least partly responsible for differences in estimates between the QFT
sample and the 2011 and 2012 quarters 3 and 4 comparison samples, significant differences in
the relevant estimates would be expected among the proxy reports and small or no differences
would be expected among the primary respondent reports. These results revealed two important
patterns among estimates that differed significantly between the QFT sample and the 2011 and
2012 quarters 3 and 4 comparison samples.
One pattern observed for several estimates was differences between the QFT and the
2011 and 2012 quarters 3 and 4 comparison samples being of similar magnitude for both proxy
and nonproxy reports. For example, the QFT estimate among all respondents aged 12 or older
(Table R-25) for having private health insurance that includes coverage for treatment of alcohol
abuse or alcoholism (item QHI08) was 73.7 percent for data reported by proxies. The QFT
proportion was significantly lower than the proxy-reported estimates for the 2011 comparison
sample (84.7 percent) and the 2012 quarters 3 and 4 comparison sample (85.1 percent). Looking
at the same estimates for data reported by the primary respondents, the QFT estimate
(76.8 percent) was similarly lower than the 2011 comparison sample (84.0 percent) and the 2012
quarters 3 and 4 comparison sample (84.2 percent). The greater use of proxies among QFT
respondents was clearly not a significant factor in explaining differences between the three
datasets for items where this pattern of results was observed.
A second pattern observed for some items was QFT proxy and nonproxy estimates being
different from each other, but still significantly different from the parallel 2011 comparison and
2012 quarters 3 and 4 comparison estimates. For example, Table R-25 shows that the QFT
proportion for receiving income from wages or pay earned from working at a job or business
(item QI05N) was 63.8 percent for data reported by proxies. The QFT proportion was
significantly lower than the proxy-reported estimates for the 2011 comparison sample
(84.9 percent) and 2012 quarters 3 and 4 comparison sample (86.3 percent). For the same
estimates for data reported by the primary respondents, the QFT estimate (71.6 percent) was
significantly higher than the QFT proxy estimates, but still significantly lower than the 2011
comparison sample (87.2 percent) and the 2012 quarters 3 and 4 comparison sample
(87.5 percent). A similar pattern was observed for receipt of food stamps (item QI07N), where
the difference between QFT estimates for proxy reports compared with the 2011 and 2012
quarters 3 and 4 comparison estimates was significantly greater than the difference in estimates
for nonproxy reports, but still significantly different. The greater use of proxies among QFT
respondents appeared to be a factor in explaining differences between the three datasets for items
where this pattern of results was observed. For these items, proxy reports exacerbated differences
between QFT estimates versus 2011 and 2012 quarters 3 and 4 comparison estimates, but did not
fully account for these differences.
Another important conclusion from Tables R-25 through R-27 is that the two patterns
identified above appeared to hold for both youth respondents aged 12 to 17 than among adult
R-60

respondents. Estimates for nonproxy reports for several of these items for respondents aged 12 to
17 were of low precision because of the low numbers of respondents in this category
(Table R-25). These low precision estimates prohibited conclusions to be reached on the
statistical significance of observed differences for youth respondents, but the proportions for
both proxy and nonproxy reports appeared to fit the two main patterns.

R-61

Table R-24 Distribution of Respondent Relationship with Proxy among Persons Aged 12 or Older Who Obtained a Proxy, by Age Group:
Percentages, and Standard Errors, 2011 Comparison, 2012 Comparison, and 2012 Questionnaire Field Test
2011
Comparison1
12-17,
Percent (SE)

2012
Comparison1,2
12-17,
Percent (SE)

2012 QFT1,3
12-17,
Percent (SE)

Father

23.7 (0.42)

23.7 (0.63)

25.1 (2.62)

6.2 (0.44)

6.4 (0.60)

4.6 (1.49)

Mother

69.7 (0.45)

69.3 (0.70)

67.8 (2.76)

22.6 (0.86)

22.9 (1.28)

23.2 (3.39)

Proxy Relationship

2011
Comparison1 2012 Comparison1,2
18 or Older,
18 or Older,
Percent (SE)
Percent (SE)

Son / Daughter

0.0 (0.00)

0.0 (0.02)

0.2 (0.16)

6.1 (1.09)

5.1 (1.22)

0.0* (0.00)

Brother / Sister

1.7 (0.15)

1.8 (0.17)

1.9 (0.72)

1.1 (0.25)

1.1 (0.34)

2.2 (1.31)

Husband / Wife

0.0* (0.00)

58.2 (1.18)

57.4 (1.85)

62.0 (4.04)

Live-in Boyfriend / Girlfriend

0.0 (0.01)

0.0 (0.02)

0.2 (0.19)

2.8 (0.47)

4.0 (0.77)

6.7 (2.60)

2012 QFT1,3
18 or Older,
Percent (SE)

R-62

Son-in-law / Daughter-in-law

0.0 (0.00)

0.4 (0.38)

0.0* (0.00)

Grandson / Granddaughter

0.0* (0.00)

0.3 (0.19)

0.3 (0.30)

0.0* (0.00)

Father-in-law / Mother-in-law

0.0* (0.00)

0.4 (0.22)

0.5 (0.36)

0.0* (0.00)

Grandfather / Grandmother

3.0 (0.17)

3.2 (0.24)

2.3 (0.62)

0.9 (0.17)

0.9 (0.18)

1.1 (0.62)

Other Adult Relative

1.9 (0.15)

2.0 (0.22)

2.6 (0.98)

1.5a (0.37)

1.0 (0.38)

0.2 (0.23)

Instrument Item
Covered by private health
Insurance? (QHI06)4,5
Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems?
(QHI10)4,5
In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5
In [YEAR], did [FILL] receive
Supplemental Security Income
or SSI? (QI03N)4,5
In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job
or business? (QI05N)4,5
In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5
At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or
county welfare program such
as [TANFFILL]? (QI08N)4,5
In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5
Before taxes and other
deductions, was [MEMBER]
total personal income from all
sources during [YEAR] more
or less than 20,000 dollars?
(QI20N)4,5
$20,000 or More
Less Than $20,000
See notes at end of table.

2011
2012
2011
2012
Comparison1 Comparison1,2 2012 QFT1,3 Comparison1 Comparison1,2 2012 QFT1,3
Proxy Percent
Proxy
Proxy
No Proxy
No Proxy
No Proxy
(SE)
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)
64.6 (0.79)

65.3 (0.96)

59.5 (3.04)

69.6a (0.49)

69.4 (0.67)

64.9 (2.19)

84.7a (0.88)

85.1a (1.05)

73.7 (5.07)

84.9a (0.52)

84.7a (0.82)

76.8 (2.13)

84.7a (0.89)

84.6a (1.04)

76.3 (3.65)

84.0a (0.53)

84.3a (0.85)

74.8 (2.26)

91.7a (0.54)

91.3a (0.74)

83.3 (3.24)

91.9a (0.32)

92.4a (0.55)

85.7 (1.80)

21.1 (0.73)

19.7 (1.18)

22.2 (2.86)

27.6 (0.53)

26.3 (0.60)

26.4 (2.06)

8.6 (0.44)

8.8 (0.53)

10.0 (1.84)

6.5a (0.23)

7.6 (0.39)

9.4 (1.18)

84.9a (0.60)

86.3a (0.79)

63.8 (2.66)

87.2a (0.42)

87.5a (0.50)

71.6 (1.90)

18.2a (0.62)

18.0a (0.74)

23.9 (2.50)

13.3 (0.36)

14.6 (0.47)

15.2 (1.67)

3.4 (0.24)

3.1 (0.26)

3.9 (0.92)

2.3 (0.13)

2.0 (0.16)

2.7 (0.59)

3.9 (0.25)

4.2 (0.34)

4.9 (1.21)

3.0 (0.15)

2.7 (0.16)

2.9 (0.58)

14.1 (0.80)
85.9 (0.80)

15.0 (0.99)
85.0 (0.99)

19.2 (2.64)
80.8 (2.64)

58.4a (0.46)
41.6a (0.46)

58.4a (0.62)
41.6a (0.62)

64.9 (1.74)
35.1 (1.74)
(continued)

R-63

Table R-25 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons
Aged 12 or Older: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, and Questionnaire Field Test Data (continued)
2011
2011
2012
2012
Comparison1 Comparison1,2 2012 QFT1,3 Comparison1 Comparison1,2 2012 QFT1,3
Proxy Percent
Proxy
Proxy
No Proxy
No Proxy
No Proxy
(SE)
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B)4,5
Less Than $1,000
60.2a (0.84)
60.1a (1.10)
53.7 (2.84)
10.5a (0.23)
10.4a (0.34)
7.6 (0.80)
$1,000-$1,999
4.1 (0.17)
4.3 (0.31)
4.5 (0.86)
1.9 (0.10)
2.0 (0.14)
2.4 (0.42)
$2,000-$2,999
3.0 (0.22)
2.7 (0.24)
1.9 (0.87)
1.6a (0.09)
1.4 (0.11)
1.0 (0.22)
$3,000-$3,999
1.9 (0.16)
2.1 (0.24)
2.1 (0.65)
1.4 (0.09)
1.5 (0.15)
1.1 (0.31)
$4,000-$4,999
1.4 (0.12)
1.4 (0.15)
2.9 (1.25)
1.3a (0.08)
1.1 (0.11)
0.7 (0.20)
$5,000-$5,999
2.0a (0.26)
1.2 (0.21)
0.9 (0.37)
1.6a (0.10)
1.4 (0.11)
0.9 (0.30)
$6,000-$6,999
1.9 (0.37)
1.1 (0.14)
0.9 (0.40)
1.4 (0.11)
1.6 (0.17)
1.0 (0.34)
$7,000-$7,999
1.4 (0.16)
1.1 (0.18)
0.5 (0.43)
1.6a (0.11)
1.6a (0.18)
0.4 (0.25)
$8,000-$8,999
1.2 (0.14)
1.5 (0.26)
1.1 (0.50)
1.8 (0.11)
1.8 (0.17)
1.3 (0.40)
$9,000-$9,999
1.6 (0.27)
1.7 (0.47)
2.1 (1.21)
1.8 (0.11)
1.8 (0.16)
2.7 (0.66)
$10,000-$10,999
1.2 (0.18)
1.4 (0.22)
3.1 (1.30)
2.2 (0.15)
2.1 (0.17)
2.2 (0.53)
$11,000-$11,999
0.7 (0.13)
1.0 (0.20)
0.5 (0.33)
1.5 (0.10)
1.8 (0.18)
1.7 (0.50)
$12,000-$12,999
1.0 (0.24)
1.4 (0.34)
0.7 (0.58)
2.2a (0.13)
2.6a (0.24)
1.3 (0.38)
$13,000-$13,999
0.8a (0.20)
1.0a (0.27)
0.2 (0.19)
1.5 (0.11)
1.3 (0.12)
1.2 (0.35)
$14,000-$14,999
0.6 (0.16)
0.5 (0.14)
0.9 (0.65)
1.5a (0.11)
1.7a (0.15)
0.9 (0.30)
$15,000-$15,999
0.5 (0.10)
0.6 (0.17)
0.3 (0.25)
1.8 (0.11)
1.6 (0.14)
2.1 (0.50)
$16,000-$16,999
0.2 (0.09)
0.4 (0.17)
1.4 (0.95)
1.2 (0.10)
1.3 (0.12)
1.6 (0.39)
$17,000-$17,999
0.8 (0.29)
0.2 (0.08)
1.3 (0.95)
1.4 (0.09)
1.2 (0.12)
1.2 (0.40)
$18,000-$18,999
0.9a (0.21)
0.8 (0.21)
0.3 (0.22)
1.8 (0.11)
1.7 (0.16)
1.9 (0.49)
$19,000-$19,999
0.8 (0.17)
0.7 (0.25)
1.5 (0.84)
1.8 (0.12)
1.7 (0.16)
2.0 (0.50)
$20,000-$24,999
2.4 (0.32)
2.6 (0.42)
4.1 (1.28)
6.8 (0.24)
6.8 (0.33)
8.5 (1.06)
$25,000-$29,999
2.3 (0.35)
1.7 (0.32)
2.7 (1.19)
6.6 (0.31)
6.2 (0.32)
6.2 (0.92)
$30,000-$34,999
1.7 (0.32)
1.8 (0.36)
2.4 (1.25)
5.9 (0.26)
5.7 (0.26)
5.3 (0.93)
$35,000-$39,999
1.2 (0.22)
1.4 (0.40)
1.0 (0.71)
5.0 (0.23)
5.0 (0.33)
7.0 (1.08)
$40,000-$44,999
1.3 (0.24)
1.7 (0.50)
1.2 (0.77)
4.4 (0.20)
4.4 (0.27)
5.3 (0.90)
$45,000-$49,999
1.1 (0.22)
1.3 (0.29)
2.3 (1.19)
4.2 (0.18)
4.8 (0.29)
6.0 (1.04)
$50,000-$74,999
2.4 (0.31)
2.4 (0.37)
2.7 (1.26)
12.0 (0.34)
12.2 (0.45)
12.2 (1.47)
$75,000-$99,999
0.8 (0.19)
0.6 (0.17)
1.9 (1.10)
5.7 (0.23)
5.5 (0.36)
5.7 (1.00)
$100,000 or More
0.4 (0.13)
1.2 (0.36)
1.0 (0.62)
7.8 (0.35)
7.5 (0.49)
8.9 (1.64)
$100,000-$149,999
-- (--)
-- (--)
1.0 (0.62)
-- (--)
-- (--)
5.1 (1.15)
$150,000 or More
-- (--)
-- (--)
0.0* (0.00* )
-- (--)
-- (--)
3.8 (1.26)
* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
-- Not available.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being intervieweradministered to self- administered.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not answer
HASJOIN, he or she is defined as not having used a proxy. Respondents who were legitimately skipped from answering
question QP01 were excluded from this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (i.e., 2011 comparison
proxy compared with 2012 QFT proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing
data.
5
Estimate is based on an edited version of the variable.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

R-64

Table R-26 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons
Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
and 2012 Questionnaire Field Test

Instrument Item
Covered by private health
Insurance? (QHI06) 4,5
Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems? (QHI10)4,5
In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5
In [YEAR], did [FILL] receive
Supplemental Security Income
or SSI? (QI03N)4,5
In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job or
business? (QI05N)4,5
In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5
At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or county
welfare program such as
[TANFFILL]? (QI08N)4,5
In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5
Before taxes and other deductions,
was [MEMBER] total personal
income from all sources during
[YEAR] more or less than
20,000 dollars? (QI20N)4,5
$20,000 or More
Less Than $20,000
See notes at end of table.

2011
2011
2012
2012
Comparison1 Comparison1,2 2012 QFT1,3 Comparison1 Comparison1,2 2012 QFT1,3
Proxy
Proxy
Proxy
No Proxy
No Proxy
No Proxy
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)
63.0 (0.58)

62.5 (0.78)

58.9 (3.06)

51.7a (1.37)

49.2a (2.04)

31.5* (5.84*)

86.8a (0.54)

87.6a (0.78)

78.0 (3.52)

64.6 (2.29)

60.4 (3.50)

43.3* (16.72*)

86.7a (0.56)

86.8a (0.81)

78.1 (3.16)

64.6 (2.34)

59.3 (3.52)

44.6* (17.16*)

92.9 (0.36)

92.8 (0.45)

88.6 (2.69)

82.7 (1.57)

81.1 (2.74)

57.9* (16.19*)

11.9 (0.41)

10.7 (0.43)

12.1 (1.88)

14.3 (0.97)

13.4 (1.12)

16.4* (4.18*)

7.5 (0.31)

8.0 (0.39)

9.4 (1.81)

8.2 (0.73)

6.2 (0.81)

14.5* (5.42*)

89.4a (0.36)

89.4a (0.47)

64.0 (2.73)

91.8a (0.73)

92.5a (0.91)

74.8* (7.17*)

20.2a (0.45)

20.4a (0.65)

26.7 (2.64)

25.0 (1.15)

26.9 (1.56)

37.9* (7.59*)

4.1 (0.23)

3.9 (0.33)

5.5 (1.20)

5.1 (0.63)

4.3 (0.62)

5.7* (3.25*)

4.2 (0.21)

4.2 (0.29)

6.3 (1.33)

5.9a (0.60)

5.5a (0.80)

0.0* (0.00*)

0.4a (0.07)
99.6a (0.07)

0.4a (0.10)
99.6a (0.10)

6.5 (1.42)
93.5 (1.42)

0.5a (0.13)
99.5a (0.13)

0.9 (0.30)
99.1 (0.30)

10.1* (4.73*)
89.9* (4.73*)
(continued)

R-65

Table R-26 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons
Aged 12 to 17: Percentages and Standard Errors, 2011 Comparison, 2012 Comparison,
and Questionnaire Field Test Data (continued)
2011
2011
2012
2012
Comparison1 Comparison1,2 2012 QFT1,3 Comparison1 Comparison1,2 2012 QFT1,3
Proxy
Proxy
Proxy
No Proxy
No Proxy
No Proxy
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B) 4,5
Less Than $1,000
85.3 (0.35) 85.8 (0.46) 82.2 (2.18)
78.6a (0.98)
78.8a (1.30)
63.6* (7.10*)
$1,000-$1,999
4.4 (0.16)
4.3 (0.29)
4.1 (1.14)
7.5 (0.64)
9.3 (0.95)
11.7* (4.46*)
$2,000-$2,999
2.4a (0.17)
2.2a (0.19)
0.8 (0.48)
4.2 (0.44)
3.5 (0.54)
2.7* (2.73*)
$3,000-$3,999
1.6 (0.13)
1.6 (0.16)
1.4 (0.65)
2.5 (0.35)
2.5 (0.48)
2.3* (2.25*)
$4,000-$4,999
1.2 (0.10)
1.1 (0.13)
1.0 (0.50)
1.4 (0.26)
1.1 (0.25)
1.3* (1.29*)
$5,000-$5,999
0.9 (0.09)
0.6 (0.10)
0.4 (0.30)
1.2a (0.28)
0.6a (0.19)
0.0* (0.00*)
$6,000-$6,999
0.8 (0.09)
0.6 (0.09)
0.8 (0.50)
1.1 (0.27)
0.9 (0.33)
1.7* (1.73*)
$7,000-$7,999
0.7a (0.08)
0.8a (0.10)
0.2 (0.18)
0.3a (0.10)
0.7a (0.22)
0.0* (0.00*)
a
a
$8,000-$8,999
0.6 (0.10)
0.7 (0.10)
0.4 (0.30)
0.4 (0.12)
0.4 (0.17)
0.0* (0.00*)
$9,000-$9,999
0.4a (0.07)
0.4a (0.09)
0.0* (0.00*)
0.3a (0.11)
0.0 (0.05)
0.0* (0.00*)
$10,000-$10,999
0.3 (0.05)
0.5 (0.08)
0.3 (0.27)
0.7 (0.16)
0.6 (0.27)
1.3* (1.36*)
$11,000-$11,999
0.2 (0.04)
0.2 (0.06)
0.2 (0.23)
0.1 (0.08)
0.3 (0.17)
0.0* (0.00*)
$12,000-$12,999
0.3 (0.09)
0.3 (0.07)
0.2 (0.20)
0.1 (0.06)
0.1 (0.06)
2.0* (1.97*)
$13,000-$13,999
0.1 (0.04)
0.1 (0.04)
0.1 (0.10)
0.1 (0.05)
0.1 (0.12)
1.5* (1.46*)
$14,000-$14,999
0.1a (0.04)
0.1a (0.05)
0.0* (0.00*)
0.1 (0.09)
0.0 (0.02)
0.0* (0.00*)
*
*
$15,000-$15,999
0.1 (0.04)
0.1 (0.05)
0.5 (0.48 )
0.5 (0.17)
0.1 (0.05)
1.4* (1.42*)
$16,000-$16,999
0.0 (0.02)
0.1 (0.04)
0.3 (0.24)
0.0 (0.03)
0.0 (0.04)
1.5* (1.53*)
$17,000-$17,999
0.0a (0.01)
0.1 (0.03)
0.0* (0.00*)
0.4a (0.17)
0.0* (0.00*)
0.0* (0.00*)
$18,000-$18,999
0.1 (0.03)
0.1 (0.04)
0.1 (0.09)
0.0 (0.04)
0.1 (0.15)
0.0* (0.00*)
$19,000-$19,999
0.1 (0.04)
0.1 (0.04)
0.5 (0.39)
0.0 (0.03)
0.0* (0.00*)
0.0* (0.00*)
a
a
$20,000-$24,999
0.1 (0.02)
0.2 (0.05)
4.2 (1.06)
0.1 (0.06)
0.3 (0.22)
2.4* (2.20*)
*
*
$25,000-$29,999
0.1 (0.03)
0.1 (0.05)
0.8 (0.45)
0.0 (0.02)
0.0 (0.00 )
0.0* (0.00*)
$30,000-$34,999
0.0 (0.02)
0.1 (0.03)
0.4* (0.44*)
0.0* (0.00*)
0.3 (0.17)
4.3* (3.07*)
*
*
*
*
$35,000-$39,999
0.0 (0.01)
0.0 (0.00 ) 0.0 (0.00 )
0.0 (0.03)
0.1 (0.07)
0.0* (0.00*)
$40,000-$44,999
0.0* (0.00*)
0.0 (0.02)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
0.0* (0.00*)
*
*
*
*
*
*
$45,000-$49,999
0.0 (0.00 )
0.0 (0.00 ) 0.2 (0.23)
0.0 (0.00 )
0.1 (0.07)
0.0* (0.00*)
$50,000-$74,999
0.1 (0.03)
0.0* (0.00*) 0.4 (0.26)
0.0 (0.03)
0.0* (0.00*)
2.1* (1.93*)
*
*
*
*
*
*
$75,000-$99,999
0.0 (0.02)
0.0 (0.00 ) 0.2 (0.24)
0.0 (0.00 )
0.0 (0.00 )
0.0* (0.00*)
a
a
*
*
$100,000 or More
0.0 (0.02)
0.1 (0.04)
0.0 (0.00 )
0.0 (0.03)
0.2 (0.10)
0.0* (0.00*)
$100,000-$149,999
-- (--)
-- (--)
0.0* (0.00*)
-- (--)
-- (--)
0.0* (0.00*)
*
*
$150,000 or More
-- (--)
-- (--)
0.0 (0.00 )
-- (--)
-- (--)
0.0* (0.00*)
* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
-- Not available.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being intervieweradministered to self- administered.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not
answer HASJOIN, he or she is defined as not having used a proxy. Respondents who were legitimately skipped from
answering question QP01 were excluded from this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (i.e., 2011 comparison
proxy compared with 2012 QFT proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing
data.
5
Estimate is based on an edited version of the variable.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

R-66

Table R-27 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons
Aged 18 or Older: Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, and 2012 Questionnaire Field Test

Instrument Item
Covered by private health
Insurance? (QHI06) 4,5
Does [MEMBER] private health
insurance include coverage for
treatment of alcohol abuse or
alcoholism? (QH108)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)4,5
Does [MEMBER] private health
insurance include coverage for
treatment for mental or
emotional problems? (QHI10)4,5
In [YEAR], did [FILL] receive
Social Security or Railroad
Retirement payments?
(QI01N)4,5
In [YEAR], did [FILL] receive
Supplemental Security Income
or SSI? (QI03N)4,5
In [YEAR], did [FILL] receive
income from wages or pay
earned while working at a job
or business? (QI05N) 4,5
In [YEAR], did [FILL] receive
food stamps? (QI07N)4,5
At any time during [YEAR], did
[FILL] receive any cash
assistance from a state or
county welfare program such as
[TANFFILL]? (QI08N)4,5
In [YEAR], because of low
income, did [FILL] receive any
other kind of nonmonetary
welfare or public assistance?
(QI10N)4,5
Before taxes and other deductions,
was [MEMBER] total personal
income from all sources during
[YEAR] more or less than
20,000 dollars? (QI20N)4,5
$20,000 or More
Less Than $20,000
See notes at end of table.

69.6 (1.84)

60.1 (5.55)

70.0 (0.50)

69.8 (0.67)

65.5 (2.24)

81.7 (1.82)

81.5 (2.27)

69.2* (8.71* )

85.1a (0.53)

85.0a (0.82)

77.0 (2.14)

81.8 (1.88)

81.3 (2.28)

74.4* (6.19* )

84.2a (0.54)

84.6a (0.85)

75.0 (2.26)

89.8a (1.28)

89.2 (1.68)

77.6* (5.92* )

92.0a (0.33)

92.5a (0.55)

85.9 (1.78)

35.4 (1.61)

33.3 (2.60)

33.7 (5.20)

27.9 (0.54)

26.6 (0.61)

26.6 (2.09)

10.2 (0.97)

10.0 (1.12)

10.7 (3.20)

6.5a (0.23)

7.6 (0.40)

9.3 (1.18)

78.0a (1.38)

81.4a (1.78)

63.5 (4.30)

87.0a (0.43)

87.4a (0.51)

71.5 (1.93)

15.2 (1.25)

14.4 (1.31)

20.7 (3.99)

13.0 (0.36)

14.3 (0.47)

14.8 (1.66)

2.3 (0.38)

2.0 (0.41)

2.1 (1.30)

2.2 (0.13)

2.0 (0.16)

2.6 (0.60)

3.5 (0.52)

4.1 (0.70)

3.3 (1.77)

3.0 (0.15)

2.6 (0.16)

2.9 (0.59)

35.5 (1.81)
64.5 (1.81)

37.6 (2.01)
62.4 (2.01)

33.7 (5.05)
66.3 (5.05)

59.8a (0.46)
40.2a (0.46)

59.7a (0.62)
40.3a (0.62)

65.8 (1.76)
34.2 (1.76)
(continued)

R-67

Table R-27 Use of Proxy in Moved Items in the 2012 Questionnaire Field Test among Persons
Aged 18 or Older, Percentages and Standard Errors, 2011 Comparison, 2012
Comparison, and Questionnaire Field Test Data (continued)
2011
2011
2012
2012
Comparison1 Comparison1,2 2012 QFT1,3 Comparison1 Comparison1,2 2012 QFT1,3
Proxy
Proxy
Proxy
No Proxy
No Proxy
No Proxy
Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE) Percent (SE)

Instrument Item
Of these income groups, which
category best represents
[MEMBER] total personal
income during [YEAR]?
(QI21A and QI21B) 4,5
Less Than $1,000
20.4 (1.24)
19.3 (1.14) 21.6 (4.06)
8.9a (0.22)
8.8a (0.34)
6.7 (0.81)
$1,000-$1,999
3.6 (0.39)
4.3 (0.62)
4.9 (1.27)
1.7 (0.10)
1.8 (0.15)
2.3 (0.42)
$2,000-$2,999
3.8 (0.50)
3.4 (0.55)
3.1 (1.83)
1.5a (0.09)
1.4 (0.12)
1.0 (0.22)
$3,000-$3,999
2.4 (0.37)
2.8 (0.54)
2.8 (1.14)
1.3 (0.09)
1.5 (0.15)
1.1 (0.32)
$4,000-$4,999
1.9 (0.27)
1.8 (0.34)
4.9* (2.75*)
1.3a (0.08)
1.1 (0.12)
0.6 (0.20)
$5,000-$5,999
3.7a (0.64)
2.1 (0.52)
1.4 (0.71)
1.6a (0.10)
1.4 (0.11)
0.9 (0.30)
$6,000-$6,999
3.7a (0.91)
1.8 (0.37)
1.1 (0.65)
1.4 (0.11)
1.7 (0.17)
0.9 (0.34)
$7,000-$7,999
2.6 (0.39)
1.7 (0.43)
0.9* (0.89*)
1.6a (0.11)
1.6a (0.18)
0.4 (0.25)
$8,000-$8,999
2.0 (0.30)
2.7 (0.66)
1.9 (1.03)
1.8 (0.11)
1.8 (0.18)
1.3 (0.41)
$9,000-$9,999
3.5 (0.67)
3.8 (1.18)
4.4* (2.58*)
1.8 (0.11)
1.8 (0.16)
2.8 (0.67)
$10,000-$10,999
2.7 (0.46)
3.0 (0.58)
6.3 (2.58)
2.3 (0.15)
2.2 (0.17)
2.2 (0.54)
$11,000-$11,999
1.5 (0.34)
2.1 (0.50)
0.9 (0.65)
1.6 (0.10)
1.8 (0.18)
1.7 (0.51)
$12,000-$12,999
2.2 (0.61)
3.3 (0.87)
1.2* (1.22*)
2.2a (0.13)
2.7a (0.25)
1.2 (0.38)
$13,000-$13,999
1.8a (0.50)
2.4a (0.70)
0.4* (0.40*)
1.6 (0.12)
1.3 (0.13)
1.1 (0.35)
$14,000-$14,999
1.5 (0.42)
1.0 (0.37)
1.9* (1.37*)
1.6a (0.11)
1.8a (0.16)
0.9 (0.30)
a
a
*
*
$15,000-$15,999
1.2 (0.25)
1.4 (0.42)
0.0 (0.00 )
1.8 (0.11)
1.7 (0.14)
2.1 (0.50)
$16,000-$16,999
0.6 (0.23)
1.0 (0.42)
2.7* (1.96*)
1.3 (0.10)
1.3 (0.12)
1.6 (0.40)
$17,000-$17,999
1.9 (0.76)
0.5 (0.21)
2.7* (1.99*)
1.4 (0.09)
1.2 (0.12)
1.2 (0.40)
$18,000-$18,999
2.2a (0.54)
1.9a (0.54)
0.5* (0.46*)
1.8 (0.11)
1.7 (0.17)
1.9 (0.50)
$19,000-$19,999
2.0 (0.44)
1.7 (0.64)
2.5* (1.72*)
1.8 (0.12)
1.8 (0.17)
2.0 (0.51)
$20,000-$24,999
6.1 (0.80)
6.6 (1.06)
4.0* (2.42*)
6.9 (0.24)
6.9 (0.34)
8.6 (1.08)
$25,000-$29,999
5.9 (0.89)
4.3 (0.81)
4.8 (2.50)
6.8 (0.32)
6.4 (0.33)
6.3 (0.94)
$30,000-$34,999
4.3 (0.83)
4.6 (0.94)
4.5* (2.56*)
6.1 (0.27)
5.9 (0.27)
5.3 (0.94)
$35,000-$39,999
3.0 (0.56)
3.7 (1.01)
2.2* (1.50*)
5.1 (0.23)
5.2 (0.33)
7.1 (1.09)
$40,000-$44,999
3.4 (0.63)
4.4 (1.25)
2.6 (1.61)
4.5 (0.21)
4.5 (0.28)
5.4 (0.91)
$45,000-$49,999
2.9 (0.56)
3.4 (0.76)
4.7* (2.52*)
4.3 (0.19)
4.9 (0.30)
6.1 (1.06)
$50,000-$74,999
6.1 (0.77)
6.3 (0.96)
5.2 (2.64)
12.3 (0.35)
12.5 (0.46) 12.4 (1.49)
$75,000-$99,999
2.2 (0.50)
1.5 (0.46)
3.8* (2.30*)
5.8 (0.24)
5.7 (0.37)
5.8 (1.02)
$100,000 or More
1.1 (0.33)
3.1 (0.92)
2.2 (1.33)
8.0 (0.36)
7.7 (0.51)
9.0 (1.67)
$100,000-$149,999
-- (--)
-- (--)
2.2 (1.33)
-- (--)
-- (--)
5.2 (1.17)
$150,000 or More
-- (--)
-- (--)
0.0* (0.00*)
-- (--)
-- (--)
3.8 (1.28)
* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = NSDUH Questionnaire Field Test.
-- Not available.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being intervieweradministered to self- administered.
NOTE: If a respondent said "yes" to HASJOIN, he or she is defined as using a proxy. If a respondent said "no" or did not answer
HASJOIN, he or she is defined as not having used a proxy. Respondents who were legitimately skipped from answering
question QP01 were excluded from this analysis.
a
Difference between estimate and corresponding QFT estimate is statistically significant at the 0.05 level (i.e., 2011 comparison
proxy compared with 2012 QFT proxy).
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews.
2
Main survey data collected in quarter 3 and quarter 4, 2012, through December 2, 2012.
3
QFT data collected from September 1 through November 3, 2012.
4
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing
data.
5
Estimate is based on an edited version of the variable.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health.

R-68

R.4

Summary and Implications

R.4.1 Summary of the Investigation of Items with Data Quality Issues
This appendix describes the data collection results and the analysis that was conducted
for sets of demographic and household questions moved from CAPI to ACASI administration in
the QFT instrument. Overall, 22 of these items were determined to have data quality issues,
either higher item missingness rates than the comparison data, significantly different estimates
from the comparison data, or both. Analysis of item missingness rates and benchmarking to
current main study data and other survey data were the two primary techniques used to examine
data quality issues for these items. For two sets of items that allowed for a proxy respondent to
answer these questions in lieu of the primary respondent—health insurance and income—the
potential impact of proxy reports on the data quality for these items was also examined.
In addition, a literature review, email requests for input via survey research listservs, direct
communication to researchers working on other Federal surveys, and input from RTI survey
methodologists were employed in the search of explanations for these findings.
The higher missingness rates for some of these items, such as receipt of food stamps and
some income items, could be viewed as counterintuitive to literature results showing that more
private modes are associated with greater willingness to report data that respondents would be
considered sensitive or private. Moving these items to ACASI provided QFT respondents with
greater privacy for responding to these questions than current National Survey on Drug Use and
Health (NSDUH) respondents who are required to provide their answers to field interviewers
(FIs). In the QFT, it is possible that some respondents provided more accurate responses than
they would have in CAPI mode, but that other respondents simply chose not to answer without
the presence of an FI. For some QFT items where missingness rates were higher than in the
CAPI data from the current main study, it is possible that the lower proportion of complete
responses provided in ACASI were more accurate overall than CAPI responses for the same
items. ACASI also provides respondents more time to think about their responses without feeling
pressure from an FI in CAPI mode to respond and move to the next question. Because these
demographic and household items were all in ACASI mode for QFT respondents and all in CAPI
mode for main study respondents, respondent reactions to answering these questions in one mode
versus the other cannot be obtained. This factor places some limits on the methods that could be
used to more clearly understand how QFT respondents might have reacted differently in ACASI
to these questions than main study respondents answering the same questions in CAPI mode.
Despite the limitations of QFT protocol and sample size, the QFT results provide credible
evidence on how missingness rates and estimates for these demographic and household items
might look when the partially redesigned protocol is implemented in 2015. For this reason,
changes were made to some of the items moved to ACASI for the 2013 Dress Rehearsal (DR).
Analysis of the item missingness rates from the QFT revealed that outdated definitions or unclear
terms could have contributed to respondent confusion on some items, so some changes involved
updates to the questions to improve clarity. In addition, two items were dropped. Some of the
key revisions to these items that were implemented for the DR included the following:
•

edited references to the F2 help boxes,

•

eliminated other F2 help boxes,
R-69

•

deleted item QD42 about the number of people working for the respondent's
employer,

•

deleted items Q105N about earning pay while working at a job or business,

•

revised the definition of SSI, and

•

reordered the list of possible income sources.

Missingness rates and estimates for these demographic and household items will be part of the
priority analyses for the DR analysis for this set of items. The following section discusses how
the QFT and DR results could inform decisions on whether to move these demographic and
household items to ACASI administration for 2015 as planned, or whether some or all of these
items should remain in the CAPI portion of the interview for 2015.
R.4.2 Implications of Possible Protocol Options for the 2015 NSDUH
To determine whether any of the survey items moved from CAPI to ACASI
administration mode in the QFT protocol should remain in ACASI portion of the interview or be
moved back to the CAPI portion for the 2015 survey, a few methodological and logistical
considerations need to be taken into account. Applying these considerations will vary based on
the specific sets of items being considered for movement from the CAPI to the ACASI portion of
the interview. Although item missingness rates and benchmarking results are not the only
indicators of data quality, several recommendations can be considered based on the QFT findings
presented in Sections R.2 and R.3. If additional analyses were undertaken, such as those
suggested by RTI methodologists in Section R.3.1.3, these analyses could also inform the
recommendations, particularly with regard to the validity of reporting.
In the data gathered during the QFT, a few sets of items showed lower item missing data
rates than in the 2011 and 2012 comparison data. These include items on workplace drug and
alcohol policies, information access, and testing (QD43 and QD44 and QD46 to QD48) and
items on private health insurance coverage for drug abuse, alcoholism, and mental health issues
(QHI08 to QHI10). If lower item missing rates are viewed as indicating higher quality data, this
viewpoint would argue for keeping these items in the ACASI portion of the instrument for the
2015 survey.
Conversely, several QFT items had higher item missing rates than in the 2011 and 2012
comparison data, and some like private health insurance and employment produced estimates
that differed significantly from comparison data for at least one age group. These include the
following:
•

marital status (QD07),

•

number of home moves in the past year (QD13) and State of residence 1 year ago
QD13a),

•

student status and school days missed (QD19 to QD21),

•

recent employment history, workdays missed, size of employing organization, and
related issues (QD26, QD33, QD36, QD38, QD39a, QD40 to QD42),

•

private health insurance coverage (QHI06), and
R-70

•

sources of income and personal income level (QI03N to QI10N, QI20N, and QI21A).

For these sets of items, three options could be considered for determining whether to assign these
items to the CAPI or ACASI portion of the 2015 instrument.
Option 1: Adopt the 2014 Main Study Protocol
One option for assigning these sets of questions to CAPI or ACASI mode would be to
adhere to the 2014 main study protocol. This approach would result in eliminating the moves
from CAPI to ACASI mode included in the QFT protocol. This approach would arguably entail
the lowest risk, in that historical data on missingness rates for these sets of items would provide
accurate expectations for the 2015 survey year. The current main study CAPI missingness rates
are lower than the QFT ACASI rates for 22 items of interest. This approach would also allow the
CAI programmers to continue to use much of the current CAI programming, thereby minimizing
the scope of the programming and testing required for the 2015 instrument.
This approach would also have implications for the audio files required for these sets of
questions. If text-to-speech technology (TTS) were to be employed starting with the 2015
protocol, this approach would eliminate sets of questions for which audio files would need to be
created. Creating audio files for some questions within the income module has proved to be
difficult to program. Keeping these sets of questions in the CAPI portion of the interview would
avoid the need to create new audio files for these items.
One outcome of this approach could be somewhat higher overall administration times for
the interview, given that interviewer-administered questions generally take longer to administer
compared with the ACASI questions. When questions are administered in ACASI, the
interviewing environment is more private and the interview is more standardized, so the
respondent experience is more consistent from question to question and from section to section.
It is also more consistent across interviews. The potential for FIs to affect responses to items is
virtually eliminated in ACASI, for better or worse. If the 2015 main study items were to be asked
in the same modes as the 2014 main study, the time efficiencies observed in the QFT protocol
would not be realized. Furthermore, this approach would affect approximately 90 questions,
based on problematic missingness rates for only 22 items, or 24 percent of these items. Despite
these concerns, the decision to adopt this approach could be justified by the observed increases in
the missingness rates for specific QFT items or the simple numbers of QFT items with an
increase in missingness rates.
Option 2: Adopt the QFT Protocol
A second approach for assigning these sets of questions to CAPI or ACASI mode would
be to continue with the QFT instrument and protocol.52 The decision on whether to adopt this
approach could be driven by some observed lower missingness rates in the QFT or by declines in
missingness rates for several ACASI items in the DR. In preparation for the DR, a number of
these sets of items were edited in ways designed to improve item response rates. If these
revisions are associated with decreases in the missingness rates for a number of these items, the
52

With the exception of the item revisions listed at the end of Section R.4.1, the 2012 QFT protocol was
also followed for the 2013 DR.

R-71

DR results would provide support for this approach. Under this approach, the programming and
logic used for the DR instrument could be carried over to the 2015 main study instrument.
If TTS were adopted to produce the audio files, TTS files would need to be created for
these items. In addition, this approach would not address observed increases in missingness rates
for 22 items in the QFT if the rates remain high for most of all of these items in the DR. As a
result, the primary risk of this approach would be the need to wait for an analysis of the DR
missingness rates to be completed and reviewed in order to make a decision.
Option 3: Adopt a Tailored Protocol Based on QFT and DR Results
A third approach would be to assign these sets of questions to either CAPI or ACASI
mode, based on the data quality results for each individual item or sets of items. Under this
approach, important considerations would include respondent burden, question order and flow,
"gate" questions for skip patterns and logical fills, and the potential for context effects based on
item placement. This approach would apply findings from both the QFT and DR to development
of the 2015 instrument.
This option could potentially mitigate increases in interview administration time, while
increasing the probability of gathering substantive responses to key items. Items that were moved
from CAPI administration in the main survey protocol to ACASI administration in the QFT
protocol would be assessed under this option. Items first introduced in either the QFT or the
DR—disability, military families, sexual orientation53—would likely not be considered for
placement in the CAPI portion of the interview.
A review of the questions that were affected by the move from CAPI to ACASI in the
QFT instrument revealed that certain sets of items were affected more than others. The impact of
ACASI administration on missingness rates for respondent and family income was inconsistent
with, and in a different direction than, what would be expected from the literature cited in
Section R.3.1. The move from CAPI to ACASI in the QFT protocol did not affect the rates of
those reporting respondent income or those reporting household income of more than $20,000.
Only those reporting a household income of less than $20,000 had higher missingness rates.
Research shows income questions typically suffer from relatively higher rates of missing data
than most other survey items (Yan, Curtin, & Jans, 2010). In the QFT, higher item missingness
rates were observed in the more private ACASI mode. This finding does not imply that overall
data quality for income items was lower in the QFT than in the main study, but it does raise
concerns about a greater amount of missing income data that would need to be addressed in the
2015 survey data.
Given the item missingness results for some questions on received income, government
program participation, employment, health insurance, and income in ACASI mode in the QFT,
this approach could lead to the following instrument structure for these sets of items:
•

Questions about moves in the past year (residency) and marital status would be
moved to the front-end CAPI section of the instrument.
Questions on sexual attraction and identity are the only new items introduced in the DR questionnaire.

R-72

•

Questions about birth country, sexual orientation, disability, and military families
would be placed at the end of the ACASI section. To accommodate differential
missingness rates, questions in the employment module would be split between the
CAPI and ACASI portions of the interview. The first two employment questions—
QD26 about work at a job or business at any time in the past week and QD27 about
having a job or business last week but not working at any time—would be moved to
ACASI. These questions ask about whether a respondent is employed and need to
precede any questions about employment. Although missingness rates for question
QD26 increased in ACASI mode in the QFT, this gate question must remain in
ACASI mode for other employment questions to be included in the module.
Employment items QD43 through QD53 on written workplace policies about
employee use of alcohol or drugs and related issues would also be administered using
ACASI. Missingness rates for these 11 questions either decreased or remained the
same in the QFT, suggesting that this module should remain in ACASI.

•

The remaining employment items—QD28 through QD41 on workdays missed, size
of employing organization, and related issues—would be asked in a back-end CAPI
module. These questions each had higher missingness rates in the QFT and therefore
would be moved back to interviewer administration.

•

The education module (items QD17 to QD21 on student status, school days missed
due to sickness or injury, and school days missed due to "skipping" or "cutting")
would follow the education questions. This module would be intervieweradministered to address the increase in missingness rates for items QD19 through
QD21 observed in ACASI in the QFT. A showcard would be needed to display the
response options for QD18. Previously, the education module has preceded the
employment module. Given that the employment module would be separated across
two portions of the questionnaire under option 3, the education module would follow
the employment module.

•

Following the education questions, the interview would resume with the modes in
place for the 2014 main study. The household roster, proxy information, health
insurance, and income modules would be administered in CAPI in order to avoid the
higher item missingness rates observed in the QFT in ACASI mode.

If changes in the placement of any of these items are implemented for the 2015 data collection,
item missingness rates should continue to be closely monitored to assess the consequences of
these moves. Similar to the second option, the decision to implement this approach would need
to wait for analysis of the DR missingness rates to be completed and reviewed. Based on higher,
similar, or lower item missingness rates for items in the QFT and DR instrument, the mode
recommendations above could be revised as needed and implemented for the 2015 partial
redesign.
This approach will likely be associated with an increased effort to update the instrument
specifications, program the instrument, and test these sections of the instrument. However, this
effort would not result in a delay in the development of the 2015 instrument. The current 2015
instrument development schedule incorporates the level of effort that would be required to
implement these specifications.

R-73

A tailored approach will be adopted for the 2015 partially redesigned instrument. Based
on the QFT results showing high item missingness rates and estimates that differed significantly
from comparison data for a number of items in the health and income modules, these two
modules will both be administered via CAPI as in the current main study instrument. All other
modules with demographic and household items that were moved from CAPI to ACASI
administration will be administered via ACASI as in the QFT and DR.

R-74

2015 NSDUH, Supporting Statement
Attachment C – Dress Rehearsal (DR) Final
Report

NATIONAL SURVEY ON DRUG
USE AND HEALTH: 2013
DRESS REHEARSAL FINAL
REPORT

Substance Abuse and Mental Health Services Administration
Center for Behavioral Health Statistics and Quality
Rockville, Maryland 20857
March 26, 2014

NATIONAL SURVEY ON DRUG
USE AND HEALTH: 2013
DRESS REHEARSAL FINAL
REPORT
Deliverable 27: Field Test Protocol
Contract No. HHSS283201000003C
RTI Project No. 0212800
RTI Authors:
Doug Currivan
Patrick Chen
Rebecca Granger
David Heller
Larry Kroutil
Patricia LeBaron
Gretchen McHenry
Allison McKamey
Andrew Moore

Project Director:
Katie Morton
Susan Myers
Hyunjoo Park
Scott Payne
Bonnie Shook-Sa
Chris Stringer
Stephanie Terrey
Kevin Wang
Hilary Zelko

David Hunter
SAMHSA Project Officer:
Peter Tice

For questions about this report, please email [email protected].
Prepared for Substance Abuse and Mental Health Services Administration,
Rockville, Maryland
Prepared by RTI International, Research Triangle Park, North Carolina
March 26, 2014
Recommended Citation: Center for Behavioral Health Statistics and Quality.
(2014). National Survey on Drug Use and Health: 2013 Dress Rehearsal
Final Report. Substance Abuse and Mental Health Services Administration,
Rockville, MD.

Acknowledgments
This report would not be possible without the guidance and input of staff from the Center for
Behavioral Health Statistics and Quality. In particular, Jonaki Bose, Grace O'Neill, Kathy
Downey, Peggy Barker, Rachel Lipari, and Dicy Painter provided useful comments. At RTI
International (a trade name of Research Triangle Institute), William Dowd, Greta Kilmer, and
Jennifer Schoden assisted with obtaining estimates from external data sources, and Debbie Bond,
Valerie Garner, Roxanne Snaauw, and Richard Straw provided report production assistance.

Table of Contents
Chapter

Page

List of Tables ..................................................................................................................... ix
List of Exhibits .............................................................................................................. xxvii
List of Figures ................................................................................................................ xxix
1.

Background and Goals .........................................................................................................1

Study Design, Field Preparations, and Data Collection Procedures ....................................5
2.1
Overview of Study Design, Field Preparations, and Data Collection
Procedures ................................................................................................................5
2.2
Study Design ............................................................................................................5
2.2.1 Target Population .........................................................................................5
2.2.2 Spanish-Language Interview Oversample ...................................................5
2.2.3 Selection of State Sampling Regions and Segments....................................6
2.2.4 Selection of Dwelling Units .........................................................................8
2.2.5 Age Group Allocations ................................................................................9
2.2.6 Selection of Persons .....................................................................................9
2.3
Field Preparations ..................................................................................................11
2.3.1 Preparing Field Equipment ........................................................................11
2.3.2 Staffing .......................................................................................................13
2.3.3 Training Procedures ...................................................................................14
2.4
Data Collection Procedures....................................................................................19
2.4.1 Questionnaire and Protocol Changes for the 2013 Dress Rehearsal..........19
2.4.2 Contacting Dwelling Units ........................................................................22
2.4.3 Dwelling Unit Screening............................................................................23
2.4.4 Interview Administration ...........................................................................23
2.4.5 Controlled Access Procedures ...................................................................25
2.4.6 Refusal Conversion Procedures .................................................................25
2.4.7 Data Collection Management and Quality Control ...................................26
2.4.8 Problems Encountered ...............................................................................28

Processing and Analysis of Dress Rehearsal and Comparison Data .................................31
3.1
Overview of Data Processing and Analysis Approach ..........................................31
3.2
Defining Usable Cases ...........................................................................................31
3.2.1 Overview of Defining Usable Cases ..........................................................31
3.2.2 Usable Case Definitions .............................................................................31
3.3
Editing and Coding Procedures .............................................................................32
3.3.1 Overview of Editing and Coding Procedures ............................................32
3.3.2 Coding of "OTHER, Specify" Data ...........................................................33
3.3.3 General Editing Principles .........................................................................33
3.3.4 Special Editing Situations ..........................................................................36
3.4
Imputation Procedures ...........................................................................................46
3.4.1 Overview of Imputation Procedures ..........................................................46
3.4.2 Imputation Methodology ...........................................................................47

iii

Table of Contents (continued)
Chapter
3.5

3.6

3.7

Page
Weighting Procedures ............................................................................................50
3.5.1 Overview of Weighting Procedures ...........................................................50
3.5.2 Weighting Procedures ................................................................................51
3.5.3 Distribution of DR Analysis Weights ........................................................57
3.5.4 Creation of Variance Estimation Strata and Replicates .............................57
Data File Preparation .............................................................................................58
3.6.1 DR Data File ..............................................................................................58
3.6.2 2012 Comparison Data File .......................................................................59
3.6.3 2013 Quarters 3 and 4 Comparison Data File ............................................59
Data Analysis Issues ..............................................................................................60
3.7.1 Primary Analytic Goals..............................................................................60
3.7.2 Comparison with Current NSDUH Data ...................................................60
3.7.3 Comparisons with Other Survey Data .......................................................63

Data Collection Outcomes and Data Quality Assessment .................................................65
4.1
Overview of Data Collection and Data Quality Outcomes ....................................65
4.2
Unit Response Rates and Sample Characteristics (Research Question 3).............65
4.2.1 Screening Response Rates (SRRs) and Number of Visits for
Completed and Noncompleted Screenings ................................................65
4.2.2 Interview Response Rates (IRRs) and Number of Visits for
Completed and Noncompleted Screenings ................................................69
4.2.3 Geographic, Demographic, and Household Characteristics for the
Complete DR Sample ................................................................................70
4.2.4 Geographic, Demographic, and Household Characteristics for the
Combined QFT-DR Sample and Comparison Samples.............................82
4.3
Imputation Rates for Common 2012 Comparison Data, 2013 Quarters 3
and 4 Comparison Data, and Dress Rehearsal Variables .......................................93
4.4
Comparisons of Item Missingness Rates for Moved DR Items with 2012
and 2013 Quarters 3 and 4 Comparison Data and Comparisons of Item
Missingness Rates for New or Revised DR Items with the QFT ........................101
4.4.1 Item Missingness Rates for Items Moved from CAPI to ACASI
Administration in English-Language Non-Hispanic Interviews..............102
4.4.2 Item Missingness Rates for Items Moved from CAPI to ACASI
Administration in Spanish-Language Interviews.....................................105
4.4.3 Item Missingness Rates for Revised or New Items in EnglishLanguage Non-Hispanic Interviews ........................................................108
4.4.4 Item Missingness Rates for Revised or New Items in SpanishLanguage Interviews ................................................................................108
4.5
Comparisons of DR English-Language and Spanish-Language Interview
Timing Results with 2012 Comparison and 2013 Quarters 3 and 4
Comparison Interviews (Research Question 2) ...................................................109
4.5.1 Overall and Module Timing Results for the 2012 Main Study, 2013
Quarters 3 and 4 Main Study, 2012 Questionnaire Field Test, and
2013 Dress Rehearsal ...............................................................................109
iv

Table of Contents (continued)
Chapter

4.6

Page
4.5.2 Overall and Module Timing Results for Affirmative Gate
Respondents from English-Language Interviews in the 2012 and
2013 Quarters 3 and 4 Comparison Data and the 2012
Questionnaire Field Test and 2013 Dress Rehearsal Data .......................138
4.5.3 Detailed Interview Timing Data for Selected Modules from the
2012 and 2013 Quarters 3 and 4 Comparison Data and the 2012
Questionnaire Field Test and 2013 Dress Rehearsal Data .......................169
Other Data Quality Indicators ..............................................................................170
4.6.1 Overview of Other Data Quality Indicators .............................................170
4.6.2 Responding to Lead Questions for "OTHER, Specify" Data ..................195
4.6.3 Patterned Responses in the Core Drug Questions for the
Comparison Data .....................................................................................197
4.6.4 Patterned Responses in the Core Drug Questions for the DR Data .........198

Assessments of the Redesigned Protocol (Research Question 1) ....................................205
5.1
Overview of DR Protocol Assessment ................................................................205
5.2
Description and Results from DR FI Training Survey ........................................205
5.2.1 Purpose and Development of the DR FI Training Survey .......................205
5.2.2 Procedures for Conducting the DR FI Training Survey ..........................205
5.2.3 Summary and Discussion of Results from the DR FI Training
Survey ......................................................................................................205
5.2.4 FI Comments on the DR FI Training Program ........................................207
5.2.5 Considerations for 2015 NSDUH Training .............................................208
5.3
Description and Results for the FI Equipment Survey ........................................208
5.3.1 Overview ..................................................................................................208
5.3.2 Feedback from the FI Equipment Survey ................................................209
5.3.3 Summary of Results .................................................................................214
5.3.4 Next Steps for Future Hardware and Software Deployment ...................215
5.4
DR FI Debriefing Item Results ............................................................................215
5.4.1 Purpose of the Debriefing Items ..............................................................215
5.4.2 Results from the Debriefing Items ...........................................................216
5.4.3 Summary and Recommendations ............................................................224
5.5
Description and Results for the DR FI Debriefing Calls .....................................225
5.5.1 Purpose of the Debriefing Calls ...............................................................225
5.5.2 Debriefing Call Procedures ......................................................................226
5.5.3 Feedback from Debriefing Calls, by Topic..............................................227
5.5.4 Summary and Recommendations ............................................................232
5.6
Field Observations of DR FIs ..............................................................................233
5.6.1 Procedures for Completing Field Observations of DR FIs ......................234
5.6.2 Summary of Results from DR Field Observations ..................................235
5.6.3 DR Field Observation Comments ............................................................239
5.6.4 DR Field Observation Summary and Recommendations ........................240

Table of Contents (continued)
Chapter

Page

Selected Core and Noncore Estimates for English- and Spanish-Language Dress
Rehearsal Data and Comparison Data (Research Question 4) ........................................241
6.1
Overview of Selected Core and Noncore Estimates for English- and
Spanish-Language Dress Rehearsal Data and Comparison Data ........................241
6.2
Analyses to Make Decisions for the 2015 Survey ...............................................241
6.2.1 Analysis of Initiation Data .......................................................................241
6.2.2 Contributions of Specific Prescription Drugs to Estimates of Use
and Misuse ...............................................................................................253
6.2.3 Height and Weight ...................................................................................265
6.3
Further Analyses Based on QFT Findings or Analyses to Explain
Anticipated Findings in 2015 ...............................................................................279
6.3.1 Core Substance Use Items Other Than Methamphetamine and
Prescription Drugs ...................................................................................279
6.3.2 Methamphetamine and Prescription Drug Items .....................................292
6.3.3 Selected Noncore Items ...........................................................................302

Selected Noncore Estimates for DR, Comparison Data, and External Data Sources ......351
7.1
Overview of Selected DR Estimates Compared with Comparison Data and
Other Survey Data................................................................................................351
7.2
Comparisons of Estimates for Items Moved from CAPI to ACASI
Administration .....................................................................................................351
7.3
Comparisons of Estimates for Items New to the QFT and DR Instruments ........358
7.4
Comparisons of Estimates from Items New to the DR Instrument .....................362
7.5
Summary of Comparisons between DR, Comparison Data, and External
Data ......................................................................................................................365

Summary and Implications ..............................................................................................367
8.1
Data Collection Outcomes and Data Quality Assessment (Research
Questions 2 and 3) ...............................................................................................367
8.1.1 Item Missingness Rates and Variable Imputation Rates..........................367
8.1.2 Interview Timing Results.........................................................................368
8.1.3 Screening and Interview Response Rates ................................................369
8.1.4 Other Data Quality Indicators ..................................................................370
8.2
Assessments of the Redesigned Protocol (Research Question 1) ........................370
8.2.1 Field Interviewer Training Survey ...........................................................370
8.2.2 Field Interviewer Equipment Survey .......................................................371
8.2.3 Field Interviewer Debriefing Items ..........................................................371
8.2.4 Debriefing Calls with Field Interviewers .................................................372
8.2.5 Field Observations of Field Interviewers .................................................372

Table of Contents (continued)
Chapter

Page

8.3

8.4

8.5

Selected Core and Noncore Estimates for English- and Spanish-Language
Dress Rehearsal Data and Comparison Data (Research Question 4) ..................373
8.3.1 Core Substance Use Estimates Other Than Methamphetamine and
Prescription Drugs (Research Question 4a) ............................................373
8.3.2 Methamphetamine, Prescription Drug, and Illicit Drug Summary
Estimates (Research Question 4b) ...........................................................374
8.3.3 Selected Noncore Estimates (Research Question 4d) ..............................375
Selected Noncore Estimates for the Dress Rehearsal, Comparison Data
and External Data Sources (Research Question 5) ..............................................375
8.4.1 Estimates for Selected Items Moved from CAPI to ACASI
Administration .........................................................................................375
8.4.2 Estimates for Items New to the QFT Questionnaire and Included in
the DR ......................................................................................................376
8.4.3 Estimates for Items New to the DR Questionnaire ..................................377
Implications for the 2015 Partially Redesigned Instrument and Protocol ...........377

References ....................................................................................................................................395
Appendix
A

Screener and Questionnaire Changes Made for the Questionnaire Field Test
(QFT) and Instrument and Protocol Revisions Made for the Dress Rehearsal
(DR) ............................................................................................................................. A-1

Item Missingness Tables for English-Language and Spanish-Language
Interviews ......................................................................................................................B-1

Dress Rehearsal Field Interviewer Training Survey Results ........................................C-1

Field Interviewer Equipment Survey Questions and Results ...................................... D-1

Dress Rehearsal Field Interviewer Debriefing Items .................................................... E-1

Moderator's Guide for the Dress Rehearsal Field Interviewer Debriefing Calls .......... F-1

Dress Rehearsal Field Observation Materials .............................................................. G-1

Estimates and Standard Errors for All New, Moved, or Revised Items in the
2012 Questionnaire Field Test and 2013 Dress Rehearsal for English-Language
Non-Hispanic Interviews among Persons Aged 12 or Older ....................................... H-1

Notes on Analysis Variables for the Dress Rehearsal ................................................... I-1

vii

viii

List of Tables
Table

Page

2.1

Expected Number of Interviews and Precision of Dress Rehearsal Estimates .................6

2.2

Number of 2013 Dress Rehearsal State Sampling Regions and Sample Sizes,
by State .............................................................................................................................7

2.3

Summary of the Dress Rehearsal Sample Design and Results .........................................9

2.4

Simulated Pair Selection Counts for Different Values of λ ............................................11

2.5

Simulated Pair Unweighted Response Rates for Different Values of λ .........................11

2.6

Tenure Distribution of 2013 Quarters 3 and 4 Main Study Field Interviewers
Compared with 2013 Dress Rehearsal Field Interviewers..............................................14

2.7

Dress Rehearsal Field Interviewer Training Program ....................................................16

3.1

Imputed Variables ...........................................................................................................48

3.2

Weight Distribution of Dress Rehearsal Analysis Weights ............................................57

3.3

Data Files Created for the 2013 Dress Rehearsal Analyses............................................59

4.1

Screenings, Interviews, and Response Rates for the 2012 Main Study, 2013
Quarters 3 and 4 Main Study, 2012 Questionnaire Field Test, and 2013 Dress
Rehearsal Estimates ........................................................................................................67

4.2

Number of Visits Made for Completed Screenings for the 2012 Main Study,
2013 Quarters 3 and 4 Main Study, and 2013 Dress Rehearsal .....................................67

4.3

Number of Visits Made for Noncompleted Screenings for the 2012 Main Study,
2013 Quarters 3 and 4 Main Study, and 2013 Dress Rehearsal .....................................68

4.4

Interview Response Rates, by Age, for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, 2012 Questionnaire Field Test, and 2013 Dress Rehearsal ..............69

4.5

Number of Visits Made for Completed Interviews for the 2012 Main Study,
2013 Quarters 3 and 4 Main Study, and 2013 Dress Rehearsal .....................................71

4.6

Number of Visits Made for Noncompleted Interviews for the 2012 Main Study,
2013 Quarters 3 and 4 Main Study, and 2013 Dress Rehearsal .....................................71

4.7a

Demographic and Geographic Characteristics among Persons Aged 12 or Older:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal .........................................................................72

4.7b

Demographic and Geographic Characteristics among Persons Aged 12 to 17:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal .........................................................................74

4.7c

Demographic and Geographic Characteristics among Persons Aged 18 to 25:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal .........................................................................76

List of Tables (continued)
Table

Page

4.7d

Demographic and Geographic Characteristics among Persons Aged 26 or Older:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal .........................................................................78

4.8a

Demographic and Geographic Characteristics among Persons Aged 12 or Older
for English-Language Non-Hispanic Interviews: Percentages, Chi-Square Test
Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal .............................................83

4.8b

Demographic and Geographic Characteristics among Persons Aged 12 to 17 for
English-Language Non-Hispanic Interviews: Percentages, Chi-Square Test
Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal .............................................85

4.8c

Demographic and Geographic Characteristics among Persons Aged 18 to 25 for
English-Language Non-Hispanic Interviews: Percentages, Chi-Square Test
Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal .............................................87

4.8d

Demographic and Geographic Characteristics among Persons Aged 26 or Older
for English-Language Non-Hispanic Interviews: Percentages, Chi-Square Test
Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal .............................................89

4.9a

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Recency of Substance Use
Variables .........................................................................................................................94

4.9b

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Past Year Initiation of Substance
Use Variables ..................................................................................................................95

4.9c

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Age at First Use for Past Year
Initiates............................................................................................................................96

4.9d

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Selected Demographic and
Socioeconomic Variables ...............................................................................................97

4.9e

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Health Insurance Variables...................98

4.9f

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, and 2013 Dress Rehearsal: Income Variables ..................................99

4.10a

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 12 or Older) .........................................114
x

List of Tables (continued)
Table

Page

4.10b

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 12 to 17) ..............................................117

4.10c

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 18 to 25) ..............................................120

4.10d

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 26 to 49) ..............................................123

4.10e

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 50 to 64) ..............................................126

4.10f

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 65 or Older) .........................................129

4.10g

Overall and Module Mean/Median Timing Data for All Interviews in the 2012
Main Study, Q3-Q4 2013 Main Study, and 2013 Dress Rehearsal in Minutes
(All Respondents Aged 12 or Older) ............................................................................132

4.11a

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 12 or Older) .........................................139

4.11b

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 12 to 17) ..............................................141

4.11c

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 18 to 25) ..............................................143

4.11d

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 26 to 49) ..............................................145

4.11e

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 50 to 64) ..............................................147
xi

List of Tables (continued)
Table

Page

4.11f

Overall and Module Mean/Median Timing Data for Spanish-Language
Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and 2013 Dress
Rehearsal in Minutes (All Respondents Aged 65 or Older) .........................................149

4.12a

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 12 or Older) ...................151

4.12b

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 12 to 17) ........................154

4.12c

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 18 to 25) ........................157

4.12d

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 26 to 29) ........................160

4.12e

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 50 to 64) ........................163

4.12f

Overall and Module Mean/Median Timing Data for English-Language
Interviews from Non-Hispanic Respondents in the 2012 Main Study, Q3-Q4
2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal in Minutes (Affirmative Gate Respondents Aged 65 or Older) ...................166

4.13a

Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes,
in Total and by Age Groups: All Respondents .............................................................171

4.13b

Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes,
in Total and by Age Groups: English-Speaking Respondents ......................................172

4.13c

Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes,
in Total and by Age Groups: Spanish-Speaking Respondents .....................................173

4.13d

Overall Interview Timing Data for the 2012 Comparison Protocol in Minutes,
in Total and by Age Groups: 2012 Comparison English-Speaking Respondents ........174

4.13e

Overall Interview Timing Data for the 2013 Comparison Protocol in Minutes,
in Total and by Age Groups: 2013 Comparison English-Speaking Respondents ........175

xii

List of Tables (continued)
Table

Page

4.13f

Overall Interview Timing Data for the Dress Rehearsal Tobacco Module
in Minutes, in Total and by Age Groups: English-Speaking Dress Rehearsal
Respondents Answering LEADCIG .............................................................................176

4.13g

Overall Interview Timing Data for the 2012 Tobacco Module in Minutes,
in Total and by Age Groups: 2012 Comparison English-Speaking Respondents
Answering LEADCIG ..................................................................................................177

4.13h

Overall Interview Timing Data for the 2013 Tobacco Module in Minutes,
in Total and by Age Groups: 2013 Comparison English-Speaking Respondents
Answering LEADCIG ..................................................................................................178

4.13i

Overall Interview Timing Data for the Dress Rehearsal Pain Reliever Screener
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................179

4.13j

Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Screener
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................180

4.13k

Overall Interview Timing Data for the Dress Rehearsal Stimulants Screener
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................181

4.13l

Overall Interview Timing Data for the Dress Rehearsal Sedatives Screener
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................182

4.13m

Overall Interview Timing Data for the Dress Rehearsal Pain Reliever Module
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................183

4.13n

Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Module
in Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents ..................................................................................................................184

4.13o

Overall Interview Timing Data for the Dress Rehearsal Stimulants Module
in Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents ..................................................................................................................185

4.13p

Overall Interview Timing Data for the Dress Rehearsal Sedatives Module
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................186

4.13q

4.13r

Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Module
in Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents ..................................................................................................................188

4.13s

Overall Interview Timing Data for the Dress Rehearsal Stimulants Module
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................189

xiii

List of Tables (continued)
Table

Page

4.13t

Overall Interview Timing Data for the Dress Rehearsal Stimulants Module
in Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents ..................................................................................................................190

4.13u

4.13v

Overall Interview Timing Data for the Dress Rehearsal Sedatives Module
in Minutes, in Total and by Age Groups: English-Speaking Respondents ..................192

4.13w

Overall Interview Timing Data for the Dress Rehearsal Sedatives Module
in Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents ..................................................................................................................193

4.13x

Overall Interview Timing Data for the Dress Rehearsal Sedatives Module
in Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents ..................................................................................................................194

4.14

Estimates and Standard Errors for New, Moved, or Revised Items in the 2013
Dress Rehearsal among Persons Aged 12 or Older ......................................................199

5.1

Field Interviewer (FI) Feedback on the FI Training Program ......................................206

5.2

Field Interviewer (FI) Expectations on Referencing the FI Handbook before
Data Collection .............................................................................................................207

5.3

All Field Interviewers' Satisfaction with the Samsung Laptop (n = 125).....................210

5.4

Bilingual Field Interviewers' Satisfaction with the Samsung Laptop (n = 39) .............211

5.5

All Field Interviewers' Frequency of Tablet Email Use (n = 125) ...............................211

5.6

Bilingual Field Interviewers' Frequency of Tablet Email Use (n = 39) .......................211

5.7

All Field Interviewers' Satisfaction with the Tablet Email Program (n = 99) ..............212

5.8

Bilingual Field Interviewers' Satisfaction with the Tablet Email Program
(n = 33) .........................................................................................................................212

5.9

Comments about the Laptop .........................................................................................216

5.10

Positive Comments about the Laptop ...........................................................................217

5.11

Negative Comments about the Laptop..........................................................................217

5.12

Comments about the Laptop, by Interview Respondent Age .......................................217

5.13

Respondent Comments on the Laptop, by Interview Respondent Education...............218

5.14

Timing of Providing Q&A Brochure ............................................................................218

xiv

List of Tables (continued)
Table

Page

5.15

Respondent Comments on the Interview Being Too Long...........................................218

5.16

Respondent Comments on the Interview Being Too Long, by Interview
Respondent Age ............................................................................................................219

5.17

Respondent Comments on the Interview Being Too Long, by Interview
Respondent Education ..................................................................................................219

5.18

Any Interview Respondent Questions or Comments on On-Screen Calendars............219

5.19

Types of Interview Respondent Questions or Comments on On-Screen
Calendars ......................................................................................................................219

5.20

Interview Respondents' Troubles with Other Questions ..............................................220

5.21

Proxy Used for Income and Health Insurance Questions .............................................220

5.22

Proxy Used for Income and Health Insurance Questions, by Interview
Respondent Age ............................................................................................................220

5.23

Interview Respondent Concerns about Revealing Answers to Proxy Respondent ......221

5.24

Interview Respondent Questions or Comments about Proxy Interview .......................221

5.25

Problems with Proxy's Use of ACASI to Answer Income and Health Insurance
Questions ......................................................................................................................222

5.26

Types of Problems with Proxy's Use of ACASI to Answer Income and Health
Insurance Questions ......................................................................................................222

5.27

Interviews Conducted at Respondent's Home for the 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................222

5.28

Interview Location Not at Respondent's Home for the 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................223

5.29

Field Interviewer Evaluation of Interview Privacy in Respondent's Home for the
2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal................................223

5.30

Field Interviewer Reports of Others Present during Interview for the 2012
Comparison, 2013 Comparison, and 2013 Dress Rehearsal.........................................224

5.31

Debriefing Call Field Interviewer Characteristics, by Call ..........................................226

5.32

DR Screening Error Rates ............................................................................................236

5.33

Dress Rehearsal-Specific Errors: Screening .................................................................237

5.34

Dress Rehearsal Interviewing Error Rates ....................................................................238

5.35

Dress Rehearsal-Specific Errors: Interviewing.............................................................238

List of Tables (continued)
Table

Page

6.1A

Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older:
Numbers in Thousands, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal................................243

6.1B

Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older:
Percentages of All Persons Aged 12 or Older, Differences, and Standard Error
of Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal .......244

6.2

Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older:
Percentages of Persons at Risk for Initiation of Illicit Drug Use, Differences,
and Standard Error of Differences, 2012 Comparison, 2013 Comparison, and
2013 Dress Rehearsal....................................................................................................246

6.3

Past Year Initiation of Use of Selected Illicit Drugs among Persons Aged 12 or
Older: Percentages of Past Year Users, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal ...........247

6.4

Mean Age at First Use of Selected Illicit Drugs among Past Year Initiates Aged
12 to 49: Differences and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................249

6.5A

Past Year Misuse of Prescription Drugs and Initiation of Misuse among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Numbers in
Thousands, Differences, and Standard Error of Differences, 2012 QFT and 2013
Dress Rehearsal.............................................................................................................251

6.5B

Past Year Misuse of Prescription Drugs and Initiation of Misuse among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Percentages
of All Persons Aged 12 or Older, Differences, and Standard Error of
Differences, 2012 QFT and 2013 Dress Rehearsal.......................................................252

6.6

Alternate Measures of Past Year Use and Misuse of Specific Pain Relievers
among Persons Aged 12 or Older for English-Language Non-Hispanic
Interviews: Numbers in Thousands and Standard Errors, Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................255

6.7

Alternate Measures of Past Year Use and Misuse of Specific Tranquilizers for
English-Language Non-Hispanic Interviews among Persons Aged 12 or Older:
Numbers in Thousands and Standard Errors, Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal ............................................................................259

6.8

Alternate Measures of Past Year Use and Misuse of Specific Stimulants for
English-Language Non-Hispanic Interviews among Persons Aged 12 or Older:
Numbers in Thousands and Standard Errors, Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal ............................................................................261

xvi

List of Tables (continued)
Table

Page

6.9

Alternate Measures of Past Year Use and Misuse of Specific Sedatives for
English-Language Non-Hispanic Interviews among Persons Aged 12 or Older:
Numbers in Thousands and Standard Errors, Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal ............................................................................263

6.10

Summary Statistics for Height in Inches among Persons Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................267

6.11

Summary Statistics for Height in Inches among Males Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................268

6.12

Summary Statistics for Height in Inches among Females Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................269

6.13

Summary Statistics for Height in Inches among Persons Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................270

6.14

Summary Statistics for Height in Inches among Males Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................271

6.15

Summary Statistics for Height in Inches among Females Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................272

6.16

Summary Statistics for Weight in Pounds among Persons Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................273

6.17

Summary Statistics for Weight in Pounds among Males Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................274

6.18

Summary Statistics for Weight in Pounds among Females Aged 12 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................275

6.19

Summary Statistics for Weight in Pounds among Persons Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................276

6.20

Summary Statistics for Weight in Pounds among Males Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................277

6.21

Summary Statistics for Weight in Pounds among Females Aged 16 or Older:
2013 Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal ....................................278

6.22

Substance Use Other Than Methamphetamine or Prescription Drugs in the
Lifetime among Persons Aged 12 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................307

6.22sp Substance Use Other Than Methamphetamine or Prescription Drugs in the
Lifetime among Persons Aged 12 or Older for Spanish-Language Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................308
xvii

List of Tables (continued)
Table

Page

6.23

Substance Use Other Than Methamphetamine or Prescription Drugs in the
Lifetime among Persons Aged 12 to 17 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................309

6.24

Substance Use Other Than Methamphetamine or Prescription Drugs in the
Lifetime among Persons Aged 18 to 25 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................310

6.25

Substance Use Other Than Methamphetamine or Prescription Drugs in the
Lifetime among Persons Aged 26 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................311

6.26

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 12 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................312

6.26sp Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 12 or Older for Spanish-Language Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................313
6.27

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 12 to 17 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................314

6.28

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 18 to 25 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................315

6.29

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Year among Persons Aged 26 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................316
xviii

List of Tables (continued)
Table
6.30

Page
Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................317

6.30sp Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 or Older for Spanish-Language Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and 2013 Dress Rehearsal ..............................................................318
6.31

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 to 17 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................319

6.32

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 18 to 25 for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................320

6.33

Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 26 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field
Test and 2013 Dress Rehearsal .....................................................................................321

6.34

Specific Hallucinogen Use in the Lifetime, by Age Group for English-Language
Non-Hispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................322

6.35

Specific Inhalant Use in the Lifetime, by Age Group for English-Language
Non-Hispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................323

6.36

Alcohol Use in the Lifetime among Persons Aged 12 or Older, by Age Group
and Gender for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal .......................................................................................................................324

xix

List of Tables (continued)
Table

Page

6.37

Alcohol Use in the Past Year among Persons Aged 12 or Older, by Age Group
and Gender for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal .......................................................................................................................325

6.38

Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal .......................................................................................................................326

6.38sp Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender for Spanish-Language Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and 2013
Dress Rehearsal.............................................................................................................327
6.39

Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age
Group and Gender for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................328

6.39sp Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age
Group and Gender for Spanish-Language Interviews: Percentages, Differences,
and Standard Error of Differences, 2012 Comparison, 2013 Comparison, and
2013 Dress Rehearsal....................................................................................................329
6.40

Lifetime Use of Felt-Tip Pens, Computer Cleaners or Other Inhalants, by Age
Group and Past Year Use of Inhalants according to Types of Inhalants Used in
the Lifetime among Persons Aged 12 or Older for English-Language NonHispanic Interviews: Percentages, Combined 2012 Questionnaire Field Test and
2013 Dress Rehearsal....................................................................................................330

6.41

Use of Hallucinogens in Lifetime among Persons Aged 12 or Older with or
without Noncore Hallucinogen Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................331

6.42

Misuse of Prescription Drugs or Methamphetamine in the Lifetime among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................332

List of Tables (continued)
Table

Page

6.42sp Misuse of Prescription Drugs or Methamphetamine in the Lifetime among
Persons Aged 12 or Older for Spanish-Language Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal .......................................................................333
6.43

Misuse of Prescription Drugs or Methamphetamine in the Past Year among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................334

6.44

Misuse of Prescription Drugs or Methamphetamine in the Past Month among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................335

6.45

Misuse of Stimulants in the Lifetime among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................336

6.46

Misuse of Stimulants in the Past Year among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................337

6.47

Misuse of Stimulants in the Past Month among Persons Aged 12 or Older with
or without Noncore Adderall® Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................338

6.48

Misuse of Sedatives in the Lifetime among Persons Aged 12 or Older with or
without Noncore Ambien® Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................339

6.49

Misuse of Sedatives in the Past Year among Persons Aged 12 or Older with or
without Noncore Ambien® Data, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................340
xxi

List of Tables (continued)
Table

Page

6.50

6.51

Substance Dependence or Abuse in the Past Year among Persons Aged 12 or
Older for English-Language Non-Hispanic Interviews, by Survey Protocol:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................342

6.52

Substance Use with a Needle in the Lifetime, Past Year, and Past Month among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................343

6.53

Perceived Great Risk of Harm Associated with Substance Use among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal .......................................................................................................................344

6.54

Number of Years Since Last Use for Selected Substances among Lifetime Users
Aged 12 to 49 for English-Language Non-Hispanic Interviews: Averages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal .......................................................................................................................345

6.55

Received Substance Use Treatment in the Lifetime and Past Year and Types of
Past Year Substance Use Treatment among Persons Aged 12 or Older for
English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and
Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal..........................346

6.56

Adult Mental Health Treatment in the Past Year and Type of Facility Where
Received Treatment among Persons Aged 18 or Older for English-Language
Non-Hispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................347

6.57

Youth Mental Health Treatment in the Past Year and Number of Nights
Received Treatment among Persons Aged 12 to 17 for English-Language NonHispanic Interviews: Percentages, Chi-Square Test Statistic and P Value, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test
and 2013 Dress Rehearsal .............................................................................................348
xxii

List of Tables (continued)
Table

Page

6.58

Selected Mental Health Measures among Persons Aged 18 or Older for EnglishLanguage Non-Hispanic Interviews: Percentages, Differences, and Standard
Error of Differences, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal ....................................................349

6.59

Adolescent Depression Characteristics among Persons Aged 12 to 17 for
English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and
Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal..........................350

6.60

Arrested and Booked in the Lifetime and Past Year for Breaking the Law
among Persons 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal.............................................................................................................350

7.1

Received Income and Program Participation among Persons Aged 12 or Older:
Percentages and Standard Errors for 2012 Comparison, 2013 Comparison, 2013
Dress Rehearsal, and Other Surveys .............................................................................353

7.2

Levels of Current Employment among Persons Aged 18 or Older: Percentages
and Standard Errors for 2012 Comparison, 2013 Comparison, 2013 Dress
Rehearsal, and CPS Data ..............................................................................................354

7.3

Unemployment Rates among Persons Aged 18 or Older, by Age Group:
Percentages and Standard Errors for 2012 Comparison, 2013 Comparison, 2013
Dress Rehearsal, and CPS Data ....................................................................................355

7.4

Health Insurance Coverage among Persons Aged 12 or Older: Percentages and
Standard Errors for 2012 Comparison Data, 2013 Comparison Data, 2013 Dress
Rehearsal, and Other Surveys .......................................................................................356

7.5

Income among Persons Aged 12 or Older: Percentages and Standard Errors for
2012 Comparison Data, 2013 Comparison Data, 2013 Dress Rehearsal, and
2013 NHIS ....................................................................................................................357

7.6

2012 NHIS, 2009-2010 NHANES, and 2011-2012 NHANES Height Statistics
among Persons Aged 16 or Older for Comparison with the 2013 Dress
Rehearsal .......................................................................................................................359

7.7

2012 NHIS, 2009-2010 NHANES, and 2011-2012 NHANES Weight Statistics
among Persons Aged 16 or Older for Comparison with the 2013 Dress
Rehearsal .......................................................................................................................359

7.8

Conditions Told to Respondent by Doctor or Other Health Professional among
Persons Aged 12 or Older: Percentages and Standard Errors, 2013 Dress
Rehearsal and 2012 National Health Interview Survey ................................................361

xxiii

List of Tables (continued)
Table

Page

7.9

Disabilities or Physical Limitations among Persons Aged 12 or Older:
Percentages and Standard Errors, 2013 Dress Rehearsal, 2012 National Health
Interview Survey, and 2012 American Community Survey .........................................361

7.10

English-Speaking Proficiency among Persons Aged 12 or Older: Percentages
and Standard Errors, 2013 Dress Rehearsal and 2012 ACS .........................................362

7.11

Sexual Identity among Persons Aged 18 or Older, by Gender: Percentages and
Standard Errors for 2013 Dress Rehearsal and 2012 GSS Data ...................................364

7.12

Sexual Identity among Persons Aged 18 to 44, by Gender: Percentages and
Standard Errors for 2013 Dress Rehearsal and 2012 GSS Data ...................................364

8.1

Issues Identified for Screener and Questionnaire Items from the QFT or DR
Analysis and Recommendation for the 2015 Partially Redesigned Questionnaire ......379

8.2

Issues Identified for Training and Materials from the QFT or DR Analysis and
Recommendation for the 2015 Partially Redesigned Protocol .....................................389

8.3

Issues Identified for Field Equipment from the QFT or DR Analysis and
Recommendation for the 2015 Partially Redesigned Protocol .....................................391

8.4

Estimates and Items Identified from the QFT or DR Analysis for Preview in the
2015 Early Data Review ...............................................................................................392

8.5

Estimates and Items Identified from the QFT or DR Analysis for Priority
Examination in the 2015 6-Month Tables ....................................................................394

A.1

Changes between the 2012 NSDUH Screener and the 2012 Questionnaire Field
Test (QFT) Screener .................................................................................................... A-1

A.2

Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire
Field Test (QFT) Questionnaire................................................................................... A-2

A.3

Changes between the 2012 Questionnaire Field Test (QFT) Screener and the
2013 Dress Rehearsal (DR) Screener .......................................................................... A-9

A.4

Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and
the 2013 Dress Rehearsal (DR) Questionnaire .......................................................... A-10

B.1

Item Missingness Rates for Moved Items for English-Language Non-Hispanic
Interviews in 2012 Comparison Data, 2013 Comparison Data, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal Data among All
Persons Aged 12 or Older .............................................................................................B-1

B.2

Item Missingness Rates for Moved Items for Spanish-Language Interviews in
2012 Comparison Data, 2013 Comparison Data, and 2013 Dress Rehearsal Data
among All Persons Aged 12 or Older .........................................................................B-10

xxiv

List of Tables (continued)
Table

Page

B.3

Item Missingness Rates for New and Revised Items for English-Language NonHispanic Interviews in the 2012 Questionnaire Field Test and 2013 Dress
Rehearsal among All Persons Aged 12 or Older ........................................................B-19

B.4

Item Missingness Rates for New and Revised Items for Spanish-Language
Interviews in the 2013 Dress Rehearsal among All Persons Aged 12 or Older .........B-20

D.1

FI Satisfaction with Laptop – All Field Interviewers .................................................. D-1

D.2

FI Satisfaction with Laptop – Bilingual Field Interviewers Only ............................... D-2

D.3

All FI Comments about Laptop, Training, and Carrying Case .................................... D-3

D.4

Frequency of FI Email Use – All Field Interviewers................................................... D-6

D.5

Frequency of FI Email Use – Bilingual Field Interviewers ......................................... D-6

D.6

FI Satisfaction with Tablet Email – All Field Interviewers ......................................... D-6

D.7

FI Satisfaction with Tablet Email – Bilingual Interviewers ........................................ D-7

D.8

FI Comments about Email, Training, or Transmitting on the Tablet ........................ D-10

H.1

Estimates and Standard Errors for New, Moved, or Revised Items in the 2012
Questionnaire Field Test and 2013 Dress Rehearsal among Persons Aged 12 or
Older ............................................................................................................................ H-1

xxv

xxvi

List of Exhibits
Exhibit

Page

2.1

Dress Rehearsal Field Interviewer Training Agenda ......................................................18

3.1

Collapsing Imputation Classes: Race..............................................................................49

3.2

Collapsing Imputation Classes: Race and Gender ..........................................................50

6.1

xxvii

xxviii

List of Figures
Figure

Page

5.1

All Field Interviewers' Tablet Keypad Preference ........................................................213

5.2

Bilingual Field Interviewers' Tablet Keypad Preference ..............................................213

5.3

All Field Interviewers' Satisfaction with the Tablet's Wireless Transmission..............214

5.4

Bilingual Field Interviewers' Satisfaction with the Tablet's Wireless
Transmission .................................................................................................................214

D.1

FI Tablet Keypad Preference – All Field Interviewers ................................................ D-8

D.2

FI Tablet Keypad Preference – Bilingual Field Interviewers ...................................... D-8

D.3

Tablet Wireless Transmission Satisfaction – All Field Interviewers........................... D-9

D.4

Tablet Wireless Transmission Satisfaction – Bilingual Field Interviewers ................. D-9

xxix

xxx

1. Background and Goals
This report summarizes the data collection and analytic methods and results for the 2013
Dress Rehearsal (DR) for the National Survey on Drug Use and Health (NSDUH). Sponsored by
the Substance Abuse and Mental Health Services Administration (SAMHSA), NSDUH is a
national survey of the U.S. civilian, noninstitutionalized population aged 12 or older.
In order to continue producing current and accurate data, SAMHSA's Center for
Behavioral Health Statistics and Quality (CBHSQ) must update NSDUH periodically to reflect
changing patterns in substance use and new mental health priorities. CBHSQ is planning to
implement changes related to a partial NSDUH redesign. These changes include use of a new
sample design in 2014 and a limited update to the interview questionnaire in 2015. The new
sample design will allow for continued national, State, and substate-level estimation comparable
with estimation from previous surveys. The sample design's improved efficiency will result in
significant cost savings. CBHSQ plans to redesign NSDUH for the 2015 survey year to achieve
two main goals: (1) revise the questionnaire to address changing policy and research data needs,
and (2) modify the survey methodology to improve the quality of estimates and the efficiency of
data collection and processing.
A Questionnaire Field Test (QFT) conducted in 2012 tested revisions to the NSDUH
respondent materials, questionnaire, procedures, and equipment associated with the 2015 partial
redesign goals (Currivan et al., 2013). Section 2.4.1 provides a complete list of the revisions that
were implemented and evaluated for the QFT. Following the QFT, the DR aims to further test
revisions made to the QFT materials, questionnaire, procedures, and equipment, as well as
further revisions made to the questionnaire and equipment specifically for the DR. Two major
differences between the QFT and the DR are the addition of Spanish-language interviews and a
test of new lightweight laptop computers. Field interviewers (FIs) used the same tablet computer
for the DR that was originally tested during the QFT for screening, respondent selection, and
case management, with a few administrative enhancements for the DR. Section 2.4.1 and
Appendix A describe and provide a complete list of the additional questionnaire and protocol
changes that were implemented for the DR. The DR provides another opportunity to further
refine and improve the redesigned questionnaire, materials, and procedures prior to any full-scale
changes for the 2015 partial redesign. Specifically, the DR presents an opportunity to do the
following:
•

assess how the partially redesigned protocol performs for Spanish-language
screenings and interviews;

•

evaluate whether problems identified in the QFT, such as data quality issues with
items moved from computer-assisted personal interviewing (CAPI) to audio
computer-assisted self-interviewing (ACASI) administration, persisted in the DR;

•

use the combined QFT and DR samples to increase the statistical power for analyses
that were inconclusive in the QFT because of limited sample size; and

•

examine new items that were introduced in the QFT and then modified for the DR.

Using multiple indicators and data sources, the primary goal of the DR is to measure the
total effect on NSDUH estimates and outcomes from all changes to the materials, questionnaire,
and procedures planned for the 2015 partial redesign. Specifically, the DR provides data to
attempt to address the following research questions, to the extent that sample sizes allow:
1. What do assessments of the DR protocol—obtained from equipment surveys,

debriefing questions, debriefing calls, and field observations of FIs—indicate about
the likely effectiveness of the 2015 partial redesign protocol?
2. What impact does the redesigned protocol, including revisions made to the DR
questions or protocol based on QFT experiences or results, have on the overall
interview timing and module timings across age groups?
3. Does the DR protocol, including changes made from the QFT protocol, meet similar
data quality standards as the QFT data collection and the current NSDUH main study,
as measured by unit nonresponse, item missingness rates, imputation rates, and other
indicators of data quality?
4. Does the DR protocol produce any significant differences in key estimates with the
QFT and the main study comparison data, both for all respondents and across age
groups and for both English-language and Spanish-language interviews?
4a. To what extent do DR estimates for core substance use items other than
methamphetamine and prescription drugs differ from the QFT (English-language
interviews only) and the main study comparison estimates (not restricted to
English-language interviews)? To what extent do these core substance use
estimates based on the combined QFT and DR English-language non-Hispanic
data differ from the corresponding main study English-language non-Hispanic
comparison estimates?
4b. To what extent do DR estimates for methamphetamine and prescription drug
items differ from the QFT (English-language interviews only) and the current
NSDUH main study (not restricted to English-language interviews)? To what
extent do these estimates based on the combined QFT and DR English-language
non-Hispanic data differ from the corresponding main study English-language
non-Hispanic comparison estimates?
4c. To what extent do DR and QFT data for individual prescription drugs contribute
to estimates of past year use or misuse for the overall category (e.g., pain
relievers) and for related prescription drugs within a category? What effect does
including or dropping data for specific drugs have on the combined QFT and DR
English-language non-Hispanic estimates?
4d. To what extent do DR estimates for selected noncore items—such as substance
dependence or abuse, substance use treatment, selected mental health measures,
mental health treatment, and demographic and household items—differ from the
QFT (English-language non-Hispanic estimates) and the current NSDUH main
study?
5. Does the DR protocol produce any significant differences in key estimates relative to
estimates from other surveys or other sources of data?

This report summarizes how the DR was conducted and the results obtained to address
the five main research questions. Chapter 2 describes the study design, field preparations, and
data collection procedures. Chapter 3 describes procedures for defining usable cases, data
editing and coding, imputation, weighting, data file preparation, and data analysis issues for the
DR data and the two NSDUH datasets that were used to compare with the DR data. This chapter
also discusses key analytic issues, especially comparisons of the DR data with the 2012 and 2013
quarters 3 and 4 NSDUH main study data. Chapter 4 addresses research questions 2 and 3 by
detailing data collection outcomes, such as screenings and interviews completed, screening and
interview response rates, overall and module interview timings, imputation rates, item
missingness rates, and other data quality indicators. Chapter 5 describes data collected from DR
interviewers through multiple methods—including an FI training survey, FI equipment survey,
FI debriefing items, debriefing calls with FIs, and field observations of FIs—to address research
question 1 about the general performance of the redesigned protocol. Chapter 6 presents
comparisons of selected core and noncore estimates for English- and Spanish-language DR data
and comparison data to address research question 4. Chapter 7 addresses research question 5
through an examination of QFT and DR estimates for moved, revised, and new items in the QFT
and DR protocols and, where applicable, comparisons with parallel estimates from the two
NSDUH main study datasets and other national survey datasets. Chapters 6 and 7 are both
organized in two sections, with the first covering priority analyses that will directly inform
decisions for the 2015 partial redesign and the second covering additional analyses that will
provide a preview of how specific estimates will look in the 2015 main study data. Finally,
Chapter 8 summarizes the key findings in the report with respect to each of the five main
research questions and the main implications of these results for finalizing the partially
redesigned questionnaire and protocol to be implemented in the 2015 NSDUH main study.

2. Study Design, Field Preparations, and
Data Collection Procedures
2.1

Overview of Study Design, Field Preparations, and Data Collection
Procedures

This chapter provides details of the design and implementation of the 2013 Dress
Rehearsal (DR). Section 2.2 describes the study design, including the sample design and
selection procedures. Section 2.3 addresses preparations made for data collection, including
preparing the field equipment, selecting the field interviewers (FIs), and training the FIs and
field supervisors (FSs). Section 2.4 describes all of the data collection procedures followed in
implementing the DR, which was fielded from September 1 through October 31, 2013.

2.2

Study Design

This section describes the target population represented by the DR, the oversampling
of Spanish-language interviews, procedures for selecting State sampling regions (SSRs) and
segments, selection of dwelling units (DUs), allocation of respondents across age groups, and
selection of persons to be respondents for the interviews.
2.2.1

Target Population

Similar to the main study of the National Survey on Drug Use and Health (NSDUH),
the respondent universe for the DR was the civilian, noninstitutionalized population aged 12 or
older. In order to control costs, persons residing in Alaska and Hawaii were excluded from the
DR. Therefore, the sample is representative of the noninstitutionalized population aged 12 or
older in the contiguous United States.
2.2.2

Spanish-Language Interview Oversample

One primary goal of the DR was to evaluate the Spanish-language questionnaire, so it
was critical to complete enough DR interviews in Spanish to allow for this evaluation.
To achieve a higher yield of Spanish-language interviews than what would be observed with a
probability proportional to size (PPS) sample, a special certainty stratum was created that
comprised the SSRs with a historically high percentage of interviews conducted in Spanish.
SSRs that had 10 percent or more of their 2011 NSDUH interviews conducted in Spanish were
assigned to the certainty stratum. The percentage of interviews conducted in Spanish was
calculated at the SSR level rather than the segment level because sample sizes at the segment
level were too small to provide reliable estimates. A total of 101 of the NSDUH SSRs fell into
the certainty stratum and were selected for the DR with certainty.
Because of the oversampling of areas with historically high concentrations of Spanishlanguage interviews, 207 of the 2,000 total interviews were expected to be completed in Spanish.
Table 2.1 presents the expected number of interviews and estimated precision of survey
estimates for the total interviews and for the Spanish-language interviews. Although this
oversampling approach led to a higher yield of Spanish-language interviews compared with a
5

design where all of the segments were selected PPS, it decreased the precision of the overall
estimates by increasing the design effects. Areas with high concentrations of Spanish-language
interviews had a much higher probability of selection under this design than they would have had
under a PPS design. This design balanced the goals of testing the Spanish-language questionnaire
and producing efficient overall estimates.
Table 2.1 Expected Number of Interviews and Precision of Dress Rehearsal Estimates

Number and Precision
Expected Number of Interviews
Standard Errors (SEs) of Estimates1
Relative Standard Errors (RSEs) of Estimates1
1

Total Interviews
2,000
1.30%
12.98%

SpanishLanguage
Interviews
207
4.08%
40.83%

SE and RSE calculations assume a design effect of 2.5 and a prevalence of p = 0.10.

2.2.3

Selection of State Sampling Regions and Segments

NSDUH is designed to yield 67,500 interviews from 7,200 segments each calendar year
(Morton & Shook-Sa, 2012). Thus, an estimated 200 segments were needed to yield
approximately 2,000 completed DR interviews. As discussed in Section 2.2.2, a special certainty
stratum was developed to ensure that a sufficient number of DR interviews would be completed
in Spanish. As mentioned in Section 2.2.2, 101 of the NSDUH SSRs fell into the certainty
stratum and were selected with certainty. To ensure national representation, the remaining 775
SSRs were stratified by census region, and 99 SSRs were selected PPS for inclusion in the DR.
Implicit stratification was achieved by sorting the frame of SSRs by the percentage urban and the
percentage of interviews completed in Spanish in 2011 prior to selecting the sample.
This design had the benefit of placing much of the sample in heavily populated areas
where a sufficient mix of FIs with various experience levels were available to meet the DR
staffing needs. As shown in Table 2.2, a large portion of the sample was selected from the eight
largest States (i.e., California, Florida, Illinois, Michigan, New York, Ohio, Pennsylvania, and
Texas). In addition, the majority of the Spanish-language interviews were expected to be
completed in States where bilingual FIs were already employed.
Within each selected SSR, a sample of DUs was drawn from the segment that was retired
from use in quarter 1 of the 2013 NSDUH. DUs that were not selected for the main study in 2012
and 2013 were eligible for selection in the field test. If an insufficient number of DUs remained
in a segment, or if significant access problems were expected, the segment was replaced with the
quarter 3 or quarter 4 2012 retired segment in the same SSR. That is, the quarter 1 2013 segment
with insufficient DUs or expected access problems was removed from the sample, and the
quarter 3 or quarter 4 2012 retired segment was included in the sample of segments prior to the
allocation of DUs across segments. One segment was replaced because it had fewer than 10 DUs
remaining, and 10 segments were replaced because of anticipated access problems in the
segments.

Table 2.2 Number of 2013 Dress Rehearsal State Sampling Regions and Sample Sizes, by State

State
CA
TX
NY
FL
IL
PA
OH
MI
GA
NC
NJ
VA
MA
WA
IN
AZ
TN
MO
WI
MD
MN
CO
AL
SC
KY
LA
OR
OK
CT
IA
MS
AR
KS
NV
UT
NM

Population
Rank
(12 or
Older)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Current
Design
3,600
3,600
3,600
3,600
3,600
3,600
3,600
3,600
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900

NSDUH
SSR Regions
48
48
48
48
48
48
48
48
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12

Number of
DR SSR
Regions/
Segments
38
23
15
13
12
4
5
3
4
4
3
4
4
1
2
7
3
2
5
2
2
3
3
1
1
2
1
1
1
1
0
1
2
5
2
4

2013 DR
Total
Interviews
640
265
142
121
96
41
49
29
23
44
19
28
42
5
19
58
22
22
51
22
20
10
32
9
8
17
12
8
2
13
0
11
19
31
11
16

2013 DR
SpanishLanguage
Interviews
60
59
11
22
8
1
2
0
0
0
0
0
0
0
0
7
0
0
1
3
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
1
(continued)

Table 2.2 Number of 2013 Dress Rehearsal State Sampling Regions and Sample Sizes, by State
(continued)

State
WV
NE
ID
ME
NH
HI
RI
MT
DE
SD
AK
VT
ND
DC
WY

Population
Rank
(12 or
Older)
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Total

Current
Design
900
900
900
900
900
900
900
900
900
900
900
900
900
900
900
67,500

NSDUH
SSR Regions
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
900

Number of
DR SSR
Regions/
Segments
3
1
1
3
0
0
3
0
1
1
0
1
0
1
1
200

2013 DR
Total
Respondents
35
7
1
34
0
0
7
0
4
15
0
14
0
4
9
2,087

2013 DR
SpanishLanguage
Respondents
0
2
0
0
0
0
0
0
2
0
0
0
0
2
0
185

DR = Dress Rehearsal; SSR = State sampling region.

2.2.4

Selection of Dwelling Units

The starting sample size and the sample allocation across the segments were determined
based on anticipated eligibility, nonresponse, and the person-level sample selection procedures.
Similar to the main study, a small reserve sample (15 percent) of DUs from each segment was
selected, and the total sample was partitioned into four probability subsamples within each
segment: 100 percent and three 5 percent partitions, for a total of 115 percent. Although the
majority of the sample (100/115) was released at the beginning of the DR data collection period,
having the additional sample partitions allowed for greater flexibility in controlling the sample
size and provided the ability to ensure that the data collection goals were attained within the field
period. No additional sample partitions were needed to achieve the target of 2,000 completed
interviews.
A total of 5,016 DUs were sampled and yielded 2,087 completed interviews (Table 2.3).
As shown in Table 2.2, 185 Spanish-language interviews were yielded from the DR sample.
The half-open interval procedure for missed DUs was implemented during the DR, but it is not
scheduled to be implemented in the 2014 or 2015 NSDUHs. Table 2.3 compares the expected
DR unweighted response rates and yields to the actual unweighted response rates and yields.

Table 2.3 Summary of the Dress Rehearsal Sample Design and Results

Statistic
State Sampling Regions
Segments
Selected Dwelling Units
Eligible Dwelling Units
Completed Screening Interviews
Selected Persons
Completed Interviews

Expected
Unweighted
Response
Total
Rate1
200
N/A
200
N/A
5,146
N/A
4,426
0.86
3,673
0.83
2,703
N/A
2,000
0.74

Actual
Unweighted
Response
Total
Rate
200
N/A
200
N/A
5,016
N/A
4,392
0.88
3,511
0.80
2,808
N/A
2,087
0.74

N/A = not applicable; NSDUH = National Survey on Drug Use and Health; QFT = Questionnaire Field Test.
1
Expected eligibility and screening rates are the observed rates from the 2012 QFT (unweighted). The expected
interview response rate is the observed rate from the QFT adjusted with 2011 NSDUH rates to account for the
oversampling of high Spanish-language interview areas.

2.2.5

Age Group Allocations

The respondent sample was allocated to the three major age groups in the following
proportions: 25 percent aged 12 to 17, 25 percent aged 18 to 25, and 50 percent aged 26 or older.
Among the 26 or older age groups, 15 percent of the sample was allocated to persons aged 26 to
34, 20 percent of the sample was allocated to persons aged 35 to 49, and 15 percent was
allocated to persons aged 50 or older. This sample allocation matched the planned allocation for
the 2014 NSDUH and the 2012 Questionnaire Field Test (QFT). One implication of the
respondent sample allocation by age groups is a potential impact on DR response rates. As with
the QFT, having a higher sampling rate for the 26 or older adults identified in DR households
compared with the NSDUH main interview had a negative effect on unweighted interview
response rates because response rates are typically lower for the 26 and older age group.
As shown in Table 4.4 in Chapter 4, both the weighted and unweighted interview response rates
for persons younger than 26 were higher than the response rates for persons aged 26 or older.
Therefore, sampling more persons 26 or older led to a lower overall unweighted interview
response rate for the DR compared with the main study. The unweighted interview response rate
for the DR sample was 74.32 percent compared with 78.01 percent for the 2012 main study
comparison sample and 80.71 percent for the 2013 quarters 3 and 4 main study comparison
sample (see Table 4.1 in Chapter 4). Weighted interview response rates are not affected by the
change in age allocation. Although a smaller proportion of 12 to 17 year olds were selected, this
age group continued to drive the number of DUs needed (i.e., relative to the total population in
this age group, the age group continued to be sampled at the highest rate). Thus, fewer DUs were
needed to yield the desired sample than would be needed under the current sample design.
2.2.6

Selection of Persons

After DUs were selected within each DR segment, an FI visited each selected DU to
obtain a roster of all persons aged 12 or older residing in the DU. This roster information was
used to select 0, 1, or 2 persons for the survey. Sampling rates were preset by segment and age
group. Roster information was entered directly into the electronic screening program, which
9

automatically implemented this stage of selection based on the segment and age group sampling
parameters. As indicated in Table 2.3, 2,808 people were selected from within 3,511 screened
and eligible DUs, which yielded 2,087 completed interviews.
The sampling algorithm in NSDUH is based on the Chromy and Penne (2002) adaptation
of a Brewer (1963, 1975) method for selecting samples of size two. The adaptation allows for
selecting samples of 0, 1, or 2 persons within a selected DU containing at least one eligible
person. Chromy and Penne (2002) also introduced a pair sampling parameter λ, which governs
the number of pairs selected. The following text describes how the sample selection algorithm
and pair sampling parameter are implemented for NSDUH.
Define the target selection probability for person i in DU h as
. Then, to ensure that
all of the pairs have a positive probability of selection, all of the person probabilities have to be
strictly less than 1; and arbitrarily, the maximum
is set to 0.99. In Brewer's (unadapted)
method of sampling pairs, the sum of the first-order inclusion probabilities is always equal to
n = 2. However, because the design calls for a selection of 0, 1, or 2 persons per DU, it is
unlikely that the sum of person probabilities within a DU sums to 2 (i.e., Sh   Phi  2 ).
i

The following adaptations were then applied to the sampling algorithm.
If Sh > 2, a multiplicative scaling factor, Fh = 2/Sh, was applied to all of the target
selection probabilities so that they were scaled down to sum to exactly 2.
If Sh < 2, the problem was remedied by creating three dummy persons and distributing the
remaining size measure (2 – Sh) to them equally (i.e., the inclusion of dummy persons in the
selection could result in the selection of 0 or 1 actual persons). Operationally, this initially
required the application of the following multiplicative scaling factor to the person probabilities:
min

0.99
max

However, a further modification was applied to this scaling factor that allowed some
flexibility in the actual number of pairs selected. This modification was governed by the pair
sampling parameter λ. Define

Then the modified multiplicative scaling factor was expressed as

Simulation analyses resulted in the selection of λ = 0.50 for the 2002 to 2013 NSDUH
sample designs. However, changes to the 2014 sample design with respect to age group and State
necessitated further simulation analyses to identify the value of λ best suited for the 2014 design.
Simulation analyses based on the 2012 screening data, modified to reflect the required 2014 age
group sample proportions (but not modified to reflect the new State proportions), were
conducted, and λ = 0.25 was selected.1 Table 2.4 displays expected pair selection counts for the
1

This 0.25 value was finalized for the 2014 NSDUH on November 25, 2013.

2014 NSDUH (scaled to sum to 67,500) for different values of λ in the simulation exercise, and
Table 2.5 displays the corresponding unweighted response rates. However, these simulation
analyses had not been conducted in time to be implemented for the QFT and DR studies;
therefore, λ = 0.50 was used for these studies. The selection of λ = 0.50 for the DR also
maintained consistency with the QFT, 2012, and 2013 DR comparison samples.
Table 2.4 Simulated Pair Selection Counts for Different Values of λ
Age Group for
Pairs
12+, 12+
12 - 17, 12 - 17
12 - 17, 18 - 25
12 - 17, 26+
18 - 25, 18 - 25
18 - 25, 26+
26+, 26+

0.00
18,054
2,951
2,170
5,211
2,728
2,962
2,032

λ=
0.50
28,630
3,169
2,517
7,317
3,606
4,908
7,113

0.25
22,752
3,041
2,326
6,208
3,185
3,833
4,160

0.75
34,047
3,340
2,671
7,726
4,142
5,629
10,538

1.00
37,809
3,489
2,775
7,956
4,576
5,867
13,146

Table 2.5 Simulated Pair Unweighted Response Rates for Different Values of λ
Age Group for
Pairs
12+, 12+
12 - 17, 12 - 17
12 - 17, 18 - 25
12 - 17, 26+
18 - 25, 18 - 25
18 - 25, 26+
26+, 26+

2.3

0.00
72.7
81.4
76.1
74.8
71.2
67.1
61.7

λ=
0.50
70.3
81.4
76.1
74.8
71.2
67.1
60.4

0.25
71.4
81.4
76.1
74.8
71.2
67.1
60.7

0.75
69.3
81.4
76.1
74.9
71.2
67.1
60.1

1.00
68.7
81.4
76.1
74.8
71.2
67.1
59.8

Field Preparations

This section describes the procedures undertaken to plan and implement the DR data
collection.
2.3.1

Preparing Field Equipment

2.3.1.1 Hardware Selection
As part of the process to resupply field staff with new data collection equipment for the
2015 NSDUH, the NSDUH team has been engaged in an ongoing equipment evaluation process.
In early 2012, after considering both a one-device and a two-device approach for equipment
resupply, the Substance Abuse and Mental Health Services Administration (SAMHSA) and RTI
decided to proceed with a two-device approach that involves the use a small mobile Android
tablet for doorway screening and a lightweight conventional Windows laptop for interviewing.
As part of the QFT, a small mobile tablet computer (Samsung Galaxy Tab 7.0") was tested as a
screening device. The tablet proved to be durable and reliable and was very well received by the
11

FIs because of the bright, large 7-inch display and the fact that the FIs felt that the touch-screen
interface was efficient and easy to use. As a result, the Samsung Galaxy Tab 7.0" was employed
for doorway screening in the DR.
The next step was to test a lightweight laptop for conducting the NSDUH interviews.
The DR presented an ideal opportunity to field test a smaller and lighter laptop. After
considering a variety of laptop models ranging in display size from 13 to 15 inches and in weight
from 2.5 to 6.0 pounds, SAMHSA and RTI narrowed the options to two lightweight models
offered by Samsung and Lenovo that were purchased for further hands-on evaluation. The
Samsung Series 9 Ultrabook was the lightest of all models (weighing 2.5 pounds) and has a
screen display size of 13.3 inches, while the Lenovo ThinkPad X1 Carbon was slightly larger
with a 14-inch display and a weight of 3.0 pounds. Both laptops include a solid-state drive
(SSD), which is typically considered faster to boot and less vulnerable to physical shock than
traditional electromechanical hard disk drives (HDD). SAMHSA decided to proceed with the
Samsung laptop for the DR field test primarily because of its brighter, crisper, and more colorful
display and its light weight.
2.3.1.2 Software Development
In preparation for the DR training and data collection, the programming team developed
screening, interview, and transmission software for all devices, as well as modified the case
management system (CMS) to accommodate the DR case assignment and transfer requirements.
The screening software developed for the Samsung Galaxy Tab 7.0" used in the QFT was used
again for the DR with several enhancements, including an integrated calendar for setting
appointments, a call distribution function that enables FIs to see their record of calls (ROC)
distributed across time of day and day of week, and a stand-alone wireless transmission
component that enabled FIs to transmit screening data independently of the NSDUH laptop.
In addition to training the DR FIs on using the default tablet keypad, a second keypad,
called the "hacker's" keypad, was loaded onto tablets as an alternative for the DR. The hacker's
keypad is more similar in layout to the iPAQ keyboard in that the main view displays the number
keys across the top of the keyboard. This feature means that FIs do not need to change the
keyboard view to enter numbers versus letters.
Also, the tablet was configured with an email program that was tested during the DR and
allowed FIs to both send and receive messages. FIs were trained that this email capability is not
for private use, but primarily for communication with their field supervisor (FS).
NSDUH's computer-assisted interviewing (CAI) software was modified to fit the
13.3-inch display of the laptop and to incorporate all of the approved changes recommended
from the QFT. Transmission software was modified to enable tethered transmission between the
Android tablet and the new Windows 7 laptop via Wi-Fi or independent transmission on each
device via Wi-Fi. More information is provided on the screening and questionnaire changes for
the DR in Section 2.4.1.
2.3.1.3 Preparing and Implementing the Equipment
The programming team prepared and quality checked the master configurations for the
field test equipment. Once the master configurations had been reviewed for quality control
12

purposes, the technical support group duplicated the masters to produce the quantity of
equipment needed for the DR training and data collection. FIs completed equipment survey
questions (Appendix D) to provide structured feedback about the new laptop, tablet email
program, tablet keypad options, and transmission. Also, calls made to the NSDUH technical
support group were monitored in order to assess any hardware, software, and transmission
problems encountered by the FIs while using the DR equipment.
There were important advantages to integrating the evaluation of the new laptop into
the DR. Using a new laptop allowed for an evaluation of the viability of the chosen device.
One significant consideration was that the Samsung laptop has a smaller display (13.3 inches)
than the current Gateway laptop used for NSDUH (15.4 inches). The results of the initial
hands-on evaluation indicated that despite its smaller size, the bright, crisp display of the
Samsung was sufficient for displaying the NSDUH interview. Information gathered in the DR
demonstrates that the smaller display is large enough to effectively present the NSDUH
interview. Also, the Samsung laptop is much lighter (2.5 pounds) than the current Gateway
laptop (6.8 pounds) and is designed to be used primarily with Wi-Fi Internet access, which offers
a significant advantage with regard to portability.
Introducing a new laptop for the DR also presented some challenges to overcome.
The CAI interview and laptop transmission software were modified to run on the new laptops,
which was configured with Windows 7 operating system (as opposed to the current operating
system, Windows XP). The QFT tablet software was enhanced to support direct Wi-Fi based
transmission of tablet data, and the tablet email client was configured to access FI email
accounts. Modifications to the Web-based CMS were necessary to accommodate the case
assignment and transfer needs of the DR. Finally, the Samsung laptop did not contain an internal
dial-up modem or Ethernet port, which meant that FIs using these devices had to use Wi-Fi
Internet or external Ethernet/dial-up adapters to transmit data back to RTI. All of the NSDUH
FIs recruited for the DR indicated that they had Wi-Fi Internet access at home or had easy access
to a Wi-Fi network. A small set of Ethernet and dial-up adapters was purchased in case the FIs
encountered significant problems using Wi-FI transmission. None of the dial-up modems were
needed during the DR, and one FI temporarily used an Ethernet adapter to transmit while she
was resolving problems with her home wireless network.
It should also be noted that equipment models change frequently. It is unlikely that the
exact equipment used in 2013 will be available for purchase for 2015 when new equipment will
be needed. However, devices with similar form factors, including similarly sized displays, will
likely be available for deployment in 2015.
2.3.2

Staffing

The field management team and structure for the 2013 DR were identical to those used
for the 2012 QFT and the 2013 main study. All of the FIs selected for the DR also collected data
during the 2013 main study's quarters 3 and 4, which overlapped with the DR field period.
FIs were chosen for the DR data collection based on several factors. Initial consideration of FIs
was determined by proximity to DR segments. Field managers analyzed the DR sample
distribution to determine which FIs would be strategic choices for consideration, taking into
account the high Spanish-speaking segments included in the sample. Location and bilingual
status, however, were not the only determining factors.
13

Length of service on NSDUH was also an important selection criterion for DR FIs.
The goal for the DR interviewing team was to have a mix of veteran and newer FIs working on
the DR data collection effort that was similar to the distribution for FIs working in quarters 3 and
4 of the main study. FIs who had attended the January 2013 new-to-project (NTP) training
session or who had attended an earlier NTP session were eligible for selection for the DR data
collection. Table 2.6 shows the distribution of the DR FIs by tenure level compared with the FIs
from the 2013 main study's quarters 3 and 4 who were collecting data at the same time.
Table 2.6 Tenure Distribution of 2013 Quarters 3 and 4 Main Study Field Interviewers Compared
with 2013 Dress Rehearsal Field Interviewers
Number of
Quarters
Worked on
NSDUH Since
2005
0-4
5-8
9 - 12
13 - 16
17 - 20
21 - 24
25 - 28
≥ 29
Total

2013 Quarters 3 and 4
NSDUH Field Interviewers
Count
147
101
78
56
27
53
34
241
737

2013 Dress Rehearsal Field
Interviewers

Percent
19.9
13.7
10.6
7.6
3.7
7.2
4.6
32.7
100.0

Count
6
17
19
15
3
14
7
52
133

Percent
4.5
12.8
14.3
11.3
2.3
10.5
5.3
39.1
100.0

Difference between
2012 QFT and 2013
DR
Percent
-15.4
-0.9
3.7
3.7
-1.4
3.3
0.7
6.4
N/A

DR = Dress Rehearsal; N/A = Not applicable; NSDUH = National Survey on Drug Use and Health; QFT =
Questionnaire Field Test.
NOTE: Percentages may not sum to 100 percent because of rounding.

Proximity to sample segments and experience level were balanced with each FI's
previous data quality and cost-efficiency results, availability, and dependability to take on the
additional DR work from September 1 through October 31, 2013. FIs who had poor data quality
ratings or high costs on their main study work were not considered for the DR data collection.
A group of alternates was also recruited as replacements in case there was any attrition among
the initially selected group of FIs. In total, 133 FIs successfully completed the DR FI training and
conducted the DR data collection (see Section 2.3.3).
2.3.3

Training Procedures

2.3.3.1 Training Materials
Using a master list of needed supplies, all of the training materials were prepared and
ordered (if necessary) in preparation for DR training activities. A detailed, near-verbatim training
guide was prepared and given to each member of the training team. Along with the training
guide, numerous printed materials were also developed:
•

DR FI handbook that contained protocols and procedures for conducting work on the
DR;
14

•

training workbook that contained necessary exercises, screening and interviewing
mock scripts, and additional instructions;

•

quality control forms specifically for the various training cases;

•

interview incentive receipts for use during the practice interviews;

•

showcard booklets for training and use during subsequent fieldwork;

•

supplies to be used during the course of training, including the lead letter, study
description, and question & answer (Q&A) brochure;

•

administrative forms providing session-specific details for proper travel
reimbursement;

•

evaluation forms used by trainers when observing FIs in class; and

•

bilingual training packets with materials for use during the bilingual training session.

Additionally, PowerPoint slides were developed to accompany the various training guide
sections, providing illustrations of the items under discussion or summarizing the main points
conveyed in the guide.
As part of the DR training plan, an electronic multimedia and interactive training
application, referred to as iLearning (which stands for independent learning), was used. Using
iLearning allowed FIs to complete an introductory DR iLearning course at their own pace and
review portions of the course again as needed. The DR iLearning course consisted of slides with
both text and graphics, an audio component providing important information and instructions,
a training video, interactive practice exercises, and an assessment portion to ensure each FI's
comprehension of the DR material presented during the course and within the DR FI handbook.
At the end of the course, assessment results were transmitted to RTI and posted to the CMS for
field management staff review. The DR iLearning course was completed by all of the FIs who
were selected for the DR prior to their attending the in-person training. All 140 DR FIs
scheduled to attend the in-person DR FI training sessions successfully completed and passed the
DR iLearning course. (See Section 2.3.3.3 for more details on the number of FIs who actually
completed the DR FI training sessions.)
2.3.3.2 Training and Materials for Trainers and Field Management Staff
The training teams for the in-person session consisted of a lead trainer (an experienced
project instrumentation or operations team member), an assistant trainer (a survey specialist), and
a technical support representative. In addition, two site leaders, a logistical assistant, and a lead
technical support manager handled operations at the training site. All of the training staff
received the same handbook and iLearning course that were sent to the FIs, as well as a nearverbatim training guide containing detailed instructions and text to ensure that all of the
necessary instructional points were covered.
To prepare trainers for their role at the in-person FI training session, a master trainers'
session was held on RTI's main campus in North Carolina on August 1, 2013. Additional RTI
staff selected to complete DR FI observations with RTI-certified bilingual DR FIs (see
Section 2.3.3.4) also attended to increase their knowledge of NSDUH and DR protocols.

The session was led by members of the instrumentation team who reviewed all portions
of the DR training guide and materials, as well as the logistics for the DR FI training and
instruction on the equipment being used. In addition, a short kickoff meeting was held with
trainers at the training site on August 23, 2013, to further review pertinent training guide details
and important reminders so that all of the trainers were fully prepared to conduct the session.
To ensure that NSDUH's regional supervisors (RSs) and field supervisors (FSs) could
appropriately manage DR FIs during the DR data collection, each RS and FS received the DR
handbook and iLearning course sent to the FIs, as well as a DR management guide outlining the
administrative duties and questions they might encounter. On August 6, 2013, members of the
management team led a special video-streamed DR management session for all of NSDUH's data
collection management staff, including FSs, RSs, and regional directors (RDs), as well as other
NSDUH team members. During the session, the leaders reviewed the DR schedule and
procedures, along with staff roles and responsibilities, and answered questions related to
managing the DR fieldwork.
2.3.3.3 Field Interviewer Training Sessions
Training for DR FIs was held in Bethesda, Maryland, during two separate sessions.
Session A was held on August 24 and 25, 2013. Session B took place on August 26 and 27,
2013. Of the 140 DR FIs scheduled to attend the in-person DR FI training, 1 FI was unable to
attend the training due to illness. Of the 139 DR FIs who attended the DR FI training sessions,
135 FIs successfully completed the training. Four FIs demonstrated significant performance
issues during the DR training session and, therefore, did not successfully complete the training.
These FIs were excused from the DR data collection, and the cases originally assigned to them
were reassigned to other FIs.
Of the 135 FIs who successfully completed the DR FI training, 62 FIs had also been
trained as FIs for the QFT. Table 2.7 summarizes the results of the DR FI training sessions.
Table 2.7 Dress Rehearsal Field Interviewer Training Program

DR FI Training Session
Session A (August 24 and 25, 2013)
Session B (August 26 and 27, 2013)
Total DR FIs Completing Training

FIs
Attending
70
69
139

FIs
Successfully
Completing
Training
68
67
135

FIs Trained
on the QFT
34
28
62

Bilingual FIs
Successfully
Completing
Training
21
20
41

DR = Dress Rehearsal; FI = field interviewer; QFT = Questionnaire Field Test.

With the inclusion of Spanish-language instruments and materials for the DR, a bilingual
FI training session was conducted at the end of day 2 for both the A and B sessions of the DR FI
training. During these bilingual training sessions, a total of 41 veteran, RTI-certified bilingual
FIs participated and successfully completed the training (see Table 2.7).
The DR FI training program included an initial self-study component (completed at home
prior to attending the in-person training) in which FIs read the DR FI handbook and completed
the DR iLearning course. During the 2-day in-person classroom training, FIs had hands-on
16

practice with the DR equipment, programs, and DR-specific procedures. In addition to detailed
instruction on specific DR procedures, FIs were reminded of key NSDUH protocols, such as
reading all screens verbatim, protecting respondent privacy, and following administrative
procedures. The 2-day DR FI training agenda is provided in Exhibit 2.1.
Day 1. Training classes began with an introduction to the DR and the FI's responsibilities
on the study. The next topic focused on the DR equipment and provided instruction in the use of
the laptop computer hardware and the basics of the tablet hardware and software, including the
screening program. After a short break, the FIs learned about locating and contacting
respondents, completed a group walk-through of a DR screening, and were able to practice
effectively answering respondent questions and dealing with nonresponse, as well as using the
tablet for screening as they completed paired mock screening exercises.
Following a lunch, FIs were introduced to the DR interview materials and procedures and
completed a group walk-through of a DR interview. The FI debriefing questions were covered,
as well as additional tips for answering DR-related respondent questions and dealing with
nonresponse. After a break, the late afternoon session was spent completing two paired mock
screening and interview exercises to gain more practice with the overall DR process. During all
of the paired mock exercises, FIs were observed by trainers and were given constructive
feedback on their performance and understanding. This was also a time when retraining could
take place and FIs could ask questions.
All of the FIs were invited to attend an evening FI laboratory session for additional
practice or assistance. For homework during the evening, FIs completed a DR screening and
interview exercise and some additional tasks on the tablet. RTI-certified bilingual FIs completed
the evening homework using the Spanish-language versions of the DR screening and interview.
Day 2. The training session on day 2 started with instruction on the transmission process
and how to troubleshoot problems with the equipment. The homework from the previous evening
was also reviewed. FIs practiced actual transmission procedures to ensure that everything was
working properly, both with a combination tablet/laptop transmission and a tablet-only
transmission. The FIs in Session A also received their assigned DR cases. The FIs in Session B
transmitted again later in the day to pick up the patch for the CAI instrument; they were told to
transmit again from home to receive their cases. The Session A FIs also transmitted again from
home to pick up the patch before beginning data collection. (See Section 2.4.8 for more
information regarding the CAI patch.)
Also in the morning on day 2 of training, FIs completed two more paired mock exercises
while their trainers observed them, and they received feedback from their trainers. Starting late in
the morning and continuing after lunch, administrative tasks were reviewed, including email on
the tablet, reporting to their FS, how to record time and expenses, and tips on organization.
During a session wrap-up in the midafternoon, key procedures and protocols of the DR were
reviewed, and FI questions were answered. FIs also completed the first installment of the FI
feedback survey covering the training topics. (See Section 5.2 for details and results of the DR FI
training survey.)

Exhibit 2.1 Dress Rehearsal Field Interviewer Training Agenda

DAY 1
8:30 (1) Introduction to the Dress Rehearsal (DR) [20 minutes]
 Introductions
 Training Agenda
 DR Overview
 DR Field Interviewer (FI) Responsibilities
8:50 (2) Introduction to the DR Equipment [1 hour, 10 minutes]
 Reviewing the Equipment Assignment and Receipt Form
(EARF)
 Laptop Hardware
 Getting Started on the Laptop
 Tablet Hardware
 Getting Started on the Tablet
 Equipment Care & Maintenance
10:00 Break
10:15 (3) Administering the DR Screening [1 hour, 45 minutes]
 Locating & Contacting Respondents
 Screening Procedures
 DR Screening - Group Walk-Through
 Answering Respondent Questions & Nonresponse
 DR Paired Screening Exercises
12:00 Lunch
1:00 (4) Administering the DR Interview [2 hours]
 Interview Materials & Procedures
 DR Interview - Group Walk-Through
 FI Debriefing Questions - Interview
 Answering Respondent Questions & Nonresponse
3:00 Break
3:15 (5) DR Paired Mocks 1 & 2 [1 hour, 45 minutes]
 Review of DR Process
 Paired Mocks 1 & 2
 Review of Paired Mocks 1 & 2
 Day 1 Question
 Day 1 Wrap-Up
5:00 Adjourn
6:00 – 8:00
Field Interviewer Lab
Homework Exercise

DAY 2
8:30 (6) Transmission & Troubleshooting [45 minutes]
 Review of Homework Exercise and Answer FI Questions
from Day 1
 Transmission Procedures (including Actual Transmission)
 Troubleshooting & Technical Support
9:15 (7) DR Paired Mocks 3 & 4 [2 hours]
 Paired Mocks 3 & 4
10:00 Break
10:15 (7) DR Paired Mocks 3 & 4 (continued)
 Review of Paired Mocks 3 & 4
11:30 (8) Administrative Tasks [1 hour, 15 minutes]
 Email on the Tablet
12:00 Lunch
1:00 (8) Administrative Tasks (continued)
 Reporting to Field Supervisor (FS)
 Recording Time & Expenses
 Organization
1:45 (9) Session Wrap-Up [45 minutes]
 Review of Key Procedures & Protocols
 Day 2 Questions
 FI Feedback Survey
 Wrap-Up
2:30 Adjourn [Bilingual FIs have a 15-minute break]
2:45 (10) DR Bilingual FI Training [1 hour, 30 minutes]
 Introductions
 Overview of Translation Process
 Review of Spanish-Language DR Materials
 Review of Spanish-Language DR Screening Instrument
 Review of Spanish-Language DR Computer-Assisted
Interviewing CAI Instrument (Audio Computer-Assisted SelfInterviewing [ACASI] Changes and Computer-Assisted
Personal Interviewing [CAPI] Questions)
 Handling Respondent Questions & Nonresponse with
Spanish-Speaking Populations
4:15 Adjourn

2.3.3.4 Bilingual Field Interviewer Training Sessions
With the inclusion of Spanish-language instruments and materials for the DR, bilingual
NSDUH DR FIs who passed certification procedures administered by RTI's Spanish-language
specialists participated in an additional bilingual FI training session at the end of day 2. For both
the A and B sessions of the DR FI training, there were two classrooms for the bilingual training.
The sessions were led by two of the Spanish-speaking RTI language specialists who
assisted with the translation and development of the DR Spanish-language questionnaire and the
development of the bilingual training guide's contents. Members of the instrumentation team
provided additional support.
Prior to attending this in-person session, bilingual FIs reviewed the Spanish-language
versions of all respondent materials and completed the day 1 homework using the Spanishlanguage DR screening and interview instruments. FIs were asked to make note of any questions
about the translations and DR changes to discuss during the training session.
During the 90-minute in-person session, bilingual FIs reviewed the Spanish-language
versions of the DR materials, screening, and CAI instruments, and they discussed issues specific
to dealing with Spanish-speaking respondents.

2.4

Data Collection Procedures

This section describes the procedures followed in conducting the DR data collection,
which was fielded from September 1 through October 31, 2013.
2.4.1

Questionnaire and Protocol Changes for the 2013 Dress Rehearsal

The DR screener and questionnaire used the QFT survey instruments as a base and made
appropriate edits to question text, response options, and routing logic. To document the changes
made to the main study screener and questionnaire from the 2012 protocol for the QFT and
further changes made to the QFT screener and questionnaire for the DR, Appendix A includes
the following tables of screener and questionnaire changes:
•

Table A.1 Changes between the 2012 NSDUH Screener and the 2012 Questionnaire
Field Test (QFT) Screener;

•

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012
Questionnaire Field Test (QFT) Questionnaire;

•

Table A.3 Changes between the 2012 Questionnaire Field Test (QFT) Screener and
the 2013 Dress Rehearsal (DR) Screener; and

•

Table A.4 Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire
and the 2013 Dress Rehearsal (DR) Questionnaire.

Table A.1 in Appendix A lists the changes made to the QFT screener and questionnaire in
preparation for the DR. These edits are not as far-reaching compared with the edits that were
made to prepare the QFT questionnaire. (The QFT edits are also included in Appendix A.)

The edits are organized by questionnaire module, and a justification for making each edit is
included where applicable. Edits were made to the questionnaire programs to reflect updated
analytic goals, to correct inconsistencies that were present in the QFT questionnaire, and to
improve questions per recommendations stemming from the QFT analysis. A selected number of
these edits are listed here:
•

A question was added in each prescription drug module to ask about initiation of the
misuse of prescription drugs more than 12 months ago if the only definite reports of
initiation occurred in the past 12 months, or all initiation data were missing. This
question was added to produce accurate estimates of recent initiation.

•

The wording of the medical marijuana questions was edited to include "or other
health care professional" for specificity. This wording change was also made in the
2013 NSDUH questions.

•

Two new questions were added to the back-end demographics module to ask about
sexual attraction and identity of adults.

•

The military family questions were edited to include the definition of "immediate
family" in the question and to provide an "Other, Specify" response to the relationship
question.

•

Questions about the size of the workplace were deleted from the employment module.
Also, a question about whether income or pay was earned while working at a job or
business was deleted from the income module.

Changes that were introduced in the QFT survey instruments were carried through to the
DR as well. These changes represent a departure from the main study NSDUH instrument and
are listed in Table A.2 in Appendix A. The majority of instrumentation changes made to the main
study NSDUH occurred in the QFT. The DR served to refine these changes, identify minor
issues that needed resolution, and implement edits to the questionnaire that were recommended
from the QFT analysis. In order to provide a comprehensive list of items that were changed
during the redesign process, compared with the main study NSDUH, this chapter also includes a
discussion of the changes that were first implemented in the QFT.
Revisions implemented and evaluated for the QFT included the following:
•

revised the contact materials that describe the survey to respondents, including the
lead letter and a Q&A brochure;

•

made general questionnaire revisions to improve questions that cause known or
suspected problems with data, to add new content to address current data needs, to
reduce errors associated with usability problems in the design and layout of the CAI
questionnaire, and to group questions about various substances in a more intuitive
manner;

•

revised the front-end demographic questions;

•

added a new methamphetamine module;

•

revised the questionnaire modules on prescription drugs to improve the measurement
of nonmedical use of prescription drugs (Colliver, Kroutil, Dai, & Gfroerer, 2006), to
ask about any use of these drugs prior to measuring misuse, to add questions about
drugs that are newly available on the market, to delete questions about drugs that are
no longer commercially available, and to add questions about any use of these drugs;

•

revised the questionnaire modules for special drugs (needle use), consumption of
alcohol, health, and back-end demographics questions;

•

deleted the industry and occupation questions;

•

moved the electronic pill images and a reference date calendar so they display on the
laptop screen where appropriate during the audio computer-assisted self-interviewing
(ACASI) portion of the interview (as opposed to the hard-copy versions of these
materials used in the current NSDUH questionnaire administration);

•

added new questions on disability status and primary language in response to
U.S. Department of Health and Human Services (HHS) data standards;

•

added new questions about military families;

•

edited the definition of binge drinking for females, which was identified during
consultations within and outside SAMHSA;

•

converted the back-end demographics module to ACASI from computer-assisted
personal interviewing (CAPI) to decrease FI burden and allow for greater respondent
privacy and shorter administration times;

•

added a new tutorial module to introduce proxy respondents to the CAI instrument
and to help them answer the new self-administered proxy questions about respondent
and household income and health insurance; and

•

modified a question about landline telephones and added a new question about
cellular phones in the home.

In addition to these changes to the materials, questionnaire, and procedures, QFT FIs tested the
use of a new Samsung Galaxy Tab 7.0", a touch-screen Android tablet computer, for screening,
interview respondent selection, and case management. Each of these changes was also included
in the DR protocol and instruments.
Additional edits are planned for implementation in the 2015 NSDUH screener and
questionnaire. Edits that have been approved for inclusion in the 2015 CAI questionnaire as of
December 19, 2013, are listed in Table A.3 in Appendix A. These changes were proposed in
response to experiences during DR training, DR field observations, recommendations from FIs,
and feedback received during the DR debriefing calls. In addition, an analysis of the levels of
item-missing data in the health insurance and income modules in the QFT revealed an increase in
"don't know" and "refused" responses for selected items that were administered using ACASI
and because there were large changes in certain estimates such as estimates of private insurance.
The change in mode from the main study was hypothesized to be correlated with this increase in
item-missing data. Because it cannot be established that these estimates are more accurate, the
2015 instrument will return these modules to an interviewer-administered mode. These changes
are listed among the approved edits in Table A.4.
21

2.4.2

Contacting Dwelling Units

The procedures for contacting respondents during the DR were the same as those used for
the NSDUH main study, with the exception of a few changes to the terminology and contact
materials used with respondents. Similar to the QFT, when contacting respondents, FIs referred
to "RTI International" (or "RTI") and the "U.S. Department of Health and Human Services
(DHHS)," as opposed to "Research Triangle Institute" and the "U.S. Public Health Service."
These updates were reflected in all field materials, including the lead letter, study description,
Q&A brochure, "Sorry I Missed You" (SIMY) card, Spanish card, interview appointment card,
"Who Uses the Data?" sheet, RTI/SAMHSA fact sheet, interview incentive receipt, certificate of
participation, and the door person letter. The majority of the DR field materials were printed on
gray paper and had the DR version number (v. DR 9.13) in the lower right corner in order to
distinguish them from the NSDUH main study materials.
Aside from any annual updates to reflect the survey year, only the Q & A brochure and
questionnaire summary were modified from the QFT. In the Q&A brochure, a picture was
replaced because the picture in the QFT version showed a respondent using a paper reference
date calendar. The questionnaire summary was updated to read in the third person, and some
terminology and wording in the tobacco products and illicit drugs sections were revised to match
the questionnaire. Additionally, because the interviews were conducted in Spanish for the DR, all
of the DR field materials were translated to Spanish. The Spanish-language versions of the DR
materials were printed on blue paper and included the DR version number.
2.4.2.1 Lead Letters
Similar to the lead letter procedures followed during the main study and the QFT, prior to
a DR FI's arrival at a sampled dwelling unit (SDU), a lead letter was mailed to the address to
briefly explain the study and request the resident's cooperation (see Appendix A). This letter was
printed on DHHS letterhead with the signature of DHHS national study director and RTI's
national field director. Upon arrival at the SDU, the FI referred the respondent to this letter and
answered any questions. If the respondent had no knowledge of the lead letter, the FI provided
another copy, explained that one was previously sent, then answered any further questions.
The lead letter was the same as the one used during the QFT. The main study lead letter
was modified for the QFT and DR with redesigned content and format changes to the FI ID and
letterhead. As previously mentioned, the "U.S. Public Health Service" reference was replaced
with the "U.S. Department of Health and Human Services" in the letter. Additionally, the letters
were preaddressed to include the county, parish, or district name as part of the address and
salutation. These changes were based on a contact materials redesign study, which included 12
English-language focus groups and five Spanish-language focus groups in five metropolitan
areas in the United States (Currivan et al., 2009).
2.4.2.2 Introduction, Study Description, and Informed Consent
When in-person contact was made with an adult resident of an SDU, the DR FIs followed
the same introductory and informed consent scripts and procedures for the screening as were
used during the NSDUH main study, with one exception. As mentioned previously, the "U.S.

Department of Health and Human Services" was identified as the sponsor of the study, and "RTI
International" was used instead of "Research Triangle Institute" in the study introduction script.
These same wording changes were made to the study description. All other informed consent
procedures remained the same for the DR, including handing a study description to the
respondent. The only change to this process between the QFT and DR was the addition of the
Office of Management and Budget (OMB) number and burden statement to the study
description.
2.4.2.3 Callbacks
FIs followed similar guidelines for callbacks during the DR as the main study and QFT,
including the use of SIMY cards and appointment cards. In cases where no one was at home
during the initial visit to the SDU, the FI left a SIMY card to inform the resident(s) that the FI
planned to make another callback at a later date and time. Appointment cards were used to
remind respondents when the FI would return to complete the interview. If the FI was unable to
contact anyone at the SDU after repeated attempts, the FI requested an unable-to-contact (UTC)
letter. During the DR, these letters were not actually sent to respondents, however, because of a
system error (see Section 2.4.8.2).
Similar to the main study, except in the case of adamant refusals, FIs attempted to make
at least four callbacks (in addition to the initial call) to each SDU in order to complete the
screening process and complete an interview. These contacts were made at varying hours on
different days of the week to increase the likelihood of completing the screening. These same
guidelines were followed as closely as possible for the DR, but the more widely dispersed
sample and the limited number of DR FIs available to travel longer distances resulted in less
flexibility for assignments and fewer staff for remote segments. For the main study, FSs were
able to generate more effective callbacks by strategically assigning and transferring cases based
on FI availability and experience.
2.4.3

Dwelling Unit Screening

DR procedures for screening at a DU were similar to those used for the NSDUH main
study. Similar to the QFT, the most significant change was that all screenings were completed on
the tablet as opposed to the iPAQ (see Section 2.3.1 for more information on the new
equipment). The introduction and informed consent scripts incorporated the changes specified
above. The information gathered from the respondent during the screening was the same as what
was collected in the main study and QFT. Unlike during the QFT, however, the DR screening
instrument was available in Spanish.
2.4.4

Interview Administration

FIs conducted the DR interviews using the same techniques as employed in the main
study; however, they were trained to answer common respondent questions based on specific DR
procedures. For example, FIs used the DR naming conventions of "RTI International" and the
"U.S. Department of Health and Human Services" rather than "Research Triangle Institute" and
the "U.S. Public Health Service." To describe the types of questions asked, the FIs provided
respondents with the DR version of the summary of the questionnaire, but FIs were instructed to

never tell respondents that they were part of a field test or provide specific sample size
information. The major change to the DR from the main study and QFT was the use of a new
lightweight Samsung laptop (described in Section 2.3.1). Also, unlike the QFT, the DR
questionnaire was translated into Spanish.
2.4.4.1 Informed Consent and Getting Started
Prior to beginning a DR interview, FIs obtained informed consent using the same
procedures employed in the main study. This included reading the DR version of the appropriate
introduction and informed consent scripts from the DR showcard booklet before the interview
began and providing the DR study description to the respondent if not already given one during
the screening. Similar to the QFT, the informed consent scripts were modified for the DR from
the main study version to ensure that respondents were accurately informed about the study.
Specifically, the main study informed consent states that the individual respondent will represent
thousands of others. Because the representativeness of each respondent differs in the DR sample,
the sample size information was removed from the DR script. In addition, the reference to the
"U.S. Public Health Service" in the introduction and informed consent scripts for respondents
aged 18 or older was replaced with the "U.S. Department of Health and Human Services."
Respondents in the DR were not informed that the interview was part of a field test.
2.4.4.2 Computer-Assisted Interviews
FIs began the DR interview with the front-end CAPI section, which contained
demographic questions similar to those on the main study with a few key differences. As with
the QFT, new questions were added regarding the respondent's prior military service, two new
categories were added to the race question ("Guamanian or Chamorro" and "Samoan"), and
response categories were adjusted in the education-level question. The new race categories were
also added to the 2013 and 2014 NSDUH main study questionnaire. As in the main study
interview, the FI introduced the respondent to the computer prior to the respondent completing
the practice session and ACASI section on his or her own. As noted in Section 2.4.1, there were
several key changes to the ACASI portion of the main study interview for the QFT and DR,
including the electronic reference calendar and on-screen pill cards. Also, for the DR only, two
new sexual orientation questions were asked of adults.
Following the ACASI section of the interview, the FI took the computer back and asked
the household roster questions. Following these questions, the FI inquired about the use of a
proxy for the health insurance and income questions. For the DR and the QFT, a second ACASI
section administered the health insurance and income questions. If a proxy was used, the FI
introduced the proxy to the computer prior to the proxy completing a short practice session and
the health insurance and income questions on his or her own. However, if the respondent
answered the questions or the proxy had previously used the computer, there was no additional
practice session.
In addition to the new sexual orientation questions, questionnaire changes between the
QFT and DR included routine updates to routing and logic, minor changes to question wording
throughout the instrument to clarify intent, and the deletion of a question in the employment
module about the number of employees who work at the respondent's business. A question about

whether the respondent earned wages or pay from working was also deleted from the income
module. A question about whether the respondent earned wages or pay from working also was
deleted from the income module, and this source was added to the list of income sources.
Differences between the main study NSDUH, the QFT, and the DR are explained in Section
2.4.1.
2.4.4.3 End of Interview Procedures
DR quality control forms were completed in the same manner as on the main study and
QFT. Minor changes were made to the main study verification screen for the DR, including
removing the word "home" in the telephone number reference to match the wording on the DR
quality control form and asking respondents to enter their current address. Text was added that
told the respondent to return the form in the sealed envelope to the FI. This verification screen
wording was also used on the QFT.
Respondents received a $30 incentive for completing the interview following the same
procedures used on the main study and QFT. At this point, if not given earlier, the FI provided
the respondent with the DR version of the Q&A brochure (see Appendix A). DR certificates of
participation were also available for youth respondents and were presented in the same way as in
the main study and QFT.
As in the QFT, the FI debriefing questions were removed from the end of the interview
because these questions were answered in the tablet upon entering a code of 70 for the completed
interview. This change allowed the FIs to answer the questions after leaving the household and
reduce the length of time in the respondent's home. The questions were answered by the FIs
based on the interview and any comments the respondent may have offered. After entering a
code 70 to document a completed interview, the FI was prompted by the tablet to complete the
debriefing questions. The questions were not read out loud to the respondent; rather, the FI
completed them on his or her own after leaving the SDU.
2.4.5

Controlled Access Procedures

Controlled access during the DR was treated similarly as for the NSDUH main study and
QFT. When controlled access situations were encountered, controlled access packets were
requested by the FS. The DR controlled access packets reflected the differences in the naming
conventions implemented for the DR. To gain access in difficult situations, FSs also transferred
cases between DR FIs. If those attempts failed, "Call-Me" letters were sent directly to a selected
household. These letters informed residents that an FI had been trying to contact them and asked
that they contact an FS by telephone.
2.4.6

Refusal Conversion Procedures

Refusal conversion procedures followed during the DR were similar to those used for the
NSDUH main study and the QFT. If a potential respondent refused, the FI attempted to address
the respondent's concerns and was trained to accept the refusal in a positive manner, thereby
avoiding the possibility of creating an adversarial relationship and precluding future
opportunities for conversion. If the potential respondent still refused to participate, a refusal
letter was requested by the FI. The refusal letter was tailored to the specific concerns expressed
25

by the potential respondent and asked him or her to reconsider participation. Based on the refusal
situation, an in-person conversion was then generally attempted by the original FI or another
DR FI available nearby or on travel assignment. In certain FS regions, another FI was not
available nearby or on travel assignment because of the small number of cases remaining in the
area. Also, refusal letters were requested by DR FIs, who thought the letters were sent to
respondents, but the letters were not sent because of a system error (see Section 2.4.8.2).
2.4.7

Data Collection Management and Quality Control

FIs and field management staff worked strategically to balance quality, cost, and
production goals for the DR, just as they do for NSDUH's main study. The case management
tools, features, and reports used by the management team to monitor fieldwork for the main
study were adapted for use during the DR.
2.4.7.1 Web-Based Case Management Reports
The Web-based CMS housed a DR reports page that mirrored the NSDUH main study
reports pages and was the same as was used for the QFT. The following daily reports were
available for case management on the DR: daily FS and State response rate report, daily status
reports, edited address reports, duplicate address reports, and recruit reports. The following
weekly reports were also available on the CMS: executive summary report (including production
and cost data), data quality summary report, missing screening data report, ROC time
discrepancies, and interview length report. These reports were the same as the main study reports
except that DR data were used. To help track the status and progress of the DR, weekly response
rate, interview completion, and cost reports comparing the DR with the QFT were also available
to project and field management for the DR.
2.4.7.2 Field Interviewer Observation Procedures
In conjunction with DR data collection, field observations of FIs were conducted by RTI
staff and SAMHSA staff members. RTI staff included language methodologists, training and
field materials team members, instrument assessment and development team members, among
others. Groups of three to four FIs were chosen for field observations in each of seven
metropolitan areas: Miami, Florida; New York, New York; Los Angeles, California; San
Francisco, California; Dallas, Texas; Houston, Texas; and Chicago, Illinois. RTI staff also
observed FIs locally in North Carolina. SAMHSA staff observed an additional four FIs in
Louisiana, Maryland, Virginia, and the District of Columbia. These observations covered
interviews completed in both English and Spanish. Spanish-speaking bilingual observers
conducted all observations of interviews conducted in Spanish. An observation was considered
complete only after a full interview was observed; therefore, observations where only screenings
or partial interviews took place were not considered complete.
Observers used the DR field observation screening checklist and the DR field observation
interviewing checklist to document their observations. A field observer reference sheet and a
field observer task list were used to help maintain consistency in planning observation
assignments and interacting with FIs and respondents (see Appendix G). Observers were asked
to ensure that a field observation FI instruction sheet was sent to each FI prior to the FI's arrival

in the field. The DR housing unit (HU) and group quarters unit (GQU) scripts and CAI
specifications for the front-end and back-end CAPI questions were provided to observers for
their use during the observations. These materials were developed specifically for the DR data
collection effort based on similar materials used for the main study and QFT field observation
process.
Observers were asked to transfer information from paper field observation screening
checklists and field observation interviewing checklists to spreadsheets designed specifically for
the DR field observations. The DR field observation manager then used the spreadsheets to
process the results of the field observation, which included issuing any appropriate disciplinary
action, creating a retraining plan to address any observed errors, and sending any comments
about the performance of the questionnaire, equipment, or materials to the appropriate RTI staff
member.2
The same standardized retraining process from the NSDUH main study was used for the
DR field observations. After the DR field observation manager reviewed each observation form
for an FI who had errors reported on his or her observation, a member of the NSDUH operations
team completed a document referred to as the FI retraining template. This template indicates the
errors the FI made, the type of retraining required, and the dates by which the retraining must be
completed. The FS used this form to provide standardized feedback and retraining (as scripted on
the template) and issued any appropriate disciplinary action as directed by the DR field
observation manager. Results of the field observations are provided in Section 5.6 in Chapter 5.
2.4.7.3 Verification of Completed Cases
Only minor changes were made to the NSDUH main study verification script for the DR.
These changes were also made to the QFT verification script. These changes included
referencing a tablet instead of an iPAQ, providing a different computer tutorial question as an
example to the respondent, and saying "U.S. Department of Health and Human Services" and
"RTI." Unlike the QFT, the DR included a Spanish-language version of the verification script.
Of the 2,087 completed DR interviews, 45 DR quality control forms were not returned.
Of the completed DR interviews, 694 cases were selected for telephone verification. No
problems were found with 434 cases, 79 cases were coded as problems, 153 cases were unable to
be contacted, and 28 cases had other issues. Of the completed DR screenings, 559 cases were
selected for telephone verification. No problems were found for 312 of the cases, 81 cases were
coded as problems, 108 cases were unable to be contacted, and 58 cases had other issues.
Problem cases were those that verified with errors, such as items the respondent did not
remember the FI performing, the respondent reported that this was not the correct phone number

FIs who committed a serious breach of protocol in DR data collection (defined as those that could
potentially violate a respondent's rights and/or significantly compromise the accuracy of the data collected) and
those observed committing four or more unrelated errors were issued a disciplinary action. All disciplinary actions
issued for the DR field observations were verbal warnings. During DR field observations, four verbal warnings were
issued. The FIs who received verbal warnings were added to the main study field observation list to be observed in
the next possible quarter.

for that address, or if the respondent said that he or she was not given the $30 incentive.3 Cases
with "other issues" were considered unresolvable and included situations in which the telephone
interviewer was never able to speak with the respondent, someone answered the phone but
refused or hung up, or an initial problem was reported but callback verification staff were not
able to recontact the respondent to confirm the issue. Staff on the callback verification team
recontacted respondents when a problem was reported and more information was needed to
confirm or clarify the situation because, during the initial call, the verification script was read
verbatim by the telephone verifiers.
2.4.8

Problems Encountered
2.4.8.1 CAI Questionnaire Issues

During DR data collection and analysis, some minor irregularities in the CAI program
were uncovered. One issue was uncovered in the English-language questionnaire: Respondents
who reported using tobacco in their lifetime, but not necessarily in the past year, were later asked
if a doctor had advised them to quit smoking in the past year in HLTH21. This routing caused an
unnecessary question to be asked of respondents who were not current smokers. The routing
logic for the comparable question in 2014 (HLTH18) was updated to prevent respondents who
reported not smoking in the past year in the tobacco module from receiving this question. This
edit will be carried over to the 2015 questionnaire as well.
A couple of additional items were uncovered in the Spanish-language questionnaire.
On the INCENT01 screen, a minor mistake in translation occurred, causing extra wording to
appear on the screen. When translated into English, this extra screen text read, "Now I will finish
some questions to show that I did the interview. Thank you very much for your help."
This additional wording was likely carried over from the specifications from the 2013 main study
instrument. This wording only appeared in the Spanish-language instrument and likely did not
affect responses to subsequent questions. There was also some confusion over a term used in the
Spanish-language instrument. The use of the phrase "heterosexual, that is, straight," in item
QD63 is problematic for some Spanish speakers, as identified in the FI debriefing calls. Some
Spanish speakers did not understand the term "heterosexual" in reference to sexual identity and
thought it meant being attracted to the same sex or was a shameful or embarrassing term. RTI
language methodologists have seen similar reactions in other surveys and are working to find a
translation that can better meet the expectations of Spanish-speaking respondents.
The team discovered an error in the specifications for both the English- and Spanishlanguage CAI instrument, which was addressed prior to fielding the instrument. The routing
logic for the sexual attraction (QD62) and sexual orientation (QD63) questions was incorrect and
was missing a reference to the age variable that restricts these questions to adult respondents.
This routing was corrected at DR FI training with an instrument patch, and only adults received
these questions during the actual 2013 DR data collection.

None of the problem or unresolved telephone verification cases for DR screenings and interviews
involved cases worked by FIs who were found to have falsified data in quarter 3 of 2012 or in subsequent quarters.
One DR FI was found to have falsified data in quarter 4 of 2013; this FI's DR cases that were not phone verified
"okay" or were completed during a field observation were removed from the dataset.

Interview length was fairly consistent across subgroups of respondents, with the
exception of Spanish-speaking respondents over the age of 65. This subgroup took significantly
longer than the rest of Spanish-speaking respondents to complete their ACASI sections of the
interview. This discrepancy is discussed in detail in Section 4.5 of this report. In general, there is
no evidence, however, that the increased completion time for Spanish-speaking respondents aged
65 or older is due to an issue with the CAI instrument or had an overall impact on data quality.
Interview timing for this age group will continue to be monitored in order to inform burden
concerns for this portion of the population.
2.4.8.2 Data Collection Issues
At the end of the DR data collection period, a problem was discovered in the lettergenerating system that prevented refusal and UTC letters from being sent. Although letters were
requested by the field, the letters were not sent to respondents from RTI as planned. This system
glitch did not occur during the QFT.
Overall, a higher percentage of refusals were finalized on the DR than the QFT,
especially for screenings. However, a smaller percentage of screenings and interviews were final
coded as unable to contact on the DR than the QFT. An investigation was conducted to
determine what the impact was of not sending the refusal and UTC letters on these DR final
dispositions.
For screening conversion, a higher percentage of screening refusals were converted on
the DR (28.73 percent) than on the QFT (27.95 percent), but a lower percentage of screening
refusals were converted on the DR than on the 2013 NSDUH (32.83 percent). When a refusal
conversion letter was requested, the refusal conversion rate was slightly higher on the DR
(27.82 percent) than on the QFT (27.40 percent), even though the DR letters were not actually
sent to respondents. If no refusal conversion letter was requested, the DR screening refusal
conversion rate (33.33 percent) was higher than the QFT's (31.03 percent) and the 2013
NSDUH's (28.96 percent). For the interview refusal conversion rates, the DR rate (15.08 percent)
was higher than the QFT's (11.04 percent) and the 2013 NSDUH's (13.76 percent), including the
rates for interview refusals where a conversion letter was and was not requested. Although a
higher percentage of refusals were finalized on the DR than on the QFT, it was difficult to
determine the causal impact of not sending the refusal conversion letters on DR response rates
because refusal conversion rates were higher on the DR. These data suggest that these letters
may be most effective at providing FIs with a confidence booster to return to the household and
an introduction to their refusal conversion efforts at the door (rather than the letter itself
converting respondents).
For cases where a UTC letter was requested (but not sent for the DR), the screening
completion rate was lower on the DR (54.30 percent) than on the QFT (74.25 percent), while the
interview completion rate was just slightly higher for the DR. Despite the lower completion rate
for screenings when a letter was requested, it was difficult to determine the causal impact of not
sending the UTC letters because the completion rates depend on whether the respondent was
home when the FI made the next contact.

3. Processing and Analysis of Dress
Rehearsal and Comparison Data
3.1

Overview of Data Processing and Analysis Approach

This chapter describes the procedures followed to process the 2013 Dress Rehearsal (DR)
data, the 2102 Questionnaire Field Test (QFT) data, the 2012 National Survey on Drug Use and
Health (NSDUH) main study comparison data, and the 2013 quarters 3 and 4 NSDUH main
study comparison data. All of the data processing procedures were developed and implemented
to provide the greatest possible degree of comparability among these three datasets to facilitate
valid comparisons. Section 3.2 describes the usable case rules followed, and Section 3.3 details
the editing and coding procedures. Section 3.4 presents the imputation procedures, while Section
3.5 describes the weighting steps followed and the creation of variance estimation strata and
replicates. Section 3.6 describes the preparation of all of the data files, and Section 3.7 discusses
important data analysis issues, especially those related to the comparison of the DR or combined
QFT and DR data with the main study data and other data sources.

3.2

Defining Usable Cases

3.2.1

Overview of Defining Usable Cases

Usable Case Definitions

The usable case criteria that were in place for the main survey were used for the 2012
main study and the 2013 quarters 3 and 4 NSDUH main study comparison data, as defined
below:
1. The lifetime cigarette gate question CG01 must be answered as "yes" or "no."
2. At least nine (9) of the following additional gates must have answers of "yes" or "no":
(a) chewing tobacco, (b) snuff, (c) cigars, (d) alcohol, (e) marijuana, (f) cocaine (in
any form), (g) heroin, (h) hallucinogens, (i) inhalants, (j) misuse of pain relievers, (k)
misuse of tranquilizers, (l) misuse of stimulants, and (m) misuse of sedatives. (For the
"multiple gate" modules for hallucinogens through misuse of sedatives, at least one
gate question in the series for that module must have an answer of "yes" or "no.")

For the DR, fully defined data for lifetime use or nonuse of cigarettes continued to be a
requirement. Because of changes to the instrument for the QFT and for the DR, the following
was the second criterion for usable cases in the DR:
•

"Usability" must be determined for at least nine (9) of the following additional
modules: (a) smokeless tobacco, (b) cigars, (c) alcohol, (d) marijuana, (e) cocaine (in
any form), (f) heroin, (g) hallucinogens, (h) inhalants, (i) methamphetamine, (j) pain
relievers, (k) tranquilizers, (l) prescription stimulants (i.e., independent of
methamphetamine), and (m) sedatives.

As in the main survey, the usability criterion for smokeless tobacco through heroin was
that lifetime use or nonuse must be determined. For the "multiple gate" modules for
hallucinogens and inhalants, at least one gate question in the series for that module was required
to have an answer of "yes" or "no."
The usability criterion for the prescription drugs in the DR required that any past year or
lifetime use or nonuse can be determined from the data. Specifically, any of the following met
the usability criteria for prescription drugs:
•

past year use of at least one specific prescription drug in a category (e.g., pain
relievers) is reported in the screener questions; or

•

lifetime use or nonuse of any prescription drugs in the category is reported; or

•

past year nonuse of all specific prescription drugs in the screener is reported,
regardless of whether lifetime use or nonuse can be determined.

In the 2012 main study, 0.06 percent of all completed interviews (including interviews
from Alaska and Hawaii) did not meet the usable case criteria.4 In the 2013 quarters 3 and 4
NSDUH main study comparison data (which excluded interviews from Alaska and Hawaii),
0.06 percent of the completed interviews also did not meet the usable case criteria. Three DR
respondents (0.14 percent of 2,092 completed interviews) did not meet the usable case criteria
and were not included for further analysis. One case had missing data for cigarettes, and the
remaining two cases ended the interview before answering a sufficient number of gate questions.

3.3

Editing and Coding Procedures

3.3.1

Overview of Editing and Coding Procedures

Data that field interviewers (FIs) transmit to RTI are processed to create a raw data file in
which no logical editing of the data has been done. The raw data file consists of one record for
each transmitted interview. Cases were eligible to be treated as final respondents if they met the
usable case criteria described in Section 3.2.
Logical editing was the first step in processing the raw DR data and the raw comparison
data from 2012 and quarters 3 and 4 of 2013. Logical editing involved using data from within a
4

The 2012 comparison dataset (excluding interviews in Alaska and Hawaii) was created from the cases in
the full survey that had already been identified as meeting the usable case criteria.

respondent's record to (a) reduce the amount of item nonresponse (i.e., missing data) in interview
records, including identification of items that were legitimately skipped; (b) make related data
elements consistent with each other; and (c) identify ambiguities or inconsistencies to be
resolved through statistical imputation procedures (see Section 3.4).
In addition, a limited set of written answers that interviewers or respondents typed for
responses that did not fit any of the listed categories or examples were assigned numeric codes to
facilitate further use of these data in creating final variables or in analysis. These are
subsequently referred to as "OTHER, Specify" data.
3.3.2

Coding of "OTHER, Specify" Data

Written answers that respondents or interviewers typed were assigned numeric codes for
the following: other Hispanic origin, other racial groups, other Asian origin, other drugs that
respondents used, and other relationships for family members who were currently serving in the
United States military (DR only).5 Except for the relationship data for other family members in
the military, typed "OTHER, Specify" responses first were compared against databases for the
relevant "OTHER, Specify" variables that contained typed entries and the associated numeric
codes. If an exact match was found between the typed response and an entry in the system, the
response was assigned the appropriate numeric code. Typed responses that did not match an
existing entry were output for manual analyst review and coding.
Coding of data for Hispanic origin, Asian origin, and race made these data available for
creating final demographic variables. Coding of "OTHER, Specify" data for drugs made these
data available for examining the quality of responses to the drug use questions.
"OTHER, Specify" data also were coded for DR respondents' relationships to other
family members in the military because of the number of respondents who reported that another
member of their immediate family was serving in the United States military (see Section 4.6.2
and Table 4.14). Unlike the data for Hispanic origin, Asian origin, race, and drugs, all of the
"OTHER, Specify" data in the DR were manually coded for relationships to other family
members who were serving in the military.
Although "OTHER, Specify" data were not coded for other variables, weighted DR
percentages were generated for affirmative reports to selected lead questions governing
"OTHER, Specify" data, such as reports of obtaining misused prescription drugs "some other
way." Findings for these additional "OTHER, Specify" data are discussed in Section 4.6 in
Chapter 4.
3.3.3

General Editing Principles

To reduce the potential for differences to be attributable to the effects of editing, data for
the main study comparison samples from 2012 and quarters 3 and 4 of 2013 (referred to in the
remainder of Section 3.3 as "comparison" data) and for the DR were edited in the same manner
5

Additional "OTHER, Specify" variables had previously been coded for the 2012 survey. These variables
were not included for the 2012 comparison data analysis because corresponding variables were not coded in the DR
or the comparison data from quarters 3 and 4 of 2013.

wherever possible. If questionnaire changes for the DR did not permit total comparability
between the editing procedures for the DR and the comparison data, the aim was to make the
procedures as comparable as possible. Also, where the questionnaire did not change between the
QFT and the DR, data for the DR were edited in the same manner as in the QFT (Currivan et al.,
2013) to allow QFT data to be combined with DR data for non-Hispanic English-language
respondents for some analyses.
One of the initial steps in the editing involved development and implementation of
procedures for identifying potential patterned responses in the data (subsequently referred to as
data "diagnostics"). Specifically, respondents may enter patterned responses in the core drug use
modules that raise questions about the validity of their answers in a particular module or in the
interview as a whole. The types of patterned responses that were reviewed in the core modules
for the comparison data are documented in the editing and coding section (Section 10) of the
2010 methodological resource book (Kroutil, Handley, & Bradshaw, 2012a). Checks were made
for these same patterns in core DR modules that did not change (or underwent minimal change)
relative to the main survey. Because the content of the new methamphetamine module in the DR
was similar to the content in the core modules for marijuana, cocaine, and heroin, the same types
of data checks in these latter modules were implemented for the methamphetamine module.
Particular attention was given to developing specifications and reviewing data for the DR
prescription drug questions because of changes to these questions for the DR. Depending on the
results, cases that otherwise met the usable case criteria could be treated as nonrespondents
because their answer patterns raised questions about the overall validity of their interview data.
Alternatively, cases could be kept as final respondents but with all variables in one or more of
their modules being assigned codes for "bad data," provided that these cases still met the usable
case criteria after the assignment of "bad data" codes (see Section 3.2); codes for "bad data" were
treated as missing values in subsequent data processing or analysis. Findings based on these data
diagnostics reviews are discussed in Section 4.6 in Chapter 4.
A key component of the editing procedures for the DR and comparison data involved
assignment of codes to indicate when it could be determined unambiguously that respondents
legitimately skipped out of questions because of their answers to previous questions. For
example, if respondents answered the lifetime alcohol use question AL01 as "no," all of the
remaining questions in the alcohol module were skipped. In this situation, the editing procedures
assigned codes to the remaining alcohol variables to indicate that the questions were not
applicable because the respondents never used alcohol. However, if respondents did not know or
refused to report whether they had ever used alcohol, the remaining questions for alcohol use
also were skipped. In this situation, the edited alcohol use variables that had been skipped
continued to have missing values. Determination of whether these respondents were lifetime
alcohol users or nonusers was handed through the imputation procedures described in
Section 3.4.
Because the DR and comparison interviews consisted of "core" sections (i.e., certain
demographic characteristics and use of cigarettes through misuse of sedatives) and noncore
sections starting with the special drugs section, a second key principle of the editing procedures
was that data from supplemental sections typically were not used to edit core data. An exception
discussed in Section 3.3.4 is that comparison data on methamphetamine use from the

supplemental special drugs module along with core data were taken into account in a special set
of edited variables for methamphetamine and stimulants.
However, core drug data could be used to edit supplemental data when respondents were
not asked supplemental questions about a given drug based on their report of most recent use of
that drug in the corresponding core module. For example, respondents in the DR or comparisons
were not asked questions about cocaine dependence or abuse in the supplemental substance
dependence and abuse module if they last used cocaine or crack cocaine more than 12 months
ago. In this situation, the edited variables for cocaine dependence or abuse were assigned codes
to indicate that respondents were not asked these questions because the questions did not apply.
In all of the core drug modules for the comparison data and in the cigarette through
methamphetamine core DR modules, respondents were asked "gate" questions to determine
lifetime use or nonuse; because of changes to the questioning strategy and routing logic in the
QFT and DR for prescription drugs, principles for editing the DR prescription drug variables are
discussed in Section 3.3.4.6 The modules for hallucinogens and inhalants in all of the datasets
and the prescription drug modules in the comparison data included multiple gate questions about
lifetime use (or misuse) of specific drugs in the category. Respondents who reported lifetime use
of the particular drug (e.g., marijuana) or any drug in the category (e.g., hallucinogens) were
asked when they last used the drug (or any drug in the category). Respondents who did not know
or refused to report when they last used were asked follow-up questions in an attempt to obtain
data on the specific period when they last used (e.g., within the past 30 days, more than 30 days
ago but within the past 12 months, or more than 12 months ago). If these respondents indicated
the specific period when they last used, the data from these follow-up questions were
incorporated into the edited variables for most recent use. If these respondents on follow-up still
did not know or refused to report when they last used, the edited variable for most recent use was
assigned a code to indicate that these respondents logically could be inferred to be users at some
point in their lifetime based on the computer-assisted interviewing (CAI) routing. A definite
period of most recent use was statistically imputed (see Section 3.4).
The CAI program included checks that alerted respondents or interviewers when an
entered answer was inconsistent with a previous answer. In this way, the inconsistency could be
resolved while the interview was in progress. In situations where a "consistency check" was
triggered during the interview, final values from these checks were incorporated into the edited
variables for drugs and selected additional measures in the DR and comparison data.
Not every inconsistency was resolved during the interviews, and the CAI program did not
include checks for every possible inconsistency that might have occurred in the data. In NSDUH
editing for the main survey, inconsistencies between related variables in core substance use
modules are flagged and the inconsistencies are resolved through statistical imputation (Kroutil
et al., 2012a). To facilitate timely data processing, however, only a limited set of additional
inconsistencies were resolved in the editing procedures. Consequently, inconsistencies could
exist between related variables in the DR or comparison data that would otherwise have been
handled in the editing procedures for the main study. However, special "flag" variables were
6

The text typically mentions "use" when referring both to prescription drugs and other substances.
For prescription drugs, however, this term means "misuse," unless otherwise indicated.

created to alert analysts to the occurrence of these inconsistencies. Findings based on these flag
variables are discussed in Section 4.6 in Chapter 4.
3.3.4

Special Editing Situations

Most editing of the DR and comparison data followed the principles discussed in Section
3.3.2. In the alcohol module, the question in the comparison data that was used to define binge
alcohol use asked both males and females about the number of days that they consumed five or
more drinks on the same occasion in the past 30 days. In the QFT and DR, males were asked
about consumption of five or more drinks on the same occasion, and females were asked about
consumption of four or more drinks on the same occasion. These binge alcohol use variables
were edited in the same manner in both the DR and comparison data. However, the edited DR
variable was given a name that was different from the name for the corresponding variable in the
comparison data to indicate the differences in content.
In addition, the following special situations were relevant to the editing of the DR or
comparison data:
•

In the comparison data, respondents were asked separate questions about their use of
snuff or their use of chewing tobacco. In the QFT and DR, respondents were asked
about their use of any smokeless tobacco product (i.e., snuff or chewing tobacco).

•

In all of the datasets, respondents could report more recent use of crack cocaine than
they reported for use of any cocaine. Respondents also could report more recent use
of specific hallucinogens (lysergic acid diethylamide [LSD], phencyclidine [PCP], or
Ecstasy in the comparison data; LSD, PCP, Ecstasy, ketamine, dimethyltryptamine
[DMT], alpha-methyltryptamine [AMT], N, N-diisopropyl-5-methoxytryptamine
[5-MeO-DIPT], or Salvia divinorum in the comparison data) than they reported for
use of any hallucinogen. In addition, respondents in the comparison data could report
more recent misuse or use of OxyContin® or methamphetamine than they reported for
any pain reliever or any stimulant, respectively.

•

In all of the datasets, respondents were asked whether they used hallucinogens,
inhalants, pain relievers, tranquilizers, stimulants, or sedatives other than those they
were asked about. Respondents were asked to specify the names of up to five
additional drugs (subsequently referred to as "OTHER, Specify" data). However,
respondents could fail to report use of specific drugs in direct questions about these
drugs and then mention these drugs in the "OTHER, Specify" data.

•

Respondents could indicate that the only prescription drugs they misused in the
lifetime period (for the comparison data) or the past year (for the QFT and DR) were
over-the-counter (OTC) medications, despite being instructed not to include use of
OTCs in answering the questions.

•

A new methamphetamine module was added for the QFT and DR. In the comparison
data, methamphetamine questions were included in the core stimulants module, and
methamphetamine was considered to be part of the general category of stimulants.
The comparison data also included methamphetamine questions in the noncore

special drugs module that were used in determining methamphetamine use, stimulant
misuse, and most recent use (or misuse).
•

The focus of the questions for specific prescription drugs in the QFT and DR was on
the past 12 months and on the lifetime period in the comparison data. In addition,
QFT and DR respondents first were asked a series of screening questions about any
use of specific prescription drugs in the past 12 months (i.e., use or misuse) or any
lifetime use if they did not report past year use. QFT and DR respondents were asked
about misuse in the past year of any of the specific prescription drugs that they
reported using in that period. In contrast, respondents in the comparison data were
asked about misuse of specific prescription drugs in the lifetime period, and questions
about more recent misuse applied to the general categories (e.g., past year or past
month misuse of any tranquilizers).

•

Questions in the QFT and DR about use of stimulants with a needle were moved from
the noncore special drugs module to the core stimulants module. These questions
applied only to use of stimulants with a needle in the past 12 months or past 30 days.

•

New questions about methamphetamine dependence or abuse were added to the
substance dependence and abuse module.

•

Sections of the interview in the comparison data that were interviewer-administered
were self-administered in the QFT and DR (e.g., health insurance, income).

Smokeless Tobacco

Editing of the DR variables for smokeless tobacco use followed the general principles
discussed previously. In the comparison data, variables for any smokeless tobacco use were
created based on the data for the use of snuff and the use of chewing tobacco. The following
principles were applied in creating the smokeless tobacco variables in the comparison data:
•

Respondents who answered "no" to both questions about lifetime use of snuff and
chewing tobacco were classified as nonusers of smokeless tobacco.

•

In general, the report of most recent use of either snuff or chewing tobacco was
chosen for the variable pertaining to most recent smokeless tobacco use. If relevant
37

variables for one of the smokeless tobacco products had missing data, special codes
were assigned for use in statistically imputing a final period of most recent use.
For example, if a respondent reported last using snuff more than 30 days ago but
within the past 12 months but did not know when he or she last used chewing
tobacco, the variable for most recent use of smokeless tobacco was assigned a code to
indicate that the respondent logically last used at some point in the past 12 months.
This respondent could have been a past month user of any smokeless tobacco if he or
she used chewing tobacco in the past month. A specific period of most recent use was
statistically imputed.
3.3.4.2

More Recent Use for General Drug Categories and Specific Drugs

For hallucinogens in the DR and comparison data and for pain relievers and stimulants in
the comparison data, consistency checks were triggered if respondents reported more recent use
of a specific type of drug in the category (e.g., Ecstasy) than they reported for their last use of
any drug in the category (e.g., any hallucinogen). As noted in the general principles (Kroutil et
al., 2012a), the editing procedures took into account data from these consistency checks.
For example, suppose a respondent reported last using any hallucinogen more than 30 days ago
but within the past 12 months and last using Ecstasy within the past 30 days. If this respondent
reported in the consistency checks that his or her last use of any hallucinogen also was in the past
30 days, the edited variable for most recent hallucinogen use reflected this change, and the data
were no longer inconsistent.
However, if the data continued to indicate more recent use of a specific drug than for use
of any drug in the category despite the respondent being given the opportunity to resolve the
inconsistency, then the editing procedures logically inferred more recent use of any drug in the
category. For example, if a respondent's answers continued to indicate last use of Ecstasy in the
past 30 days and last use of any hallucinogen more than 30 days ago but within the past
12 months, the respondent was logically inferred to have last used any hallucinogen in the past
30 days; a special code was assigned to the variable for most recent hallucinogen use to indicate
that this edit had been performed.
In the comparison data, these principles applied to editing of the variable for most recent
use of any hallucinogen relative to reports of most recent use of LSD, PCP, or Ecstasy.
Questions in the comparison data about most recent use of the hallucinogens ketamine, DMT,
AMT, or 5-MeO-DIPT ("Foxy"), and Salvia divinorum were in the supplemental special drugs
module and therefore were not used in editing the data for most recent use of any hallucinogen.
For the DR, questions about these three additional hallucinogens were moved from the special
drugs module to the core hallucinogens module. The hallucinogens module for the DR also
included consistency checks that were triggered if respondents reported more recent use of any
of these three hallucinogens than was reported for most recent use of any hallucinogen.
Consequently, data on most recent use of these additional hallucinogens, along with data on most
recent use of LSD, PCP, or Ecstasy, were used in editing the data for most recent use any
hallucinogen in the DR. The same principles applied to editing the DR data when respondents
reported more recent use of any of these additional hallucinogens compared with reports of most
recent use of any hallucinogen.

The cocaine and crack cocaine modules in the DR and comparison data did not include
consistency checks if respondents reported more recent use of crack cocaine than for cocaine in
general. Consequently, data on the most recent use of crack were used to infer more recent use of
cocaine in general, as per the example discussed previously for hallucinogens. Additional issues
related to the editing of the data for most recent use of methamphetamine and misuse of any
stimulant are discussed in the methamphetamine section.
3.3.4.3

"OTHER, Specify" Data for Drugs

For hallucinogens and inhalants in all three datasets and for prescription drugs in the
comparison data, questions about lifetime use (or misuse) were logically inferred to be "yes" if
respondents originally did not report use of these drugs in the direct questions but reported them
in the "OTHER, Specify" data. Additional details about these editing procedures for the
comparison data are provided in the editing and coding section of the 2010 methodological
resource book (Kroutil et al., 2012a).
As noted previously, DR respondents were asked about the use of specific prescription
drugs in the past year and misuse of those drugs that they used in the past year. Consistent with
the structure of questions in the comparison data, DR respondents who reported that they
misused "any other" drug in the category (e.g., any other prescription pain reliever) in the past
12 months could specify past year misuse of up to five individual drugs. If a respondent reported
past year use of a specific drug (e.g., the generic pain reliever hydrocodone), did not report
misusing the drug in the past year, but then reported it in the "OTHER, Specify" data, the
response in the edited variable for past year misuse was logically inferred to be "yes"; no editing
needed to be done for the variable pertaining to any use in the past year. If the respondent
reported misuse of a particular drug in the "OTHER, Specify" data but did not report using it in
the past year (and therefore was not asked about past year misuse of the drug), both the variable
for any past year use and the variable for past year misuse of that drug were assigned codes to
indicate that the respondent used and misused that drug in the past year.
3.3.4.4

OTC Misuse

One way that persons can misuse prescription drugs is by taking them without having
their own prescription. Because OTC drugs by definition are available without a prescription,
respondents in both the DR and the comparison data interviews were instructed not to include
OTCs when answering the prescription drug questions. For the comparison data, respondents
who specified that they misused OTCs were logically inferred never to have misused any of the
prescription drugs in the overall category (e.g., pain relievers) if they reported never misusing
any of the specific prescription drugs in the gate questions and the only other "prescription"
drugs they reported misusing in their lifetime were OTCs.
A similar principle was applied to the editing of the DR prescription drug data, except
that these edits focused on the misuse of prescription drugs in the past year. Specifically, DR
respondents were logically inferred not to have misused any of the prescription drugs in that
category in the past year if they did not use or misuse any of the drugs in that category except for
"any other" drug, and the only other drugs they reported misusing in the past year were OTCs.
However, no editing was done to the screening question about any use of other drugs in that

category in the past year (which resulted in respondents being routed to the question about
misuse of any other drug in the category) because respondents could have used other
prescription drugs in the past year that they did not misuse.
3.3.4.5

Methamphetamine Use

Editing of the methamphetamine variables in the comparison data took into account the
placement of the methamphetamine questions in the core stimulants module. Specifically, the
CAI program for the comparison data required answers to questions about methamphetamine use
to be consistent with answers to related questions about misuse of stimulants in general. As noted
previously, for example, a consistency check was triggered if respondents reported more recent
use of methamphetamine than they reported for the most recent misuse of any prescription
stimulant. In keeping with the general editing principles for the comparison data, the editing
procedures took answers in these consistency checks into account when creating the edited
methamphetamine and general stimulant variables. Furthermore, the editing procedures for the
comparison data required misuse of any stimulant always to be as recent as or more recent than
the last use of methamphetamine.
Since 2005, questions about methamphetamine use have been included in the
supplemental special drugs module for respondents who did not previously report
methamphetamine use in the core stimulants module. Because methamphetamine in recent years
has typically been manufactured illegally rather than through the legitimate pharmaceutical
industry, methamphetamine users may fail to report their use when questions about the drug are
asked in the context of questions about misuse of stimulants that are (or have been) available by
prescription in the United States. Data from these methamphetamine questions in the special
drugs module were used to create "core-plus-noncore" (CPN) measures of lifetime and most
recent use of methamphetamine in the comparison data. For example, if respondents in the
comparison data did not report methamphetamine use in the core stimulants module because they
did not think of it as a prescription drug but they reported use in the special drugs module, their
reports for their most recent use of methamphetamine in the special drugs module were
incorporated into the CPN variable for most recent use. In addition, if these respondents who did
not think of methamphetamine as a prescription drug reported more recent use of
methamphetamine in the special drugs module than they reported for their most recent misuse of
any stimulant, the edited CPN variable for most recent stimulant misuse reflected the special
drugs data for methamphetamine.
Editing of the DR data for lifetime and most recent use of methamphetamine followed the
general principles described in Section 3.3.3. Because the methamphetamine use questions in the
DR were placed in a module separate from questions about misuse of prescription stimulants, the
edited data for use or most recent use of methamphetamine were not required to be consistent
with data from the core stimulants module. For example, DR respondents could report lifetime
use of methamphetamine without reporting misuse of prescription stimulants in their lifetime;
these responses were not considered to be inconsistent.

3.3.4.6

Prevalence of Prescription Drug Misuse

Editing of the prescription drug variables in the comparison data generally followed the
overall principles described in Section 3.3.3. Editing of these variables also included the special
situations for "OTHER, Specify" data and reports of misuse of only OTC drugs that were
described previously in Sections 3.3.4.3 and 3.3.4.4.
In the DR, respondents first were asked to report any use of a series of prescription drugs
in that psychotherapeutic category (e.g., pain relievers) in the past 12 months (subsequently
referred to in this section as "screener" questions). Respondents who did not report past year use
of any prescription drug in that category (including use of "any other" prescription drug) were
asked whether they ever used any prescription drug in that category. Respondents who endorsed
use of one or more specific prescription drugs in the past 12 months in the screener questions
were asked about past year misuse of the prescription drugs that they reported using in that
period. If respondents reported misuse of any prescription drugs in a given category in the past
12 months, they were asked whether they misused any prescription drugs in that category in the
past 30 days. Thus, unlike the 12-month questions, misuse in the past 30 days applied only to the
broad prescription drug category rather than to specific prescription drugs. If respondents used
prescription drugs in a given category in the past 12 months but they did not report misuse, they
were asked about lifetime misuse of any prescription drugs in that category. Similarly,
respondents who reported lifetime but not past year use of any prescription drugs in that category
were asked about lifetime misuse. Thus, as for misuse in the past 30 days, lifetime misuse
applied only to the broad prescription drug category.
Consistent with the general editing principles described in Section 3.3.3, an important
component of editing the prescription drug variables in the DR for determining the prevalence of
use or misuse involved assignment of codes to indicate when respondents were not asked
questions that were not applicable. For example, if respondents did not report use of a particular
drug in the past 12 months, then the corresponding edited variables for misuse of that drug in the
past 12 months were assigned codes to indicate that the questions did not apply.
As an exception to the general principle of retaining missing values when respondents
answered a question governing a skip pattern as "don't know" (DK) or "refused" (REF),
DR respondents who had responses of DK or REF in their screener data for past year use of
specific prescription drugs and reported no past year use of other drugs in the screener could
answer the question about lifetime use of any prescription drugs in the category as "no." In this
situation, the report of no lifetime use of any prescription drug in the category took precedence
over the responses of DK or REF in editing the DR prescription drug variables. Similarly, if
respondents answered one or more questions about past year misuse of specific prescription
drugs as DK or REF and answered questions about past year misuse of other prescription drugs
as "no" (or were skipped out of the past year misuse questions because they did not report any
past year use of these drugs), they were asked whether they ever misused any prescription drug
in that category in their lifetime. Again, if these respondents answered this lifetime misuse
question as "no," this report overruled the responses of DK or REF in editing the past year
misuse variables.

Because of the structure of the prescription drug questions in the DR, respondents were
not asked a specific question for their most recent misuse of any prescription drugs in that
category. Rather, variables for the most recent misuse of prescription pain relievers,
tranquilizers, stimulants, and sedatives were created from respondents' answers to questions
about the misuse of any prescription drug in the category in the past 30 days, misuse of specific
prescription drugs in a given category in the past 12 months, and lifetime misuse of any
prescription drug in the category. The following general principles were applied in creating the
variables for the most recent use of any prescription drugs in a given category in the DR data:
•

Respondents who reported misuse of prescription drugs7 in the past 30 days were
classified as having last misused prescription drugs in the past 30 days.

•

3.3.4.7

Initiation of Use of Illicit Drugs Other Than Prescription Drugs

For marijuana through inhalants in the comparison data and for marijuana through
methamphetamine in the DR, lifetime users were asked to report the age when they first used the
drug, and respondents who first used within 1 year of their current age were asked to report the
year and month when they first used. The age, year, and month data were used to establish
whether the respondent initiated use in the past 12 months. Specifically, respondents were
defined as a past year initiate if any of the following occurred:
•

They first used the drug at their current age.

•

They first used at the age that was 1 year prior to their current age, but they first used
in the current calendar year (e.g., 2013 for DR interviews).

•

They first used at the age that was 1 year prior to their current age and in the year
prior to the current calendar year, but their month of first use was unambiguously
within 12 months of the interview date (e.g., first use in October, November, or
December 2012 for DR interviews in September 2013).

In this text, "prescription drugs" refers to any prescription drugs in a given category (e.g., any prescription
pain reliever).

Because all lifetime users were asked to report the age at first use (AFU), a special AFU
did not need to be created for past year initiates. At a minimum, all past year initiates would have
an AFU.
As noted in Section 3.3.3, however, inconsistencies could exist between related variables
in the DR or comparison data that would otherwise have been handled in the editing procedures
for the main study. In particular, some respondents who were classified as past year initiates
based on their age, year, or month of first use could have reported that they last used a drug
"more than 12 months ago." In these situations, neither the data for the most recent use or
initiation were edited to be consistent with one another. However, if respondents had missing
data for initiation (e.g., the AFU was answered as DK or REF) but they reported last using the
drug more than 12 months ago, the recency variable took precedence, and the respondents were
classified as not being past year initiates.
Respondents were defined as being unknown for past year initiation if they last used a
drug in the past 12 months (including use in the past 30 days) or at some point in their lifetime
and any of the following occurred:
•

Their age at first use was answered as DK or REF.

•

They first used at the age that was 1 year prior to their current age, and their year of
first use was answered as DK or REF.

•

They first used at the age that was 1 year prior to their current age, their year of first
use was in the previous calendar year, and their month of first use was answered as
DK or REF.

•

They first used at the age that was 1 year prior to their current age, their year of first
use was in the previous calendar year, and their month of first use was the same
month as the interview month (e.g., first use in September 2012 for a DR interview in
September 2013).

In the main survey, a full date of first use (DFU) is imputed for all lifetime users.
Therefore, if respondents reported first use in the same month in the calendar year that was
1 year prior to the interview, they could be imputed to be past year initiates or not to be past year
initiates depending on whether the imputed DFU was within 12 months of the interview date or
more than 12 months prior to the interview date. For this reason, the final initiation status for this
last group of respondents was handled through imputation.
3.3.4.8

Initiation of Prescription Drug Misuse

In the comparison data, respondents who reported lifetime misuse of prescription drugs
were asked about initiation of prescription drug misuse in the same manner as for the other
substances in the core modules. Therefore, the procedures that were described in Section 3.3.4.7
also applied to identification of past year initiates or determination of unknown initiation status
for misusers of prescription drugs in the comparison data.
In the QFT and DR, respondents were asked about initiation of misuse only for the
individual prescription drugs that they had misused in the past 12 months. However, a limitation

of the QFT questions for measuring past year initiation of misuse of any prescription drug in an
overall category (e.g., pain relievers) was that respondents who reported only past year initiation
for the prescription drugs that they misused in the past 12 months could have initiated misuse of
other drugs in the category more than 12 months ago. Consequently, it could be determined that
these respondents were past year initiates for the specific prescription drugs that they misused in
the past year, but past year initiation for the overall category could not be determined.
Therefore, the DR questionnaire was modified so that respondents who reported only past
year initiation of the drugs they misused in that period were asked a follow-up question to
determine whether they ever misused any drugs in that category more than 12 months prior to the
interview.8 There was no need to ask this follow-up question if respondents reported initiation
more than 12 months ago for any of the prescription drugs that they misused in the past year
because these respondents by definition were not past year initiates.
Creation of the edited variables for past year initiation of misuse of any pain reliever,
tranquilizer, stimulant, or sedative involved reverse coding of the answers for these follow-up
questions. Specifically, if DR respondents were routed to a given follow-up question and
answered it as "yes," then they were defined as not being past year initiates for the overall
category. Respondents who answered the question as "no" were defined as past year initiates for
the overall entire category. If respondents were skipped out of the follow-up question because
they reported initiation of misuse of some prescription drugs more than 12 months ago, the
edited variable was assigned a code of 4 (Not a past year initiate LOGICALLY ASSIGNED).
If respondents answered the follow-up question as DK or REF, the edited variable
retained a missing value. Their status as a past year initiate (or not) was resolved through
imputation (see Section 3.4). Respondents also could have missing data for the edited past year
initiation variable for a given prescription drug category for the following reasons:
•

Respondents reported misuse of a given category of prescription drugs at some point
in their lifetime, but their most recent period of misuse was unknown.

•

Respondents reported initiation of misuse of some prescription in the past 12 months,
but they reported initiation of misuse of the remaining drugs in the same month as the
interview month but in the previous calendar year (e.g., September 2012 for
respondents who were interviewed in September 2013).

For this second group of respondents, the CAI logic did not consider initiation in the
same month 1 year ago as being a potential indication of past year initiation. Consequently, these
respondents were not asked the follow-up questions about first misuse of any prescription drugs
in the category more than 12 months ago. These respondents' status as a past year initiate (or not)
was resolved through imputation.
In addition, if the edited variable indicated that the respondent was a past year initiate, the
initiation data were checked for the individual drugs that respondents misused in the past
12 months to determine an AFU among past year initiates. Generally, this was the minimum
8

Respondents also were asked the follow-up question if the sum of the reports of past year initiation plus
missing data for initiation equaled the number of specific drugs that they misused in the past year.

AFU (i.e., either current age or current age minus 1 year) among the AFUs for the individual
prescription drugs that respondents misused. If some AFU questions were answered as DK or
REF, it also was possible to infer logically that first misuse happened at the age that was 1 year
prior to a respondent's current age if at least one AFU was reported to be at the respondent's
current age minus 1 year. If the respondent was a past year initiate of misuse but all AFUs had
missing data or there was a combination of missing AFU data and remaining AFUs at the
respondent's current age, then the edited AFU for the overall prescription drug category retained
a missing value.
3.3.4.9

Needle Use

Editing of the needle use data in the DR and comparison samples principally involved
assignment of the appropriate codes to indicate when respondents were not asked questions that
did not apply. For example, respondents were not asked the needle use questions for a given drug
(e.g., cocaine) if they reported in the corresponding core module that they never used the drug.
Respondents also were not asked the follow-up questions in the special drugs module about most
recent use of a drug with a needle if they used the drug in their lifetime but never used a needle
to inject it.
In addition, "OTHER, Specify" data on the use of other drugs with a needle were used to
edit needle use data within the special drugs module. For example, if respondents did not report
using cocaine with a needle but they specified it as some "other" drug they used with a needle,
the edits inferred that these respondents used cocaine with a needle at some point in their
lifetime.
Consistent with editing in the core modules (and with general principles of editing
described previously), however, data on needle use from the special drugs module were not used
in editing drug use data from the corresponding core module. For example, if respondents
reported more recent use of cocaine with a needle in the special drugs module compared with
their reports of most recent use of cocaine (including any reports of crack cocaine), the editing
procedures for both the DR and comparison data did not resolve this inconsistency.
As noted previously, the needle use questions for stimulants in the DR were moved from
the special drugs module to the core stimulants module. In addition, the questions about use of
stimulants with a needle applied to stimulants that respondents misused in the past 12 months.
Even if the editing procedures allowed editing of core data based on data in the special drugs
module, reports of lifetime use of prescription stimulants with a needle in the "OTHER, Specify"
data for special drugs could not be used to infer past year use of stimulants with a needle or to
infer past year misuse of specific stimulants in the core stimulants module.
3.3.4.10

Methamphetamine and Prescription Stimulant Dependence or Abuse

Consequently, DR respondents who reported methamphetamine use in the past year but who did
not report past year misuse of prescription stimulants were asked dependence and abuse
questions for methamphetamine but were not asked corresponding questions for stimulants.
DR respondents who reported past year use of methamphetamine and past year misuse of
prescription stimulants were asked both sets of dependence and abuse questions. For these
respondents, no editing was done to the methamphetamine dependence or abuse variables based
on respondents' answers to questions about corresponding symptoms of dependence or abuse for
prescription stimulants. Similarly, no editing was done to the stimulant dependence or abuse
variables based on respondents' answers to questions about corresponding symptoms of
dependence or abuse for methamphetamine.
3.3.4.11

Interviewer-Administered versus Self-Administered Data

The basic content of the DR variables for marital status, employment status, health
insurance, and income underwent little or no change relative to the variables in the comparison
data, except that they were self-administered instead of being interviewer-administered.
Consequently, little or no change to the editing procedures for these variables in the DR were
required relative to the procedures for editing these variables in the comparison data. Editing of
these variables in all three datasets principally involved assignment of codes to indicate when it
could be determined unambiguously that respondents were not asked questions that did not
apply.

3.4

Imputation Procedures

3.4.1

Overview of Imputation Procedures

This section describes the imputation procedures that were implemented for the 2013 DR
data and the two comparison datasets―the 2012 main study data and the 2013 quarters 3 and 4
main study data. The advantages of performing imputation include the following: (1) reducing
bias due to differential nonresponse, (2) allowing all cases to be used for analysis, and
(3) improving the quality of data at the subdomain level. The small DR sample sizes and sparse
donor pools made it difficult to implement the standard NSDUH imputation methods. Because
the comparison of the DR data with the main study data was performed at a fairly aggregate
level, a simple mean imputation procedure satisfies the needs of the DR and could be
implemented within the relatively shorter data processing period for the DR. The two main study
comparison datasets―all four quarters from 2012 and quarters 3 and 4 from 2013―were
imputed using the same approach. One of the simplest methods of imputing for missing data is to
replace each missing value with the weighted mean of the observed values for a variable within a
class of respondents containing the respondent with the missing value. This method provides an
unbiased estimate of the overall variable mean either if the probability of the value being missing
is the same for every respondent in a class or if values within a class are not related to their
probabilities of being missing. If neither of these conditions holds, the estimated variable mean
after imputation is biased, but the bias is likely to be less than if no imputation had taken place,
which is equivalent to treating the entire sample as a single imputation class.

3.4.2

Imputation Methodology

Variables that were imputed include demographics, health insurance, income, recency of
drug use, an indicator of past year initiation of drug use, and age of first drug use for past year
initiates. The noncore variables associated with drug abuse were not imputed.9 Table 3.1 lists the
variables that were imputed for each of the three sets of data. As was done in the main study,
imputation indicators were created for each imputed variable. For the drug use variables, three
variables indicating lifetime use, past year use, and past month use were created from the
imputed recency of use variables. In addition to misuse, the DR instrument asked about any use
of prescription drugs. These variables were not imputed for this analysis. Questions about
lifetime and past month use of OxyContin® were not included in the DR instrument; therefore,
only the past year indicator variable for OxyContin® misuse was imputed for the DR data. The
DR instrument contained separate modules for methamphetamine and prescription stimulants.
Therefore, an additional recency of misuse of stimulants excluding methamphetamine was
imputed for the DR only. For the 2012 and 2013 quarters 3 and 4 comparison data, the CPN
measures for methamphetamine and misuse of stimulants were created to compare with the
combined stimulants and methamphetamine variables in the DR.
For categorical variables (including both nominal and ordinal), the weighted percentage
for each variable level within an imputation class was used to impute the missing values.
Imputation classes were based, where possible, on categorical age (12 to 17 years old, 18 to 25
years old, and 26 years old or older), gender, and four-level race (white, black, Hispanic, and
other). For the race variable imputation, only age group and gender were used to create
imputation classes. For the continuous variable WELMOS—number of months on welfare—the
weighted mean was computed within an imputation class, then used to impute the missing
values. Weighted means were computed using PROC DESCRIPT from SUDAAN®
(RTI International, 2008), and weighted percentages were computed using PROC CROSSTAB.
As an example, assume that among white females aged 26 or older the marital status variable has
a complete case weighted distribution as follows: married (65 percent), widowed (10 percent),
divorced (15 percent), and never married (10 percent). If 20 cases within this imputation class
have missing values, then 13 cases would be imputed as married, 2 cases as widowed, 3 cases as
divorced, and 2 cases as never been married. Rounding was used when the percentages did not
result in exact numbers of cases and when there were fewer records with missing values than
there were levels of the imputed value. For example, an imputation class for the four-level
recency variable may have had only two records requiring imputation. In these cases, the
distribution of imputed cases may have looked very different from the distribution of complete
cases. However, the rounding algorithm was such that the distribution of imputed values would
match the weighted distribution of complete values in expectation.

Variables that regularly undergo imputation, but did not for the DR include the following: roster
variables; roster pair variables; Hispanic group and immigrant status; personal income variables; "old method"
insurance variables; daily cigarette use, cigar, pipe, chewing tobacco, and snuff use variables; core-only stimulants
and methamphetamine use variables; 12-month and 30-day frequency of drug use variables; age at first drug use
variables; and nicotine dependence variables.

Table 3.1 Imputed Variables
Race
Hispanic Indicator
Marital Status

Demographic Variables
Education
Employment Status

Income Variables
Food Stamps
Welfare Payments
Welfare Services
Number of Months on Welfare
Health Insurance Variables
Medicaid/CHIP (Children's Health Insurance Program)
Private Health Insurance
Medicare
Other Health Insurance
CHAMPUS (Civilian Health and Medical Program of
the Uniformed Services)
Drug Use Variables
Cigarette Use
Inhalant Usea
Smokeless Tobacco Use
Marijuana Usea
Alcohol Use
Core plus Noncore Stimulant Misusea
Binge Alcohol Use (Past Month Only)
Core plus Noncore Stimulant Misuse, Excluding
Cocaine Usea
Methamphetamine Use (DR Only) a
a
Crack Use
Core plus Noncore Methamphetamine Misusea
a
Hallucinogen Use
Pain Reliever Misusea
a
LSD Use (Lysergic Acid Diethylamide)
OxyContin® Misuse (DR: Past Year Only) a
a
PCP Use (Phencyclidine)
Sedative Misusea
a
Ecstasy Use
Tranquilizer Misusea
a
Heroin Use
DR = Dress Rehearsal.
a
Imputed values indicating past year initiation and age at first use for past year initiates were also calculated. For
OxyContin® misuse, these measures were only created for the 2012 and 2013 comparison files.
Family Income
Wages (2012/2013 Comparison Only)
Social Security
Supplemental Security

For the age at first use (AFU) variables, imputations were performed only for past year
initiates. Therefore, all nonrespondents had to initiate at either their current age or at age – 1. If
they initiated prior to this point, then they would not have been classified as a past year initiate.
For each past year initiate with a missing AFU, the year preceding the interview date was
divided into two parts based on the birth date of the individual. The proportion of the year prior
to the interview in which the individual was at his or her current age and at his or her current
age – 1 was then calculated. Each individual requiring imputation was next assigned a random
number between zero and one from a uniform distribution, and this value was used to determine
the final imputed AFU. For example, suppose a past year initiate of marijuana with a missing age
of first marijuana use was 25 years old at the time of the interview and he or she was 25 years
old for 292 days (80 percent) in the year prior to the interview and 24 years old for 73 days
(20 percent) in the year prior to the interview. For this individual, there was an 80 percent chance
that the imputed age of first marijuana use would equal his or her current age (25 years old) and
a 20 percent chance that the imputed age of first marijuana use would equal age – 1 (24 years
old).
Imputation was occasionally restricted to a few categories when partial information about
the nonrespondent was known or in order to maintain consistency with other variables.
For example, when imputing employment status, if the nonrespondent was known to be

employed, but the level of employment (full time or part time) was not known, the weighted
percentages were calculated among employed respondents in each imputation class, and
imputation was restricted to full- or part-time employment.
In a few cases, the imputation class contained only nonrespondents. When this happened,
imputation classes were collapsed by race, then by gender, then by age until at least one
respondent was in the imputation class. For example, Exhibit 3.1 shows the imputation classes
for the 12- to 17-year-old age category. If the nonrespondent was a 15-year-old, Hispanic, and
female, and no respondents were in the imputation class for 12- to 17-year-old, Hispanic
females, that class would be merged with the class containing 12- to 17-year-old females of other
races. Collapsing would continue up the hierarchy until at least one respondent was in the
imputation class. Continuing the example above, it may have been necessary to collapse all races
or both genders. Note that if collapsing was necessary, care was taken to collapse as few classes
as possible. As shown in Exhibit 3.1, if collapsing of the race categories was only necessary
among females, parallel collapsing was not done among males. Similarly, if collapsing was only
necessary among 12- to 17-year-olds, no collapsing was done within the other age categories
(see Exhibit 3.2).
Exhibit 3.1 Collapsing Imputation Classes: Race

Exhibit 3.2 Collapsing Imputation Classes: Race and Gender

3.5

Weighting Procedures

3.5.1

Overview of Weighting Procedures

Estimates and measures of data quality from the DR sample were compared with those
from the 2013 main study during the same quarters (2013 quarters 3 and 4) and from the full year
for the 2012 main study. Analysis weights for those three samples needed to be developed for the
DR analysis. This section discusses the methods used to develop sample weights for the DR
analysis.
For some research questions, DR respondents were compared with the 2013 quarters 3
and 4 and the 2012 NSDUH respondents. To increase the efficiency of the comparisons by
removing the impact of differences between the demographic characteristics of the three samples
caused by random sampling, then exacerbated by nonresponse, nonresponse-adjusted weights
were calibrated for the DR sample and the 2013 quarters 3 and 4 main study sample to
distributions of demographic variables from the 2012 sample. Instead of the full process (Chen et
al., 2014) that was used in developing 12-month analysis weights, where five adjustment steps
were implemented, a shortened process was used similar to producing weights for the 6-month
detailed tables. That is, the design weights were computed for both the DR sample and the 2013
quarters 3 and 4 main sample in a manner consistent with 2012 NSDUH weighting procedures.
The design weights were then adjusted for nonresponse at the dwelling unit and person level,
followed by a poststratification adjustment where nonresponse-adjusted weights were further
poststratified to the sum of the analysis weights from the 2012 NSDUH sample for selected
demographic domains.

The weight distributions were calculated for the final analysis weights of the 2012
comparison data, DR data, and 2013 quarters 3 and 4 main study data. The weight distribution
results are presented in Table 3.2 in Section 3.5.3.
3.5.2

Weighting Procedures

This section discusses in detail the procedures used to develop the analysis weights for
the three samples and summarizes the distribution of the DR analysis weights.
3.5.2.1

2012 NSDUH Sample Weights

The analysis weights (ANALWT) for the 2012 NSDUH sample had 15 weight
components, and among them 5 were adjustment factors at both the dwelling and person levels
(Chen et al., 2014). The generalized exponential model (GEM) (Folsom & Singh, 2000) was
used for the nonresponse and poststratification adjustments within nine model groups
corresponding to nine census divisions. ANALWT is the product of all 15 weight components.
After removing respondents from Hawaii and Alaska, analysis weights for the remaining
respondents in the 2012 NSDUH were used for the DR analyses. The domain-level sums of the
ANALWT for these retained respondents were used as control totals in the poststratification for
the DR sample and the 2013 quarters 3 and 4 main study sample, as discussed in the following
section.
3.5.2.2

2013 Quarters 3 and 4 Main Study Sample Weights

Design-based weights were computed for the 2013 quarters 3 and 4 main study sample in
a manner consistent with standard NSDUH weighting procedures. To facilitate timely
completion of the DR analyses, quarter 4 screenings and interviews completed after December 1,
2013, were considered nonrespondents. After December 1, 2013, an additional 3,222 screenings
and 715 interviews were completed that would have been included in the 2013 quarters 3 and 4
main study comparison data had the December 1, 2013, cutoff date not been implemented.
The nonresponse adjustments at both the dwelling unit level (DUNR) and person level (PRNR)
for the 2013 quarters 3 and 4 main study sample were similar to those used to develop the regular
6-month analysis weights. However, the person-level poststratification (PRPS) for the 2013
quarters 3 and 4 main study sample was different from the regular 6-month analysis weights,
where the nonresponse-adjusted weights were adjusted to the census population estimates.
For the DR analyses, the person-level poststratification adjusted the weights to match ANALWT
sums for eligible respondents from the 2012 NSDUH sample. GEM was used to implement all
three adjustment steps.
The final analysis weights for the 2013 quarters 3 and 4 main study sample were the
product of various design weights and three adjustment factors. The various design weights were
as follows:
•

inverse probability of selecting census tracts;

•

inverse probability of selecting segments;

•

quarter segment weight adjustment;
51

•

subsegmentation inflation adjustment;

•

inverse probability of selecting dwelling units;

•

added/subsampled dwelling unit adjustment;

•

dwelling unit sample release adjustment;

•

dwelling unit-level nonresponse adjustment;

•

inverse probability of selecting a person from a dwelling unit;

•

person-level nonresponse adjustment; and

•

person-level poststratification adjustment.

The three adjustment factors were as follows:
•

Dwelling Unit-Level Nonresponse Adjustment (DUNR). One model was used to
account for the failure to obtain screening interviews from eligible dwelling units.
The proposed variables in the model are listed below, and they were all kept in the
final model.
– State,
– quarter,
– population density (metropolitan statistical area [MSA], ≥ 1 million; MSA, < 1
million; non-MSA, urban; non-MSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO) (> 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB) (> 50 percent; 10 to 50
percent; < 10 percent),
– percent of Hispanics in a segment (CH) (> 50 percent; 10 to 50 percent; < 10
percent),
– segment combined median rent and housing value (CV) (1st quintile; 2nd
quintile; 3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB, and
– CV * CH.

•

– quarter,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban; nonMSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO) (> 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB) (> 50 percent; 10 to 50
percent; < 10 percent),
– percent of Hispanics in a segment (CH) (> 50 percent; 10 to 50 percent; < 10
percent),
– segment combined median rent and housing value (CV) (1st quintile; 2nd
quintile; 3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB,
– CV * CH,
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender,
– Hispanicity * gender,
– age group * Race3 * Hispanicity,
– age group * Race3 * gender,
– age group * Hispanicity * gender, and
– Race3 * Hispanicity * gender.
•

Person-Level Poststratification Adjustment (PRPS). The respondents in the 2013
quarters 3 and 4 main sample from Hawaii and Alaska and interviews completed with
the Spanish-language questionnaire were removed before the PRPS. One model was
used to force the weights of the 2013 quarters 3 and 4 main study sample to sum up to

the ANALWT totals for eligible respondents in the 2012 NSDUH by the following
proposed demographic domains (all proposed variables were kept in the final model):
– State,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 to 64; 65 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender,
– Hispanicity * gender,
– age group * Race3 * Hispanicity,
– age group * Race3 * gender,
– age group * Hispanicity * gender, and
– Race3 * Hispanicity * gender.
3.5.2.3

2013 DR Sample Weights

Design-based weights for the 2013 quarters 3 and 4 DR sample were computed in a
manner consistent with standard NSDUH weighting procedures. The three adjustment steps (i.e.,
DUNR, PRNR, and PRPS) were implemented in a similar fashion as for the 2013 quarters 3 and
4 main study sample weights using GEM. The differences were that fewer variables in the GEM
models were used to develop DR sample weights because of the relatively small DR sample.
The final analysis weights for the 2013 quarters 3 and 4 DR sample were the product of
various design weights and three adjustment factors. The various design weights were as follows:
•

inverse probability of selecting DR State sampling regions (SSRs);

•

inverse probability of selecting census tracts;

•

inverse probability of selecting segments;

•

quarter segment weight adjustment;

•

subsegmentation inflation adjustment;

•

inverse probability of selecting dwelling units;

•

added or subsampled dwelling unit adjustment;

•

dwelling unit sample release adjustment;

•

dwelling unit-level nonresponse adjustment;

•

inverse probability of selecting a person from a dwelling unit;

•

person-level nonresponse adjustment; and

•

person-level poststratification adjustment.

The three adjustment factors were as follows:
•

Dwelling Unit-Level Nonresponse Adjustment (DUNR). One model was used to
account for the failure to obtain screening interviews from eligible dwelling units.
The variables in the model are listed below, and some two-way interactions of
segment-level variables (CO, CB, CH, and CV) were collapsed in order to get a
convergent model:
– State,
– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban; nonMSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO) (> 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB) (> 50 percent; 10 to 50
percent; < 10 percent),
– percent of Hispanics in a segment (CH) (> 50 percent; 10 to 50 percent; < 10
percent),
– segment combined median rent and housing value (CV) (1st quintile; 2nd
quintile; 3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB, and
– CV * CH.

•

Person-Level Nonresponse Adjustment (PRNR). One model was used to adjust
person-level nonresponse, and the proposed variables in the model are listed as
follows (they were all kept in the final model):
– State,
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),

– population density (MSA, ≥ 1 million; MSA, < 1 million; non-MSA, urban; nonMSA, rural),
– group quarters (college dorm; other group quarters; non-group quarters),
– percent of owner-occupied dwelling units in a segment (CO) (> 50 percent; 10 to
50 percent; < 10 percent),
– percent of blacks or African Americans in a segment (CB) (> 50 percent; 10 to 50
percent; < 10 percent),
– percent of Hispanics in a segment (CH) (> 50 percent; 10 to 50 percent; < 10
percent),
– segment combined median rent and housing value (CV) (1st quintile; 2nd
quintile; 3rd quintile; 4th quintile; 5th quintile),
– CO * CB,
– CO * CH,
– CO * CV,
– CV * CB,
– CV * CH,
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,
– Race3 * gender, and
– Hispanicity * gender.
•

Person-Level Poststratification Adjustment (PRPS). One model was used to force the
weights of the 2013 quarters 3 and 4 DR sample to sum up to ANALWT totals for
eligible respondents in the 2012 NSDUH by the following proposed demographic
domains (all variables were kept in the final model):
– age group (12 to 17; 18 to 25; 26 to 34; 35 to 49; 50 to 64; 65 or older),
– race (white; black; Native American; Asian; multiple races),
– Hispanicity (Hispanic; non-Hispanic),
– gender (male; female),
– age group * Race3 (white; black; others),
– age group * Hispanicity,
– age group * gender,
– Race3 * Hispanicity,

– Race3 * gender, and
– Hispanicity * gender.
3.5.3

Distribution of DR Analysis Weights

The distribution of analysis weights for the 2012 NSDUH sample, 2013 quarters 3 and 4
DR sample, and 2013 quarters 3 and 4 main study sample are summarized in Table 3.2.
Table 3.2 Weight Distribution of Dress Rehearsal Analysis Weights

Statistics
100% Maximum
99%
95%
90%
75% Quarter 3
50% Median
25% Quarter 1
10%
5%
1%
0% Minimum
n
Mean
Sum of Weights
Unequal Weighting Effect (UWE)1
1

2012 NSDUH
Sample Weights
133,926
29,474
15,629
10,182
4,181
1,590
749
342
203
78
1
66,542
3,883
258,349,358
3.5016

2013 Quarters 3
and 4 Dress
Rehearsal Sample
Weights
1,648,168
765,303
478,051
313,542
150,227
66,812
22,715
12,390
7,758
3,384
2,310
2,089
123,671
258,349,358
2.6950

2013 Quarters 3 and 4
Main Study Sample
Weights
200,068
54,299
32,194
21,434
8,966
3,365
1,838
905
519
241
42
32,162
8,033
258,349,358
3.0258

UWE measures the variation in weights.

3.5.4

Creation of Variance Estimation Strata and Replicates

The nature of the stratified, clustered sampling design of the NSDUH main study, DR,
and QFT samples requires that the design structure be taken into consideration when computing
variances of survey estimates. Because the DR and QFT samples are assumed to be independent,
two sets of key nesting variables (pseudo-strata and replicates) were utilized in the analyses.
One set captured the design structure of the DR, and the other captured the design structure of
the QFT. Both sets of nesting variables were mapped to the main study comparison data to allow
for comparisons with the DR and QFT samples. The development of the QFT nesting variables is
further discussed in the 2012 QFT final report (Currivan et al., 2013). This section outlines the
creation of the DR nesting variables.
To allow for comparisons between the DR and main study samples, a common set of
stratification and clustering variables were defined. Because State sampling regions (SSRs) serve
as strata for the main study samples and as primary sampling units (PSUs) for the DR sample,

there was no direct way of capturing the covariance between the samples and using the entire
main study sample. Instead, the approach used for the 1999 paper-and-pencil interviewing
(PAPI) and CAI mode of analysis was followed in developing a design structure that could be
used to simultaneously analyze all three samples (Gfroerer, Eyerman, & Chromy, 2002).
This methodology was also used to create the nesting variables for the QFT. Steps in the process
were as follows:
•

Within the five DR sampling strata (high Spanish certainty stratum and four
noncertainty census region strata), variance pseudo-strata were formed by assigning
two sequential DR-selected SSRs to the same variance pseudo-strata on the sorted
sampling frame. Each sampled SSR was then assigned to a replicate (1 or 2).
However, there were three DR SSRs per variance pseudo-strata for four randomly
selected pseudo-strata. This was necessary because an odd number of DR SSRs were
selected in four of the strata. Within these four pseudo-strata, the third SSR was
randomly assigned to either replicate 1 or replicate 2. This led to a total of 98 DR
variance pseudo-strata, with two replicates per pseudo-strata.

•

The main study SSRs that comprised the DR certainty stratum received the same
pseudo-strata assignments as the DR certainty SSRs. For the noncertainty SSRs, the
main study SSRs not selected for the DR were assigned to DR sampling pseudo-strata
sequentially on the sorted SSR frame, in accordance with the assignments of selected
DR SSRs. These assignments kept the number of SSRs per pseudo-strata as equal as
possible given the distribution of DR-sampled SSRs within the sorted SSR frame.
For both certainty and noncertainty SSRs, the original replicate assignments of either
replicate 1 or replicate 2 were maintained for the main study. A further discussion of
the assignment of main study replicates can be found in the 2012 sample design
report (Morton, Martin, Shook-Sa, Chromy, & Hirsch, 2013).

Although this approach to design structure variables does not fit the main study perfectly, it does
capture the total variance and allows for taking advantage of any covariance induced by the
overlapping SSRs between the DR and main study samples.

3.6

Data File Preparation

Three data files were prepared for the DR analysis. In order to evaluate the DR results
and estimates, two comparison data files from all four quarters of 2012 and quarters 3 and 4 of
2013 were created from the main study cases.
3.6.1

DR Data File

The DR data file was comprised of interviews conducted from September 1, 2013,
through October 31, 2013. No interviews in Alaska and Hawaii were conducted, and these data
underwent the normal data quality checks and telephone verification. Falsification was detected
after the initial DR data file was produced for data processing and analysis. Two such cases were
excluded from the final set of respondents, but the imputation and weighting were not redone
once these cases were dropped from the data file. The final DR analysis data file resulted in
2,087 respondents.

3.6.2

2012 Comparison Data File

The 2012 comparison data file was created from the 2012 main study analysis file.
The full set of respondents was reduced to 66,542 cases because of the exclusion of screenings
and interviews conducted in Alaska and Hawaii.
3.6.3

2013 Quarters 3 and 4 Comparison Data File

The 2013 comparison data file was created using most of the 2013 main study cases
fielded in quarters 3 and 4. As was done for the 2012 comparison file, screenings and interviews
conducted in Alaska and Hawaii were also excluded. In order to allow time for analysis under
the DR schedule, the 2013 comparison file only included cases with a completed date prior to
December 2, 2013. Because this time frame was prior to completing verification on the full 2013
main study sample, some decisions were made to exclude cases undergoing field verifications at
the time, based on the following criteria:
•

Cases completed by quarter 3 or 4 FIs found to have been falsified as of December 1,
2013. In addition to cases that were determined to have some form of falsification,
cases completed by these same FIs were dropped whenever it could not be
determined whether the interview was actually completed or whether informed
consent was completed. This second set of cases usually resulted from being unable
to contact the respondent.

•

Quarter 4 cases that were worked by FIs whose work was still being field verified as
of December 1, 2013.

•

Quarter 3 interviews for FIs whose work was still being field verified as of December
1, 2013. If falsification of quarter 4 cases was found, previous 2013 work completed
by these FIs needed to be field verified.

Interviews scheduled for telephone verification that were not finalized by close of
business on December 2, 2013, and met any of the three exclusion criteria above were not
included in the 2013 quarters 3 and 4 comparison data file.
Additional falsification was detected after the initial 2013 comparison data file was
produced for data processing and analysis. A total of 48 cases were excluded from the final set of
respondents, but the imputation and weighting were not redone once these cases were dropped
from the data file. The resulting 2013 quarters 3 and 4 comparison data file contained 32,162
interviews (see Table 3.3).
Table 3.3 Data Files Created for the 2013 Dress Rehearsal Analyses
Data File
Dress Rehearsal (DR)
2012 Comparison
2013 Comparison

Data Collection Period
9/1/2013 – 10/31/2013
1/1/2012 – 12/31/2012
7/1/2013 – 12/1/2013

Number of Respondents
2,087
66,542
32,162

3.7

Data Analysis Issues

3.7.1

Primary Analytic Goals

The QFT provided evidence on the potential effects of changes to the protocol planned
for the 2015 partial redesign on NSDUH estimates. The primary goal of the DR was to provide
additional evidence to support the QFT results, and especially whether these results were
generalizable to a data collection effort that includes Spanish-language interviews. The main
focus of the statistical analysis is the measurement of how the collective set of protocol changes
could affect key NSDUH estimates—overall and by the three major age groups—when the new
protocol is implemented in 2015. The DR sample size was not large enough to permit
quantitative assessments of the impact of individual changes in the protocol because such
analyses would require dedicated samples for assessing each change. To carry out such a design
to estimate the effects of each protocol change would be prohibitively costly and infeasible.
Also, the resources needed to carry out such extensive testing would have risked having an
impact on the main 2013 survey's estimates by affecting the availability of FIs to work on the
main study.
3.7.2

Comparison with Current NSDUH Data

Most of the analyses in this report compare outcomes and estimates from the combined
QFT and DR data from English-language interviews with non-Hispanic respondents with
outcomes and estimates from the 2012 main study and 2013 quarters 3 and 4 main study for the
same subgroups. These comparisons are limited to English-language interviews, given that
Spanish-language interviews were not conducted in the QFT. In addition, comparisons are
limited to non-Hispanics because the absence of a Spanish-language protocol could have affected
the likelihood of participation in the QFT for Hispanics relative to the DR. (This factor would
have any impact on the likelihood of response for non-Hispanics.) Hispanics who chose to
participate in the QFT despite the lack of a Spanish-language instrument cannot be considered
comparable with Hispanics who chose to participate in the DR in English, unless the
participation decision was completely independent from the availability of a Spanish-language
instrument. Including Hispanics in comparisons between the combined QFT and DR data and the
2012 and 2013 quarters 3 and 4 main study data from English-language interviews would require
an assumption that those who participated in the QFT would still have chosen to participate in
English had a Spanish-language version been available. Given that this is a questionable
assumption, English-language interview data from Hispanic respondents were removed from the
datasets to provide more comparable sets of respondents. For Spanish-language interviews, only
outcomes and estimates from the DR could be compared with Spanish-language interviews from
the 2012 main study and 2013 quarters 3 and 4 main study because the QFT interviews were
completed in English only. Where appropriate and useful, comparisons examining both QFT and
DR separately in relation to the two main study datasets were also presented.
Comparing the combined QFT and DR data or the DR data separately with data from
quarters 3 and 4 from 2013 allowed for estimating the effects of the overall protocol change over
approximately the same time period, with the DR being conducted during the last month of
quarter 3 and the first month of quarter 4 of the 2013 main study. An additional point of
comparison is provided by estimates from the 2012 main study. Use of the 2012 main study
60

provides additional sample with which to compare against the combined QFT and DR sample
and the DR sample. Rather than relying solely on comparisons with the 2013 quarters 3 and 4
sample, the survey designers felt that it would be informative to compare estimates from the
combined QFT and DR sample and the DR with the 2012 main study sample as well. The 2012
main study provides another data point with a larger sample size for these comparisons. This
provides assurance that differences in estimates between either the combined QFT and DR data
or the DR data alone and the 2013 quarters 3 and 4 sample are not unique to that comparison.
Comparisons using the pooled field test samples (QFT and DR) among non-Hispanics
who were interviewed in English with the 2012 and 2013 quarters 3 and 4 comparison samples
(also limited to non-Hispanics interviewed in English) were particularly useful for items that
were modified for the QFT and retained in the DR questionnaire. If the difference between
estimates for the QFT and DR samples are not statistically significant, these two datasets can be
pooled to create an estimate with a larger sample size for comparison with the main study
comparison datasets.
In addition to comparisons of estimates between the DR and 2012 and 2013 quarters 3
and 4 main study samples, two other types of analyses were completed in order to identify or rule
out potential confounders of comparisons between the DR and the 2013 quarters 3 and 4
samples.
3.7.2.1

Comparison of DR Data and 2013 Quarters 3 and 4 Data to Assess
"Seasonality" Effects on Estimates

In principle, the 2013 DR and comparison cases from quarters 3 and 4 of the 2013
NSDUH generally cover the same time period—late summer and early fall. Estimates from
quarter 3 in the 2013 NSDUH were compared with estimates from quarter 4 in the 2013 NSDUH
as a check for differences in estimates between the two quarters. Because the DR was conducted
in only 2 months out of the 6 months of quarters 3 and 4, there was concern that the particular
months chosen for the DR sample (September and October 2013) may not be representative of
all 6 months in the last half of 2013, particularly if there were differences in estimates between
quarters 3 and 4. If there were underlying changes in behavior taking place throughout the
6 months of quarters 3 and 4, the ideal design would involve collecting data using the redesigned
instrument throughout the same time period. However, because of resource constraints, the DR
sample could not be fielded in all of the 6 months of quarters 3 and 4 in 2013. If estimates in
quarter 3 were similar to those in quarter 4 and there was no underlying change in the behaviors
estimated by NSDUH, the time point at which the DR was fielded would be of less concern.
Given that the DR was conducted during a 2-month period, an assumption needed to be
made that the net impact of the protocol changes will not be different for the 2 months of the
field test than for the other 10 months of the year. This does not imply an assumption that drug
and mental health reporting cannot be affected by the month of data collection, only that the net
impact of the changes in the redesign protocol will not be affected by the particular month or
season chosen.
Comparisons were carried out for the following lifetime use measures between quarters 3
and 4 data in the 2013 comparison data: marijuana, cocaine, crack, heroin, hallucinogens, LSD,

PCP, Ecstasy, inhalants, cigarettes, smokeless tobacco, and alcohol. The same comparisons were
carried out for lifetime misuse of pain relievers, tranquilizers, sedatives, stimulants (based on the
standard definition that includes methamphetamine), methamphetamine, any illicit drug
(standard definition that includes prescription drug misuse and methamphetamine), and any illicit
drug other than marijuana. For 13 of the 19 measures, differences were not statistically
significant. Six measures showed statistically significant differences between estimates from
quarters 3 and 4: cocaine, PCP, inhalants, stimulant misuse (standard definition),
methamphetamine, and any illicit drug (standard definition that includes prescription drug
misuse and methamphetamine). These differences were all in the direction of higher prevalence
rates in quarter 4 than in quarter 3. Consequently, DR estimates may be slightly underestimated
if it is assumed that conducting more DR interviews later in the year would result in higher DR
estimates. The magnitudes of the differences, however, suggest that this underestimate would be
very small.
3.7.2.2

Comparison of DR Outcomes with 2013 Quarters 3 and 4 Main Study
Outcomes to Assess Level of Effort Effects on Estimates

Another concern with comparing estimates from the QFT sample with those from the
2013 quarters 3 and 4 main study sample is that that field efforts for NSDUH are not distributed
equally across the 3 months of each quarter. Typically, many interviews are conducted in the
first month of each quarter, fewer are conducted in the second month, and fewer still in the third
month. First-month responses may be systematically different from third-month responses, given
differences in the level of effort required to screen households and interview selected
respondents in the first month versus the third month. Analyses of the relationship between
indicators related to length of time in the field, such as interview visits, have shown that
respondents requiring more calls to complete the interview may have higher self-reported rates
of illicit drug use (Biemer & Wang, 2006). Given that the DR data were collected in a
compressed 2-month time period, a reduced calling effort may lead to differences between
estimates from the DR sample and the 2013 quarters 3 and 4 samples.
To investigate this possibility, estimates for a limited number of measures were examined
by the number of visits required to complete the interview for both the DR and 2013 quarters 3
and 4 samples. Estimates examined for both the DR sample and the 2013 comparison sample
were for the lifetime use measures for a number of substances were examined, including
marijuana, cocaine, crack, heroin, hallucinogens, inhalants, cigarettes, smokeless tobacco, and
alcohol, as well as the misuse of pain relievers, tranquilizers, sedatives, and stimulants (measures
including and excluding methamphetamine) and methamphetamine use. Overall, there was little
evidence of strong differences in estimates by the number of visits and little indication that any
such patterns differed by sample. Overall, estimates for the lifetime use (or misuse) of these
substances were not strongly correlated with the number of visits required to complete the
interview in either sample. In addition, the results indicated that these patterns did not differ in
any meaningful way across the two samples. For the 2013 quarters 3 and 4 sample, estimates for
lifetime misuse of prescription drugs (and methamphetamine use) were not available when this
report was produced. As with the DR sample, estimates of lifetime use for other substances were
mostly uncorrelated with the numbers of visits needed to complete the interviews.

3.7.3

Comparisons with Other Survey Data

Estimates from the DR sample were also compared with estimates from other appropriate
sources, primarily for the purpose of providing further evidence on differences in estimates for
items moved from computer-assisted personal interviewing (CAPI) to audio computer-assisted
self-interviewing (ACASI) administration. External data sources were also used for benchmark
comparisons for items that were introduced in the QFT and repeated for the DR, as well as items
that were introduced for the first time in the DR. Comparisons between NSDUH estimates and
those from benchmark external data sources are typically shown in the NSDUH national findings
report, such as in Appendix C from the 2012 NSDUH national findings report (Center for
Behavioral Health Statistics and Quality [CBHSQ], 2013). Such comparisons provide relevant
evidence on the effects of changes in the NSDUH data collection protocol. As noted in the 2012
national findings report, the results of such comparisons may be difficult to interpret given
differences between NSDUH and other data collection systems in a number of areas, including
the population of interest, sample design, data collection periods, screening and interviewing
protocols, and estimation procedures.
The following data sources were used in these comparisons:
•

National Health Interview Survey (NHIS), which includes measures of program
participation, income, health insurance coverage, height and weight, and health
conditions, and disabilities and physical limitations;

•

National Health and Nutrition Examination Survey (NHANES), which includes selfreported and direct measures of height and weight;

•

American Community Survey (ACS), which provides estimates on program
participation and health insurance coverage and English proficiency;

•

Current Population Survey (CPS), which provides estimates on employment;

•

General Social Survey (GSS), which includes an item on sexual identity; and

•

National Survey of Family Growth (NSFG), which includes an item on sexual
identity.

Results for these comparisons are presented and discussed in Chapter 7.

4. Data Collection Outcomes and
Data Quality Assessment
4.1

Overview of Data Collection and Data Quality Outcomes

This chapter presents a variety of indicators used to assess the quality of the 2013 Dress
Rehearsal (DR) data. Where feasible and appropriate, data quality outcomes for the DR data are
assessed in relation to the 2012 main study comparison data and the 2013 quarters 3 and 4 main
study comparison data. Examining these indicators identifies the potential impact of the
questionnaire and protocol revisions implemented for the DR on data quality when the partial
redesign is implemented in the 2015 National Survey on Drug Use and Health (NSDUH).
Section 4.2 presents both screening and interviewing unit response rates for all three
datasets, the number of field interviewer (FI) visits for both completed and noncompleted
screenings and completed and noncompleted interviews, and comparisons of demographic and
geographic characteristics among the datasets. Section 4.3 details the imputation rates for the
variables that were common to the 2012 comparison data, the 2013 quarters 3 and 4 comparison
data, and the DR data, while Section 4.4 details the missing data rates for moved, revised, or new
items in the DR questionnaire. Section 4.5 presents the overall and module-specific interview
timing results, including comparisons among the datasets where appropriate. Section 4.6
describes other data quality indicators for the new prescription drug modules included in the DR
questionnaire.

4.2

Unit Response Rates and Sample Characteristics (Research Question 3)
4.2.1

Screening Response Rates (SRRs) and Number of Visits for Completed and
Noncompleted Screenings

The screening response rate (SRR) is the total number of completed screenings divided
by the total eligible dwelling units (DUs). The eligible DUs are computed by subtracting the
number of sample dwelling units (SDUs) not eligible to be included in NSDUH from the total
number of SDUs. Ineligibles include a vacant unit, not a primary residence, not a DU, a group
quarters unit (GQU) listed as housing unit (HU), an HU listed as a GQU, only military, listing
errors, other ineligibles, and those SDUs where the residents will live there less than half of the
quarter.
SRRs were calculated for the 2012 main study comparison sample, the 2013 quarters 3
and 4 main study comparison sample, the 2012 Questionnaire Field Test (QFT) sample, and the
2013 DR sample. Response rates for 2012 were calculated using final 2012 main study data.
Data for Alaska and Hawaii were removed to make rates more comparable with the 2012 QFT
and 2013 DR samples. SRRs for the 2013 comparison sample were calculated based on the
preliminary results for quarters 3 and 4 of 2013, with Alaska and Hawaii removed. Screeners
associated with FIs that were subject to field verification at the time the preliminary data were
obtained were considered nonrespondents to minimize the risk of introducing falsified cases onto
the comparison file. Because the 2013 comparison data were based on the data collected as of
65

December 1, 2013, quarter 4 screenings completed after that date were considered
nonrespondents for the purposes of the DR analysis. Similarly, any screener completions that
were later recoded as screener incompletes were treated as screener completions for the purposes
of the DR analysis. An exception to this rule was that cases in the 2013 quarters 3 and 4
comparison data file that were determined (or suspected) to be falsified were removed from the
2013 quarters 3 and 4 comparison data file (see Section 3.6.3).
Table 4.1 lists the sample totals and the national screening and interviewing response
rates for the 2012 main study comparison file, the 2013 quarters 3 and 4 main study comparison
file, the QFT, and the DR. This table provides both the weighted and unweighted screening and
interviewing response rates for each sample. The weighted screening response rates for the 2012
main study comparison file, the 2013 quarters 3 and 4 main study comparison file, the QFT, and
the DR were 86.09 percent, 79.23 percent, 83.58 percent, and 81.70 percent, respectively.
One difference between the QFT sample and the three other samples (two main study
samples and the DR sample) that could not be accounted for is the language used to complete
screenings. For the main study and the DR, screenings could be completed in English or Spanish,
and the FI had the ability to switch languages as needed. As a result, the language that was used
for each screening could not be determined. For the QFT, a Spanish-language version of the
screening interview was not available, so households that could not complete the screening in
English were treated as nonrespondents. An additional factor that could affect SRRs was
improvements to the QFT and DR lead letters, which were expected to improve SRRs.
Table 4.2 presents data on the number of visits made for successfully completed
screenings in each of the three samples. The overall distribution of visits for completed
screenings in the DR sample looked quite similar to the distributions for the 2012 and 2013
quarters 3 and 4 comparison samples, with only slight differences for a few number of visit
categories. These distributions indicate that there were no significant differences in the number
of visits required to complete screenings in the DR data collection compared with the 2012 and
2013 quarters 3 and 4 comparison samples.
For comparison, Table 4.3 presents data on the number of visits made to DUs that were
not successfully screened for each of the three samples. This further comparison allows for an
assessment of how the DR screening results might have differed from the 2012 and 2013
quarters 3 and 4 comparison samples when screening efforts were not successful. In general, a
smaller percentage of the DR sample cases were finalized as noncompleted screenings with one
or two visits than noncompleted screenings in the 2012 and 2013 quarters 3 and 4 comparison
samples. Only 10.1 percent of noncompleted screenings in the DR sample were finalized on the
first visit compared with 18.2 percent and 16.5 percent of noncompleted screenings in the 2012
and 2013 quarters 3 and 4 samples, respectively, and these were finalized on the first visit.
Similarly, only 13.0 percent of noncompleted screenings in the DR sample were finalized in two
visits compared with 19.0 percent and 17.4 percent of noncompleted screenings in the 2012 and
2013 quarters 3 and 4 samples, respectively. In the highest category, 10 or more screening visits,
the results were reversed. In the DR sample, 27.3 percent of noncompleted screenings were
finalized after 10 or more visits compared with 20.9 percent and 22.2 percent of noncompleted
screenings in the 2012 and 2013 quarters 3 and 4 samples, respectively.

Table 4.1 Screenings, Interviews, and Response Rates for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, 2012 Questionnaire
Field Test, and 2013 Dress Rehearsal Estimates
Sample Totals and Rates
Eligible Dwelling Units
Completed Screenings
Screening Response Rate
Selected Persons
Completed Interviews
Interviewing Response Rate
Overall Response Rate

2012 Main Study
Comparison Sample
173,956
149,992
Unweighted
Weighted
86.22%
86.09%
85,295
66,542
Unweighted
Weighted
78.01%
73.06%
Unweighted
Weighted
67.27%
62.89%

2013 Quarters 3 and 4 Main
Study Comparison Sample
93,509
74,960
Unweighted
Weighted
80.16%
79.23%
39,850
32,162
Unweighted
Weighted
80.71%
77.26%
Unweighted
Weighted
64.70%
61.21%

2012 Questionnaire Field
Test
4,623
3,837
Unweighted
Weighted
83.00%
83.58%
2,823
2,044
Unweighted
Weighted
72.41%
69.04%
Unweighted
Weighted
60.09%
57.71%

2013 Dress Rehearsal
4,392
3,511
Unweighted
Weighted
79.94%
81.70%
2,808
2,087
Unweighted
Weighted
74.32%
70.55%
Unweighted
Weighted
59.41%
57.64%

Table 4.2 Number of Visits Made for Completed Screenings for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal

Visits
1
2
3
4
5 to 9
10 or More
Unknown
Total

2012 Main Study Comparison Sample
Cumulative
Screenings
Percent
Percent
54,959
36.6
36.6
30,964
20.6
57.3
18,436
12.3
69.6
11,998
8.0
77.6
23,843
15.9
93.4
9,792
6.5
100.0
0
0.0
100.0
149,992

2013 Quarters 3 and 4 Main Study
Comparison Sample
Cumulative
Screenings
Percent
Percent
28,315
36.2
36.2
16,031
20.5
56.7
9,543
12.2
68.9
6,223
8.0
76.9
12,660
16.2
93.1
5,407
6.9
100.0
0
0.0
100.0
78,179

2013 Dress Rehearsal
Cumulative
Screenings
Percent
Percent
1,289
36.5
36.5
716
20.3
56.8
439
12.4
69.2
317
9.0
78.2
533
15.0
93.2
238
6.7
100.0
0
0.0
100.0
3,532

Table 4.3 Number of Visits Made for Noncompleted Screenings for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013
Dress Rehearsal

Visits
1
2
3
4
5 to 9
10 or More
Unknown
Total

2012 Main Study Comparison Sample
Noncompleted
Cumulative
Screenings Percent
Percent
10,676
18.2
18.2
11,130
19.0
37.2
6,546
11.2
48.3
4,845
8.3
56.6
13,230
22.5
79.2
12,261
20.9
100.0
1
0.0
100.0
58,689

2013 Quarters 3 and 4 Main Study
Comparison Sample
Noncompleted
Cumulative
Screenings
Percent
Percent
5,469
16.5
16.5
5,772
17.4
33.9
3,853
11.6
45.5
2,835
8.6
54.1
7,874
23.8
77.9
7,351
22.2
100.0
0
0.0
100.0
33,154

2013 Dress Rehearsal
Noncompleted
Screenings
150
203
219
139
368
405
0
1,484

Percent
10.1
13.7
14.8
9.4
24.9
27.3
0.0

Cumulative
Percent
10.1
23.8
38.5
47.9
72.9
100.0
100.0

Overall, the results presented in Tables 4.2 and 4.3 suggest that completed screeners in
the DR sample were finalized based on similar numbers of visits as those in the 2012 and 2013
quarters 3 and 4 comparison samples. One notable difference among the three datasets was that a
smaller proportion of noncompleted screeners in the DR were finalized with only one or two
visits than in the 2012 and 2013 quarters 3 and 4 comparison samples, and a larger proportion of
noncompleted screeners in the DR were finalized with 10 or more visits than in the 2012 and
2013 quarters 3 and 4 comparison samples.
4.2.2

Interview Response Rates (IRRs) and Number of Visits for Completed and
Noncompleted Screenings

The interviewing response rate (IRR) is the number of completed interviews divided by
the total number of eligible respondents chosen through screening. Any ineligible respondents
(younger than 12 or actually in the military) were subtracted from the total. For the 2013 main
study comparison sample, interview status was determined based on the December 1, 2013,
preliminary results. Cases that were undergoing field verification at that time were treated as
nonrespondents. Cases that resulted in interview completions after this date were treated as
nonrespondents, and cases that were classified as interviews on this date that were later recoded
as noncompletes were treated as completed interviews for the purposes of the DR analysis, with
the exception of interviews dropped because they were determined (or suspected) to have been
falsified (see Section 3.6.3). To make the 2012 main study and the 2013 quarters 3 and 4 main
study more comparable with the DR and QFT samples, interviews completed in Alaska and
Hawaii were excluded. One difference between the QFT and the three other samples (two main
study samples and the DR sample) is the language used to complete the interview. Unlike the
other three samples, the QFT did not allow for interviews to be completed in Spanish. Persons
selected for the QFT who could not complete the interview in English were treated as eligible
nonrespondents, while bilingual interviewers were available to interview respondents in Spanish
for the other three studies.
Table 4.4 presents the unweighted and weighted IRRs by age group for all four samples.
The weighted IRRs for the 2012 main study, the 2013 quarters 3 and 4 main study, the QFT, and
the DR were 73.06 percent, 77.26 percent, 69.04 percent, and 70.58 percent, respectively.
Table 4.4 Interview Response Rates, by Age, for the 2012 Main Study, 2013 Quarters 3 and 4 Main
Study, 2012 Questionnaire Field Test, and 2013 Dress Rehearsal
Age
Category
12 to 17
18 to 25
26 to 34
35 to 49
50 to 64
65 or Older

2012
83.03
79.43
75.71
73.51
70.40
66.55

Unweighted Percent
2013 Quarters
3 and 4
QFT
85.10
82.05
81.83
75.71
78.30
68.07
77.39
66.25
74.39
67.25
69.15
63.68

DR
79.62
81.54
72.11
69.07
64.34
68.39

2012
82.88
79.24
75.31
73.05
69.17
65.64

Weighted Percent
2013 Quarters
3 and 4
QFT
85.67
82.25
81.84
75.26
78.80
68.91
77.44
66.32
75.68
66.78
68.95
63.48

DR
78.03
81.82
72.00
68.86
62.70
71.07

NOTE: Cases where respondents provided only the age category 50 or older were counted in the 65 or older category.

Table 4.5 presents data on the number of visits made for completed interviews for the DR
sample and the 2012 and 2013 quarters 3 and 4 comparison samples. Similar to the results on the

number of visits for completed screenings (see Table 4.2), the percentage of completed
interviews in each category of the number of visits followed a similar pattern across the three
samples. A lower percentage of DR interviews were completed on the first visit than in the 2012
and 2013 quarter 3 and 4 comparison samples. DR interviews were less likely to be completed
"on the spot," that is, at the same time the household was screened and one or more respondents
were selected. However, by the second visit, the cumulative percentages of cases with two or
fewer visits required to complete the interview were all close to 70 percent for all three samples.
Overall, the distribution of completed interviews by the number of visits made for the DR sample
was similar to the 2012 and 2013 quarters 3 and 4 comparison samples.
Table 4.6 presents results for the number of visits made for selected respondents who
were not successfully interviewed for each of the three samples. This further comparison allows
for an assessment of how the DR interviewing results might have differed from the 2012 and
2013 quarters 3 and 4 comparison samples when attempts to interview selected respondents were
unsuccessful. In general, the proportion of noninterviews for the DR sample across the categories
of visits followed a similar pattern as the 2012 and 2013 quarters 3 and 4 comparison samples.
The percentage of noninterviews finalized after the first two calls in the DR sample
(17.4 percent) was more similar to that of the 2012 comparison sample (17.9 percent) than to that
of the 2013 quarter 3 and 4 sample (20.4 percent). By the ninth call, the cumulative proportions
of finalized noninterviews were similar for the DR sample (73.1 percent) and the 2013 quarters 3
and 4 comparison sample (74.1 percent). In the 2012 comparison sample, the proportion of
noninterviews was slightly lower (70.3 percent) than in the other two samples. Overall, these
results indicate some relatively small and inconsistent variation in the distribution of completed
noninterviews by the number of visits made for the DR sample relative to the 2012 and 2013
quarters 3 and 4 comparison samples.
4.2.3

Geographic, Demographic, and Household Characteristics for the Complete DR
Sample

Another way to assess the relative representativeness of the full DR sample is to compare
demographic and household estimates for the combined QFT-DR data and the comparison data.
Tables 4.7a through 4.7d present estimates for selected geographic, demographic, and household
characteristics across age groups for both English-language and Spanish-language interviews for
the DR dataset and the 2012 and 2013 quarters 3 and 4comparison datasets. To assess the
significance of any differences, the tables also provide a chi-square statistic and p value for both
weighted and unweighted comparisons.10

Differences in estimates for the selected demographic and household items between the combined
QFT-DR data and the main study comparison data presented in Tables 4.7a through 4.7d could result from either
(1) differences in the composition of the combined QFT-DR and comparison samples or (2) differences in how
respondents reported these items in the QFT and DR interviews versus the main study comparison interviews.
Because these demographic and household questions were administered via audio computer-assisted selfinterviewing (ACASI) for all QFT and DR respondents and via computer-assisted personal interviewing (CAPI) for
all comparison sample respondents, the potentially confounding effects of sample differences and mode differences
cannot be directly assessed from this study design.

Table 4.5 Number of Visits Made for Completed Interviews for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal

Visits
1
2
3
4
5 to 9
10 or More
Unknown
Total

2012 Main Study Comparison Sample
Completed
Cumulative
Interviews
Percent
Percent
23,898
35.9
35.9
22,821
34.3
70.2
7,698
11.6
81.8
3,574
5.4
87.1
6,162
9.1
96.3
2,283
3.4
99.8
106
0.2
100.0
66,542

2013 Quarters 3 and 4 Main Study
Comparison Sample
Completed
Cumulative
Interviews
Percent
Percent
12,659
39.4
39.4
10,290
32.0
71.4
3,520
10.9
82.3
1,737
5.4
87.7
2,876
8.9
96.6
1,020
3.2
99.8
60
0.2
100.0
32,162

2013 Dress Rehearsal
Completed
Cumulative
Interviews
Percent
Percent
886
42.5
42.5
586
28.1
70.5
240
11.5
82.0
107
5.1
87.2
185
8.8
96.0
81
3.9
99.9
2
0.1
100.0
2,087

Table 4.6 Number of Visits Made for Noncompleted Interviews for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013
Dress Rehearsal

Visits
1
2
3
4
5 to 9
10 or More
Unknown
Total

2012 Main Study Comparison Sample
Noncompleted
Cumulative
Interviews
Percent
Percent
1,242
6.4
6.4
2,218
11.5
17.9
1,895
9.8
27.7
1,733
9.0
36.7
6,491
33.6
70.3
5,579
28.9
99.1
174
0.9
100.0
19,332

2013 Quarters 3 and 4 Main Study
Comparison Sample
Noncompleted
Cumulative
Interviews
Percent
Percent
549
7.5
7.5
936
12.8
20.4
856
11.7
32.1
727
10.0
42.1
2,338
32.1
74.1
1,864
25.6
99.7
25
0.3
100.0
7,295

2013 Dress Rehearsal
Noncompleted
Cumulative
Interviews
Percent
Percent
40
5.3
5.3
91
12.1
17.4
84
11.2
28.6
62
8.3
36.9
272
36.2
73.1
199
26.5
99.6
3
0.4
100.0
751

Table 4.7a Demographic and Geographic Characteristics among Persons Aged 12 or Older: Percentages, Chi-Square Test Statistic, and
P Value, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal
2012 Comparison
(n = 66,542)1,2

Characteristic
REGION
Northeast
Midwest
South
West
COUNTY TYPE
Large Metro
Small Metro
Nonmetro
EDUCATION5
< High School
High School Graduate
Some College
College Graduate
CURRENTLY
EMPLOYED5
EMPLOYMENT5
Full-Time
Part-Time
Unemployed
Other6
OVERALL HEALTH7
Excellent
Very Good
Good
Fair/Poor
COVERED BY ANY
HEALTH INSURANCE
See notes at end of table.

Unwtd
n

2013 Comparison
(n = 32,162)1,3

Unwtd
Wtd
Percent Percent

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 DR
(n = 2,087)1,4

Unwtd
n

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
Wtd
Wtd
Percent Percent
1.62, 0.1888
14.4
19.5
16.3
25.8
31.3
34.5
38.0
20.2
1.07, 0.3471
64.6
51.1
23.9
30.5
11.5
18.4
5.76, 0.0011c
18.4
16.4
27.8
28.8
33.5
31.8
20.3
23.1

13,773
19,142
20,886
12,741

20.7
28.8
31.4
19.1

18.3
21.6
37.3
22.8

6,784
9,145
10,169
6,064

21.1
28.4
31.6
18.9

18.3
21.6
37.3
22.8

301
340
653
793

30,691
22,925
12,926

46.1
34.5
19.4

55.3
30.1
14.6

15,022
11,047
6,093

46.7
34.3
18.9

54.0
31.3
14.7

1,348
499
240

6,604
14,368
13,344
10,269

14.8
32.2
29.9
23.0

14.6
29.6
26.6
29.2

2,839
7,047
6,624
4,886

13.3
32.9
31.0
22.8

13.0
30.1
27.1
29.8

291
440
530
321

30,342

68.1

64.0

14,812

69.2

65.1

1,031

65.2

62.6

21,770
8,572
3,720
10,523

48.8
19.2
8.3
23.6

50.0
14.0
5.8
30.2

10,714
4,098
1,610
4,974

50.1
19.2
7.5
23.2

51.0
14.0
4.5
30.5

776
255
136
415

49.1
16.1
8.6
26.2

48.6
14.0
6.2
31.2

18,465
26,899
16,004
5,166

27.8
40.4
24.1
7.8

23.0
37.1
27.3
12.7

8,891
12,954
7,863
2,448

27.6
40.3
24.5
7.6

23.1
36.8
27.6
12.5

494
805
553
233

23.7
38.6
26.5
11.2

19.8
37.0
28.2
15.0

56,355

84.7

84.8

27,359

85.1

85.5

1,642

78.7

83.2

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
1.61, 0.1929

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
20.68, 0.0000c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
21.39, 0.0000c

0.70, 0.5009

20.21, 0.0000c

18.39, 0.0000c

6.58, 0.0004c

9.28, 0.0000c

12.40, 0.0000c

0.36, 0.5486
0.18, 0.9118

1.27, 0.2618
1.76, 0.1602

4.21, 0.0429c
3.25, 0.0250c

8.74, 0.0039c
4.30, 0.0068c

2.12, 0.1024

2.26, 0.0859

13.35, 0.0000c

14.04, 0.0000c

0.95, 0.3314

2.35, 0.1289

24.53, 0.0000c

29.26, 0.0000c
(continued)

Characteristic
FAMILY INCOME
< $20,000
$20,000-$49,999
$50,000-$74,999
≥ $75,000
PARTICIPATED IN
GOVERNMENT
PROGRAM8
RECEIVED INCOME
Social Security
Supplemental Security
Income
Food Stamps
Welfare Payments
BETTER PROVIDER
OF INFORMATION7
USED PROXY

Unwtd
n

2013 Comparison
(n = 32,162)1,3

Unwtd
Wtd
Percent Percent

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 DR
(n = 2,087)1,4

Unwtd
n

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
Wtd
Wtd
Percent Percent
10.51, 0.0000c
30.1
28.0
34.0
33.3
13.7
14.9
22.2
23.8

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
13.03, 0.0000c

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
5.33, 0.0019c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
6.45, 0.0005c

15,763
21,677
10,549
18,553

23.7
32.6
15.9
27.9

18.8
32.2
16.4
32.6

7,507
10,350
4,997
9,308

23.3
32.2
15.5
28.9

18.6
31.6
17.0
32.8

629
710
285
463

17,106

25.7

21.1

8,468

26.3

21.2

621

29.8

24.7

5.78, 0.0181c

4.94, 0.0286c

6.07, 0.0155c

4.26, 0.0416c

9,887

14.9

26.7

5,035

15.7

26.8

316

15.1

23.6

1.87, 0.1744

1.87, 0.1747

0.07, 0.7892

0.20, 0.6519

4,928
14,153
2,502

7.4
21.3
3.8

7.6
16.4
2.5

2,546
7,032
1,120

7.9
21.9
3.5

7.7
16.3
2.1

174
502
138

8.3
24.1
6.6

8.0
19.9
3.2

0.17, 0.6773
6.15, 0.0149c
1.76, 0.1877

0.08, 0.7800
6.17, 0.0147c
5.77, 0.0182c

1.75, 0.1889
2.97, 0.0880
11.34, 0.0011c

0.33, 0.5675
1.84, 0.1776
14.22, 0.0003c

24,852
22,787

43.9
34.2

19.1
13.8

12,668
11,547

46.0
35.9

21.2
14.9

680
606

38.2
29.0

24.5
16.4

9.73, 0.0024c
4.07, 0.0463c

3.69, 0.0577
1.21, 0.2733

19.83, 0.0000c
18.49, 0.0000c

38.65, 0.0000c
33.08, 0.0000c

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Sample does not include Alaska or Hawaii.
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Education and employment estimates are based only on respondents aged 18 or older. Sample sizes for respondents 18 or older are n = 44,585 for 2012 comparison, n = 21,396 for 2013 comparison, and n
= 1,582 for 2013 DR.
6
The Other Employment category includes students, persons keeping house or caring for children full time, retired or disabled persons, or other persons not in the labor force.
7
Respondents with unknown data were excluded.
8
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI), cash assistance (Temporary Assistance for Needy Families, TANF),
noncash assistance, or food stamps.
2

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.7b Demographic and Geographic Characteristics among Persons Aged 12 to 17: Percentages, Chi-Square Test Statistic, and
P Value, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal
2012 Comparison
(n = 21,957)1,2

Characteristic
REGION
Northeast
Midwest
South
West
COUNTY TYPE
Large Metro
Small Metro
Nonmetro
OVERALL HEALTH5
Excellent
Very Good
Good
Fair/Poor
COVERED BY ANY
HEALTH INSURANCE
FAMILY INCOME
< $20,000
$20,000-$49,999
$50,000-$74,999
≥ $75,000
See notes at end of table.

Unwtd
n

2013 Comparison
(n = 10,766)1,3

Unwtd
Wtd
Unwtd Unwtd
Wtd
Percent Percent
n
Percent Percent

2013 DR
(n = 505)1,4

Unwtd
n

Unwtd
Wtd
Percent Percent

4,421
6,387
6,964
4,185

20.1
29.1
31.7
19.1

17.1
21.9
37.6
23.5

2,291
3,051
3,373
2,051

21.3
28.3
31.3
19.1

17.8
22.2
36.8
23.2

66
61
163
215

13.1
12.1
32.3
42.6

19.2
25.2*
32.7
22.9

10,211
7,426
4,320

46.5
33.8
19.7

56.2
29.7
14.1

5,097
3,601
2,068

47.3
33.4
19.2

55.6
30.5
14.0

314
137
54

62.2
27.1
10.7

46.5
36.6
17.0

7,405
9,267
4,452
832

33.7
42.2
20.3
3.8

35.1
41.0
20.3
3.7

3,572
4,579
2,220
395

33.2
42.5
20.6
3.7

33.3
42.8
20.3
3.6

147
230
100
27

29.2
45.6
19.8
5.4

30.0
48.5
16.1
5.3

20,545

93.6

93.0

10,087

93.7

93.0

467

92.5

94.9

4,073
6,788
3,718
7,378

18.5
30.9
16.9
33.6

17.8
30.0
16.6
35.6

2,027
3,333
1,759
3,647

18.8
31.0
16.3
33.9

18.6
30.4
15.8
35.1

153
174
67
111

30.3
34.5
13.3
22.0

25.4
35.9
13.2
25.4

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
0.43, 0.7349

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
0.32, 0.8138

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
23.48, 0.0000c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
25.92, 0.0000c

1.08, 0.3420

0.98, 0.3807

11.60, 0.0000c

10.44, 0.0001c

2.13, 0.1011

1.91, 0.1323

1.96, 0.1251

1.92, 0.1308

1.12, 0.2921
4.06, 0.0092c

1.08, 0.3003
3.17, 0.0277c

0.74, 0.3931
8.69, 0.0000c

0.91, 0.3416
8.05, 0.0001c

(continued)

Characteristic
PARTICIPATED IN
GOVERNMENT
PROGRAM6
RECEIVED INCOME
Social Security
Supplemental Security
Income
Food Stamps
Welfare Payments
BETTER PROVIDER
OF INFORMATION5
USED PROXY

2013 Comparison
(n = 10,766)1,3

2013 DR
(n = 505)1,4

Unwtd
Wtd
Percent Percent

Unwtd
n

Unwtd
Wtd
Percent Percent

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

6,221

28.3

27.2

3,182

29.6

28.7

191

37.8

33.2

2.37, 0.1271

1.17, 0.2812

10.27, 0.0018c

7.20, 0.0086c

2,575

11.7

11.6

1,340

12.4

11.4

12.9

13.0

0.30, 0.5833

0.41, 0.5235

0.56, 0.4575

0.07, 0.7950

1,829
5,178
950

8.3
23.6
4.3

8.0
22.5
4.1

942
2,679
480

8.7
24.9
4.5

8.0
24.2
4.2

54
161
53

10.7
31.9
10.5

10.1
28.3
6.1

1.10, 0.2970
2.53, 0.1149
2.29, 0.1336

1.07, 0.3028
1.08, 0.3011
2.10, 0.1504

2.93, 0.0901
8.69, 0.0040c
16.48, 0.0001c

1.94, 0.1665
5.73, 0.0186c
15.91, 0.0001c

19,681
18,804

90.0
85.6

89.3
84.9

9,759
9,323

91.0
86.6

90.7
86.7

454
438

90.1
86.7

91.4
88.3

1.29, 0.2595
2.18, 0.1434

0.15, 0.7018
0.49, 0.4855

0.00, 0.9780
0.49, 0.4856

0.48, 0.4918
0.01, 0.9321

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.7c Demographic and Geographic Characteristics among Persons Aged 18 to 25: Percentages, Chi-Square Test Statistic, and
P Value, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal
2012 Comparison
(n = 21,943)1,2

Characteristic
REGION
Northeast
Midwest
South
West
COUNTY TYPE
Large Metro
Small Metro
Nonmetro
EDUCATION
< High School
High School Graduate
Some College
College Graduate
CURRENTLY
EMPLOYED
EMPLOYMENT
Full-Time
Part-Time
Unemployed
Other5
OVERALL HEALTH6
Excellent
Very Good
Good
Fair/Poor
COVERED BY ANY
HEALTH INSURANCE
See notes at end of table.

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 Comparison
(n = 10,436)1,3

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 DR
(n = 529)1,4

Unwtd
n

Unwtd
Percent

Wtd
Percent

4,666
6,220
6,941
4,116

21.3
28.3
31.6
18.8

17.9
21.4
37.1
23.6

2,164
3,019
3,324
1,929

20.7
28.9
31.9
18.5

17.8
21.8
37.7
22.7

86
86
150
207

16.3
16.3
28.4
39.1

23.8*
20.6
35.5
20.1

10,097
7,851
3,995

46.0
35.8
18.2

56.2
31.2
12.6

4,868
3,710
1,858

46.6
35.6
17.8

54.3
32.4
13.3

352
135
42

66.5
25.5
7.9

51.7
36.2
12.1

3,446
7,752
7,504
3,241

15.7
35.3
34.2
14.8

15.4
34.0
35.7
14.8

1,422
3,886
3,682
1,446

13.6
37.2
35.3
13.9

13.2
36.9
35.3
14.6

91
184
205
49

17.2
34.8
38.8
9.3

15.6
35.4
38.6
10.4

14,690

66.9

66.1

7,174

68.7

67.5

335

63.3

65.3

8,851
5,839
2,565
4,688

40.3
26.6
11.7
21.4

38.9
27.2
11.9
22.0

4,431
2,743
1,158
2,104

42.5
26.3
11.1
20.2

41.0
26.6
10.9
21.5

209
126
73
121

39.5
23.8
13.8
22.9

39.8
25.4
11.9
22.9

6,191
9,317
5,174
1,259

28.2
42.5
23.6
5.7

29.5
41.5
23.4
5.6

2,939
4,401
2,495
600

28.2
42.2
23.9
5.7

28.4
42.4
23.8
5.4

136
211
140
42

25.7
39.9
26.5
7.9

25.0
45.7
24.1
5.3

17,018

77.6

77.3

8,093

77.5

76.9

360

68.1

74.7

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
0.61, 0.6125

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
0.58, 0.6303

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
11.88, 0.0000c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
12.32, 0.0000c

0.39, 0.6802

0.24, 0.7872

14.72, 0.0000c

14.05, 0.0000c

0.60, 0.6189

1.09, 0.3551

2.75, 0.0467c

2.99, 0.0346c

0.07, 0.7923
0.10, 0.9588

0.52, 0.4706
0.19, 0.9012

2.80, 0.0974
1.31, 0.2757

6.16, 0.0147c
2.31, 0.0806

1.41, 0.2443

0.75, 0.5221

1.88, 0.1374

1.79, 0.1547

0.78, 0.3802

0.52, 0.4727

18.63, 0.0000c

19.45, 0.0000c
(continued)

Characteristic
FAMILY INCOME
< $20,000
$20,000-$49,999
$50,000-$74,999
≥ $75,000
PARTICIPATED IN
GOVERNMENT
PROGRAM7
RECEIVED INCOME
Social Security
Supplemental Security
Income
Food Stamps
Welfare Payments
BETTER PROVIDER
OF INFORMATION6
USED PROXY

Unwtd
n

2013 Comparison
(n = 10,436)1,3

Unwtd
Wtd
Percent Percent

Unwtd
n

2013 DR
(n = 529)1,4

Unwtd
Wtd
Percent Percent

Unwtd
n

Unwtd
Wtd
Percent Percent

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
4.53, 0.0051c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
3.33, 0.0226c

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
2.48, 0.0656

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
3.56, 0.0170c

7,793
7,446
2,783
3,921

35.5
33.9
12.7
17.9

33.0
33.3
13.3
20.4

3,687
3,466
1,309
1,974

35.3
33.2
12.5
18.9

34.8
33.0
12.6
19.7

216
187
57
69

40.8
35.3
10.8
13.0

43.9
34.0
9.8
12.4

5,947

27.1

25.3

2,885

27.6

26.4

168

31.8

29.1

1.51, 0.2223

0.73, 0.3936

4.04, 0.0473c

3.13, 0.0802

2,025

9.2

9.6

1,040

10.0

9.6

8.7

8.6

0.36, 0.5513

0.34, 0.5613

0.26, 0.6101

1.34, 0.2501

1,374
5,040
964

6.3
23.0
4.4

6.2
21.0
4.1

743
2,446
398

7.1
23.4
3.8

6.4
22.1
3.3

34
134
39

6.4
25.3
7.4

6.2
23.2
4.9

0.00, 0.9724
0.45, 0.5059
0.51, 0.4753

0.02, 0.8913
0.11, 0.7428
2.20, 0.1410

0.03, 0.8674
1.17, 0.2822
7.07, 0.0092c

0.42, 0.5169
0.78, 0.3789
11.09, 0.0012c

3,764
2,885

22.5
13.1

22.8
13.6

2,149
1,631

26.4
15.6

27.6
16.2

117
83

26.7
15.7

29.0
15.5

3.44, 0.0668
0.80, 0.3725

0.16, 0.6939
0.07, 0.7908

3.50, 0.0642
2.71, 0.1031

0.01, 0.9123
0.00, 0.9713

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Sample does not include Alaska or Hawaii.
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
The Other Employment category includes students, persons keeping house or caring for children full time, retired or disabled persons, or other persons not in the labor force.
6
Respondents with unknown data were excluded.
7
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI), cash assistance (Temporary Assistance for Needy Families, TANF),
noncash assistance, or food stamps.
2

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.7d Demographic and Geographic Characteristics among Persons Aged 26 or Older: Percentages, Chi-Square Test Statistic, and
P Value, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal
2012 Comparison
(n = 22,642)1,2

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 Comparison
(n = 10,960)1,3

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 DR
(n = 1,053)1,4

Unwtd
n

Unwtd
Wtd
Percent Percent

4,686
6,535
6,981
4,440

20.7
28.9
30.8
19.6

18.5
21.7
37.3
22.6

2,329
3,075
3,472
2,084

21.3
28.1
31.7
19.0

18.4
21.6
37.3
22.8

149
193
340
371

14.2
18.3
32.3
35.2

18.8
26.8
34.6
19.8

10,383
7,648
4,611

45.9
33.8
20.4

55.0
29.9
15.1

5,057
3,736
2,167

46.1
34.1
19.8

53.7
31.3
15.0

682
227
144

64.8
21.6
13.7

51.6
28.7
19.7

3,158
6,616
5,840
7,028

13.9
29.2
25.8
31.0

14.5
28.9
25.0
31.6

1,417
3,161
2,942
3,440

12.9
28.8
26.8
31.4

13.0
28.9
25.7
32.4

200
256
325
272

19.0
24.3
30.9
25.8

16.5
27.6
30.6
25.3

15,652

69.1

63.6

7,638

69.7

64.6

696

66.1

62.2

12,919
2,733
1,155
5,835

57.1
12.1
5.1
25.8

51.9
11.7
4.7
31.7

6,283
1,355
452
2,870

57.3
12.4
4.1
26.2

52.8
11.9
3.4
32.0

567
129
63
294

53.8
12.3
6.0
27.9

50.2
12.0
5.2
32.6

4,869
8,315
6,378
3,075

21.5
36.7
28.2
13.6

20.3
35.8
28.8
15.0

2,380
3,974
3,148
1,453

21.7
36.3
28.7
13.3

20.9
35.0
29.2
14.8

211
364
313
164

20.1
34.6
29.8
15.6

17.7
34.1
30.4
17.8

18,792

83.0

85.0

9,179

83.8

86.1

815

77.4

83.2

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
1.70, 0.1716

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
1.81, 0.1510

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
12.64, 0.0000c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
13.59, 0.0000c

0.90, 0.4083

0.71, 0.4960

16.41, 0.0000c

15.58, 0.0000c

4.66, 0.0044c

4.97, 0.0030c

11.65, 0.0000c

10.85, 0.0000c

0.32, 0.5723
0.21, 0.8906

1.01, 0.3167
1.69, 0.1740

2.88, 0.0926
1.23, 0.3045

4.18, 0.0435c
2.86, 0.0407c

1.49, 0.2227

1.81, 0.1498

1.31, 0.2739

1.55, 0.2066

0.97, 0.3279

2.61, 0.1097

11.63, 0.0009c

14.07, 0.0003c
(continued)

2013 Comparison
(n = 10,960)1,3

Unwtd Unwtd
Wtd
Unwtd Unwtd
Wtd
n
Percent Percent
n
Percent Percent

2013 DR
(n = 1,053)1,4

Unwtd
n

Unwtd
Wtd
Percent Percent

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
8.22, 0.0001c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Wtd
11.40, 0.0000c

2012
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
6.65, 0.0004c

2013
Comparison
vs. DR
Chi-Square
Statistic,
P Value
Unwtd
8.97, 0.0000c

3,897
7,443
4,048
7,254

17.2
32.9
17.9
32.0

16.5
32.3
17.0
34.3

1,793
3,551
1,929
3,687

16.4
32.4
17.6
33.6

15.8
31.5
17.9
34.8

260
349
161
283

24.7
33.1
15.3
26.9

25.6
32.9
16.0
25.6

4,938

21.8

19.7

2,401

21.9

19.4

262

24.9

22.9

3.88, 0.0518

4.35, 0.0397c

3.24, 0.0749

3.09, 0.0817

5,287

23.4

31.5

2,655

24.2

31.6

205

19.5

27.5

2.30, 0.1323

2.33, 0.1300

4.88, 0.0294c

6.54, 0.0121c

1,725
3,935
588

7.6
17.4
2.6

7.7
14.9
2.1

861
1,907
242

7.9
17.4
2.2

7.8
14.3
1.6

86
207
46

8.2
19.7
4.4

8.0
18.2
2.5

0.06, 0.8129
5.56, 0.0203c
0.50, 0.4832

0.02, 0.8935
7.07, 0.0092c
3.01, 0.0857

0.39, 0.5333
1.88, 0.1736
6.70, 0.0111c

0.11, 0.7437
1.90, 0.1714
11.77, 0.0009c

1,407
1,098

7.8
4.8

7.7
5.0

760
593

8.8
5.4

9.2
5.8

109
85

13.0
8.1

12.6
7.5

11.76, 0.0009c
6.48, 0.0124c

5.31, 0.0233c
2.65, 0.1066

26.47, 0.0000c
18.65, 0.0000c

15.63, 0.0001c
11.16, 0.0012c

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Sample does not include Alaska or Hawaii.
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
The Other Employment category includes students, persons keeping house or caring for children full time, retired or disabled persons, or other persons not in the labor force.
6
Respondents with unknown data were excluded.
7
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI), cash assistance (Temporary Assistance for Needy Families, TANF),
noncash assistance, or food stamps.
2

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Among respondents aged 12 or older, Table 4.7a provides comparisons of the
geographic, demographic, and household estimates between the DR data and the 2012 or 2013
comparison data:
•

For both region and county type, the only significant differences observed between
the DR data and the 2012 and 2013 comparison data involved unweighted estimates.
No significant differences were observed among the weighted estimates for these two
variables. These results indicate that the weights applied to the DR data and the 2012
and 2013 comparison data (see Section 3.5) produced similar distributions for
geographic region and county type.

•

The estimated weighted proportion of having less than a high school education as the
highest level of education was higher in the DR data (16.4 percent) than in the 2012
(14.6 percent) and 2013 (13.0 percent) comparison data.11 The difference between the
2012 and 2013 comparison data for having less than a high school education was also
statistically significant. In addition, the estimated weighted proportion of having some
college as the highest level of education appeared to be higher in the DR data
(31.8 percent) than in the 2012 (26.6 percent) and 2013 (27.1 percent) comparison
data. The estimated weighted proportion of having a college degree was lower in the
DR data (23.1 percent) than in the 2012 (29.2 percent) and the 2013 (29.8 percent)
comparison data.

•

No significant differences were observed between the weighted estimates for being
currently employed in the DR sample (62.6 percent) versus the 2012 (64.0 percent)
and 2013 (65.1 percent) comparison samples. In addition, no significant differences
were observed between the weighted estimates for employment categories in the DR
data versus the 2012 and 2013 comparison data.

•

No significant differences were observed between the weighted estimates for overall
health status (excellent, good, fair, or poor) in the DR sample versus the 2012 and
2013 comparison samples. In addition, no significant differences were observed
between the weighted estimates for being covered by any type of health insurance in
the DR sample (83.2 percent) versus the 2012 (84.8 percent) and 2013 (85.5 percent)
comparison samples.

•

The estimated weighted proportion of family income of less than $20,000 was higher
in the DR data (28.0 percent) than in the 2012 (18.8 percent) and 2013 (18.6 percent)
comparison data. The estimated proportion of family income greater than $75,000
was lower in the DR data (23.8 percent) than in the 2012 (32.6 percent) and 2013
(32.8 percent) comparison data.

•

The estimated weighted proportion for participating in government programs was
significantly higher in the DR data (24.7 percent) than in the 2012 comparison data
(21.1 percent) and the 2013 comparison data (21.2 percent).

•

Receipt of income from government programs was higher in the DR data than in both
comparison datasets for one source—food stamps. The estimated weighted proportion

Table 4.7b does not include estimates for education level because these data are not produced for
respondents aged 12 to 17.

receiving food stamps was significantly higher in the DR data (19.9 percent) than in
the 2012 (16.4 percent) and the 2013 (16.3 percent) comparison data.
•

Identification and use of a proxy reporter for the health insurance and income items
was higher in the DR data than in the 2012 comparison data, but not significantly
different from the 2013 data. The weighted proportion indicating there was a better
provider of information on health insurance and income was significantly higher for
the DR data (24.5 percent) than the 2012 comparison data (19.1 percent) ), but not the
2013 comparison data (21.2 percent). Actual use of a proxy respondent for the health
insurance and income items was also significantly higher for the DR data (16.4
percent) than the 2012 comparison data (13.8 percent. Although the proportion for
use of a proxy in the DR data was also higher than for the 2013 comparison data (14.9
percent), this difference was not statistically significant. Differences in the weighted
estimates for identification and use of a proxy reporter between the 2012 and 2013
comparison data were statistically significant, with the 2013 estimates for both items
being higher than the 2012 estimates.

Tables 4.7b through 4.7d provide the same geographic, demographic and household
estimates for the DR dataset and the 2012 and 2013 comparison datasets for three specific age
groups: 12 to 17, 18 to 25, and 26 or older. Many of the significant differences in these estimates
between the DR data and the 2012 and 2013 comparison data for all of respondents aged 12 or
older were also observed across these three specific age groups, with the following exceptions:
•

The estimated proportions for the four-category education variable did not differ
significantly between the DR data and the 2012 and 2013 comparison data for the 18
to 25 age group (Table 4.7c). Given that differences in the distribution of the four
education categories were significant between the DR data and both the 2012 and
2013 comparison data for all respondents, this difference can be attributed to
significant differences in the 26 or older age group.

•

The significant differences between the DR data and the 2012 and 2013 comparison
data for receipt of food stamps among all respondents were not observed for the 12 to
17 year old age group (Table 4.7b) or the 18 to 25 year old age group (Table 4.7c).
Given that the difference in the receipt of food stamps was significant between the
DR data and both the 2012 and 2013 comparison data for all respondents, this
difference can be attributed to significant differences in the 26 or older age group

•

Identification and use of a proxy reporter for the health insurance and income items
was not significantly higher in the DR data than in the 2012 comparison data for the
12 to 17 year old age group (Table 4.7b) or the 18 to 25 year old age group
(Table 4.7c). As with all respondents aged 12 or older, however, identification of a
better reporter for the health insurance and income items was significantly higher in
the DR data than in the 2012 and 2013 comparison data for the 26 or older age group.
Use of a proxy reporter also appeared to be higher in the DR data than in the 2012
and 2013 comparison data for the 26 or older age group, but only the higher
proportion in the DR data compared with the 2012 data was statistically significant.
As with the estimates for all respondents aged 12 or older, differences in the weighted
estimates for identification and use of a proxy reporter between the 2012 and 2013
81

comparison data were statistically significant for the 26 or older age group. For both
items, the 2013 estimates were higher than the 2012 estimates.
Although most of the same differences in estimates for demographic and household items
between the combined QFT-DR data and the 2012 or 2013 comparison data were observed
across age groups, variation in the degree and significance of differences was observed for some
of these items for specific age groups.
4.2.4

Geographic, Demographic, and Household Characteristics for the Combined
QFT-DR Sample and Comparison Samples

To assess the representativeness of the combined QFT-DR sample relative to the 2012
and 2013 comparison datasets, estimates for the same set of geographic, demographic, and
household items included in Tables 4.7a through 4.7d are presented in Tables 4.8a through 4.8d.
These tables present both weighted and unweighted estimates across four age groups for Englishlanguage non-Hispanic interviews for the combined QFT-DR dataset and the 2012 and 2013
quarters 3 and 4 comparison datasets.12 To assess the significance of any differences, the tables
also provide chi-square statistics and p values for both weighted and unweighted comparisons.
Based on data from all respondents aged 12 or older, Table 4.8a presents comparisons of
the geographic, demographic, and household estimates between the combined QFT-DR data and
the 2012 and 2013 comparison data:
•

For both region and county type, the only significant differences observed between
the combined QFT-DR data and the 2012 and 2013 comparison data involved
unweighted estimates. No significant differences were observed among the weighted
estimates for these two variables.

•

The estimated weighted proportion of having some college as the highest level of
education was higher in the combined QFT-DR data (32.3 percent) than in the 2012
(26.9 percent) and 2013 (28.0) comparison data.13 In addition, the estimated weighted
proportion of having less than a high school education appeared to be slightly higher
in the combined QFT-DR data and the estimated weighted proportion of having a
college degree appeared to be slightly lower in the combined QFT-DR data, in
relation to these same estimates for the 2012 and 2013 comparison data sets. The
weighted estimates for the four-category education variable also differed significantly
between the 2012 and 2013 comparison data. The two categories with the greatest
difference appeared to be less than a high school education, where the 2012 estimate
was 11.5 percent and the 2013 estimate was 9.9 percent, and some college, where the
2012 estimate was 26.9 percent and the 2013 estimate was 28.0 percent.

Table 4.8b does not include employment estimates because these data are not collected for all
respondents aged 12 to 17.
13
Table 4.8b does not include estimates for education level because these data are not produced for
respondents aged 12 to 17.

Table 4.8a Demographic and Geographic Characteristics among Persons Aged 12 or Older for English-Language Non-Hispanic
Interviews: Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal
2012 Comparison
(n = 55,232)1,2

Unwtd
n

Unwtd
Wtd
Percent Percent

2013 Comparison
(n = 26,617)1,3

Combined 2012 QFT and
2013 DR
(n = 3,012)1,4

Unwtd
n

Unwtd
Percent

Wtd
Percent

Unwtd
n

2012
2013
2012
2013
Comparison Comparison Comparison Comparison
vs. Combined vs. Combined vs. Combined vs. Combined
QFT and DR QFT and DR QFT and DR QFT and DR
Chi-Square
Chi-Square
Chi-Square
Chi-Square
Statistic,
Statistic,
Statistic,
Statistic,
P Value
P Value
P Value
P Value
Unwtd
Wtd
Wtd
Wtd
Unwtd
Unwtd
Percent Percent
1.75, 0.1572
1.26, 0.2884 13.23, 0.0000c 14.12, 0.0000c
18.4
20.4
23.5
26.9
34.1
35.2
24.0
17.5
1.30, 0.2745
0.83, 0.4373
7.26, 0.0009c 5.60, 0.0043c
52.5
49.0
29.6
31.0
17.9
20.0
8.36, 0.0000c 8.12, 0.0000c
8.77, 0.0000c 8.51, 0.0000c
11.7
12.3
27.5
27.3
34.8
32.3
26.0
28.1

11,814
17,437
17,085
8,896

21.4
31.6
30.9
16.1

19.0
24.0
37.5
19.6

5,830
8,346
8,351
4,090

21.9
31.4
31.4
15.4

19.2
24.3
37.5
19.0

553
709
1,027
723

23,681
19,744
11,807

42.9
35.7
21.4

52.6
31.2
16.1

11,615
9,410
5,592

43.6
35.4
21.0

51.4
32.3
16.3

1,582
891
539

4,639
12,010
11,483
9,527

12.3
31.9
30.5
25.3

11.5
29.7
26.9
31.9

1,928
5,861
5,745
4,473

10.7
32.5
31.9
24.8

9.9
30.0
28.0
32.1

270
633
803
599

25,682

68.2

63.6

12,431

69.0

64.4

1,545

67.0

63.6

18,350
7,332
3,048
8,929

48.7
19.5
8.1
23.7

49.5
14.1
5.6
30.8

8,956
3,475
1,332
4,244

49.7
19.3
7.4
23.6

50.3
14.2
4.3
31.2

1,162
383
177
583

50.4
16.6
7.7
25.3

49.4
14.2
5.7
30.7

15,521
22,939
12,687
4,082

28.1
41.5
23.0
7.4

23.2
38.3
26.4
12.1

7,408
11,032
6,264
1,907

27.8
41.5
23.5
7.2

23.0
38.2
27.1
11.7

765
1,220
750
277

25.4
40.5
24.9
9.2

20.8
39.3
27.2
12.7

48,270

87.4

87.7

23,341

87.7

88.5

2,570

85.3

86.8

0.00, 0.9759
0.02, 0.9965

0.26, 0.6133
2.30, 0.0788

1.02, 0.3144
3.55, 0.0154c

2.99, 0.0851
3.27, 0.0221c

1.82, 0.1441

1.59, 0.1921

6.54, 0.0003c

6.53, 0.0003c

0.86, 0.3558

3.43, 0.0655

8.66, 0.0036c 11.85, 0.0007c
(continued)

Unwtd
n

Unwtd
Wtd
Percent Percent

2012
2013
2012
2013
Comparison
Comparison
Comparison
Comparison
vs. Combined vs. Combined vs. Combined vs. Combined
QFT and DR QFT and DR QFT and DR QFT and DR
Chi-Square
Chi-Square
Chi-Square
Chi-Square
Statistic,
Statistic,
Statistic,
Statistic,
P Value
P Value
P Value
P Value
Wtd
Unwtd Unwtd
Wtd
Wtd
Wtd
Unwtd
Unwtd
Percent
n
Percent Percent
9.74, 0.0000c
3.53, 0.0158c
4.46, 0.0047c
8.19, 0.0000c
17.3
792
26.3
23.1
29.8
915
30.4
31.9
17.5
434
14.4
16.0
35.4
871
28.9
29.1

2013 Comparison
(n = 26,617)1,3

Unwtd
n

Unwtd
Percent

Combined 2012 QFT and
2013 DR
(n = 3,012)1,4

12,464
16,797
9,109
16,862

22.6
30.4
16.5
30.5

17.5
30.6
17.0
34.9

5,810
8,090
4,323
8,394

21.8
30.4
16.2
31.5

13,134

23.8

19.5

6,537

24.6

19.4

816

27.1

23.4

11.59, 0.0008c

10.74, 0.0012c

6.69, 0.0104c

3.51, 0.0624

8,690

15.7

28.7

4,400

16.5

28.7

540

17.9

26.6

1.72, 0.1906

1.64, 0.2015

6.49, 0.0116c

2.30, 0.1306

4,016
10,731
1,977

7.3
19.4
3.6

7.3
14.9
2.3

2,123
5,322
818

8.0
20.0
3.1

7.6
14.5
1.7

261
634
138

8.7
21.0
4.6

8.7
17.6
3.2

3.93, 0.0487c
5.83, 0.0166c
5.61, 0.0188c

1.87, 0.1732
8.04, 0.0050c
20.95, 0.0000c

5.30, 0.0223c
1.63, 0.2036
4.75, 0.0304c

1.17, 0.2809
0.63, 0.4271
12.12, 0.0006c

20,282
18,578

43.7
33.6

18.7
13.2

10,277
9,381

45.7
35.2

20.7
14.3

927
829

37.5
27.5

22.1
14.9

9.19, 0.0028c
4.28, 0.0399c

1.69, 0.1945
0.48, 0.4884

30.53, 0.0000c
32.84, 0.0000c

53.44, 0.0000c
51.07, 0.0000c

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; n = number; QFT = Questionnaire Field Test; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the interview in English also have been excluded for these
comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Education and employment estimates are based only on respondents aged 18 or older. Sample sizes for respondents 18 or older are n = 37,659 for 2012 comparison, n = 18,007 for 2013 comparison,
and n = 2,305 for combined 2012 QFT and 2013 DR.
6
The Other Employment category includes students, persons keeping house or caring for children full time, retired or disabled persons, or other persons not in the labor force.
7
Respondents with unknown data were excluded.
8
Government Assistance is defined as one or more household family members having received Supplemental Security Income (SSI), cash assistance (Temporary Assistance for Needy Families, TANF),
noncash assistance, or food stamps.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.8b Demographic and Geographic Characteristics among Persons Aged 12 to 17 for English-Language Non-Hispanic Interviews:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal
2012 Comparison
(n = 17,573)1,2

2013 Comparison
(n = 8,610)1,3

Wtd
Unwtd
Percent
n

Unwtd
Percent

Combined 2012 QFT and
2013 DR
(n = 707)1,4

Unwtd
n

Unwtd
Percent

Wtd
Unwtd Unwtd
Wtd
Percent
n
Percent Percent

3,650
5,697
5,514
2,712

20.8
32.4
31.4
15.4

18.4
25.3
38.3
18.1

1,930
2,694
2,698
1,288

22.4
31.3
31.3
15.0

19.4
25.7
37.5
17.3

109
160
264
174

15.4
22.6
37.3
24.6

17.8
28.0
37.6
16.6

7,540
6,189
3,844

42.9
35.2
21.9

53.1
31.1
15.8

3,828
2,932
1,850

44.5
34.1
21.5

52.8
31.3
15.9

346
239
122

48.9
33.8
17.3

45.4
37.4
17.2

6,046
7,610
3,324
593

34.4
43.3
18.9
3.4

35.8
42.2
18.8
3.2

2,931
3,738
1,656
285

34.0
43.4
19.2
3.3

34.0
43.8
18.9
3.2

233
308
129
37

33.0
43.6
18.2
5.2

32.6
44.2
18.1
5.1

16,762

95.4

94.9

8,223

95.5

95.0

670

94.8

95.4

2,837
4,925
3,128
6,683

16.1
28.0
17.8
38.0

15.6
26.5
17.4
40.5

1,380
2,451
1,520
3,259

16.0
28.5
17.7
37.9

15.9
27.4
17.0
39.6

167
211
96
233

23.6
29.8
13.6
33.0

23.0
31.2
13.1
32.7

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Wtd
0.21, 0.8917

2013
2012
Comparison
Comparison
vs. Combined vs. Combined
QFT and DR QFT and DR
Chi-Square
Chi-Square
Statistic,
Statistic,
P Value
P Value
Wtd
Unwtd
0.18, 0.9117
11.81, 0.0000c

1.44, 0.2388

1.38, 0.2533

2.35, 0.0980

1.49, 0.2280

1.94, 0.1242

1.73, 0.1618

2.40, 0.0690

2.57, 0.0551

0.20, 0.6579
7.26, 0.0001c

0.15, 0.7035
6.32, 0.0004c

0.62, 0.4317
8.54, 0.0000c

0.93, 0.3349
8.55, 0.0000c

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
13.60, 0.0000c

(continued)

Characteristic
PARTICIPATED IN
GOVERNMENT
PROGRAM6
RECEIVED INCOME
Social Security
Supplemental Security
Income
Food Stamps
Welfare Payments
BETTER PROVIDER
OF INFORMATION5
USED PROXY

Unwtd
n

2012
2013
Comparison
Comparison
vs. Combined vs. Combined
QFT and DR QFT and DR
Chi-Square
Chi-Square
Statistic,
Statistic,
P Value
P Value
Wtd
Unwtd Unwtd
Wtd
Wtd
Wtd
Percent
n
Percent Percent

2013 Comparison
(n = 8,610)1,3

Unwtd
Wtd
Unwtd
Percent Percent
n

Unwtd
Percent

Combined 2012 QFT and
2013 DR
(n = 707)1,4

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd

4,538

25.8

24.9

2,334

27.1

26.1

209

29.6

29.5

2.79, 0.0963

1.37, 0.2432

3.05, 0.0822

1.20, 0.2744

2,139

12.2

12.3

1,093

12.7

11.7

13.0

12.5

0.02, 0.8865

0.22, 0.6433

0.33, 0.5645

0.04, 0.8348

1,460
3,706
714

8.3
21.1
4.1

8.1
20.3
3.8

767
1,920
331

8.9
22.3
3.8

8.2
21.6
3.4

65
176
40

9.2
24.9
5.7

9.6
25.2
5.2

0.97, 0.3253
3.74, 0.0544
1.81, 0.1804

0.69, 0.4062
1.80, 0.1814
3.77, 0.0537

0.58, 0.4483
3.33, 0.0693
3.89, 0.0501

0.05, 0.8174
1.42, 0.2353
5.50, 0.0200c

15,906
15,208

90.9
86.5

90.2
85.8

7,874
7,521

91.9
87.4

91.6
87.4

647
618

92.4
87.4

92.0
87.2

1.55, 0.2149
0.64, 0.4256

0.08, 0.7829
0.00, 0.9466

1.71, 0.1925
0.37, 0.5428

0.26, 0.6119
0.00, 0.9670

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; n = number; QFT = Questionnaire Field Test; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.8c Demographic and Geographic Characteristics among Persons Aged 18 to 25 for English-Language Non-Hispanic Interviews:
Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal
2012 Comparison
(n = 18,029)1,2

Combined 2012 QFT and
2013 DR
(n = 702)1,4

2012
2013
Comparison
Comparison
vs. Combined vs. Combined
QFT and DR QFT and DR
Chi-Square
Chi-Square
Statistic,
Statistic,
P Value
P Value
Wtd
Unwtd Unwtd
Wtd
Wtd
Wtd
Percent
n
Percent Percent
1.63, 0.1829
1.22, 0.3038
18.9
147
20.9
25.1
25.2
175
24.9
25.1
38.5
220
31.3
34.5
17.4
160
22.8
15.2
0.34, 0.7103
0.21, 0.8086
50.5
371
52.8
49.8
34.2
223
31.8
36.1
15.2
108
15.4
14.1
0.59, 0.6245
0.99, 0.3967
11.4
89
12.7
13.3
35.4
242
34.5
33.5
36.9
279
39.7
39.1
16.3
92
13.1
14.1

2013 Comparison
(n = 8,532)1,3

Unwtd
n

Unwtd
Percent

Wtd
Unwtd
Percent
n

Unwtd
Percent

3,978
5,623
5,634
2,794

22.1
31.2
31.2
15.5

18.8
24.6
37.7
18.9

1,811
2,756
2,700
1,265

21.2
32.3
31.6
14.8

7,664
6,721
3,644

42.5
37.3
20.2

52.8
32.9
14.3

3,668
3,160
1,704

43.0
37.0
20.0

2,525
6,233
6,311
2,960

14.0
34.6
35.0
16.4

13.9
32.9
36.3
16.9

1,006
3,107
3,118
1,301

11.8
36.4
36.5
15.2

12,161

67.5

66.5

5,853

68.6

67.0

455

64.8

66.7

7,221
4,940
2,075
3,793

40.1
27.4
11.5
21.0

38.5
28.0
11.9
21.5

3,572
2,281
952
1,727

41.9
26.7
11.2
20.2

40.2
26.8
11.2
21.8

276
179
96
151

39.3
25.5
13.7
21.5

40.8
25.9
12.5
20.8

5,196
7,840
4,042
951

28.8
43.5
22.4
5.3

30.4
42.5
22.0
5.1

2,415
3,719
1,933
464

28.3
43.6
22.7
5.4

28.4
44.2
22.3
5.2

202
290
175
35

28.8
41.3
24.9
5.0

28.2
43.1
23.9
4.8

14,626

81.1

6,904

80.9

552

78.6

80.2

2012
2013
Comparison Comparison
vs. Combined vs. Combined
QFT and DR QFT and DR
Chi-Square
Chi-Square
Statistic,
Statistic,
P Value
P Value
Unwtd
Unwtd
3.89, 0.0099c 4.66, 0.0036c

5.41, 0.0052c

4.46, 0.0127c

1.94, 0.1239

1.13, 0.3365

0.00, 0.9497
0.48, 0.6960

0.02, 0.8983
0.36, 0.7803

1.84, 0.1765
1.19, 0.3163

3.81, 0.0523
1.80, 0.1482

0.65, 0.5842

0.38, 0.7701

0.94, 0.4219

0.99, 0.3982

0.22, 0.6366

0.11, 0.7440

2.65, 0.1052

2.07, 0.1515
(continued)

Combined 2012 QFT and
2013 DR
(n = 702)1,4

2013 Comparison
(n = 8,532)1,3

Wtd
Unwtd
Percent
n

Unwtd
Percent

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
3.89, 0.0099c

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
4.93, 0.0025c

Unwtd
n

Unwtd
Percent

6,526
5,700
2,344
3,459

36.2
31.6
13.0
19.2

33.7
30.6
13.5
22.1

3,014
2,697
1,083
1,738

35.3
31.6
12.7
20.4

4,617

25.6

24.0

2,260

26.5

24.9

231

32.9

29.9

6.66, 0.0106c

4.34, 0.0385c

11.95, 0.0007c

8.24, 0.0045c

1,685

9.3

10.0

869

10.2

9.6

9.0

0.55, 0.4606

0.21, 0.6464

0.11, 0.7449

0.99, 0.3211

1,102
3,895
792

6.1
21.6
4.4

6.1
20.0
4.1

604
1,904
304

7.1
22.3
3.6

6.4
20.8
3.0

63
174
43

9.0
24.8
6.1

9.0
22.0
5.3

4.98, 0.0267c
0.85, 0.3578
1.77, 0.1848

3.63, 0.0583
0.27, 0.6014
8.11, 0.0048c

6.85, 0.0095c
2.23, 0.1373
3.44, 0.0652

2.50, 0.1154
1.25, 0.2649
8.76, 0.0034c

3,175
2,440

23.7
13.5

24.6
14.3

1,765
1,357

27.2
15.9

28.1
16.3

142
102

27.2
14.5

29.4
15.4

3.33, 0.0696
0.47, 0.4938

0.22, 0.6381
0.21, 0.6496

3.07, 0.0815
0.52, 0.4725

0.00, 0.9958
0.83, 0.3636

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; n = number; QFT = Questionnaire Field Test; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Table 4.8d Demographic and Geographic Characteristics among Persons Aged 26 or Older for English-Language Non-Hispanic
Interviews: Percentages, Chi-Square Test Statistic, and P Value, 2012 Comparison, 2013 Comparison, and Combined 2012
Questionnaire Field Test and 2013 Dress Rehearsal
2012 Comparison
(n = 19,630)1,2

Combined 2012 QFT and
2013 DR
(n = 1,603)1,4

2012
2013
2012
Comparison
Comparison
Comparison
vs. Combined vs. Combined vs. Combined
QFT and DR QFT and DR QFT and DR
Chi-Square
Chi-Square
Chi-Square
Statistic,
Statistic,
Statistic,
P Value
P Value
P Value
Wtd
Unwtd Unwtd
Wtd
Wtd
Wtd
Unwtd
Percent
n
Percent Percent
1.40, 0.2441
1.10, 0.3511
9.17, 0.0000c
19.2
297
18.5
19.9
24.0
374
23.3
27.0
37.3
543
33.9
35.1
19.5
389
24.3
18.0
1.39, 0.2516
1.06, 0.3476
7.22, 0.0009c
51.3
865
54.0
49.3
32.1
429
26.8
29.5
16.5
309
19.3
21.2
6.64, 0.0003c 12.25, 0.0000c
7.15, 0.0001c
9.6
181
11.3
12.2
29.1
391
24.4
26.3
26.6
524
32.7
31.2
34.7
507
31.6
30.3

2013 Comparison
(n = 9,475)1,3

Unwtd
n

Unwtd
Percent

Wtd
Unwtd
Percent
n

Unwtd
Percent

4,186
6,117
5,937
3,390

21.3
31.2
30.2
17.3

19.0
23.8
37.3
19.8

2,089
2,896
2,953
1,537

22.0
30.6
31.2
16.2

8,477
6,834
4,319

43.2
34.8
22.0

52.5
31.0
16.5

4,119
3,318
2,038

43.5
35.0
21.5

2,114
5,777
5,172
6,567

10.8
29.4
26.3
33.5

11.1
29.2
25.4
34.3

922
2,754
2,627
3,172

9.7
29.1
27.7
33.5

13,521

68.9

63.1

6,578

69.4

64.0

1,090

68.0

63.1

11,129
2,392
973
5,136

56.7
12.2
5.0
26.2

51.2
11.9
4.6
32.3

5,384
1,194
380
2,517

56.8
12.6
4.0
26.6

51.9
12.2
3.3
32.7

886
204
81
432

55.3
12.7
5.1
26.9

50.8
12.4
4.6
32.2

4,279
7,489
5,321
2,538

21.8
38.2
27.1
12.9

20.7
37.2
28.0
14.2

2,062
3,575
2,675
1,158

21.8
37.8
28.2
12.2

20.9
36.6
28.8
13.7

330
622
446
205

20.6
38.8
27.8
12.8

18.3
38.2
28.7
14.8

16,882

86.0

87.9

8,214

86.7

89.0

1,348

84.1

86.8

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
9.33, 0.0000c

6.36, 0.0021c

9.29, 0.0000c

0.00, 0.9856
0.06, 0.9788

0.25, 0.6183
1.96, 0.1205

0.39, 0.5335
0.35, 0.7925

1.02, 0.3142
1.37, 0.2546

1.10, 0.3506

1.50, 0.2154

0.41, 0.7479

0.64, 0.5913

0.90, 0.3430

3.90, 0.0497c

3.50, 0.0627

6.56, 0.0112c
(continued)

2013 Comparison
(n = 9,475)1,3

Wtd Unwtd
Percent
n

Unwtd
Percent

Combined 2012 QFT and
2013 DR
(n = 1,603)1,4

Wtd Unwtd Unwtd
Wtd
Percent
n
Percent Percent

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Wtd
5.14, 0.0019c

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Wtd
7.08, 0.0002c

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
3.70, 0.0126c

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value
Unwtd
4.99, 0.0023c

Unwtd
n

Unwtd
Percent

3,101
6,172
3,637
6,720

15.8
31.4
18.5
34.2

15.2
31.1
17.5
36.3

1,416
2,942
1,720
3,397

14.9
31.1
18.2
35.9

14.6
29.9
18.4
37.1

308
515
256
524

19.2
32.1
16.0
32.7

19.8
32.4
17.0
30.7

3,979

20.3

18.1

1,943

20.5

17.8

376

23.5

21.7

8.21, 0.0046c

9.42, 0.0024c

5.90, 0.0160c

5.00, 0.0265c

4,866

24.8

33.5

2,438

25.7

33.6

385

24.0

31.0

1.84, 0.1766

1.93, 0.1665

0.33, 0.5642

1.46, 0.2277

1,454
3,130
471

7.4
15.9
2.4

7.4
13.5
1.9

752
1,498
183

7.9
15.8
1.9

7.8
12.7
1.3

133
284
55

8.3
17.7
3.4

8.6
16.0
2.6

2.02, 0.1569
4.83, 0.0292c
3.44, 0.0651

0.78, 0.3793
9.21, 0.0027c
14.75, 0.0002c

1.42, 0.2355
2.06, 0.1526
5.26, 0.0228c

0.21, 0.6490
2.48, 0.1170
13.09, 0.0004c

1,201
930

7.7
4.7

7.7
4.8

638
503

8.6
5.3

9.2
5.8

138
109

11.0
6.8

10.9
6.6

10.31, 0.0015c
7.06, 0.0085c

2.49, 0.1161
1.29, 0.2582

15.78, 0.0001c
13.80, 0.0003c

7.57, 0.0065c
5.76, 0.0173c

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; n = number; QFT = Questionnaire Field Test; Unwtd = unweighted; Wtd = weighted.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

•

No significant differences were observed between the weighted estimates for being
currently employed in the combined QFT-DR sample (63.6 percent) versus the 2012
(63.6 percent) and 2013 (64.4 percent) comparison samples. In addition, no
significant differences were observed between the weighted estimates for
employment categories in the combined QFT-DR data versus the 2012 and 2013
comparison data.

•

No significant differences were observed between the weighted estimates for overall
health status (excellent, good, fair, or poor) in the combined QFT-DR sample versus
the 2012 and 2013 comparison samples. In addition, no significant differences were
observed between the weighted estimates for being covered by any type of health
insurance in the combined QFT-DR sample (86.8 percent) versus the 2012
(87.7 percent) and 2013 (88.5 percent) comparison samples.

•

The estimated weighted proportion of family income of less than $20,000 was higher
in both the combined QFT-DR data (23.1 percent) than in the 2012 (17.5 percent) and
2013 (17.3 percent) comparison data. The estimated weighted proportion of family
income greater than $75,000 was lower in the combined QFT-DR data (29.1 percent)
than in both the 2012 (34.9 percent) and 2013 (35.4 percent) comparison data.

•

The estimated weighted proportion participating in government programs was
significantly higher in the combined QFT-DR data (23.4 percent) than in the 2012
comparison data (19.5 percent) and the 2013 comparison data (19.4 percent).

•

Receipt of income from government programs was higher in the combined QFT-DR
data than in both comparison datasets for two sources—food stamps and welfare
payments. The estimated weighted proportion receiving food stamps was significantly
higher in the combined QFT-DR data (17.6 percent) than in the 2012 comparison data
(14.9 percent) and the 2013 comparison data (14.5 percent). The estimated weighted
proportion receiving welfare payments was significantly higher in the combined
QFT-DR data (3.2 percent) than in the 2012 comparison data (2.3 percent) and the
2013 comparison data (1.7 percent). The difference between the 2012 and 2013
comparison data in the weighted estimates for receipt of welfare payments was also
statistically significant.

•

In addition, the weighted estimate for receipt of Supplemental Security Income (SSI)
was significantly higher in the combined QFT-DR data (8.7 percent) than in the 2012
comparison data (7.3 percent). The difference between the combined QFT-DR
weighted estimate and the 2013 comparison weighted estimate (7.6 percent) was not
statistically significant.

•

Identification and use of a proxy reporter for the health insurance and income items
were higher in the combined QFT-DR data than in the 2012 comparison data, but the
differences for these two weighted items between the combined QFT-DR data and the
2013 comparison data were not statistically significant. The weighted proportion
indicating there was a better provider of information on health insurance and income
was significantly higher for the combined QFT-DR data (22.1 percent) than for the
2012 comparison data (18.7 percent), but not significantly higher than for the 2013
comparison data (20.7 percent). The difference between the 2012 and 2013
comparison data in the weighted estimates for a better provider of information on
91

health insurance and income was also statistically significant. Actual use of a proxy
respondent for the health insurance and income items (weighted) was significantly
higher for the combined QFT-DR data (14.9 percent) than for the 2012 comparison
data (13.2 percent). Although the weighted proportion for the use of a proxy in the
combined QFT-DR data was also slightly higher than for the 2013 comparison data
(14.3 percent), this difference was not statistically significant. The difference between
the 2012 and 2013 comparison data in the weighted estimates for the use of a proxy
was statistically significant.
Tables 4.8b through 4.8d provide the same geographic, demographic and household
estimates for the combined QFT-DR data and the 2012 and 2013 comparison data for three
specific age groups: 12 to 17, 18 to 25, and 26 or older. Many of the significant differences in
these estimates between the combined QFT-DR data and the 2012 and 2013 comparison data for
all of respondents aged 12 or older were also observed across these three specific age groups,
with the following exceptions:
•

The estimated proportions for the four-category education variable did not differ
significantly between the combined QFT-DR data and the 2012 and 2013 comparison
data for the 18 to 25 age group (Table 4.8c). Given that differences in the distribution
of the four education categories were significant between the combined QFT-DR data
and both the 2012 and 2013 comparison data for all respondents, this difference can
be attributed to differences in the 26 or older age group.

•

Not all observed differences between the combined QFT-DR data and the 2012 and
2013 comparison data for receipt of SSI, food stamps, and welfare payments among
all respondents were observed across all age groups. For respondents aged 12 to 17
(Table 4.8b), none of the weighted estimates for these three items differed
significantly between the combined QFT-DR data and the 2012 and 2013 comparison
data. Only two of these comparisons produced significantly different estimates among
respondents aged 18 to 25 (Table 4.8c). The weighted estimate for receipt of SSI in
the combined QFT-DR data was significantly higher than in the 2012 comparison
data (only), and the weighted estimate for receipt of welfare payments in the
combined QFT-DR data was significantly higher than in the 2013 comparison data
(only). The difference between the 2012 and 2013 comparison data in the weighted
estimates for receipt of welfare payments was statistically significant for respondents
aged 18 to 25. Among respondents aged 26 or older (Table 4.8d), most of the
significant differences between the combined QFT-DR data and the 2012 and 2013
comparison data observed for all respondents were significant for this age group, with
two exceptions. The difference in weighted estimates for receipt of SSI between the
combined QFT-DR data and the 2012 comparison data was not statistically
significant. Likewise, the difference in weighted estimates for receipt of welfare
payments between the combined QFT-DR data and the 2012 comparison data was not
statistically significant. The difference between the 2012 and 2013 comparison data in
the weighted estimates for receipt of welfare payments was statistically significant for
respondents aged 26 or older.

•

Identification and use of a proxy reporter for the health insurance and income items
(weighted) was not significantly higher in the combined QFT-DR data than in the
92

2012 comparison data for the 12 to 17 year old age group (Table 4.8b) or for the 18 to
25 year old age group (Table 4.8c). As with all respondents aged 12 or older,
however, identification of a better reporter for the health insurance and income items
and use of a proxy reporter were significantly higher in the combined QFT-DR data
than in the 2012 comparison data for the 26 or older age group (Table 4.8d).
Together, these findings indicate the overall difference for identification and use of a
proxy reporter was driven mostly by differences in the 26 or older age group.
In addition, for the weighted estimates for both a better provider of information on
health insurance and income and use of a proxy reporter, differences between the
2012 and 2013 comparison data in the weighted estimates were statistically
significant for respondents aged 26 or older.
Although many of the same differences in estimates for geographic, demographic, and household
items between the combined QFT-DR data and the 2012 or 2013 comparison data were also
observed across age groups, the degree and significance of these differences varied across
specific age groups.

4.3

Imputation Rates for Common 2012 Comparison Data, 2013 Quarters 3
and 4 Comparison Data, and Dress Rehearsal Variables

Another indicator of the quality of the DR data is the proportion of cases for which
imputation was required prior to using specific variables for analysis. For the DR data, 2012
comparison data, and 2013 quarters 3 and 4 comparison data, records with missing data were
subject to the same imputation procedures. However, when the values of other nonmissing
variables could be used to determine the value of the missing variable, the value was "logically
assigned" instead of imputed.
Tables 4.9a through 4.9f provide rates of imputation and logical assignment that selected
variables underwent in processing the 2012 comparison data, the 2013 quarters 3 and 4
comparison data, and the DR data. (Section 3.4 in Chapter 3 describes these imputation
procedures.) These tables include the following columns for the variables of interest:
•

number of respondents in domain (unweighted),

•

number of respondents whose values were imputed or logically assigned, and

•

weighted percentage (relative to their domain size) of respondents whose values were
imputed or logically assigned.

Table 4.9a Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Recency of Substance Use Variables

Recency Variable
Marijuana
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
Cigarettes
Smokeless Tobacco
Alcohol
Binge Alcohol Use
Pain Relievers
OxyContin®5
OxyContin® Past
Year Use5
Tranquilizers
Sedatives
Core-Plus-Noncore
Stimulants
Core-Plus-Noncore
Methamphetamine
Stimulants Excluding
Methamphetamine5

2012 Comparison Data1
Unweighted
Weighted
Respondents
Frequency
Percentage
in Domain4
26,388
64
0.2
6,910
42
0.5a
1,591
4
0.2
919
2
0.1
8,471
129
1.0
4,226
30
0.4
960
24
1.5
4,632
32
0.5a
5,467
107
1.3
33,085
34
0.0
11,010
30
0.4
46,257
78
0.1a
29,423
805
2.8
10,216
224
1.9
2,221
40
1.1

2013 Comparison Data1,2
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
12,585
28
0.2
3,145
13
0.3
731
2
0.3
467
1
0.2
3,999
51
0.9
2,068
17
0.6
414
7
0.4
2,221
12
0.4a
2,412
41
0.8
15,398
13
0.1
5,180
11
0.3
21,889
35
0.1a
13,905
359
2.1
4,537
87
1.7
1,048
19
1.7

N/A
5,639
1,084

N/A
40
13

N/A
0.7
0.4

N/A
2,578
494

N/A
20
6

4,714

114

1.9a

2,202

2,153

2.0a

1,005

N/A

N/A
0.6
0.6

2013 Dress Rehearsal1,3
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
831
2
0.2
267
0
0.0
70
0
0.0
32
0
0.0
316
7
1.6
158
2
1.2
38
1
2.8
164
0
0.0
225
7
1.0
1,054
1
0.1
256
2
0.3
1,543
0
0.0
919
23
2.6
257
4
1.3
N/A
N/A
N/A
32
96
48

6
2
0

19.6
1.7
0.0

1.6a

169

0.0

1.8a

107

0.0

N/A

DR = Dress Rehearsal; LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

The domain for all variables except Binge Alcohol Use includes all lifetime users of the drug in question. The domain for Binge Alcohol Use includes all past month users of alcohol.

OxyContin® recency was only available for the 2012 and 2013 comparison files; the DR only asked about past year use. Stimulant misuse excluding methamphetamine was only available on the DR.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.9b Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Past Year Initiation of Substance Use Variables

Past Year Initiation
Variable
Marijuana
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
Pain Relievers
OxyContin®5
Tranquilizers
Sedatives
Stimulants
Methamphetamine
Stimulants Excluding
Methamphetamine 5

2012 Comparison Data1
Unweighted
Weighted
Respondents
Frequency
Percentage
in Domain4
26,388
218
0.4
6,910
35
0.3
1,591
6
0.3
919
8
0.5a
8,471
88
0.5
4,226
43
0.4
960
23
1.4
4,632
49
0.5
5,467
176
1.7
10,216
391
3.3a
2,221
49
1.2
5,639
105
1.4a
1,084
23
0.4
4,320
74
0.8
1,652
28
0.8
N/A

N/A

2013 Comparison Data1,2
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
12,585
136
0.4
3,145
16
0.3
731
4
0.4
467
8
0.5a
3,999
41
0.5
2,068
25
0.8
414
7
0.4
2,221
19
0.4
2,412
75
1.2
4,537
161
2.8
1,048
22
3.1
2,578
50
1.4a
494
11
0.8
2,028
46
1.5
781
16
1.6
N/A

N/A

2013 Dress Rehearsal1,3
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
831
4
0.3
267
1
0.1
70
0
0.0
32
0
0.0
316
2
0.6
158
2
1.2
38
2
3.1
164
1
0.7
225
8
1.2
257
7
1.4
N/A
N/A
N/A
96
3
0.3
48
1
0.2
169
2
0.8
107
1
1.0
83

DR = Dress Rehearsal; LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine; PY = past year.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

The domain for all variables includes all lifetime users of the drug in question.

OxyContin® initiation was only available for the 2012 and 2013 comparison files. Stimulant misuse excluding methamphetamine was only available on the DR.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

0.3

Table 4.9c Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Age at First Use for Past Year Initiates

Age at First Use
Variable
Marijuana
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
Pain Relievers
OxyContin®5
Tranquilizers
Sedatives
Stimulants
Methamphetamine
Stimulants Excluding
Methamphetamine5

2012 Comparison Data1
Unweighted
Weighted
Respondents
Frequency
Percentage
in Domain4
1,853
9
0.4a
417
2
0.7
57
0
0.0
90
0
0.0
795
10
1.0a
319
6
1.8a
62
3
1.6
584
2
0.2
533
39
5.9a
1,184
68
4.2
241
3
0.5
672
17
2.2a
101
6
4.6
397
8
1.0a
81
3
1.7
N/A

N/A

2013 Comparison Data1,2
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
827
5
0.3
198
0
0.0
21
1
1.9
39
0
0.0
333
3
0.3
155
1
0.5
7
0
0.0
236
3
0.7
183
11
4.4a
463
29
4.1
93
1
0.4
274
6
1.4a
27
0
0.0
162
4
2.0
39
1
1.8
N/A

N/A

2013 Dress Rehearsal 1,3
Respondents
Unweighted
Weighted
in Domain4
Frequency
Percentage
35
0
0.0
12
0
0.0
0
0
0.0
2
0
0.0
18
0
0.0
7
0
0.0
0
0
0.0
10
0
0.0
13
0
0.0
26
5
13.8
N/A
N/A
N/A
11
2
45.9
7
1
1.2
16
0
0.0
4
0
0.0
14

DR = Dress Rehearsal; AFU = age at first use; LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

The domain for all variables includes past year initiates of the drug in question.

OxyContin® initiation was only available for the 2012 and 2013 comparison files. Stimulant misuse excluding methamphetamine was only available on the DR.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

0.0

Table 4.9d Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Selected Demographic and Socioeconomic Variables

Variable (Domain)
Detailed Race: 15 Levels
Hispanic or Latino Origin
Education Level
Marital Status (Age 15+)
Employment Status
(Age 15+)
Employment Status
(Age 18+)

2012 Comparison Data1
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage

2013 Comparison Data1,2

2013 Dress Rehearsal1,3

Respondents
in Domain

Unweighted
Frequency

Weighted
Percentage

Respondents
in Domain

Unweighted
Frequency

Weighted
Percentage

66,542
66,542
66,542
55,642

2,991
164
9
7

4.5
0.1a
0.0
0.0a

32,162
32,162
32,162
26,783

1,498
92
2
7

4.5
0.1a
0.0
0.0a

2,087
2,087
2,087
1,853

193
3
1
11

4.9
0.0
0.0
0.3

55,642

0.1a

26,783

0.0a

1,853

0.4

44,585

0.1

21,396

0.0a

1,582

0.3

DR = Dress Rehearsal.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.9e Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Health Insurance Variables

2012 Comparison Data1
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

2013 Comparison Data1,2
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

2013 Dress Rehearsal1,3
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

66,542

595

0.4

32,162

318

0.4

2,087

0.9

66,542
66,542

495
253

0.4
0.2

32,162
32,162

267
138

0.4
0.3

2,087
2,087

24
17

0.8
0.4

66,542
66,542

271
642

0.2
0.5

32,162
32,162

166
348

0.3
0.6

2,087
2,087

16
26

0.7
1.0

11,823

291

1.3

5,552

157

1.2

541

2.4

CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Department of Veteran's Affairs; CHIP = Children's Health
Insurance Program; DR = Dress Rehearsal; VA = Department of Veteran's Affairs.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.9f

Cases Imputed or Logically Assigned for the 2012 Main Study, 2013 Quarters 3 and 4 Main Study, and 2013 Dress
Rehearsal: Income Variables

Variable (Domain)
Total Family Income
> or < $20,000
Total Family Income
– Finer Categories
Source of Family
Income
Social Security or
Railroad
Retirement
Payments
Wages4
Public Assistance
Supplemental
Security
Income

Food Stamps
Welfare/Job
Placement/
Child Care
Number of
Months on
Welfare (Family
Receives Public
Assistance or
Welfare/Job
Placement/
Child Care)

2012 Comparison Data1
Respondents Unweighted
Weighted
in Domain
Frequency
Percentage

2013 Comparison Data1,2
Respondents in
Unweighted
Weighted
Domain
Frequency
Percentage

2013 Dress Rehearsal1,3
Respondents
Unweighted
Weighted
in Domain
Frequency
Percentage

66,542

2,904

4.0

32,162

1,461

4.0

2,087

134

4.4

66,542

7,880

14.9

32,162

3,937

15.4

2,087

317

13.8

66,542
66,542
66,542

671
199
509

0.6
0.2
0.4

32,162
32,162
32,162

353
99
271

0.6
0.2
0.5

2,087
N/A
2,087

33
N/A
32

1.2
N/A
0.9

66,542
66,542

903
316

0.8
0.3

32,162
32,162

495
145

1.0
0.3

2,087
2,087

42
26

1.4
0.9

66,542

388

0.3a

32,162

180

0.3a

2,087

0.8

4,687

210

4.0

2,173

110

3.5a

182

11.0

DR = Dress Rehearsal.
a

Difference between estimate and DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii.

Main survey data collected in quarter 3 and quarter 4, 2013, through December 2, 2013.

DR data collected from September 1 through November 3, 2013.

Family income from wages was only available for the 2012 and 2013 comparison files.

Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

As Table 4.9a shows, the weighted percentages of cases that were either imputed or
logically assigned were similar in all three datasets for the recency of substance use variables.
The following substances showed a statistically significant difference between the percentage
imputed or logically assigned in the DR dataset and the percentage imputed or logically assigned
in either the 2012 comparison dataset and/or the 2013 quarters 3 and 4 comparison dataset:
•

cocaine recency,

•

Ecstasy recency,

•

alcohol recency,

•

core-plus-noncore (CPN) stimulant recency, and

•

CPN methamphetamine recency.

In the pain relievers module of the main survey, respondents are asked about their most
recent use of OxyContin®. However, in the DR, respondents are asked whether or not they
misused OxyContin® in the past year. Therefore, direct comparisons between the 2012 and 2013
quarters 3 and 4 comparison data and the 2013 DR data were not possible for these measures.
However, as in the DR, respondents in the QFT were asked to report whether or not they
misused OxyContin® in the past year, so a direct comparison between the rates of imputation or
logical assignment can be made between the DR and the QFT. The weighted percentage of
imputed or logically assigned cases for past year misuse of OxyContin® in the QFT was
36.7 percent compared with 19.6 percent in the DR (see Table 4.9a). Although this difference
and the individual rates appear large in comparison with other measures, both the QFT and DR
estimates suffered from a small domain size and lack of power. As a result, this difference
between the QFT and DR for past year misuse of OxyContin® was not statistically significant at
5 percent. However, in all of these instances, no cases required logical assignment or imputation
in the DR data. This fact is likely attributable to the small domain sizes in the DR, and the
significance result therefore merely indicates that the estimate from the comparison dataset is
different from zero.
In Table 4.9b, the weighted percentage of cases that required either imputation or logical
assignment for past year initiation status was similar across all three datasets. For pain relievers
and tranquilizers, the weighted percentage of cases imputed or logically assigned was
significantly lower in the DR compared with the 2012 comparison data and/or the 2013 quarters
3 and 4 data. Statistically significant differences between the DR data and the comparison
datasets were also observed for heroin, although there were no cases in the DR data requiring
imputation or logical assignment.
As shown in Table 4.9c, the majority of the age at first use variables in the DR data
required no logical assignment or imputation. This occurrence is likely due to the very small
domains resulting from only defining the age at first use variables for past year initiates.
The very large weighted percentage of cases requiring imputation or logical assignment for the
pain relievers (13.8 percent) age at first use variables and the tranquilizers (45.9 percent) age at
first use variables in the DR data are an artifact of both small domain sizes and high variability in
the person-level weights in the DR.

100

The weighted percentages of cases that were either imputed or logically assigned in all
three datasets were low for most of the demographic variables presented in Table 4.9d. These
rates were generally similar across all three datasets and for all but one variable were below
0.5 percent. The weighted percentages of imputed or logically assigned cases for the detailed
race variable ranged between 4.5 and 4.9 percent across the three datasets.
In Table 4.9e, the weighted percentages of cases for the health insurance variables that
were either imputed or logically assigned in all three datasets were similar, and no statistically
significant differences were observed between the DR dataset and either the 2012 comparison
data or the 2013 quarters 3 and 4 data. Although no statistically significant differences were
observed, the weighted percentages of the imputed or logically assigned cases were higher for all
of these variables in the DR data than in the 2012 comparison data and the 2013 quarters 3 and 4
data. The health insurance questions were among the set of items moved from CAPI to ACASI
in the DR instrument, so the higher imputation rates observed could have resulted from DR
respondents being more likely to not answer this question. This outcome could also provide an
explanation for other questionnaire items moved from CAPI to ACASI in the DR instrument.
(See Section 4.4 for the complete results and a discussion of item missingness rates in the DR
data and the 2012 and 2013 quarters 3 and 4 comparison data.)
Weighted percentages for cases that were either imputed or logically assigned in all three
datasets for the income variables are shown in Table 4.9f. Not surprisingly, the weighted
percentages for some of the income variables were relatively high, such as the total family
income's finer categories and the number of months on welfare. For all three datasets, the rates
for the total family income's finer categories were similar, and all were greater than 13.8 percent.
This DR percentage was similar to the one in the QFT where 14.1 percent of cases were imputed
or logically assigned for the total family income's finer categories variable. For the number of
months on welfare variable and the indicator of whether the family received welfare, job
placement, or child care services, the percentage of cases requiring imputation or logical
assignment was significantly higher in the DR data. Similar results for this set of variables were
also observed in the QFT where 6.8 percent of cases were imputed or logically assigned for the
indicator of whether or not the family received welfare, job placement, or child services and
9.3 percent of cases were imputed or logically assigned for the number of months on welfare
variable. The differences observed between the QFT and the DR were not statistically significant
for any of the income variables shown in Table 4.9f. The questions about source of income and
total family income were among the items moved from CAPI to ACASI in the QFT and DR
instruments. For this reason, the differences in imputation rates when compared with the main
survey data could be attributed partially to the revised mode of administration. (See Section 4.4
for the complete results and a discussion of item missingness rates in the DR and the 2012 and
2013 quarters 3 and 4 comparison data.)

4.4

Comparisons of Item Missingness Rates for Moved DR Items with 2012
and 2013 Quarters 3 and 4 Comparison Data and Comparisons of Item
Missingness Rates for New or Revised DR Items with the QFT

To examine data quality among survey items in the DR questionnaire, this section
examines item missingness rates for three types of items in the DR:

101

•

items that were moved from CAPI to ACASI administration in the QFT and DR,

•

items that were introduced in the QFT and then revised between the QFT and DR,
and

•

items that were new to the DR questionnaire.

To consider potential data quality issues for these items in the 2015 NSDUH, this section
focuses primarily on DR questions with (1) missingness rates that were significantly higher than
the comparison datasets; (2) missingness rates that were not significantly higher than the
comparison datasets, but exhibited similar patterns as the significant items; and (3) notably high
missingness rates observed in DR that were not observed in the QFT. Given that none of these
items had high missingness rates that were not observed in the QFT, this section focuses on
highlighting the first two types of items.
Appendix B provides missingness rates for the following sets of DR items and the
datasets indicated in the tables' titles:
•

Table B.1. Item Missingness Rates for Moved Items for English-Language NonHispanic Interviews in 2012 Comparison Data, 2013 Comparison Data, and
Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal Data among All
Persons Aged 12 or Older;

•

Table B.2. Item Missingness Rates for Moved Items for Spanish-Language
Interviews in 2012 Comparison Data, 2013 Comparison Data, and 2013 Dress
Rehearsal Data among All Persons Aged 12 or Older;

•

Table B.3. Item Missingness Rates for New and Revised Items for English-Language
Non-Hispanic Interviews in the 2012 Questionnaire Field Test and 2013 Dress
Rehearsal among All Persons Aged 12 or Older; and

•

Table B.4. Item Missingness Rates for New and Revised Items for Spanish-Language
Interviews in the 2013 Dress Rehearsal among All Persons Aged 12 or Older.

All four tables in Appendix B provide unweighted numbers of cases with missing data and
weighted item missingness rates for each data set.
4.4.1

Item Missingness Rates for Items Moved from CAPI to ACASI Administration
in English-Language Non-Hispanic Interviews

As Table B.1 indicates, the following DR items moved to ACASI administration showed
higher missingness rates in the combined QFT-DR data than the same items in CAPI mode from
the comparison datasets, which was consistent with the pattern that was previously observed for
these items when comparing just the QFT data with the main study comparison datasets:
•

Item QD07 on marital status had a significantly higher item missingness rate in the
combined QFT and DR data (0.3 percent) than in both the 2012 comparison data
(0.0 percent) and the 2013 quarters 3 and 4 comparison data (0.1 percent). This item
is planned for ACASI administration in the partially redesigned 2015 questionnaire,
so this item will be examined closely in the 2015 6-month tables.

102

•

Item QD13, which asks about the number of home moves in the past year, had a
significantly higher item missingness rate in the combined QFT and DR data
(0.8 percent) than in both the 2012 comparison data (0.0 percent) and the 2013
quarters 3 and 4 comparison data (0.0 percent). This item is planned for ACASI
administration in the partially redesigned 2015 questionnaire, so this item will be
examined closely in the 2015 6-month tables.

•

Item QD19 on full- or part-time student status had a significantly higher item
missingness rate in the combined QFT and DR data (0.8 percent) than in both the
2012 comparison data (0.0 percent) and the 2013 quarters 3 and 4 comparison data
(0.1 percent). This item is planned for ACASI administration in the partially
redesigned 2015 questionnaire, so this item will be examined closely in the 2015 6month tables.

•

Item QD20 on missing school due to illness or injury and item QD21 on skipping
school days both had higher item missingness rates in the combined QFT and DR
data (0.7 and 0.6 percent, respectively) than in both the 2012 comparison data
(0.2 and 0.3 percent, respectively) or the 2013 quarters 3 and 4 comparison data
(0.1 and 0.2 percent, respectively). For both items, however, only the differences
between the combined QFT and DR data and the 2012 comparison data were
statistically significant. Furthermore, the missingness rates for QD20 and QD21 in the
DR were significantly lower than they were in the QFT. The higher missingness rates
for these two items shown in Table B.1 can therefore be attributed mainly to the
higher rates in the QFT data included in the combined QFT-DR dataset. This item is
planned for ACASI administration in the partially redesigned 2015 questionnaire, so
this item will be examined closely in the 2015 6-month tables.

•

Item QD26, which asks about work at a job or business at any time in the past week,
had a significantly higher item missingness rate in the combined QFT and DR data
(0.2 percent) than in both the 2012 comparison data (0.0 percent) and the 2013
quarters 3 and 4 comparison data (0.0 percent). This item is planned for ACASI
administration in the partially redesigned 2015 questionnaire, so this item will be
examined closely in the 2015 6-month tables.

•

Several items that ask about recent employment history, missing workdays, size of
employing organization, and related issues—QD31, QD33, QD36, QD38, QD39a,
QD40, QD41, and QD42—had significantly higher item missingness rates in the DR
data than in the 2012 or 2013 quarters 3 and 4 comparison data. The lower
missingness rates for all of these items were quite similar in the 2012 and 2013
quarters 3 and 4 comparison data. These items are planned for ACASI administration
in the partially redesigned 2015 questionnaire, so these items will be examined
closely in the 2015 6-month tables.

•

Two items asking about health insurance coverage, QHI02 on Medicaid plans and
QHI03 on military plans, had higher item missingness rates in the DR data than in the
2012 and 2013 quarters 3 and 4 comparison datasets. For item QHI03, however, only
the difference between the 2012 comparison data and the combined QFT-DR data
was statistically significant. The missingness rates for QHI02 were 0.4 percent for
both the 2012 and 2013 quarters 3 and 4 comparison data, but the rate was 0.9 percent

103

for the combined QFT-DR data. For item QHI03, the missingness rate was 0.2
percent for the 2012 comparison data, 0.3 percent for the 2013 quarters 3 and 4
comparison data, and 0.6 percent for the combined QFT-DR data. These two items
are planned to be moved back to CAPI administration in the partially redesigned 2015
questionnaire, so missingness rates for these two items will not be a concern in the
2015 data.
•

Furthermore, although the difference in missingness rates for item QHI06 (private
health insurance) followed a similar pattern previously observed for this item when
comparing just the QFT data with the main study comparison datasets, the differences
between the DR missingness rate (0.7 percent) and the two comparison datasets (both
0.4 percent) were not statistically significant. This item is planned to be moved back
to CAPI administration in the partially redesigned 2015 questionnaire, so missingness
rates for this item will not be a concern in the 2015 data.

•

Some of the items asking about receipt of various sources of income or participation
in government assistance programs—QI03N, QI08N, and QI10N—had significantly
higher item missingness rates in the DR data than in the 2012 or 2013 quarters 3 and
4 comparison data. Missingness rates for all of these items were quite similar in the
2012 and 2013 quarters 3 and 4 comparison data. Although the difference in
missingness rates for item QI06N (receipt of food stamps) followed a similar pattern
previously observed for this item when comparing just the QFT data with the main
study comparison datasets, the differences between the DR missingness rate (0.6
percent) and the two comparison datasets (both 0.3 percent) were not statistically
significant. These items are planned to be moved back to CAPI administration in the
partially redesigned 2015 questionnaire, so missingness rates for these items will not
be a concern in the 2015 data.

•

Two items on personal income levels—QI20N and QI21A—had higher item
missingness rates in the QFT data than in the 2012 or 2013 quarters 3 and 4
comparison data. For item QI21A, however, only the difference between the 2012
comparison data and the combined QFT-DR data was statistically significant. The
missingness rate for QI20N was close to 2 percent in both the 2012 and 2013 quarters
3 and 4 comparison data, but the rate was 3.4 percent for combined QFT-DR data.
For item QI21A, the missingness rate was 2.3 percent for the 2012 comparison data,
3.2 percent for the 2013 quarters 3 and 4 comparison data, and 4.5 percent for the
combined QFT-DR data. These two items are planned to be moved back to CAPI
administration in the partially redesigned 2015 questionnaire, so missingness rates for
these two items will not be a concern in the 2015 data.

The following two sets of items administered in ACASI for both the QFT and DR had
significantly lower missingness rates than the CAPI estimates for the 2012 and 2013 quarters 3
and 4 comparison data:
•

Items QD43, QD44, QD45, QD46, QD47, QD48, QD49, QD50, and QD51 on issues
related to workplace drug and alcohol use policies, including testing for drug and
alcohol use, had lower item missingness rates in the combined QFT-DR data
compared with the 2012 or 2013 quarters 3 and 4 comparison data. For item QD45,
however, only the difference between the 2013 quarters 3 and 4 comparison data and
104

the combined QFT-DR data was statistically significant. The higher missingness rates
for all of these items were generally quite similar in the 2012 and 2013 quarters 3 and
4 comparison data.
•

Several items asking about health insurance coverage, including whether private
health insurance was obtained through work (QHI07), coverage for treatment of
alcohol abuse (QHI08), coverage for treatment of drug abuse (QHI09), coverage for
treatment of mental health issues (QHI10), and the amount of time elapsed since
having any kind of health care coverage, had lower item missingness rates in the
combined QFT-DR data than in the 2012 or 2013 quarters 3 and 4 comparison data.
Missingness rates for QHI08 and QHI09 showed especially large differences, where
the missingness rates ranged from about 44 or 45 percent in the 2012 and 2013
quarters 3 and 4 comparison data, but were only 26 to 27 percent in the combined
QFT-DR data.

Overall, observed differences in missingness rates for items moved to ACASI
administration in the combined QFT-DR data followed very similar patterns previously observed
for the same items when comparing just the QFT data with the main study comparison datasets.
This result was true both for items that had higher missingness rates in the combined QFT-DR
data in relation to parallel CAPI items in the main study comparison datasets and those items that
produced relatively lower missingness rates in the combined QFT-DR data. A few items varied
in the degree of difference between the combined QFT-DR data in relation to parallel CAPI
items in the main study comparison datasets. For example, a few comparisons that were
significantly different when comparing just the QFT data with the main study comparison
datasets were not significant when the combined QFT-DR data were compared. Although these
findings suggest a slight decrease in missingness rates in the DR for a few items that were
significantly higher when comparing just the QFT data with the main study comparison datasets,
the overall patterns of missingness rates in the DR data were quite similar to the patterns
observed in the DR. None of the items moved to ACASI administration had notably higher
missingness rates in the DR data than were observed in the QFT data.
4.4.2

Item Missingness Rates for Items Moved from CAPI to ACASI Administration
in Spanish-Language Interviews

As Table B.2 indicates, relatively small sample sizes for the DR data and small sample
sizes for some moved items in the 2012 comparison data and the 2013 quarters 3 and 4 data
resulted in many missingness rate estimates with low precision (denoted by an asterisk). These
small sample sizes greatly limit the ability to draw clear conclusions about missingness rates in
the Spanish-language interviews for items moved to ACASI administration in the DR. Given
these limitations, only the following five items moved to ACASI administration showed
significantly higher missingness rates in the DR data from Spanish-language interviews than the
same items in CAPI mode from Spanish-language interviews in the comparison datasets:
•

Item QD13, which asks about the number of home moves in the past year, had a
significantly higher item missingness rate in the DR data (0.2 percent) than in both
the 2012 comparison data (0.0 percent) and the 2013 quarters 3 and 4 comparison
data (0.0 percent). The estimated missingness rate for the DR Spanish-language
interviews for item QD13 was based on only 8 cases with missing data from a total
105

set of 185 respondents and, therefore, was a low precision estimate. With this caveat,
this finding was consistent with the missingness rate pattern for this item based on the
data from English-language non-Hispanic interviews noted in Section 4.4.1.
•

Item QD38, which asks about the length of unemployed time during the past year,
had a significantly higher item missingness rate in the DR data (32.1 percent) than in
both the 2012 comparison data (0.0 percent) and the 2013 quarters 3 and 4
comparison data (0.0 percent). The estimated missingness rate for the DR Spanishlanguage interviews for item QD38 was based on only 5 cases with missing data from
a total set of 20 respondents and, therefore, was a low precision estimate. With this
caveat, this finding was consistent with the missingness rate pattern for this item
based on the data from English-language non-Hispanic interviews noted in Section
4.4.1.

•

Item QD40, which asks about the number of whole work days missed due to illness or
injury in the past month, had a significantly higher item missingness rate in the DR
data (6.2 percent) than in both the 2012 comparison data (0.3 percent) and the 2013
quarters 3 and 4 comparison data (0.0 percent). The estimated missingness rate for the
DR Spanish-language interviews for item QD40 was based on only 7 cases with
missing data from a total set of 80 respondents and, therefore, was a low precision
estimate. With this caveat, this finding was consistent with the missingness rate
pattern for this item based on the data from English-language non-Hispanic
interviews noted in Section 4.4.1.

•

Item QD41, which asks about the number of whole work days missed due to not
wanting to be at work in the past month, had a significantly higher item missingness
rate in the DR data (5.1 percent) than in both the 2012 comparison data (0.4 percent)
and the 2013 quarters 3 and 4 comparison data (0.0 percent). The estimated
missingness rate for the DR Spanish-language interviews for item QD41 was based
on only 6 cases with missing data from a total set of 80 respondents and, therefore,
was a low precision estimate. With this caveat, this finding was consistent with the
missingness rate pattern for this item based on the data from English-language nonHispanic interviews noted in Section 4.4.1.

•

Item QI12BN, which asks about the number of months receiving any type of welfare
or public assistance (not including food stamps), had a significantly higher item
missingness rate in the DR data (27.7 percent) than in both the 2012 comparison data
(2.3 percent) and the 2013 quarters 3 and 4 comparison data (3.4 percent). The
estimated missingness rate for the DR Spanish-language interviews for item QI12BN
was based on only 5 cases with missing data from a total set of 16 respondents and,
therefore, was a low precision estimate. With this caveat, this finding was consistent
with the missingness rate pattern for this item based on the data from Englishlanguage non-Hispanic interviews noted in Section 4.4.1. In the data from Englishlanguage non-Hispanic interviews, the missingness rate for this item appeared to be
higher in the combined QFT-DR data (7.2 percent) than in the 2012 (3.8 percent) or
2013 quarters 3 and 4 comparison data (2.9 percent), although these differences were
not statistically significant.

106

Similarly, only the following five items administered in ACASI for the DR had
significantly lower missingness rates than the CAPI estimates for the 2012 and 2013 quarters 3
and 4 comparison data based on Spanish-language interviews:
•

Item QD45, which asks about receiving any educational information regarding the
use of alcohol or drugs, had a significantly lower item missingness rate in the DR data
(0.0 percent) than in both the 2012 comparison data (0.6 percent) and the 2013
quarters 3 and 4 comparison data (0.2 percent). The estimated missingness rate for the
DR Spanish-language interviews for item QD45 was based on a total sample size of
80 respondents and, therefore, was a low precision estimate. With this caveat, this
finding was consistent with the missingness rate pattern for this item based on the
data from English-language non-Hispanic interviews noted in Section 4.4.1.

•

Item QD53, which asks about the likelihood of wanting to work for an employer that
randomly tests for drug or alcohol use, had a significantly lower item missingness rate
in the DR data (0.0 percent) than in both the 2012 comparison data (1.0 percent) and
the 2013 quarters 3 and 4 comparison data (0.8 percent). The estimated missingness
rate for the DR Spanish-language interviews for item QD53 was based on a total
sample size of 80 respondents and, therefore, was a low precision estimate. This
finding was not consistent with the missingness rate pattern for this item based on the
data from English-language non-Hispanic interviews noted in Section 4.4.1, where
the DR missingness rate was not significantly higher than the missingness rates for
the comparison samples.

•

Item QHI02, which asks about health insurance coverage through a Medicaid plan,
had a significantly lower item missingness rate in the DR data (0.0 percent) than in
the 2012 comparison data (0.6 percent) but not in the 2013 quarters 3 and 4
comparison data (0.0 percent). The estimated missingness rate for the DR Spanishlanguage interviews for item QHI02 was a low-precision estimate. This finding was
not consistent with the missingness rate pattern for this item based on the data from
English-language non-Hispanic interviews noted in Section 4.4.1, where the DR
missingness rate was significantly higher than the missingness rates for both
comparison samples.

•

Two items asking about health insurance coverage, including health insurance
coverage for treatment of alcohol abuse (QHI08) and coverage for treatment of drug
abuse (QHI09), both had significantly lower item missingness rate in the DR data
than in both the 2012 comparison data and the 2013 quarters 3 and 4 comparison data.
For both items, the missingness rates in the DR data were 9.1 percent. The
missingness rates for these two items were 34.9 and 36.1 percent, respectively, in the
2012 comparison data and 41.9 and 43.8 percent, respectively, in the 2013 quarters 3
and 4 comparison data. The estimated missingness rate for the DR Spanish-language
interviews for items QHI08 and QHI09 was based on only 3 cases with missing data
from a total set of 37 respondents and, therefore, was a low precision estimate. With
this caveat, these findings were consistent with the missingness rate pattern for this
item based on the data from English-language non-Hispanic interviews noted in
Section 4.4.1.

107

4.4.3

Item Missingness Rates for Revised or New Items in English-Language NonHispanic Interviews

As Table B.3 indicates, the 11 items that were either new or revised for the QFT or the
DR had relatively low missingness rates based on the data from English-language non-Hispanic
interviews. Of the 11 items, 5 had no missing data, and the small number of cases produced low
precision estimates for some items. Item QD10e asking about which specific immediate family
member was serving in the United States military had the highest weighted item missingness rate
among these items at 8.9 percent of the QFT and 2.7 percent for the DR, although this was a low
precision estimate in the DR data because only a small number of respondents were routed to this
item. In the DR, item CG26 on the first use of "smokeless" tobacco had the next highest
weighted missingness rate at 1.0 percent, followed by item QD10d asking about immediate
family members serving in the United States military at 0.3 percent. The two items new to the
DR instrument on sexual identity (QD62) and sexual orientation (QD63) had missingness rates
of 0.2 and 0.3 percent, respectively. Based on the English-language non-Hispanic interviews in
the QFT and DR, missing data appeared to be a potential data quality issue for only one of these
items, QD10e, but this conclusion is limited by the small number of DR respondents routed to
this item.
4.4.4

Item Missingness Rates for Revised or New Items in Spanish-Language
Interviews

As Table B.4 indicates, the 11 items that were either new or revised for the QFT or the
DR had relatively low missingness rates based on the data from the DR Spanish-language
interviews, consistent with the data from English-language non-Hispanic interviews shown in
Table B.3. Of the 11 items, 6 had no missing data, and the small number of cases produced low
precision estimates for some items. The two items new to the DR instrument on sexual attraction
(QD62) and sexual identity (QD63) had two of the highest weighted missingness rates among
this set of items at 1.3 and 7.8 percent, respectively. Item QD10d asking about immediate family
members serving in the United States military had a weighted missingness rate at 2.1 percent,
although this was a low precision estimate small number of respondents were routed to this item.
Item QD05 on race had a weighted missingness rate of 0.3 percent, and item QD11 on the
highest grade or year of school completed had a weighted missingness rate of 0.2 percent.
Overall, missing data did not appear to be a data quality issue for these items based on the
Spanish-language interviews in the DR, except for new item QD63 on sexual identity. Given
these results, item QD63 should be on the list of items to examine closely among Spanishlanguage interview data in the 2015 Early Data Review (EDR) and/or in the 2015 6-month
tables.

108

4.5

Comparisons of DR English-Language and Spanish-Language
Interview Timing Results with 2012 Comparison and 2013 Quarters 3
and 4 Comparison Interviews (Research Question 2)
4.5.1

Overall and Module Timing Results for the 2012 Main Study, 2013 Quarters 3
and 4 Main Study, 2012 Questionnaire Field Test, and 2013 Dress Rehearsal

4.5.1.1

Overall and Module Timing Data for English-Language Interviews from
Non-Hispanic Respondents

To assess interview timing for the partially redesigned DR instrument administered in
English, Tables 4.10a through 4.10f provide mean and median timing results by module for the
2012 main study comparison data, the 2013 quarters 3 and 4 comparison data,14 the QFT data,
and the DR data.15 These tables also include combined timing data for the QFT and DR.
The comparisons include timing results for all respondents in each of the three sets of interviews
and separate timing results for five age categories (i.e., 12 to 17, 18 to 25, 26 to 49, 50 to 64, and
65 or older). The age group timing results provide data on how age is related to interview
duration for the partially redesigned DR questionnaire and how this information compares with
the QFT and current main study timing results. Respondents with an overall administration time
of less than 30 minutes or greater than 240 minutes were classified as outliers and were excluded
from the timing results.16
Administration times for all four datasets were calculated according to the standard
NSDUH timing data calculation procedures. One necessary variation to the timing calculations
was creating an "administrative residual" category to capture small amounts of additional
interviewing time that did not clearly fall within a defined interview section. Because the
administrative residual timings differed in the revised QFT and DR protocols compared with the
2012 main study and the 2013 quarters 3 and 4 protocol, accounting for this time in all of the
datasets allowed for more direct and accurate comparisons of overall and section timings across
the datasets. In addition, the administrative residual category provides the ability to add mean
section timings to produce the mean overall timing. For each of the four sets of respondents,

The timing data for the 2013 quarters 3 and 4 comparison dataset in the draft DR report include
interviews through December 1, 2013.
15
For readability, Tables 4.10a through 4.10f appear together at the end of this discussion in
Section 4.5.1.1.
16
Because the DR interviews included a higher number of cases with extreme values, which were excluded
from this Table 4.13 series of tables (as indicated in footnote 1), the overall mean and median timings for the DR,
QFT, 2012 comparison data, and 2013 comparison data interviews were also calculated with the extreme values
included. Including the extreme cases had minimal impact on the overall mean and median interview times for the
2012 and 2013 comparison data. Among extreme cases, the number of interviews and differences in timing for
shorter interviews had a slightly greater impact on the mean overall timing than the number of interviews and the
differences in timing for the longer interviews. The impact on the overall mean and median interview times for the
DR was somewhat greater, resulting in decreases of about 0.5 minute in the mean and about 0.7 minute in the
median timing. Given that including the extreme cases resulted in slightly decreased overall mean and median
interview times for the DR, including the extreme cases would lead to similar conclusions as those drawn from
comparing the DR timing data with the QFT, 2012, and 2013 comparison data interviews with the extreme cases
excluded.

109

the mean overall interview time can be calculated by adding the following mean section times,
which are presented in boldface in Tables 4.10a through 4.10f:
•

introduction,

•

core demographics,

•

calendar,

•

beginning ACASI,

•

tutorial,

•

total core substances,

•

special drugs to consumption of alcohol,

•

back-end demographics,

•

household roster,

•

proxy information/decision,

•

proxy tutorial,

•

health insurance,

•

income, and

•

verification.

Table 4.10a shows that overall interview times were somewhat lower for all DR
respondents aged 12 or older (mean 59.24, median 55.60) compared with all 2012 respondents
(mean 60.69, median 57.87) and the 2013 quarters 3 and 4 respondents (mean 62.03, median
58.75). However, the overall interview times were somewhat higher for the DR respondents aged
12 or older than they were for all QFT respondents (mean 59.13, median 55.60). Overall
interview times were lower or similar for DR respondents compared with the 2012 respondents,
the 2013 quarters 3 and 4 respondents, and the QFT respondents for most age groups, as shown
in Tables 4.10b through 4.10f. There were two exceptions to this pattern. The overall timing for
English-speaking, non-Hispanic DR respondents aged 26 to 49 was slightly higher than it was
for QFT respondents aged 26 to 49. Also, a larger gap was evident between DR and QFT
respondents aged 65 or older in that DR respondents aged 65 or older had a mean administration
time of 82.60 minutes, while QFT respondents in the same age group had a mean time of 80.47
minutes. Patterns of overall interview timing across the five age groups were generally similar
for the four sets of respondents, where respondents aged 12 to 17 and those aged 50 or older had
higher overall timings than those aged 18 to 49. For all sets of respondents, the mean and median
overall interview times were greatest for those aged 65 or older.
The first five sections in the partially redesigned DR questionnaire—introduction, core
demographics, calendar, beginning ACASI, and tutorial—took less or similar time to administer
for most respondents compared with the 2012, 2013, and QFT questionnaire. Timings for these
sections varied, so a few exceptions to this general pattern were observed. These exceptions can
be seen in Tables 4.10b through 4.10f.

110

As expected, the average timing for the total core substance use sections for all
respondents aged 12 or older was higher for the DR respondents (mean 13.27, median 11.58)
than it was for the 2012 respondents (mean 12.16, median 11.03) and the 2013 quarters 3 and 4
respondents (mean 12.09, median 10.87). The average timing for the total core substance use
sections for all respondents aged 12 or older was only slightly lower for the DR respondents than
it was for the QFT respondents (mean 13.57, median 11.68). Additions and revisions to the
hallucinogens, inhalants, and prescription drug sections in the partially redesigned DR
questionnaire contributed to higher administration times among DR respondents for the core
substance use modules when compared with the main study data. Combining the smokeless
tobacco items appeared to contribute to lower average timings for the tobacco section for the DR
respondents compared with the 2012 and 2013 quarters 3 and 4 respondents across all age
groups. This pattern also held between the QFT and DR data across all age groups. Timing
differences between the DR respondents versus the 2012, 2013 quarters 3 and 4, and QFT
respondents for the remaining core substance use modules—alcohol, marijuana, cocaine and
crack, and heroin—were generally small and inconsequential.
Timings for the redesigned prescription drug modules are of particular interest, given the
considerable changes made to these modules in the QFT and DR questionnaires. The average
total timing for the four prescription drug modules for the DR respondents aged 12 or older
(mean 5.88, median 4.77) was clearly higher than it was for the 2012 respondents (mean 5.31,
median 4.75) and the 2013 quarters 3 and 4 respondents (mean 5.30, median 4.68). The average
total timing for these sections in the DR was lower than it was for the QFT respondents (mean
5.96, median 4.92). Among the redesigned prescription drug modules, the pain relievers module
accounted for the higher administration times for the DR respondents compared with the 2012
and 2013 quarters 3 and 4 respondents. Administering the pain relievers module to the QFT
respondents took longer than administering it to the DR respondents (3.00 QFT, 2.92 DR).
Average timings for the other three prescription drug modules—tranquilizers, stimulants, and
sedatives—were similar or lower among the four sets of respondents. Administration times
varied across age groups among the DR, QFT, 2012, and 2013 quarters 3 and 4 respondents.
For example, Table 4.10b shows that DR respondents aged 12 to 17 and DR respondents aged 18
to 25 had lower overall total prescription drug timing results than did respondents in the same
age groups in the 2012, 2013, and QFT comparison samples. The overall average timing for the
prescription drug modules was increased among the DR respondents by higher administration
times for adult respondents aged 26 or older. In addition, the timing differences between the DR
respondents and the 2012, 2013 quarters 3 and 4, and QFT respondents increased steadily across
the four adult age groups, so that differences among the four sets of respondents were most
pronounced among those aged 65 or older (Table 4.10f).
For questionnaire sections from special drugs to consumption of alcohol, administration
times for DR respondents aged 12 or older varied in relation to the section timings for the 2012
respondents and the 2013 quarters 3 and 4 respondents. Sections with lower DR timings
compared with the 2012 and 2013 quarters 3 and 4 interviews included special drugs,
risk/availability, prior substance use, youth experiences, youth mental health service utilization,
adolescent depression, and consumption of alcohol. The lower administration times for special
drugs and prior substance use appeared to result from the deletion of one or more items from
these sections in the DR questionnaire. However, the QFT and DR timing of the youth
experiences module was about 1 minute shorter compared with timing results from the annual
111

survey. One brief item (YE04, number of times moved in the past 5 years) was deleted from the
module, so the reason for this decrease is unclear. DR administration times were higher than
those in the 2012 and 2013 quarters 3 and 4 interviews for the following sections, despite few
changes to these sections in the DR questionnaire: substance dependence and abuse, adult mental
health services utilization, social environment, parenting experiences, and mental health. The DR
modules had similar, but higher, administration times compared with the QFT for each of these
modules, except for substance dependence and abuse.
For the remaining sections from special drugs to consumption of alcohol, administration
times for DR respondents were generally similar to the section timings for the 2012 and 2013
quarters 3 and 4 respondents. Administration times in the DR instrument were similar to
administration times in the QFT instrument across all modules.
Section timings for the remaining back-end modules also varied for all respondents aged
12 or older when comparing DR respondents with 2012 and 2013 quarters 3 and 4 respondents,
based mostly on changes made to the DR questionnaire. Administration times between the DR
and QFT instruments were similar across modules. For example, under back-end demographics,
the average times for the DR respondents compared with 2012 and 2013 quarters 3 and 4
respondents were almost identical for education, but lower for employment. These findings are
consistent with the changes to the DR questionnaire, such as deleting questions on industry and
occupation from the employment section. As expected, these results were similar in the QFT and
DR data.
For the health insurance section, a higher average administration time was observed for
the DR respondents compared with the 2012 respondents and the 2013 quarters 3 and 4
respondents. The average administration times for this section were similar in the QFT and DR
data. The only change to this section in the DR questionnaire, relative to the main study
instrument, was moving these questions from CAPI to ACASI administration. Only the 18 to 25
year old age group had lower administration times compared with the annual NSDUH. One
possible explanation for the increased time among the DR respondents was that proxy reporters
answered these questions in the DR and the health insurance module is the first section after the
proxy tutorial. One consequence of this sequence is that DR proxy reporters might have used
additional time getting accustomed to the interview protocol, including the relationship fills.
The income section was also moved from CAPI to ACASI administration in the DR
questionnaire, and a new question on household telephone service was added to this section.
These changes corresponded with lower timings for the DR respondents compared with the 2012
respondents and the 2013 quarters 3 and 4 respondents for those aged 12 to 64. The 65 or older
age group had higher timings compared with the timing results of their counterparts from the
annual NSDUH. Comparisons with the QFT results revealed timings that were largely similar
between the two field tests.

112

Spanish-speaking respondents were oversampled in the DR to allow for adequate testing
of the Spanish questionnaire. Experience with the current NSDUH questionnaire shows that
interviews administered in Spanish produce longer overall average administration times. This
finding was replicated in the DR, where the mean overall interview time in Spanish was about 24
minutes longer than mean interview time in English.
To avoid artificially inflating the overall administration time of the DR interviews due to
the oversample of Spanish-speaking respondents, the DR timing results provided in Tables 4.10a
through 4.10f were based only on English-language interviews with non-Hispanic respondents.
To provide a sense of the expected interview length for both English-language and Spanishlanguage interviews in 2015, Table 4.10g presents timing results for all DR respondents,
regardless of language of administration. Average administration times for the DR questionnaire
among all respondents was 61.77 minutes, which was very similar to the average administration
times in the 2012 and 2013 main study comparison data. Despite the large difference between the
overall average timings for the English-language and Spanish-language interviews, the overall
average timing for the entire DR sample was only about 2 minutes greater than the average time
for the English-language interviews with non-Hispanic respondents. Taking into account the
oversample of Spanish-speaking respondents, the average interview administration time for the
partially redesigned 2015 NSDUH is expected to be about 60 minutes.

113

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

114

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.80
2.13
1.66
2.38
3.42
12.16
1.98
2.12
0.49
0.21
0.10
0.81
1.14
N/A
5.31
N/A
N/A
N/A
N/A

1.65
1.77
1.50
2.18
3.23
11.03
1.68
1.98
0.37
0.13
0.08
0.62
0.90
N/A
4.75
N/A
N/A
N/A
N/A

1.69
2.16
1.64
2.35
3.46
12.09
1.94
2.10
0.49
0.21
0.10
0.81
1.14
N/A
5.30
N/A
N/A
N/A
N/A

1.55
1.78
1.48
2.13
3.25
10.87
1.63
1.93
0.35
0.13
0.08
0.62
0.88
N/A
4.68
N/A
N/A
N/A
N/A

1.56
2.03
1.14
2.24
3.35
13.57
1.86
2.23
0.51
0.22
0.10
1.16
1.32
0.20
5.96
2.40
0.89
0.92
0.80

1.45
1.63
1.17
2.03
3.13
11.68
1.48
2.05
0.38
0.13
0.08
0.90
1.05
0.15
4.92
2.02
0.70
0.75
0.63

1.56
2.04
1.15
2.20
3.42
13.27
1.77
2.14
0.51
0.24
0.10
1.12
1.30
0.21
5.88
2.38
0.90
0.93
0.81

1.42
1.63
1.13
1.95
3.18
11.58
1.40
1.98
0.37
0.13
0.08
0.85
1.00
0.15
4.77
1.97
0.68
0.73
0.62

1.56
2.03
1.15
2.22
3.38
13.44
1.82
2.19
0.51
0.23
0.10
1.14
1.31
0.20
5.92
2.39
0.89
0.92
0.81

1.43
1.63
1.15
2.00
3.15
11.63
1.45
2.02
0.38
0.13
0.08
0.88
1.02
0.15
4.85
2.00
0.68
0.73
0.63

2.06

1.87

2.06

1.85

3.00

2.43

2.92

2.35

2.97

2.40

1.14

0.98

1.14

0.97

1.05

0.77

1.06

0.75

1.06

0.75

1.16

0.97

1.16

0.95

1.03

0.78

1.02

0.75

1.03

0.77

0.94

0.75

0.94

0.75

0.87

0.67

0.87

0.65

0.87

0.65
(continued)

115

Special Drugs to
Consumption of Alcohol
Special Drugs
Risk/Availability
Blunts
Substance Dependence and
Abuse
Market Information for
Marijuana
Prior Substance Use
Special Topics, Drug
Treatment
Health Care
Adult Mental Health Service
Utilization
Social Environment
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health
Service Utilization
Adolescent Depression
Consumption of Alcohol
Back-End Demographics
(Moves, Born in U.S.,
Disability, Education and
Employment)
Education
Employment

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

21.62
1.58
2.91
0.27

19.90
1.43
2.63
0.20

23.34
1.59
2.95
0.30

21.57
1.42
2.65
0.20

20.43
0.57
2.86
0.28

18.79
0.52
2.54
0.20

20.24
0.57
2.86
0.29

18.63
0.50
2.60
0.20

20.35
0.57
2.86
0.28

18.70
0.52
2.57
0.20

2.18

1.62

2.08

1.52

2.30

1.73

2.20

1.70

2.26

1.72

0.28
1.24

0.00
0.95

0.27
1.18

0.00
0.88

N/A
1.13

N/A
0.97

N/A
1.10

N/A
0.88

N/A
1.11

N/A
0.93

1.61
1.29

1.32
1.08

1.60
2.96

1.32
2.62

1.66
2.79

1.37
2.45

1.61
2.76

1.33
2.43

1.64
2.78

1.35
2.43

0.82
0.97
0.14
2.61
2.14
1.14

0.65
1.02
0.00
0.00
1.87
0.32

0.78
0.94
0.14
2.70
2.04
1.10

0.62
0.98
0.00
0.00
1.72
0.30

0.88
0.95
0.19
1.96
2.33
1.22

0.73
0.95
0.00
0.00
2.05
0.38

0.93
0.99
0.24
1.68
2.44
1.32

0.73
1.00
0.00
0.00
2.18
0.42

0.90
0.97
0.21
1.84
2.38
1.26

0.73
0.97
0.00
0.00
2.12
0.40

0.60
0.52
0.54

0.00
0.00
0.47

0.63
0.57
0.52

0.00
0.00
0.43

0.45
0.41
0.46

0.00
0.00
0.40

0.38
0.42
0.44

0.00
0.00
0.42

0.42
0.42
0.45

0.00
0.00
0.40

4.55
0.57
3.62

4.58
0.43
3.87

4.36
0.55
3.53

4.38
0.42
3.75

3.93
0.83
1.78

3.62
0.66
1.70

4.37
0.84
1.68

4.02
0.70
1.55

4.13
0.84
1.74

3.78
0.68
1.63
(continued)

Table 4.10a Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 12 or Older) (continued)
Overall
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative
Residual
Overall Questionnaire

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.63

1.40

1.64

1.40

1.45

1.25

1.55

1.22

1.49

1.23

0.56
N/A
1.39
3.58
3.12

0.33
N/A
1.28
3.22
2.67

0.61
N/A
1.38
3.50
3.41

0.35
N/A
1.27
3.10
2.87

0.57
0.70
1.57
3.10
3.34

0.43
0.00
1.35
2.70
2.85

0.63
0.69
1.56
3.00
3.40

0.38
0.00
1.35
2.52
2.85

0.60
0.70
1.57
3.06
3.37

0.42
0.00
1.35
2.62
2.85

0.68
60.69

N/A
57.87

0.38
62.03

N/A
58.75

0.14
59.13

N/A
55.60

0.16
59.24

N/A
55.60

0.15
59.18

N/A
55.60

116

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** QFT and DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .26 to .39
minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

117

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.88
2.00
1.65
2.41
3.65
11.70
1.70
1.56
0.45
0.17
0.10
0.86
1.34
N/A
5.53
N/A
N/A
N/A
N/A

1.77
1.68
1.52
2.25
3.55
10.75
1.43
1.32
0.32
0.13
0.08
0.70
1.10
N/A
5.03
N/A
N/A
N/A
N/A

1.73
2.00
1.64
2.37
3.73
11.74
1.69
1.54
0.45
0.17
0.10
0.86
1.34
N/A
5.59
N/A
N/A
N/A
N/A

1.58
1.68
1.50
2.20
3.62
10.68
1.43
1.27
0.32
0.13
0.08
0.70
1.10
N/A
5.05
N/A
N/A
N/A
N/A

1.58
1.88
1.20
2.22
3.38
11.81
1.42
1.62
0.49
0.16
0.09
1.21
1.46
0.21
5.14
2.33
0.79
0.82
0.72

1.52
1.53
1.22
2.12
3.36
10.54
1.13
1.32
0.38
0.13
0.08
0.98
1.19
0.18
4.45
2.01
0.65
0.70
0.60

1.67
1.79
1.21
2.21
3.50
10.75
1.39
1.36
0.45
0.17
0.09
1.01
1.34
0.21
4.74
2.20
0.72
0.77
0.64

1.65
1.58
1.20
2.00
3.32
10.00
1.10
1.05
0.32
0.13
0.07
0.83
1.12
0.18
4.18
1.98
0.62
0.67
0.53

1.62
1.85
1.21
2.21
3.43
11.38
1.41
1.52
0.47
0.16
0.09
1.13
1.41
0.21
4.98
2.28
0.76
0.80
0.69

1.58
1.55
1.20
2.07
3.33
10.35
1.12
1.20
0.35
0.13
0.08
0.93
1.17
0.18
4.32
2.00
0.63
0.70
0.57

2.13

2.00

2.15

2.00

2.65

2.28

2.49

2.23

2.58

2.25

1.19

1.05

1.20

1.05

0.85

0.66

0.75

0.62

0.81

0.65

1.21

1.03

1.22

1.03

0.90

0.72

0.84

0.68

0.88

0.72

1.01

0.83

1.02

0.83

0.75

0.60

0.67

0.55

0.71

0.58
(continued)

118

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

21.88
1.66
2.95
0.25

20.48
1.57
2.73
0.20

23.79
1.68
3.04
0.27

22.38
1.57
2.80
0.20

20.39
0.53
2.77
0.28

19.04
0.50
2.50
0.20

19.97
0.50
2.67
0.26

19.02
0.50
2.57
0.20

20.22
0.52
2.73
0.27

19.02
0.50
2.52
0.20

0.88

0.00

0.79

0.00

0.86

0.00

0.78

0.00

0.83

0.00

0.20
0.56

0.00
0.00

0.19
0.50

0.00
0.00

N/A
0.48

N/A
0.00

N/A
0.43

N/A
0.00

N/A
0.46

N/A
0.00

1.33
1.33
N/A

1.15
1.17
N/A

1.34
3.02
N/A

1.13
2.73
N/A

1.27
2.71
N/A

1.08
2.44
N/A

1.26
2.66
N/A

1.10
2.52
N/A

1.26
2.69
N/A

1.08
2.48
N/A

N/A
N/A
8.17
N/A
N/A

N/A
N/A
7.73
N/A
N/A

N/A
N/A
8.11
N/A
N/A

N/A
N/A
7.67
N/A
N/A

N/A
N/A
7.81
N/A
N/A

N/A
N/A
7.25
N/A
N/A

N/A
N/A
7.58
N/A
N/A

N/A
N/A
7.13
N/A
N/A

N/A
N/A
7.72
N/A
N/A

N/A
N/A
7.18
N/A
N/A

1.88
1.62
0.28

1.58
0.63
0.00

1.88
1.70
0.25

1.57
0.65
0.00

1.79
1.65
0.24

1.50
0.60
0.00

1.70
1.90
0.23

1.57
0.67
0.00

1.75
1.75
0.23

1.52
0.62
0.00

2.59
0.89
1.39

1.75
0.83
0.30

2.49
0.84
1.42

1.63
0.78
0.28

3.33
1.27
0.76

3.04
1.16
0.49

3.69
1.30
0.76

3.42
1.20
0.62

3.48
1.28
0.76

3.22
1.17
0.55
(continued)

Table 4.10b Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 12 to 17) (continued)
12 to 17
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

2.13

1.88

2.14

1.87

1.92

1.73

1.88

1.62

1.90

1.67

1.01
N/A
1.39
3.79
3.20
0.51
59.78

0.77
N/A
1.28
3.45
2.75
N/A
57.73

1.05
N/A
1.39
3.65
3.51
0.35
61.58

0.78
N/A
1.27
3.25
2.97
N/A
58.93

0.90
2.04
1.73
3.35
3.15
0.12
59.02

0.78
2.25
1.53
2.95
2.78
N/A
56.85

1.18
2.28
1.83
3.36
3.40
0.12
58.84

0.82
2.60
1.68
2.90
2.78
N/A
56.35

1.01
2.14
1.77
3.35
3.25
0.12
58.95

0.78
2.38
1.58
2.93
2.78
N/A
56.60

119

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

120

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.70
2.07
1.64
2.29
2.95
11.41
1.98
2.21
0.54
0.20
0.09
0.71
0.90
N/A
4.78
N/A
N/A
N/A
N/A

1.60
1.73
1.48
2.10
2.78
10.32
1.68
2.07
0.38
0.12
0.07
0.50
0.72
N/A
4.30
N/A
N/A
N/A
N/A

1.60
2.05
1.62
2.26
2.97
11.29
1.92
2.20
0.54
0.20
0.09
0.72
0.89
N/A
4.74
N/A
N/A
N/A
N/A

1.52
1.73
1.47
2.05
2.78
10.17
1.62
2.02
0.37
0.12
0.07
0.50
0.70
N/A
4.18
N/A
N/A
N/A
N/A

1.36
1.85
0.95
2.17
2.73
12.04
1.82
2.10
0.55
0.20
0.09
0.96
0.99
0.15
5.17
1.91
0.69
0.69
0.58

1.42
1.58
0.92
2.03
2.63
10.75
1.55
2.03
0.37
0.12
0.07
0.67
0.83
0.11
4.33
1.73
0.57
0.62
0.53

1.33
2.01
0.94
2.10
2.73
11.01
1.61
2.00
0.55
0.20
0.08
0.88
0.94
0.15
4.59
1.86
0.65
0.71
0.56

1.33
1.59
0.88
1.91
2.50
9.76
1.38
1.90
0.39
0.10
0.07
0.62
0.72
0.10
3.91
1.60
0.55
0.60
0.48

1.35
1.92
0.94
2.14
2.73
11.60
1.73
2.06
0.55
0.20
0.08
0.93
0.97
0.15
4.92
1.89
0.67
0.70
0.57

1.40
1.58
0.90
1.98
2.58
10.20
1.50
1.98
0.38
0.10
0.07
0.63
0.78
0.10
4.12
1.68
0.57
0.60
0.52

1.96

1.73

1.92

1.70

2.65

2.16

2.30

1.92

2.50

2.08

1.02

0.87

1.02

0.85

0.93

0.61

0.81

0.57

0.88

0.60

1.03

0.85

1.02

0.83

0.96

0.67

0.90

0.65

0.93

0.67

0.78

0.63

0.79

0.63

0.53

0.58

0.50

0.61

0.53
(continued)

121

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

19.95
1.42
2.50
0.31

18.32
1.28
2.30
0.22

21.22
1.41
2.53
0.36

19.48
1.27
2.30
0.23

17.83
0.49
2.36
0.33

16.47
0.43
2.15
0.23

16.98
0.48
2.34
0.36

15.64
0.42
2.12
0.23

17.46
0.49
2.35
0.35

16.32
0.43
2.13
0.23

3.01

2.37

2.84

2.22

3.05

2.30

2.59

2.08

2.85

2.18

0.46
1.44

0.00
1.17

0.45
1.37

0.00
1.10

N/A
1.15

N/A
1.00

N/A
1.12

N/A
0.94

N/A
1.14

N/A
0.98

1.58
1.01

1.30
0.88

1.56
2.46

1.28
2.25

1.56
2.19

1.27
2.05

1.42
2.15

1.19
1.98

1.50
2.18

1.21
2.02

1.04
1.27
0.01

0.80
1.15
0.00

1.00
1.26
0.01

0.77
1.13
0.00

0.96
1.03
0.01

0.76
0.98
0.00

0.93
1.04
0.00

0.73
0.93
0.00

0.95
1.03
0.00

0.75
0.95
0.00

2.85
1.52

2.67
0.47

2.80
1.50

2.60
0.45

2.65
1.53

2.51
0.48

2.63
1.41

2.40
0.43

2.64
1.48

2.45
0.45

N/A
N/A
0.70

N/A
N/A
0.58

N/A
N/A
0.67

N/A
N/A
0.57

N/A
N/A
0.52

N/A
N/A
0.43

N/A
N/A
0.49

N/A
N/A
0.42

N/A
N/A
0.51

N/A
N/A
0.42

5.80
0.65
4.72

5.67
0.53
4.72

5.68
0.63
4.68

5.53
0.50
4.63

3.84
0.74
1.86

3.58
0.65
1.73

4.17
0.74
1.74

3.91
0.65
1.54

3.98
0.74
1.81

3.68
0.65
1.67
(continued)

Table 4.10c Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 18 to 25) (continued)
18 to 25
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

2.13

1.88

2.14

1.87

1.92

1.73

1.88

1.62

1.90

1.67

0.39
N/A
1.41
3.52
3.00
0.59
58.24

0.23
N/A
1.33
3.17
2.65
N/A
55.58

0.45
N/A
1.40
3.46
3.34
0.36
59.26

0.27
N/A
1.30
3.10
2.88
N/A
56.28

0.53
0.31
1.39
2.58
3.30
0.15
52.47

0.42
0.00
1.25
2.37
2.93
N/A
50.23

0.47
0.49
1.38
2.77
3.13
0.16
51.59

0.35
0.00
1.18
2.25
2.87
N/A
48.41

0.50
0.39
1.39
2.66
3.23
0.15
52.09

0.38
0.00
1.22
2.32
2.89
N/A
49.52

122

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
**QFT and DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .26 to .39
minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

123

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.79
2.23
1.65
2.30
3.25
12.07
2.06
2.35
0.46
0.24
0.10
0.76
1.02
N/A
5.08
N/A
N/A
N/A
N/A

1.60
1.85
1.45
2.10
3.07
10.98
1.78
2.17
0.35
0.13
0.08
0.58
0.83
N/A
4.55
N/A
N/A
N/A
N/A

1.67
2.27
1.62
2.27
3.26
12.08
2.05
2.39
0.46
0.24
0.10
0.76
1.03
N/A
5.06
N/A
N/A
N/A
N/A

1.52
1.85
1.45
2.05
3.07
10.87
1.78
2.18
0.35
0.13
0.08
0.58
0.83
N/A
4.47
N/A
N/A
N/A
N/A

1.59
2.05
1.06
2.09
2.99
13.03
1.90
2.30
0.49
0.23
0.10
1.05
1.16
0.19
5.60
2.21
0.84
0.88
0.74

1.40
1.60
1.05
1.92
2.87
11.37
1.65
2.15
0.35
0.13
0.08
0.85
0.95
0.13
4.85
1.97
0.70
0.75
0.63

1.47
2.01
1.09
2.01
3.14
13.26
1.87
2.32
0.49
0.27
0.10
1.11
1.23
0.21
5.66
2.23
0.84
0.85
0.76

1.37
1.60
1.10
1.82
2.95
11.92
1.57
2.18
0.33
0.15
0.08
0.87
0.97
0.13
4.87
1.92
0.68
0.72
0.63

1.53
2.03
1.07
2.05
3.06
13.13
1.89
2.31
0.49
0.25
0.10
1.07
1.19
0.20
5.63
2.22
0.84
0.86
0.75

1.38
1.60
1.08
1.88
2.90
11.63
1.62
2.17
0.35
0.13
0.08
0.87
0.97
0.13
4.87
1.95
0.70
0.73
0.63

2.00

1.78

1.99

1.77

2.86

2.40

2.86

2.37

2.86

2.39

1.10

0.93

1.09

0.92

1.01

0.78

1.05

0.77

1.03

0.77

1.11

0.93

1.10

0.92

0.93

0.77

0.92

0.75

0.93

0.77

0.87

0.72

0.87

0.72

0.80

0.67

0.83

0.67

0.82

0.67
(continued)

124

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

21.69
1.54
2.87
0.26

19.73
1.38
2.63
0.18

23.58
1.54
2.92
0.28

21.50
1.35
2.65
0.20

20.17
0.57
2.73
0.26

18.65
0.50
2.48
0.20

20.34
0.61
2.80
0.29

18.63
0.52
2.60
0.20

20.25
0.58
2.76
0.27

18.64
0.52
2.52
0.20

2.78

2.22

2.82

2.27

2.78

2.20

2.84

2.22

2.81

2.21

0.23
1.65

0.00
1.35

0.24
1.64

0.00
1.33

N/A
1.34

N/A
1.18

N/A
1.30

N/A
1.15

N/A
1.33

N/A
1.17

1.78
1.22

1.45
1.07

1.82
2.85

1.45
2.57

1.79
2.58

1.45
2.33

1.75
2.48

1.43
2.37

1.77
2.53

1.45
2.34

1.22
1.39
0.49

0.93
1.25
0.00

1.19
1.37
0.52

0.92
1.23
0.00

1.15
1.21
0.49

0.87
1.07
0.00

1.15
1.20
0.58

0.92
1.10
0.00

1.15
1.21
0.53

0.88
1.08
0.00

3.10
1.78

2.88
0.52

3.05
1.75

2.77
0.52

3.04
1.77

2.73
0.50

3.01
1.84

2.85
0.55

3.02
1.80

2.76
0.53

N/A
N/A
0.62

N/A
N/A
0.57

N/A
N/A
0.62

N/A
N/A
0.57

N/A
N/A
0.48

N/A
N/A
0.47

N/A
N/A
0.48

N/A
N/A
0.47

N/A
N/A
0.48

N/A
N/A
0.47

5.61
0.23
5.03

5.45
0.13
4.93

5.41
0.21
4.91

5.25
0.13
4.80

4.02
0.59
2.19

3.68
0.47
2.02

4.45
0.64
2.00

4.10
0.57
1.83

4.21
0.62
2.11

3.92
0.52
1.94
(continued)

Table 4.10d Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 26 to 49) (continued)
26 to 49
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.42

1.23

1.41

1.22

1.33

1.17

1.38

1.15

1.35

1.17

0.30
N/A
1.33
3.39
2.96
0.88
60.88

0.22
N/A
1.23
3.00
2.53
N/A
57.93

0.33
N/A
1.32
3.36
3.26
0.41
62.26

0.23
N/A
1.20
2.92
2.72
N/A
58.68

0.40
0.22
1.40
2.91
3.20
0.12
56.59

0.35
0.00
1.23
2.48
2.75
N/A
53.52

0.45
0.11
1.40
2.56
3.17
0.19
57.03

0.33
0.00
1.23
2.27
2.67
N/A
53.92

0.43
0.17
1.40
2.75
3.19
0.16
56.79

0.35
0.00
1.23
2.40
2.73
N/A
53.63

125

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

126

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.92
2.36
1.72
2.53
4.08
14.18
2.36
2.73
0.52
0.29
0.12
0.91
1.28
N/A
5.95
N/A
N/A
N/A
N/A

1.67
1.90
1.50
2.27
4.02
12.83
1.97
2.50
0.42
0.18
0.10
0.70
1.02
N/A
5.33
N/A
N/A
N/A
N/A

1.89
2.54
1.71
2.51
4.03
14.09
2.28
2.69
0.53
0.30
0.13
0.93
1.29
N/A
5.93
N/A
N/A
N/A
N/A

1.56
1.97
1.48
2.17
3.97
12.65
1.95
2.47
0.42
0.18
0.10
0.72
1.02
N/A
5.25
N/A
N/A
N/A
N/A

1.72
2.22
1.40
2.42
4.31
16.77
2.34
2.84
0.52
0.29
0.14
1.40
1.56
0.23
7.44
2.92
1.17
1.23
1.12

1.49
1.83
1.48
2.08
4.19
14.62
1.87
2.45
0.47
0.20
0.12
1.11
1.26
0.18
6.25
2.41
0.97
0.93
0.86

1.67
2.19
1.31
2.49
4.08
15.90
1.95
2.48
0.54
0.33
0.12
1.34
1.48
0.27
7.40
2.79
1.14
1.21
1.06

1.35
1.70
1.38
2.13
3.95
13.65
1.65
2.35
0.43
0.22
0.10
1.03
1.10
0.18
5.90
2.20
0.87
0.93
0.77

1.70
2.21
1.36
2.45
4.21
16.37
2.16
2.68
0.53
0.31
0.13
1.37
1.53
0.25
7.42
2.86
1.16
1.22
1.09

1.43
1.77
1.45
2.08
4.08
14.15
1.70
2.40
0.47
0.20
0.10
1.07
1.22
0.18
6.10
2.28
0.90
0.93
0.83

2.18

1.98

2.17

1.93

3.59

3.02

3.60

2.83

3.59

2.93

1.30

1.13

1.28

1.11

1.34

1.00

1.36

1.02

1.35

1.00

1.35

1.13

1.34

1.13

1.29

0.97

1.25

0.97

1.27

0.97

1.12

0.93

1.13

0.92

1.22

0.93

1.18

0.83

1.20

0.90
(continued)

127

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

24.02
1.74
3.44
0.21

21.67
1.55
3.13
0.18

26.18
1.77
3.49
0.23

23.77
1.57
3.15
0.18

22.47
0.66
3.49
0.22

20.63
0.62
3.22
0.20

21.66
0.66
3.26
0.23

20.20
0.57
2.85
0.18

22.10
0.66
3.39
0.22

20.28
0.58
3.12
0.20

2.54

2.08

2.46

2.10

2.72

2.21

2.39

1.75

2.57

2.02

0.14
1.79

0.00
1.50

0.14
1.81

0.00
1.50

N/A
1.52

N/A
1.33

N/A
1.39

N/A
1.28

N/A
1.46

N/A
1.30

1.98
1.72

1.63
1.48

1.98
3.74

1.62
3.37

1.93
3.54

1.63
3.27

1.80
3.06

1.50
2.88

1.87
3.32

1.57
3.12

1.45
1.64
0.26

1.10
1.47
0.00

1.47
1.63
0.24

1.10
1.45
0.00

1.27
1.51
0.19

1.02
1.40
0.00

1.41
1.46
0.27

0.97
1.28
0.00

1.33
1.49
0.22

1.02
1.35
0.00

3.59
2.04

3.30
0.58

3.49
2.01

3.18
0.60

3.22
1.55

2.92
0.54

3.48
1.78

3.08
0.65

3.34
1.65

3.00
0.57

N/A
N/A
0.67

N/A
N/A
0.62

N/A
N/A
0.66

N/A
N/A
0.62

N/A
N/A
0.64

N/A
N/A
0.53

N/A
N/A
0.49

N/A
N/A
0.48

N/A
N/A
0.57

N/A
N/A
0.50

5.20
0.19
4.73

5.13
0.12
4.75

4.92
0.18
4.61

4.92
0.12
4.68

4.58
0.67
2.48

4.17
0.55
2.29

4.81
0.75
2.08

4.52
0.63
2.03

4.69
0.71
2.30

4.32
0.58
2.15
(continued)

Table 4.10e Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 50 to 64) (continued)
50 to 64
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.10

0.90

1.02

0.83

1.09

0.97

1.31

0.78

1.19

0.90

0.31
N/A
1.36
3.42
3.30
1.01
66.53

0.22
N/A
1.23
3.02
2.68
N/A
63.13

0.32
N/A
1.43
3.37
3.40
0.48
67.88

0.23
N/A
1.20
2.90
2.85
N/A
64.35

0.50
0.19
1.71
3.34
3.85
0.17
66.76

0.38
0.00
1.50
3.01
2.95
N/A
62.60

0.41
0.15
1.62
3.22
3.59
0.17
64.55

0.30
0.00
1.42
2.63
2.82
N/A
62.22

0.46
0.17
1.67
3.29
3.73
0.17
65.76

0.35
0.00
1.45
2.83
2.90
N/A
62.33

128

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

129

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.91
2.64
1.85
3.03
4.88
17.42
2.86
3.18
0.48
0.26
0.17
1.21
1.88
N/A
7.38
N/A
N/A
N/A
N/A

1.70
2.15
1.60
2.67
4.75
16.00
2.35
2.92
0.45
0.22
0.15
0.95
1.45
N/A
6.63
N/A
N/A
N/A
N/A

1.90
2.91
1.79
2.94
4.89
17.15
2.75
3.19
0.51
0.28
0.17
1.16
1.83
N/A
7.27
N/A
N/A
N/A
N/A

1.67
2.48
1.55
2.53
4.70
15.57
2.32
2.87
0.43
0.22
0.15
0.90
1.43
N/A
6.50
N/A
N/A
N/A
N/A

1.73
2.66
1.53
2.91
5.37
22.04
2.59
3.47
0.61
0.32
0.16
1.78
2.28
0.30
10.55
4.27
1.68
1.69
1.60

1.57
2.30
1.57
2.32
5.13
19.45
2.20
3.27
0.52
0.23
0.15
1.45
1.72
0.23
8.28
3.05
1.27
1.27
1.25

2.05
2.65
1.61
2.89
5.32
22.31
2.54
3.36
0.57
0.27
0.17
1.82
2.19
0.31
11.08
4.32
1.95
1.89
1.78

1.73
2.04
1.63
2.56
4.83
18.96
2.22
3.12
0.60
0.27
0.15
1.48
1.84
0.30
8.88
3.31
1.26
1.29
1.21

1.88
2.66
1.57
2.90
5.35
22.17
2.57
3.42
0.59
0.29
0.17
1.80
2.24
0.30
10.80
4.29
1.80
1.78
1.68

1.65
2.27
1.60
2.50
4.95
19.28
2.20
3.18
0.53
0.25
0.15
1.45
1.75
0.27
8.48
3.15
1.27
1.27
1.22

2.49

2.33

2.51

2.30

5.10

3.73

4.96

4.09

5.04

3.83

1.64

1.48

1.62

1.43

1.92

1.43

2.22

1.54

2.06

1.47

1.70

1.45

1.64

1.42

1.74

1.27

1.93

1.31

1.83

1.28

1.54

1.28

1.50

1.23

1.79

1.30

1.97

1.35

1.87

1.32
(continued)

130

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

26.78
2.06
4.59
0.23

24.58
1.88
4.05
0.20

29.23
2.05
4.47
0.23

26.88
1.85
3.93
0.20

26.73
0.75
4.34
0.25

24.08
0.67
3.85
0.20

27.15
0.77
4.42
0.26

25.80
0.65
4.08
0.20

26.93
0.76
4.38
0.26

24.48
0.67
3.90
0.20

1.82

1.37

1.87

1.58

2.08

1.85

1.95

2.03

2.02

1.95

0.02
1.58

0.00
1.33

0.02
1.59

0.00
1.27

N/A
1.68

N/A
1.35

N/A
1.53

N/A
1.43

N/A
1.61

N/A
1.40

2.20
2.52

1.92
2.17

2.15
5.26

1.87
4.65

2.35
4.76

1.95
4.35

2.17
5.38

2.02
4.38

2.27
5.05

1.98
4.35

1.80
2.25
0.05

1.37
2.00
0.00

1.71
2.21
0.05

1.28
1.95
0.00

1.75
1.94
0.08

1.33
1.77
0.00

1.77
1.98
0.03

1.25
1.73
0.00

1.76
1.96
0.05

1.28
1.77
0.00

4.59
1.61

4.13
0.67

4.32
1.56

3.85
0.63

4.68
1.40

4.35
0.68

4.50
1.76

3.58
0.77

4.60
1.57

4.10
0.73

N/A
N/A
0.70

N/A
N/A
0.65

N/A
N/A
0.67

N/A
N/A
0.63

N/A
N/A
0.67

N/A
N/A
0.62

N/A
N/A
0.60

N/A
N/A
0.65

N/A
N/A
0.62

3.01
0.16
2.57

1.85
0.12
1.40

2.67
0.16
2.40

1.63
0.12
1.35

4.94
0.90
2.09

4.40
0.68
1.75

5.77
0.92
2.04

4.93
0.78
1.47

5.33
0.91
2.07

4.58
0.75
1.63
(continued)

Table 4.10f Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (All Respondents Aged 65 or Older) (continued)
65 or Older
2012 Main Study
Questionnaire Module
Household Roster
Proxy Information/
Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

0.89

0.67

0.76

0.58

0.95

0.75

0.72

0.61

0.84

0.68

0.31
N/A
1.49
3.76
3.87
0.93
72.79

0.20
N/A
1.32
3.32
3.08
N/A
68.97

0.36
N/A
1.47
3.61
3.94
0.52
74.14

0.20
N/A
1.28
3.15
3.17
N/A
69.60

0.48
0.33
2.15
4.41
4.06
0.18
80.47

0.42
0.00
1.95
3.90
3.18
N/A
74.62

0.71
0.25
1.92
4.25
4.83
0.15
82.60

0.31
0.00
1.71
3.48
3.41
N/A
77.01

0.59
0.29
2.05
4.33
4.42
0.16
81.45

0.37
0.00
1.82
3.68
3.30
N/A
75.32

131

Table 4.10g Overall and Module Mean/Median Timing Data for All Interviews in the 2012 Main Study, Q3-Q4 2013 Main Study, and
2013 Dress Rehearsal in Minutes (All Respondents Aged 12 or Older)
Overall
2012 Main Study

Q3-Q4 2013 Main Study

132

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

1.79
2.24
1.67
2.40
3.50
12.47
2.00
2.18
0.50
0.21
0.10
0.84
1.20
N/A
5.45
N/A
N/A
N/A
N/A
2.11
1.18
1.19
0.97

1.62
1.85
1.50
2.18
3.30
11.27
1.70
2.00
0.37
0.13
0.08
0.65
0.93
N/A
4.85
N/A
N/A
N/A
N/A
1.92
1.00
0.98
0.77

1.69
2.22
1.66
2.36
3.52
12.32
1.96
2.14
0.49
0.21
0.10
0.83
1.18
N/A
5.40
N/A
N/A
N/A
N/A
2.09
1.16
1.18
0.97

1.55
1.83
1.48
2.15
3.32
11.02
1.65
1.97
0.37
0.13
0.08
0.63
0.92
N/A
4.77
N/A
N/A
N/A
N/A
1.88
1.00
0.97
0.77

1.51
2.20
1.23
2.19
3.57
13.95
1.76
2.29
0.51
0.24
0.11
1.24
1.42
0.22
6.17
2.55
0.96
0.99
0.87
3.08
1.10
1.07
0.92

1.37
1.82
1.20
1.95
3.30
11.83
1.38
2.12
0.38
0.15
0.08
0.92
1.05
0.17
4.83
2.03
0.70
0.73
0.63
2.43
0.75
0.77
0.65

See notes at end of table.

(continued)

133

Special Drugs to Consumption of Alcohol
Special Drugs
Risk/Availability
Blunts
Substance Dependence and Abuse
Market Information for Marijuana
Prior Substance Use
Special Topics, Drug Treatment
Health Care
Adult Mental Health Service Utilization
Social Environment
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health Service Utilization
Adolescent Depression
Consumption of Alcohol
Back-End Demographics (Moves, Born in
U.S., Disability, Education and
Employment)
Education
Employment
See notes at end of table.

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

22.04
1.63
3.00
0.27
2.15
0.27
1.23
1.65
1.31
0.82
0.99
0.15
2.74
2.15
1.12
0.63
0.56
0.54

20.28
1.47
2.70
0.20
1.55
0.00
0.93
1.35
1.10
0.65
1.02
0.00
0.00
1.82
0.32
0.00
0.00
0.45

23.61
1.63
3.01
0.30
2.07
0.27
1.18
1.63
3.00
0.79
0.96
0.15
2.73
2.06
1.08
0.63
0.57
0.52

21.80
1.45
2.70
0.22
1.48
0.00
0.88
1.33
2.63
0.62
0.98
0.00
0.00
1.70
0.30
0.00
0.00
0.43

20.80
0.60
3.02
0.30
2.16
N/A
1.07
1.68
2.87
0.93
1.05
0.26
1.88
2.40
1.24
0.41
0.48
0.45

18.95
0.53
2.68
0.22
1.67
N/A
0.87
1.38
2.47
0.73
1.02
0.00
0.00
2.02
0.40
0.00
0.00
0.41

4.53
0.59
3.58

4.50
0.47
3.77

4.36
0.56
3.52

4.37
0.43
3.72

4.74
0.91
1.77

4.18
0.77
1.58
(continued)

Q3-Q4 2013 Main Study

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Household Roster
Proxy Information/ Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

1.69
0.57
N/A
1.40
3.68
3.17
0.71
61.86

1.45
0.33
N/A
1.28
3.25
2.68
N/A
58.82

1.68
0.62
N/A
1.40
3.53
3.44
0.40
62.79

1.42
0.35
N/A
1.27
3.12
2.88
N/A
59.27

1.62
0.64
0.82
1.70
3.29
3.38
0.15
61.77

1.30
0.42
0.00
1.42
2.70
2.85
N/A
57.33

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from 0.26 to 0.39 minute.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

134

4.5.1.2

Overall and Module Timing Data for Spanish-Language Interviews

To assess interview timing for the partially redesigned DR instrument administered in
Spanish, Tables 4.11a through 4.11f provide mean and median timing results by module for the
2012 main study comparison data, the 2013 quarters 3 and 4 comparison data, and the DR data.17
The QFT was not administered in Spanish; therefore, no QFT data are available for inclusion in
this set of tables. The Spanish-language comparisons include timing results for all respondents in
each of the three sets of interviews and separate timing results for five age categories (i.e., 12 to
17, 18 to 25, 26 to 49, 50 to 64, and 65 or older). The age group timing results provide data on
how age is related to interview duration for the partially redesigned DR questionnaire and how
this information compares with current main study timing results. As was the case with the
timing data for English-language interviews among non-Hispanic respondents, the Spanishlanguage respondents with an overall administration time of less than 30 minutes or greater than
240 minutes were classified as outliers and were excluded from the timing results.
Administrative residual timings and means and medians were calculated via the same process for
the Spanish-language interviews as was used for the English-language interviews among nonHispanic respondents and discussed in Section 4.5.1.1.
Although the overall interview timing for the English-language DR interview decreased
compared with the timing for the main study, Table 4.11a shows that overall interview times
were somewhat higher for Spanish-language DR respondents. Specifically, the mean timing for
all respondents was 83.94 minutes (median 79.32) compared with a mean of 83.54 minutes for
the Spanish-language 2012 respondents (median 79.78) and a mean administration time of 82.97
minutes for the Spanish-language 2013 quarters 3 and 4 respondents (median 79.23). The
differences in DR interview times across age groups steadily increased as the respondent age
increased. A similar pattern appeared in the annual Spanish-language DR timing data. However,
in the annual Spanish-language NSDUH, the 65 or older age group had shorter timing figures
compared with the 50 to 64 year old age group, as shown in Tables 4.11e and 4.11f.
The Spanish-language DR questionnaire was shorter than the Spanish-language 2012
main study questionnaire for the younger age groups (i.e., for those aged 12 to 17, 18 to 25, or 26
to 49). However, the duration of the DR questionnaire was longer than the 2012 main study
questionnaire for the oldest Spanish-speaking NSDUH respondents. The Spanish-language DR
questionnaire took longer to administer to the 26 to 49 age group compared with the 2013
Spanish-language questionnaire for quarters 3 and 4. The 2013 questionnaire took longer for the
50 to 64 age group.
The only age group where the Spanish-language DR timing data were considerably
higher than either the Spanish-language 2012 main study or the Spanish-language 2013 quarters
3 and 4 was respondents aged 65 or older. For this age group, the mean and median times
(109.71 and 100.19 minutes, respectively) were considerably higher for this age group in the
2012 main study and in the 2013 quarters 3 and 4. Despite this larger gap in average times for
respondents aged 65 or older, the overall timing pattern was consistent with the English-language
DR results.
17

For readability, Tables 4.11a through 4.11f appear together at the end of this discussion in
Section 4.5.1.2.

135

The first five sections in the partially redesigned Spanish-language DR questionnaire—
introduction, core demographics, calendar, beginning ACASI, and tutorial—took less time to
administer overall compared with the Spanish-language 2012 main study and the Spanishlanguage 2013 quarters 3 and 4 questionnaire. This result was also consistent with the DR timing
results for interviews administered in English. Timings for these sections varied, so a few
exceptions to this general pattern were observed. These exceptions can be seen in Tables 4.11b
through 4.11f.
Similar to the English-language DR timing data, the average timing for the total core
substance use sections for all respondents aged 12 or older was higher for the Spanish-speaking
DR respondents (mean 21.26, median 17.40) than for the Spanish-speaking 2012 respondents
(mean 18.33, median 17.20) and the Spanish-speaking 2013 quarters 3 and 4 respondents (mean
18.20, median 17.23). Additions and revisions to the hallucinogens, inhalants, and prescription
drug sections in the partially redesigned DR questionnaire contributed to higher administration
times among DR respondents for the core substance use modules when compared with the main
study data. Combining the smokeless tobacco items appeared to contribute to lower average
timings for the tobacco section for DR respondents compared with the 2012 and 2013 quarters 3
and 4 respondents across all age groups. Again, similar to the English-language timing results,
differences between Spanish-speaking DR respondents versus the Spanish-speaking 2012
respondents and Spanish-speaking 2013 quarters 3 and 4 respondents for the remaining core
substance use modules—alcohol, marijuana, cocaine and crack, and heroin—were generally
small and inconsequential.
The Spanish-language DR questionnaire is the first test of the revised prescription drug
modules in Spanish, and the timing results for this section are important for understanding how
these questions result in cognitive burden on Spanish speakers. The mean total timing for the
four prescription drug modules for Spanish-speaking DR respondents aged 12 or older (mean
10.16, median 7.80) was clearly higher than it was for the Spanish-speaking 2012 respondents
(mean 7.83, median 7.32) and the Spanish-speaking 2013 quarters 3 and 4 respondents (mean
7.99, median 7.52). There were smaller differences between the median timings of the
questionnaires. The higher mean timing for the DR appears to be affected by extreme interview
timings, particularly for Spanish-language respondents aged 65 or older. Further, the timing data
for Spanish-language respondents aged 65 or older were based on a small number of respondents
who might not be representative of a larger sample of Spanish-language respondents in this age
group in the main survey.
As with the English-language questionnaire, the pain relievers module accounted for the
higher administration times for Spanish-speaking DR respondents compared with the Spanishspeaking 2012 respondents and the Spanish-speaking 2013 quarters 3 and 4 respondents.
Average timings for the other three prescription drug modules—tranquilizers, stimulants, and
sedatives—were similar, but slightly higher, among the three sets of respondents. Administration
times did vary across age groups among the respondents DR, 2012, and 2013 quarters 3 and 4
respondents. For example, Table 4.11b shows that Spanish-speaking DR respondents aged 12 to
17 took less time to complete the four prescription drug modules than did Spanish-speaking
adolescent respondents in the 2012 and 2013 comparison samples. The timing of the Spanishspeaking respondents aged 18 to 25 was very similar across the three samples. The overall
average timing for the prescription drug modules was increased among Spanish-speaking DR
136

respondents by higher administration times for older respondents. In addition, the timing
differences between the Spanish-speaking DR respondents and the Spanish-speaking 2012 and
2013 quarters 3 and 4 respondents increased steadily across the four adult age groups, so that
differences among the three sets of respondents were most pronounced among those aged 65 or
older (Table 4.11f). This example is similar to findings from the English-language questionnaire
timing tables.
For questionnaire sections from special drugs to consumption of alcohol, administration
times for Spanish-speaking DR respondents aged 12 or older were lower than for the Spanishspeaking 2012 respondents and the Spanish-speaking 2013 quarters 3 and 4 respondents. The
timing for these sections increased across age groups. Individual sections with lower DR timings
compared with the 2012 and 2013 quarters 3 and 4 interviews included special drugs,
risk/availability, prior substance use, special topics, social environment, youth experiences,
youth mental health service utilization, and consumption of alcohol. Similar to the Englishlanguage questionnaire findings, the lower administration times for the special drugs and prior
substance use sections appeared to result from the deletion of one or more items from these
sections in the DR questionnaire. The youth experiences module was markedly shorter in the
Spanish-language DR instrument as well compared with the timing results for the annual survey.
The reasons for this difference are not immediately clear given that only one short question was
deleted from the module.
For the remaining sections from special drugs to consumption of alcohol, administration
times for Spanish-speaking DR respondents were generally similar to the section timings for the
Spanish-speaking 2012 respondents and the Spanish-speaking 2013 quarters 3 and 4 respondents.
Only the parenting experiences and the adult depression modules had longer timing in the
Spanish-language DR compared with both the 2012 and 2013 comparison data, and the
differences were minimal.
Section timings for the remaining back-end modules varied for all Spanish-speaking
respondents aged 12 or older when comparing Spanish-speaking DR respondents with the
Spanish-speaking 2012 respondents and the Spanish-speaking 2013 quarters 3 and 4 respondents,
based mostly on changes made to the Spanish-language DR questionnaire. Interestingly, the
overall average time for Spanish-speaking respondents aged 65 or older to complete the backend demographics module was lower than the completion time for Spanish-speaking respondents
aged 50 to 64 and very similar to the time for Spanish-speaking respondents aged 26 to 49. Also
under the back-end demographics, the average times for the Spanish-speaking DR respondents
compared with the Spanish-speaking 2012 respondents and the Spanish-speaking 2013 quarters 3
and 4 respondents were higher for education, but lower for employment. These findings are
consistent with the changes to the Spanish-language DR questionnaire, such as adding new items
on disability to the education section and deleting questions on industry and occupation from the
employment section, and with the English-language questionnaire finding of longer times for
employment.
For the health insurance section, a higher average administration time was observed for
Spanish-speaking DR respondents compared with the Spanish-speaking 2012 respondents and
the Spanish-speaking 2013 quarters 3 and 4 respondents. As with the English-language
questionnaire, the only change to this section, relative to the main study questionnaire, was
137

moving these questions from CAPI to ACASI administration. One possible explanation for the
increased timing among Spanish-speaking DR respondents was that a higher number of proxy
reporters answered these questions in the DR and the health insurance module is the first section
after the proxy tutorial. One consequence of this sequence is that DR proxy reporters might have
used additional time getting accustomed to the interview protocol, including the relationship fills.
In addition, the specifications for the health insurance module in the DR included two additional
variables that were not included in this module in the 2012 or 2013 data.
The income section was also moved from CAPI to ACASI administration in the Spanishspeaking DR questionnaire, and a new question on household telephone service was added to
this section. Compared with the annual Spanish-language NSDUH, the administration of the
Spanish-language DR income section was either similar (age categories 18 to 25 and 50 to 64) or
longer (age categories 12 to 17, 26 to 49, and 65 or older). None of the age groups in the
English-speaking group took longer to complete the DR income module compared with the
annual NSDUH income module. These timing data indicate that the income module may be
more burdensome on Spanish-speaking respondents.
4.5.2

Overall and Module Timing Results for Affirmative Gate Respondents from
English-Language Interviews in the 2012 and 2013 Quarters 3 and 4
Comparison Data and the 2012 Questionnaire Field Test and 2013 Dress
Rehearsal Data

Tables 4.12a through 4.12f display mean and median timings by module for respondents
who responded affirmatively to at least one gate question within the core substance use
questions, or respondents whose prior responses directed them to complete a particular module.18
For example, only respondents who reported smoking part or all of a cigarette in their lifetime
were included in the timing reports for the tobacco module. Similarly, only respondents who
were administered the parenting experiences module contributed to the mean timing for that
module. These tables present results for respondents whose administration times for a module
were beyond the minimal time taken by respondents who had no data to report for a given
module. These respondents reported behavior that led to additional questions. An important
difference in the affirmative gate timing data compared with the overall timing data discussed in
Section 4.5.1 is that the module timings should not be expected to add up to the appropriate
section timings because different sets of affirmative gate respondents can be included for each
module in each section.
Of respondents who reported use and misuse of prescription drugs, DR timings of the
prescription drug modules are similar to those in the 2012 and 2013 comparison samples.
The greatest difference was observed among those aged 65 or older, which was 1 minute and
26 seconds longer than the 2102 comparison data. When reviewing these tables overall, the
timing results do not indicate a need to revise these modules to reduce administration times.

For readability, Tables 4.12a through 4.12f appear together at the end of this discussion in Section 4.5.2.

138

Table 4.11a Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 12 or Older)
Overall
2012 Main Study

Q3-Q4 2013 Main Study

139

Questionnaire Module

Mean

Median

Mean

Median

Mean

Introduction
Core Demographics
Calendar
Beginning ACASI
Tutorial
Total Core Substances
Tobacco
Alcohol
Marijuana
Cocaine and Crack
Heroin
Hallucinogens
Inhalants
Methamphetamine
Total Prescription Drugs
Pain Relievers (Screener)
Tranquilizers (Screener)
Stimulants (Screener)
Sedatives (Screener)
Pain Relievers (screener plus main module)
Tranquilizers (screener plus main module)
Stimulants (screener plus main module)
Sedatives (screener plus main module)
Special Drugs to Consumption of Alcohol
Special Drugs
Risk/Availability
Blunts
Substance Dependence and Abuse
Market Information for Marijuana
Prior Substance Use
Special Topics, Drug Treatment
Health Care
Adult Mental Health Service Utilization

2.25
3.24
2.03
2.98
5.03
18.33
2.75
3.23
0.55
0.31
0.18
1.38
2.11
N/A
7.83
N/A
N/A
N/A
N/A
2.75
1.76
1.78
1.54
29.63
2.68
4.71
0.29
2.10
0.06
1.29
2.50
1.98
1.56

1.88
2.78
1.85
2.67
5.08
17.20
2.38
2.78
0.52
0.25
0.15
1.18
1.75
N/A
7.32
N/A
N/A
N/A
N/A
2.63
1.60
1.57
1.28
27.72
2.53
4.37
0.27
0.00
0.00
0.90
2.03
1.72
1.25

2.02
3.05
1.93
2.87
5.09
18.20
2.61
3.11
0.54
0.29
0.17
1.39
2.11
N/A
7.99
N/A
N/A
N/A
N/A
2.77
1.78
1.85
1.59
31.25
2.72
4.80
0.31
2.00
0.06
1.17
2.41
4.60
1.42

1.82
2.67
1.82
2.64
5.07
17.23
2.32
2.75
0.50
0.25
0.15
1.27
1.73
N/A
7.52
N/A
N/A
N/A
N/A
2.65
1.67
1.67
1.36
29.40
2.53
4.31
0.27
0.00
0.00
0.80
2.00
4.15
1.20

1.63
2.80
1.78
2.49
4.94
21.26
1.97
3.16
0.52
0.27
0.16
2.15
2.50
0.37
10.16
4.52
1.72
1.74
1.52
4.96
1.81
1.83
1.56
25.64
0.88
4.38
0.32
2.02
N/A
1.01
2.15
4.13
1.48

Median
1.37
2.65
1.88
2.25
4.92
17.40
1.57
2.77
0.55
0.23
0.15
1.78
2.02
0.30
7.80
3.42
1.23
1.18
1.00
4.00
1.28
1.23
1.02
23.10
0.87
4.20
0.25
0.00
N/A
0.75
1.83
3.63
1.23
(continued)

140

Social Environment
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health Service Utilization
Adolescent Depression
Consumption of Alcohol
Back-End Demographics (Moves, Born in U.S.,
Disability, Education and Employment)
Education
Employment
Household Roster
Proxy Information/Decision
Proxy Tutorial
Health Insurance**
Income
Verification
Administrative Residual
Overall Questionnaire

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

1.95
0.65
2.03
3.37
1.28
0.44
0.42
0.57

1.97
0.00
0.00
2.33
0.60
0.00
0.00
0.38

1.85
0.60
2.16
3.06
1.00
0.50
0.39
0.53

1.92
0.00
0.00
2.12
0.55
0.00
0.00
0.35

1.73
0.62
1.54
3.09
1.33
0.31
0.17
0.47

1.85
0.00
0.00
1.93
0.57
0.00
0.00
0.37

5.39
0.63
4.31
2.24
0.61
N/A
1.62
5.06
4.06
1.06
83.54

4.82
0.32
4.03
1.80
0.33
N/A
1.45
3.97
3.43
N/A
79.78

4.87
0.58
3.93
2.24
0.74
N/A
1.70
4.09
4.13
0.79
82.97

4.38
0.30
3.66
1.80
0.38
N/A
1.48
3.73
3.57
N/A
79.23

7.57
1.25
2.60
1.84
0.74
1.43
2.72
5.24
3.76
0.11
83.94

6.47
1.13
2.33
1.52
0.57
0.00
2.27
4.32
3.45
N/A
79.32

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.11b Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 12 to 17)
12 to 17
2012 Main Study

Q3-Q4 2013 Main Study

141

Questionnaire Module

Mean

Median

Mean

Median

Mean

2.12
2.59
1.75
2.48
4.32
14.32
2.07
2.12
0.47
0.20
0.13
1.13
1.76
N/A
6.42
N/A
N/A
N/A
N/A
2.46
1.37
1.40
1.19
25.90
2.09
3.63
0.28
0.86
0.15
0.62
1.77
1.47
N/A

1.79
2.18
1.67
2.30
4.32
13.47
1.87
1.88
0.42
0.17
0.12
0.94
1.48
N/A
5.97
N/A
N/A
N/A
N/A
2.36
1.23
1.18
0.98
24.45
1.93
3.43
0.23
0.00
0.00
0.00
1.51
1.28
N/A

1.93
2.73
1.71
2.46
4.44
13.86
1.87
1.77
0.46
0.20
0.13
1.16
1.69
N/A
6.58
N/A
N/A
N/A
N/A
2.42
1.42
1.49
1.25
26.97
2.11
3.80
0.26
0.44
0.09
0.36
1.58
3.66
N/A

1.58
2.32
1.50
2.22
4.25
13.47
1.78
1.58
0.40
0.20
0.12
1.05
1.47
N/A
6.18
N/A
N/A
N/A
N/A
2.35
1.28
1.28
1.05
25.57
2.03
3.52
0.22
0.00
0.00
0.00
1.33
3.32
N/A

1.38
2.62
1.22
1.92
3.63
11.13
1.27
1.39
0.40
0.16
0.14
1.20
1.64
0.22
4.73
2.20
0.78
0.78
0.66
2.49
0.79
0.78
0.66
18.89
0.58
2.92
0.32
0.47
N/A
0.55
1.33
2.59
N/A

Median
0.77
2.40
1.28
1.80
3.20
9.40
1.08
0.92
0.22
0.12
0.08
0.98
1.28
0.18
4.10
2.02
0.70
0.73
0.60
2.02
0.70
0.73
0.60
18.17
0.58
2.82
0.20
0.00
N/A
0.00
1.23
2.42
N/A
(continued)

142

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

N/A
N/A
9.49
N/A
N/A
2.06
1.97
0.30

N/A
N/A
9.00
N/A
N/A
1.81
0.74
0.00

N/A
N/A
9.40
N/A
N/A
2.18
1.72
0.15

N/A
N/A
9.22
N/A
N/A
1.93
0.77
0.00

N/A
N/A
7.60
N/A
N/A
1.54
0.83
0.17

N/A
N/A
7.35
N/A
N/A
1.37
0.48
0.00

2.23
0.88
0.94
2.29
1.33
N/A
1.63
5.73
3.74
0.77
71.19

1.50
0.78
0.00
1.85
0.87
N/A
1.42
4.38
3.28
N/A
68.11

2.01
0.87
0.83
2.61
1.62
N/A
1.60
4.21
4.07
0.53
70.75

1.47
0.82
0.00
2.13
0.95
N/A
1.40
3.85
3.30
N/A
67.07

3.69
1.15
0.64
2.12
0.93
4.40
3.44
6.77
3.34
0.10
65.57

3.33
1.18
0.58
1.97
0.87
4.73
3.38
5.93
3.15
N/A
63.40

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.11c Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 18 to 25)
18 to 25
2012 Main Study

Q3-Q4 2013 Main Study

143

Questionnaire Module

Mean

Median

Mean

Median

Mean

2.06
3.18
2.08
3.07
4.96
18.20
2.91
3.34
0.54
0.33
0.17
1.35
2.00
N/A
7.56
N/A
N/A
N/A
N/A
2.73
1.67
1.71
1.45
28.05
2.76
4.57
0.30
2.59
0.09
1.53
2.67
1.85
1.70

1.85
2.83
1.88
2.85
5.03
17.32
2.42
2.82
0.43
0.23
0.13
1.17
1.65
N/A
7.12
N/A
N/A
N/A
N/A
2.57
1.53
1.48
1.18
25.65
2.58
4.15
0.27
0.00
0.00
1.15
2.07
1.65
1.38

2.06
3.00
2.01
2.77
4.86
18.04
2.71
3.43
0.64
0.33
0.15
1.32
1.96
N/A
7.51
N/A
N/A
N/A
N/A
2.73
1.64
1.70
1.44
28.51
2.53
4.31
0.34
3.03
0.16
1.42
2.48
4.10
1.47

1.93
2.73
1.87
2.58
4.87
17.35
2.35
2.92
0.52
0.25
0.12
1.08
1.52
N/A
6.87
N/A
N/A
N/A
N/A
2.57
1.48
1.48
1.20
26.55
2.47
4.00
0.27
2.12
0.00
1.17
2.05
3.63
1.30

1.89
2.89
1.61
3.30
4.98
18.17
1.96
3.54
0.62
0.19
0.13
1.73
2.16
0.35
7.50
3.84
1.22
1.07
0.95
4.26
1.22
1.07
0.95
22.18
0.80
3.98
0.39
2.89
N/A
0.99
1.90
3.10
1.65

Median
1.75
2.77
1.63
2.70
4.50
13.80
1.62
2.97
0.60
0.20
0.12
1.37
1.77
0.28
5.63
2.92
0.98
0.83
0.78
3.02
0.98
0.83
0.78
19.23
0.68
3.82
0.25
2.72
N/A
1.00
1.67
3.12
1.17
(continued)

144

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

2.17
0.04
N/A
3.99
1.50
N/A
N/A
0.63

2.03
0.00
N/A
3.28
0.70
N/A
N/A
0.52

2.04
0.00
0.00
3.47
1.18
N/A
N/A
0.67

1.87
0.00
0.00
3.02
0.63
N/A
N/A
0.53

1.65
0.00
0.00
3.00
1.06
N/A
N/A
0.77

1.47
0.00
0.00
2.62
0.62
N/A
N/A
0.58

7.08
1.22
5.31
2.23
0.46
N/A
1.63
5.06
4.29
1.00
83.34

6.62
1.02
5.08
1.85
0.28
N/A
1.43
3.95
3.52
N/A
79.78

6.49
1.05
5.01
2.32
0.59
N/A
1.60
4.24
4.00
0.80
81.31

6.13
0.82
4.82
2.03
0.33
N/A
1.57
3.83
3.65
N/A
79.10

6.14
1.22
2.14
1.99
0.72
1.88
2.60
4.67
3.91
0.12
77.04

5.47
1.17
1.73
1.53
0.73
0.00
2.18
4.05
3.77
N/A
72.18

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.11d Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 26 to 49)
26 to 49
2012 Main Study

Q3-Q4 2013 Main Study

145

Questionnaire Module

Mean

Median

Mean

Median

Mean

2.38
3.48
2.11
3.07
5.39
19.90
2.93
3.70
0.58
0.33
0.21
1.47
2.24
N/A
8.44
N/A
N/A
N/A
N/A
2.85
1.94
1.95
1.69
31.33
2.89
5.11
0.29
2.45
0.02
1.43
2.73
2.18
2.04

1.93
3.00
1.92
2.75
5.42
19.12
2.43
2.98
0.60
0.28
0.17
1.33
1.85
N/A
7.87
N/A
N/A
N/A
N/A
2.75
1.73
1.75
1.47
29.52
2.82
4.72
0.28
0.00
0.00
1.10
2.22
1.95
1.60

2.05
3.18
1.95
3.01
5.38
19.54
2.89
3.55
0.50
0.29
0.17
1.42
2.24
N/A
8.48
N/A
N/A
N/A
N/A
2.82
1.91
1.98
1.76
33.32
3.06
5.24
0.30
2.38
0.01
1.40
2.65
5.00
1.87

1.82
2.82
1.88
2.87
5.48
19.42
2.40
3.03
0.53
0.27
0.15
1.33
1.89
N/A
8.24
N/A
N/A
N/A
N/A
2.77
1.84
1.85
1.56
31.23
2.93
4.76
0.28
0.00
0.00
1.08
2.23
4.63
1.53

1.69
2.73
1.96
2.46
5.13
23.20
2.16
3.55
0.51
0.32
0.16
2.33
2.59
0.41
11.17
4.94
1.93
1.95
1.67
5.37
2.03
2.08
1.69
27.16
0.96
4.63
0.30
2.58
N/A
1.08
2.36
4.51
1.75

Median
1.58
2.63
2.17
2.35
5.12
20.35
1.60
2.90
0.58
0.25
0.15
2.02
2.10
0.37
8.35
3.97
1.40
1.38
1.25
4.42
1.42
1.43
1.25
24.33
0.97
4.50
0.27
2.00
N/A
0.85
2.05
3.90
1.37
(continued)

146

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

2.51
1.34

2.30
0.00

4.12
1.58
N/A
N/A
0.66

3.35
0.75
N/A
N/A
0.57

2.48
1.35
0.00
3.85
1.09
N/A
N/A
0.62

2.28
0.00
0.00
3.11
0.70
N/A
N/A
0.55

2.15
1.04
0.00
3.69
1.64
N/A
N/A
0.49

1.97
0.00
0.00
2.17
0.67
N/A
N/A
0.42

6.08
0.28
5.37
2.34
0.38
N/A
1.61
4.57
3.86
1.10
87.60

5.80
0.15
5.17
1.87
0.27
N/A
1.50
3.67
3.40
N/A
84.27

5.58
0.24
4.95
2.08
0.42
N/A
1.77
3.77
4.13
0.85
87.01

5.33
0.13
4.73
1.64
0.27
N/A
1.48
3.53
3.48
N/A
84.63

8.40
1.28
3.23
1.73
0.67
0.43
2.38
4.70
3.88
0.11
86.64

7.77
1.07
2.87
1.45
0.43
0.00
2.08
4.13
3.30
N/A
85.17

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.11e Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 50 to 64)
50 to 64
2012 Main Study

Q3-Q4 2013 Main Study

147

Questionnaire Module

Mean

Median

Mean

Median

Mean

2.63
3.61
2.08
3.52
5.30
21.26
3.06
3.53
0.65
0.35
0.20
1.60
2.57
N/A
9.30
N/A
N/A
N/A
N/A
3.06
2.11
2.19
1.93
33.49
3.00
5.49
0.30
2.34
0.00
1.58
2.64
2.42
2.46

2.05
3.02
1.92
2.92
5.52
20.53
2.52
3.20
0.65
0.33
0.20
1.70
2.63
N/A
9.30
N/A
N/A
N/A
N/A
2.92
2.17
2.18
2.05
31.77
3.27
5.45
0.30
0.00
0.00
1.12
2.27
2.27
2.15

2.11
3.07
1.97
3.30
6.04
21.55
2.88
3.72
0.57
0.33
0.23
1.76
2.63
N/A
9.45
N/A
N/A
N/A
N/A
2.97
2.22
2.28
1.98
37.62
3.10
6.19
0.42
1.88
0.00
1.47
2.96
6.12
2.45

1.90
2.79
1.87
3.10
5.99
21.59
2.47
3.65
0.63
0.33
0.20
1.73
2.49
N/A
9.88
N/A
N/A
N/A
N/A
3.17
2.23
2.23
1.88
37.88
2.91
5.59
0.30
0.00
0.00
1.28
2.56
5.51
1.79

1.68
3.05
2.09
2.34
5.50
26.38
1.87
3.86
0.64
0.29
0.18
2.57
2.80
0.38
13.80
5.59
2.34
2.42
2.17
6.25
2.58
2.61
2.36
29.11
1.06
4.97
0.40
1.98
N/A
1.47
2.43
4.94
2.30

Median
1.45
2.74
2.23
2.48
5.93
22.52
1.65
3.41
0.68
0.32
0.18
2.33
2.11
0.30
11.09
4.33
1.55
1.59
1.62
5.32
1.66
2.03
1.89
26.62
0.98
4.50
0.28
0.81
N/A
1.13
2.17
4.41
1.72
(continued)

148

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

3.18
0.74
N/A
5.47
1.85
N/A
N/A
0.64

2.62
0.00
N/A
4.87
0.85
N/A
N/A
0.60

2.79
0.48
0.00
5.06
2.26
N/A
N/A
0.58

2.58
0.00
0.00
4.04
0.78
N/A
N/A
0.58

2.43
0.56
0.00
4.36
1.48
N/A
N/A
0.73

2.20
0.00
0.00
4.39
0.78
N/A
N/A
0.74

6.28
0.27
5.70
1.96
0.48
N/A
1.64
5.57
4.63
1.45
93.90

6.05
0.15
5.07
1.58
0.28
N/A
1.43
4.18
3.78
N/A
91.77

5.51
0.26
5.00
1.95
0.38
N/A
1.92
4.41
4.32
1.11
95.27

5.48
0.13
5.01
1.28
0.31
N/A
1.57
3.92
3.88
N/A
93.87

11.43
1.27
3.51
1.86
0.80
1.35
2.19
4.98
3.30
0.10
96.17

8.13
1.16
3.08
1.48
0.47
0.00
2.11
3.83
3.38
N/A
94.75

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table 4.11f Overall and Module Mean/Median Timing Data for Spanish-Language Interviews in the 2012 Main Study, Q3-Q4 2013
Main Study, and 2013 Dress Rehearsal in Minutes (All Respondents Aged 65 or Older)
65 or Older
2012 Main Study

Q3-Q4 2013 Main Study

149

Questionnaire Module

Mean

Median

Mean

Median

Mean

1.95
3.82
2.12
3.21
4.88
18.29
2.87
2.86
0.48
0.36
0.24
1.49
2.37
N/A
7.62
N/A
N/A
N/A
N/A
2.75
1.77
1.61
1.49
33.98
2.59
5.58
0.26
1.44
0.00
1.24
2.77
2.68
2.22

1.68
3.29
1.83
3.20
4.80
15.82
2.51
2.13
0.38
0.28
0.19
1.50
2.20
N/A
7.11
N/A
N/A
N/A
N/A
2.33
1.58
1.37
1.28
32.53
2.33
5.28
0.27
0.00
0.00
0.95
2.00
2.03
1.67

1.78
3.85
2.40
3.33
5.08
22.87
2.98
3.12
0.64
0.36
0.26
1.97
2.91
N/A
10.62
N/A
N/A
N/A
N/A
4.07
2.26
2.45
1.85
36.30
3.09
5.91
0.28
1.40
0.00
1.37
3.05
5.22
2.34

1.74
3.08
2.11
3.04
5.27
24.43
2.38
2.92
0.70
0.38
0.23
1.97
3.22
N/A
10.59
N/A
N/A
N/A
N/A
3.21
2.30
2.51
1.93
37.61
3.23
5.66
0.29
0.00
0.00
1.30
2.53
5.16
2.08

1.54
3.49
1.82
3.59
6.48
33.13
2.66
3.77
0.62
0.35
0.26
3.54
4.48
0.59
16.86
7.58
2.68
2.82
2.73
8.26
2.85
2.92
2.83
33.54
1.07
6.56
0.31
1.03
N/A
1.25
2.92
6.02
2.29

Median
0.44
3.50
2.28
3.15
6.50
33.72
1.75
3.41
0.68
0.38
0.27
4.01
5.73
0.73
14.83
6.59
2.13
2.63
2.30
7.44
2.13
2.63
2.30
27.15
1.21
7.58
0.34
0.00
N/A
1.18
2.99
6.36
1.88
(continued)

150

Q3-Q4 2013 Main Study

Mean

Median

Mean

Median

Mean

Median

2.78
0.27

2.66
0.00

5.68
2.43
N/A
N/A
0.50

3.83
0.89
N/A
N/A
0.17

2.76
0.34
0.00
5.69
1.81
N/A
N/A
0.75

2.63
0.00
0.00
4.19
0.87
N/A
N/A
0.37

2.59
0.00
0.00
5.92
3.12
N/A
N/A
0.48

2.43
0.00
0.00
4.10
0.88
N/A
N/A
0.53

2.73
0.21
2.21
1.40
0.45
N/A
1.75
5.89
5.57
2.01
88.04

1.42
0.12
1.06
0.84
0.28
N/A
1.13
4.45
3.38
N/A
82.59

2.93
0.18
2.60
1.95
0.53
N/A
1.70
4.89
4.91
0.92
93.44

1.82
0.14
1.40
1.63
0.31
N/A
1.38
3.82
3.91
N/A
93.38

8.93
1.24
2.73
1.53
0.65
0.00
4.23
6.13
4.58
0.09
109.71

7.76
1.28
1.68
0.99
0.41
0.00
3.13
4.54
4.20
N/A
100.19

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
Note: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
** DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .21 to .42 minutes.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

151

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.79
2.24
1.67
2.40
3.50
12.73
2.65
2.61
0.81
0.72
0.49
1.42
1.68
N/A
6.47
N/A
N/A
N/A
N/A

1.62
1.85
1.50
2.18
3.30
11.45
2.32
2.33
0.67
0.57
0.33
1.20
1.38
N/A
5.87
N/A
N/A
N/A
N/A

1.69
2.22
1.66
2.36
3.52
12.58
2.72
2.60
0.81
0.72
0.56
1.43
1.68
N/A
6.48
N/A
N/A
N/A
N/A

1.55
1.83
1.48
2.15
3.32
11.20
2.40
2.32
0.67
0.57
0.35
1.20
1.35
N/A
5.80
N/A
N/A
N/A
N/A

1.56
2.03
1.14
2.24
3.35
13.91
2.48
2.61
0.79
0.68
0.51
1.70
1.66
0.43
6.46
2.40
0.89
0.92
0.80

1.45
1.63
1.17
2.03
3.13
11.98
2.10
2.33
0.65
0.55
0.32
1.43
1.40
0.37
5.40
2.02
0.70
0.75
0.63

1.56
2.04
1.15
2.20
3.42
13.59
2.35
2.49
0.78
0.69
0.41
1.64
1.71
0.50
6.33
2.38
0.90
0.93
0.81

1.42
1.63
1.13
1.95
3.18
11.77
2.10
2.28
0.63
0.57
0.28
1.43
1.38
0.35
5.28
1.97
0.68
0.73
0.62

1.56
2.03
1.15
2.22
3.38
13.77
2.42
2.56
0.78
0.68
0.46
1.67
1.68
0.46
6.41
2.39
0.89
0.92
0.81

1.43
1.63
1.15
2.00
3.15
11.93
2.10
2.32
0.65
0.57
0.29
1.43
1.40
0.37
5.35
2.00
0.68
0.73
0.63

3.10

2.80

3.12

2.78

3.39

2.87

3.28

2.68

3.34

2.78

1.86

1.63

1.82

1.58

1.81

1.40

1.76

1.40

1.79

1.40

1.97

1.75

1.99

1.73

1.77

1.33

1.52

1.25

1.66

1.28

1.88

1.58

1.90

1.63

1.47

1.15

1.37

0.97

1.42

1.07
(continued)

Table 4.12a Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 12 or Older) (continued)
Overall
2012 Main Study
Questionnaire Module

152

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

22.04
1.63
3.00
0.54

20.28
1.47
2.70
0.45

23.61
1.63
3.01
0.64

21.80
1.45
2.70
0.55

20.43
0.57
2.86
0.58

18.79
0.52
2.54
0.50

20.24
0.57
2.86
0.60

18.63
0.50
2.60
0.52

20.35
0.57
2.86
0.59

18.70
0.52
2.57
0.50

3.78

3.03

3.70

2.93

3.72

2.97

3.55

2.78

3.65

2.90

1.48
1.65

1.37
1.32

1.48
1.61

1.37
1.28

N/A
1.42

N/A
1.20

N/A
1.35

N/A
1.15

N/A
1.39

N/A
1.18

1.65
1.31

1.35
1.10

1.63
3.00

1.33
2.63

1.66
2.79

1.37
2.45

1.61
2.76

1.33
2.43

1.64
2.78

1.35
2.43

2.25
1.47
2.66
8.28
3.67
3.27

1.88
1.28
2.23
7.85
3.23
1.35

2.19
1.44
2.55
8.14
3.56
3.24

1.85
1.27
2.15
7.72
3.15
1.38

2.16
1.27
2.40
7.81
3.62
3.16

1.84
1.13
1.93
7.25
3.18
1.39

2.19
1.28
2.64
7.58
3.61
3.15

1.85
1.13
2.25
7.13
3.17
1.43

2.17
1.27
2.51
7.72
3.61
3.16

1.85
1.13
2.10
7.18
3.17
1.42

3.15
2.67
0.78

2.73
1.05
0.63

3.07
2.71
0.76

2.70
1.07
0.62

3.09
2.60
0.62

2.90
0.95
0.53

2.61
3.08
0.58

2.40
1.10
0.50

2.89
2.79
0.60

2.63
1.04
0.52

4.53
0.59
4.28

4.50
0.47
4.37

4.36
0.56
4.23

4.37
0.43
4.32

3.93
0.83
2.03

3.62
0.67
1.85

4.37
0.84
1.86

4.02
0.70
1.70

4.13
0.84
1.96

3.78
0.68
1.78
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.69

1.45

1.68

1.42

1.45

1.25

1.55

1.22

1.49

1.23

0.57
N/A
1.40
3.68
3.17
0.71
61.86

0.33
N/A
1.28
3.25
2.68
N/A
58.82

0.62
N/A
1.40
3.53
3.44
0.40
62.79

0.35
N/A
1.27
3.12
2.88
N/A
59.27

0.57
0.68
1.57
3.10
3.34
0.14
59.13

0.43
0.00
1.35
2.70
2.85
N/A
55.60

0.63
0.00
1.56
3.00
3.40
0.16
59.24

0.38
0.00
1.35
2.52
2.85
N/A
55.60

0.60
0.42
1.57
3.06
3.37
0.15
59.18

0.42
0.00
1.35
2.62
2.85
N/A
55.60

153

ACASI = audio computer-assisted self-interviewing; DR = Dress Rehearsal; N/A = not applicable; Q = quarter; QFT = Questionnaire Field Test.
** QFT and DR timings for the health insurance module included two additional variables (HINSINT and TOPROXY) whose mean administration times ranged from .26 to .39
minutes.
NOTE: Analysis excludes extreme records that have an interview length of less than 30 minutes or more than 240 minutes.
NOTE: Some module rows are shown in bold for consistency with Tables 4.10a to 4.10f for all respondents. However, mean affirmative gate timings in this table for modules in
bold are not necessarily mutually exclusive and are not intended to sum to the overall mean questionnaire or section timing.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

154

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.85
2.12
1.65
2.40
3.70
12.24
2.89
2.47
1.20
1.15
0.66
1.95
2.26
N/A
6.72
N/A
N/A
N/A
N/A

1.72
1.77
1.52
2.23
3.60
11.16
2.57
2.25
1.08
0.98
0.66
1.72
1.98
N/A
6.10
N/A
N/A
N/A
N/A

1.73
2.07
1.65
2.38
3.75
11.97
3.05
2.44
1.18
1.16
1.04
1.85
2.35
N/A
6.93
N/A
N/A
N/A
N/A

1.57
1.73
1.50
2.20
3.65
10.87
2.75
2.20
1.07
0.95
0.97
1.62
2.00
N/A
6.20
N/A
N/A
N/A
N/A

1.58
1.88
1.20
2.22
3.38
11.88
2.48
2.43
1.14
0.92
N/A
2.10
1.96
0.41
5.69
2.33
0.79
0.82
0.72

1.52
1.53
1.22
2.12
3.36
10.73
2.02
2.15
0.98
0.92
N/A
1.73
1.71
0.42
4.93
2.01
0.65
0.70
0.60

1.67
1.79
1.21
2.21
3.50
10.95
2.83
2.21
1.25
1.78
0.00
1.53
1.95
0.00
5.30
2.20
0.72
0.77
0.64

1.65
1.58
1.20
2.00
3.32
10.27
2.63
2.08
1.17
1.78
0.00
1.43
1.63
0.00
4.78
1.98
0.62
0.67
0.53

1.62
1.85
1.21
2.21
3.43
11.51
2.60
2.34
1.18
1.49
0.00
1.95
1.95
0.41
5.53
2.28
0.76
0.80
0.69

1.58
1.55
1.20
2.07
3.33
10.45
2.22
2.13
1.08
1.55
0.00
1.69
1.69
0.42
4.85
2.00
0.63
0.70
0.57

3.45

3.15

3.59

3.28

3.12

2.80

3.06

2.70

3.10

2.78

2.08

1.88

2.09

1.85

1.90

1.50

1.80

1.72

1.87

1.52

2.17

1.87

2.17

1.90

2.02

1.63

1.53

1.25

1.79

1.38

2.08

1.83

2.09

1.77

1.51

1.20

1.35

0.85

1.44

1.17
(continued)

Table 4.12b Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 12 to 17) (continued)
12 to 17
2012 Main Study
Questionnaire Module

155

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

22.32
1.69
3.02
0.69

20.90
1.60
2.80
0.60

23.91
1.69
3.06
0.83

22.50
1.58
2.83
0.75

20.39
0.53
2.77
0.78

19.04
0.50
2.50
0.70

19.97
0.50
2.67
0.83

19.02
0.50
2.57
0.78

20.22
0.52
2.73
0.80

19.02
0.50
2.52
0.71

3.77

3.02

3.59

2.85

3.68

2.90

3.93

3.05

3.77

3.00

1.47
1.37

1.37
1.05

1.48
1.30

1.38
1.02

N/A
1.12

N/A
0.97

N/A
1.05

N/A
0.83

N/A
1.09

N/A
0.92

1.37
1.34

1.17
1.17

1.35
3.03

1.15
2.75

1.27
2.71

1.08
2.44

1.26
2.66

1.10
2.52

1.26
2.69

1.08
2.48

N/A
N/A
N/A
8.28
N/A
N/A

N/A
N/A
N/A
7.85
N/A
N/A

N/A
N/A
N/A
8.14
N/A
N/A

N/A
N/A
N/A
7.72
N/A
N/A

N/A
N/A
N/A
7.81
N/A
N/A

N/A
N/A
N/A
7.25
N/A
N/A

N/A
N/A
N/A
7.58
N/A
N/A

N/A
N/A
N/A
7.13
N/A
N/A

N/A
N/A
N/A
7.72
N/A
N/A

N/A
N/A
N/A
7.18
N/A
N/A

3.15
2.67
0.85

2.73
1.05
0.55

3.07
2.71
0.82

2.70
1.07
0.53

3.09
2.60
0.68

2.90
0.95
0.42

2.61
3.08
0.71

2.40
1.10
0.43

2.89
2.79
0.69

2.63
1.04
0.42

2.54
0.89
2.65

1.75
0.83
1.43

2.46
0.85
2.76

1.62
0.78
1.48

3.33
1.27
1.48

3.04
1.17
1.18

3.69
1.30
1.38

3.42
1.20
1.07

3.48
1.28
1.44

3.22
1.17
1.14
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

2.16

1.90

2.17

1.88

1.92

1.73

1.88

1.62

1.90

1.67

1.01
N/A
1.40
3.90
3.23
0.55
60.83

0.77
N/A
1.28
3.47
2.75
N/A
58.68

1.06
N/A
1.40
3.67
3.53
0.35
61.97

0.80
N/A
1.27
3.27
2.98
N/A
59.28

0.90
2.04
1.73
3.35
3.15
0.12
59.02

0.78
2.25
1.53
2.95
2.78
N/A
56.85

1.18
0.00
1.83
3.36
3.40
0.12
58.84

0.82
0.00
1.68
2.90
2.78
N/A
56.35

1.01
1.60
1.77
3.35
3.25
0.12
58.95

0.78
0.00
1.58
2.93
2.78
N/A
56.60

156

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

157

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.67
2.17
1.65
2.30
3.04
11.96
2.58
2.51
0.82
0.76
0.53
1.43
1.48
N/A
6.13
N/A
N/A
N/A
N/A

1.57
1.82
1.48
2.12
2.85
10.82
2.28
2.25
0.70
0.58
0.39
1.20
1.25
N/A
5.62
N/A
N/A
N/A
N/A

1.59
2.11
1.63
2.27
3.03
11.77
2.65
2.48
0.82
0.76
0.62
1.48
1.44
N/A
6.02
N/A
N/A
N/A
N/A

1.52
1.78
1.48
2.07
2.82
10.57
2.37
2.22
0.70
0.62
0.44
1.23
1.20
N/A
5.43
N/A
N/A
N/A
N/A

1.36
1.85
0.95
2.17
2.73
12.36
2.36
2.37
0.81
0.76
0.53
1.77
1.50
0.43
5.95
1.91
0.69
0.69
0.58

1.42
1.58
0.92
2.03
2.63
11.02
2.06
2.17
0.70
0.65
0.35
1.45
1.37
0.40
5.08
1.73
0.57
0.62
0.53

1.33
2.01
0.94
2.10
2.73
11.07
2.11
2.19
0.77
0.68
0.56
1.58
1.68
0.80
5.18
1.86
0.65
0.71
0.56

1.33
1.59
0.88
1.91
2.50
9.83
1.93
2.02
0.65
0.62
0.32
1.33
1.20
0.55
4.45
1.60
0.55
0.60
0.48

1.35
1.92
0.94
2.14
2.73
11.80
2.25
2.29
0.79
0.73
0.54
1.69
1.57
0.55
5.63
1.89
0.67
0.70
0.57

1.40
1.58
0.90
1.98
2.58
10.58
1.98
2.12
0.67
0.63
0.33
1.43
1.27
0.43
4.78
1.68
0.57
0.60
0.52

2.95

2.68

2.91

2.63

3.20

2.60

2.78

2.32

3.03

2.47

1.74

1.55

1.70

1.50

2.11

1.55

1.73

1.47

1.97

1.50

1.89

1.70

1.85

1.65

1.97

1.50

1.69

1.38

1.86

1.40

1.85

1.62

1.75

1.52

1.24

0.94

0.90

0.78

1.12

0.88
(continued)

Table 4.12c Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 18 to 25) (continued)
18 to 25
2012 Main Study
Questionnaire Module

158

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

20.32
1.47
2.61
0.52

18.63
1.32
2.37
0.44

21.43
1.45
2.58
0.62

19.63
1.28
2.33
0.55

17.83
0.49
2.36
0.56

16.47
0.43
2.15
0.48

16.98
0.48
2.34
0.57

15.64
0.42
2.12
0.53

17.46
0.49
2.35
0.56

16.32
0.43
2.13
0.50

3.95

3.23

3.84

3.12

3.83

3.03

3.40

2.92

3.65

2.98

1.44
1.62

1.33
1.32

1.43
1.55

1.33
1.25

N/A
1.30

N/A
1.12

N/A
1.23

N/A
1.03

N/A
1.27

N/A
1.10

1.63
1.04

1.33
0.90

1.59
2.50

1.30
2.27

1.56
2.19

1.27
2.05

1.42
2.15

1.19
1.98

1.50
2.18

1.21
2.02

2.03
1.32
2.88
N/A
3.23
2.89

1.75
1.18
2.25
N/A
2.93
1.18

1.99
1.28
2.13
N/A
3.13
2.85

1.68
1.15
1.73
N/A
2.83
1.18

1.85
1.03
1.61
N/A
2.95
2.80

1.55
0.98
1.61
N/A
2.73
1.23

1.87
1.04
N/A
N/A
2.83
2.64

1.66
0.93
N/A
N/A
2.52
1.31

1.86
1.03
1.61
N/A
2.90
2.73

1.57
0.95
1.61
N/A
2.65
1.25

N/A
N/A
0.82

N/A
N/A
0.68

N/A
N/A
0.79

N/A
N/A
0.65

N/A
N/A
0.62

N/A
N/A
0.52

N/A
N/A
0.57

N/A
N/A
0.47

N/A
N/A
0.60

N/A
N/A
0.50

5.80
0.67
4.70

5.67
0.55
4.68

5.68
0.64
4.67

5.53
0.52
4.63

3.84
0.74
1.86

3.58
0.65
1.73

4.17
0.74
1.74

3.91
0.65
1.54

3.98
0.74
1.81

3.68
0.65
1.67
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.58

1.35

1.60

1.35

1.45

1.24

1.92

1.35

1.65

1.30

0.40
N/A
1.42
3.61
3.05
0.63
59.37

0.25
N/A
1.33
3.20
2.67
N/A
56.45

0.45
N/A
1.40
3.49
3.36
0.37
59.93

0.27
N/A
1.32
3.12
2.88
N/A
56.75

0.53
0.31
1.39
2.58
3.30
0.15
52.47

0.42
0.00
1.25
2.37
2.93
N/A
50.23

0.47
0.01
1.38
2.77
3.13
0.16
51.59

0.35
0.00
1.18
2.25
2.87
N/A
48.41

0.50
0.19
1.39
2.66
3.23
0.15
52.09

0.38
0.00
1.22
2.32
2.89
N/A
49.52

159

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

160

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.81
2.35
1.67
2.35
3.42
12.67
2.47
2.59
0.64
0.66
0.45
1.27
1.40
N/A
6.48
N/A
N/A
N/A
N/A

1.60
1.93
1.48
2.12
3.17
11.40
2.15
2.30
0.48
0.52
0.30
1.10
1.18
N/A
5.85
N/A
N/A
N/A
N/A

1.68
2.34
1.64
2.31
3.40
12.57
2.54
2.60
0.64
0.63
0.48
1.24
1.37
N/A
6.48
N/A
N/A
N/A
N/A

1.52
1.93
1.47
2.08
3.15
11.20
2.20
2.30
0.48
0.52
0.28
1.07
1.18
N/A
5.82
N/A
N/A
N/A
N/A

1.59
2.05
1.06
2.09
2.99
13.16
2.32
2.44
0.68
0.62
0.57
1.53
1.54
0.40
5.85
2.21
0.84
0.88
0.74

1.40
1.60
1.05
1.92
2.87
11.46
1.98
2.22
0.52
0.52
0.27
1.27
1.34
0.30
5.03
1.97
0.70
0.75
0.63

1.47
2.01
1.09
2.01
3.14
13.36
2.23
2.43
0.68
0.67
0.35
1.62
1.56
0.50
5.90
2.23
0.84
0.85
0.76

1.37
1.60
1.10
1.82
2.95
11.98
2.00
2.27
0.50
0.55
0.28
1.38
1.35
0.32
5.18
1.92
0.68
0.72
0.63

1.53
2.03
1.07
2.05
3.06
13.25
2.28
2.44
0.68
0.64
0.45
1.58
1.55
0.45
5.87
2.22
0.84
0.86
0.75

1.38
1.60
1.08
1.88
2.90
11.68
1.98
2.23
0.50
0.53
0.28
1.30
1.34
0.32
5.08
1.95
0.70
0.73
0.63

3.03

2.70

3.06

2.73

3.07

2.62

3.07

2.57

3.07

2.60

1.84

1.62

1.78

1.57

1.49

1.16

1.56

1.26

1.52

1.22

1.92

1.68

1.94

1.68

1.27

1.15

1.32

1.14

1.29

1.14

1.85

1.53

1.91

1.65

1.18

0.99

1.18

0.89

1.18

0.96
(continued)

Table 4.12d Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 26 to 29) (continued)
26 to 29
2012 Main Study
Questionnaire Module

161

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

22.42
1.64
3.05
0.47

20.35
1.43
2.73
0.42

24.20
1.63
3.08
0.54

21.95
1.40
2.73
0.47

20.17
0.57
2.73
0.50

18.65
0.50
2.48
0.45

20.34
0.61
2.80
0.55

18.63
0.52
2.60
0.47

20.25
0.58
2.76
0.52

18.64
0.52
2.52
0.47

3.62

2.87

3.62

2.83

3.58

2.87

3.58

2.67

3.58

2.80

1.58
1.76

1.45
1.43

1.57
1.74

1.42
1.40

N/A
1.44

N/A
1.23

N/A
1.38

N/A
1.20

N/A
1.41

N/A
1.23

1.85
1.29

1.50
1.10

1.88
2.98

1.48
2.63

1.79
2.58

1.45
2.33

1.75
2.48

1.43
2.37

1.77
2.53

1.45
2.34

2.24
1.48
2.61
N/A
3.74
3.49

1.90
1.30
2.20
N/A
3.37
1.58

2.15
1.44
2.54
N/A
3.65
3.49

1.83
1.27
2.13
N/A
3.25
1.65

2.17
1.21
2.34
N/A
3.52
3.44

1.86
1.07
1.88
N/A
3.08
1.86

2.10
1.20
2.63
N/A
3.51
3.29

1.82
1.10
2.15
N/A
3.23
1.68

2.14
1.21
2.47
N/A
3.51
3.37

1.85
1.08
2.06
N/A
3.17
1.77

N/A
N/A
0.69

N/A
N/A
0.60

N/A
N/A
0.69

N/A
N/A
0.60

N/A
N/A
0.53

N/A
N/A
0.48

N/A
N/A
0.52

N/A
N/A
0.50

N/A
N/A
0.52

N/A
N/A
0.48

5.64
0.24
5.05

5.45
0.13
4.93

5.42
0.22
4.91

5.27
0.13
4.80

4.02
0.59
2.19

3.68
0.47
2.02

4.45
0.64
2.00

4.10
0.57
1.83

4.21
0.62
2.11

3.92
0.52
1.94
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.49

1.27

1.46

1.23

1.33

1.17

1.38

1.15

1.35

1.17

0.31
N/A
1.35
3.49
3.04
0.92
62.94

0.22
N/A
1.23
3.03
2.57
N/A
59.39

0.34
N/A
1.35
3.38
3.31
0.44
63.81

0.23
N/A
1.22
2.95
2.73
N/A
59.63

0.40
0.16
1.40
2.91
3.20
0.12
56.59

0.35
0.00
1.23
2.48
2.75
N/A
53.52

0.45
0.00
1.40
2.56
3.17
0.19
57.03

0.33
0.00
1.23
2.27
2.67
N/A
53.92

0.43
0.09
1.40
2.75
3.19
0.16
56.79

0.35
0.00
1.23
2.40
2.73
N/A
53.63

162

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

163

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.95
2.45
1.73
2.58
4.15
14.53
2.80
2.92
0.70
0.72
0.38
1.46
1.62
N/A
7.42
N/A
N/A
N/A
N/A

1.68
1.97
1.52
2.30
4.08
13.08
2.35
2.68
0.53
0.58
0.33
1.25
1.36
NA
6.77
N/A
N/A
N/A
N/A

1.90
2.57
1.73
2.54
4.12
14.35
2.78
2.87
0.71
0.75
0.45
1.49
1.83
N/A
7.50
N/A
N/A
N/A
N/A

1.57
2.02
1.48
2.20
4.02
12.88
2.40
2.63
0.53
0.60
0.35
1.30
1.40
N/A
6.58
N/A
N/A
N/A
N/A

1.72
2.22
1.40
2.42
4.31
16.71
2.86
3.04
0.71
0.63
0.33
1.81
2.09
0.57
7.45
2.92
1.17
1.23
1.12

1.49
1.83
1.48
2.08
4.19
14.85
2.33
2.63
0.63
0.58
0.34
1.64
1.80
0.38
6.35
2.41
0.97
0.93
0.86

1.67
2.19
1.31
2.49
4.08
15.47
2.37
2.62
0.73
0.72
0.38
1.84
1.92
0.39
7.30
2.79
1.14
1.21
1.06

1.35
1.70
1.38
2.13
3.95
13.53
2.14
2.48
0.62
0.63
0.47
1.63
1.55
0.37
5.92
2.20
0.87
0.93
0.77

1.70
2.21
1.36
2.45
4.21
16.15
2.64
2.85
0.72
0.67
0.35
1.83
2.01
0.46
7.38
2.86
1.16
1.22
1.09

1.43
1.77
1.45
2.08
4.08
14.11
2.23
2.53
0.62
0.60
0.37
1.63
1.65
0.38
6.13
2.28
0.90
0.93
0.83

3.34

2.99

3.48

3.06

3.70

3.13

3.69

2.96

3.69

3.07

2.16

1.95

2.19

1.94

1.75

1.53

1.60

1.37

1.67

1.41

2.16

1.97

2.40

2.08

1.98

1.38

1.41

1.20

1.68

1.26

1.73

1.43

1.81

1.59

1.37

1.56

1.10

1.57

1.25
(continued)

Table 4.12e Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 50 to 64) (continued)
50 to 64
2012 Main Study
Questionnaire Module

164

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

24.43
1.80
3.55
0.63

22.00
1.58
3.18
0.52

26.71
1.84
3.62
0.69

24.15
1.60
3.22
0.59

22.47
0.66
3.49
0.78

20.63
0.62
3.22
0.63

21.66
0.66
3.26
0.84

20.20
0.57
2.85
0.78

22.10
0.66
3.39
0.81

20.28
0.58
3.12
0.73

3.61

2.82

3.52

2.78

3.98

3.10

3.69

2.87

3.85

3.08

1.82
1.91

1.68
1.57

1.74
1.92

1.47
1.60

N/A
1.63

N/A
1.44

N/A
1.49

N/A
1.37

N/A
1.57

N/A
1.40

2.02
1.76

1.65
1.50

2.03
3.83

1.65
3.43

1.93
3.54

1.63
3.27

1.80
3.06

1.50
2.88

1.87
3.32

1.57
3.12

2.74
1.71
2.81
N/A
4.53
4.16

2.33
1.50
2.46
N/A
4.10
1.67

2.67
1.68
2.57
N/A
4.42
4.15

2.20
1.48
2.28
N/A
3.95
1.95

2.48
1.51
2.85
N/A
4.17
3.56

2.11
1.40
2.68
N/A
3.64
1.66

2.40
1.46
2.75
N/A
4.38
3.41

2.17
1.28
2.72
N/A
4.05
1.70

2.44
1.49
2.79
N/A
4.27
3.48

2.15
1.35
2.68
N/A
3.80
1.68

N/A
N/A
0.74

N/A
N/A
0.65

N/A
N/A
0.73

N/A
N/A
0.65

N/A
N/A
0.74

N/A
N/A
0.57

N/A
N/A
0.55

N/A
N/A
0.52

N/A
N/A
0.65

N/A
N/A
0.55

5.24
0.19
4.76

5.13
0.12
4.75

4.96
0.18
4.64

4.93
0.12
4.68

4.58
0.67
2.48

4.17
0.55
2.29

4.81
0.75
2.08

4.52
0.63
2.03

4.69
0.71
2.30

4.32
0.58
2.15
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

1.14

0.92

1.07

0.85

1.09

0.97

1.31

0.78

1.19

0.90

0.32
N/A
1.38
3.51
3.34
1.03
67.76

0.22
N/A
1.23
3.03
2.70
N/A
63.78

0.32
N/A
1.45
3.41
3.44
0.50
69.19

0.23
N/A
1.22
2.93
2.88
N/A
64.98

0.50
0.19
1.71
3.34
3.85
0.17
66.76

0.38
0.00
1.50
3.01
2.95
N/A
62.60

0.41
0.00
1.62
3.22
3.59
0.17
64.55

0.30
0.00
1.42
2.63
2.82
N/A
62.22

0.46
0.10
1.67
3.29
3.73
0.17
65.76

0.35
0.00
1.45
2.83
2.90
N/A
62.33

165

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

166

Questionnaire Module

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

1.94
2.70
1.86
3.02
4.88
17.52
3.38
3.48
0.86
0.78
0.72
2.02
2.32
N/A
9.49
N/A
N/A
N/A
N/A

1.70
2.18
1.62
2.67
4.75
15.95
2.98
3.18
0.72
0.68
0.37
1.43
2.03
N/A
8.36
N/A
N/A
N/A
N/A

1.90
2.95
1.81
2.95
4.90
17.42
3.45
3.54
0.93
1.15
0.72
2.21
3.50
N/A
9.79
N/A
N/A
N/A
N/A

1.67
2.50
1.57
2.55
4.70
15.68
3.07
3.18
0.73
0.78
0.72
1.52
1.77
N/A
8.62
N/A
N/A
N/A
N/A

1.73
2.66
1.53
2.91
5.37
22.37
3.02
3.83
1.09
1.09
0.39
2.02
1.66
0.53
10.77
4.27
1.68
1.69
1.60

1.57
2.30
1.57
2.32
5.13
19.56
2.46
3.75
0.80
0.88
0.39
2.25
1.66
0.42
8.83
3.05
1.27
1.27
1.25

2.05
2.65
1.61
2.89
5.32
22.90
3.13
3.70
0.90
0.61
0.23
1.65
2.17
0.47
10.75
4.32
1.95
1.89
1.78

1.73
2.04
1.63
2.56
4.83
19.27
2.62
3.43
0.80
0.55
0.23
1.36
2.16
0.37
9.04
3.31
1.26
1.29
1.21

1.88
2.66
1.57
2.90
5.35
22.61
3.07
3.77
1.00
0.92
0.34
1.79
2.00
0.49
10.76
4.29
1.80
1.78
1.68

1.65
2.27
1.60
2.50
4.95
19.52
2.58
3.50
0.80
0.78
0.32
1.58
1.93
0.41
8.94
3.15
1.27
1.27
1.22

4.40

4.01

4.62

3.94

5.46

4.17

4.91

4.32

5.21

4.22

3.43

2.90

3.00

2.83

2.61

2.08

2.99

1.88

2.78

1.97

2.96

2.40

2.63

2.33

2.86

2.15

2.32

2.07

2.60

2.15

2.61

1.80

2.95

2.62

2.66

2.27

2.46

2.12

2.57

2.19
(continued)

Table 4.12f Overall and Module Mean/Median Timing Data for English-Language Interviews from Non-Hispanic Respondents in the
2012 Main Study, Q3-Q4 2013 Main Study, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal in
Minutes (Affirmative Gate Respondents Aged 65 or Older) (continued)
65 or Older
2012 Main Study
Questionnaire Module

167

Q3-Q4 2013 Main Study

QFT

Combined QFT and
Dress Rehearsal

Mean

Median

Mean

Median

Mean

Median

Mean

Median

Mean

Median

27.02
2.09
4.64
0.64

24.82
1.90
4.07
0.53

29.39
2.08
4.51
0.75

27.03
1.87
3.98
0.67

26.73
0.75
4.34
0.84

24.08
0.67
3.85
0.64

27.15
0.77
4.42
1.07

25.80
0.65
4.08
1.02

26.93
0.76
4.38
0.95

24.48
0.67
3.90
0.78

3.54

2.98

3.50

2.93

3.67

2.90

3.34

2.82

3.51

2.85

2.34
1.76

1.57
1.45

1.86
1.77

1.80
1.40

N/A
1.84

N/A
1.47

N/A
1.66

N/A
1.56

N/A
1.75

N/A
1.50

2.23
2.52

1.92
2.17

2.17
5.25

1.87
4.65

2.35
4.76

1.95
4.35

2.17
5.38

2.02
4.38

2.27
5.05

1.98
4.35

3.54
2.28
3.88
N/A
5.98
3.87

2.87
2.02
3.63
N/A
5.35
1.35

3.48
2.23
4.10
N/A
5.65
3.75

2.80
1.97
3.48
N/A
5.07
1.35

3.47
1.94
4.80
N/A
5.66
2.58

3.19
1.77
4.80
N/A
4.93
1.07

3.26
1.98
2.80
N/A
5.60
3.54

2.67
1.73
2.80
N/A
5.18
1.23

3.35
1.96
4.13
N/A
5.63
3.05

2.83
1.77
4.00
N/A
5.02
1.08

N/A
N/A
0.85

N/A
N/A
0.72

N/A
N/A
0.82

N/A
N/A
0.72

N/A
N/A
0.80

N/A
N/A
0.68

N/A
N/A
0.73

N/A
N/A
0.67

N/A
N/A
0.77

N/A
N/A
0.68

3.00
0.16
2.55

1.83
0.12
1.38

2.67
0.16
2.40

1.65
0.12
1.35

4.94
0.90
2.09

4.40
0.68
1.75

5.77
0.92
2.04

4.93
0.78
1.47

5.33
0.91
2.07

4.58
0.75
1.63
(continued)

Mean

Median

Q3-Q4 2013 Main Study
Mean

QFT

Combined QFT and
Dress Rehearsal

Median

Mean

Median

Mean

Median

Mean

Median

0.91

0.67

0.80

0.60

0.95

0.75

0.72

0.61

0.84

0.68

0.31
N/A
1.50
3.83
3.93
0.98
73.32

0.20
N/A
1.30
3.32
3.10
N/A
69.43

0.36
N/A
1.48
3.66
3.96
0.55
74.70

0.20
N/A
1.28
3.17
3.20
N/A
70.13

0.48
0.33
2.15
4.41
4.06
0.18
80.47

0.42
0.00
1.95
3.90
3.18
N/A
74.62

0.71
0.00
1.92
4.25
4.83
0.15
82.60

0.31
0.00
1.71
3.48
3.41
N/A
77.01

0.59
0.18
2.05
4.33
4.42
0.16
81.45

0.37
0.00
1.82
3.68
3.30
N/A
75.32

168

4.5.3

Detailed Interview Timing Data for Selected Modules from the 2012 and 2013
Quarters 3 and 4 Comparison Data and the 2012 Questionnaire Field Test and
2013 Dress Rehearsal Data

Administration times for the 2012 and 2013 quarters 3 and 4 comparison samples and the
DR were calculated according to standard timing data calculation procedures for a number of
specific questionnaire sections. Tables 4.13a through 4.13x present unweighted overall DR
timing results and results for selected modules for all non-Hispanic respondents who opted to
take the interview in English and for five separate age groups.19 Overall DR timing data for
interviews conducted in Spanish are included in Table 4.13a as well. Timing results by age
group for each section are presented in separate tables for the DR interviews, the 2012
comparison interviews, and the 2013 quarters 3 and 4 comparison interviews. For each age group
category, these tables provide the number of interviews, the number of extreme or missing
records, summary statistics, quartiles, percentiles, and the highest and lowest extreme cases.
Respondents with an overall interview administration time of less than 30 minutes or greater than
240 minutes were classified as outliers and were excluded from these timing results.
As noted in Section 4.5.1.1, the partially redesigned DR instrument took less than
60 minutes on average to administer among English-speaking respondents aged 12 or older, as
shown in Table 4.10a. Examining English-speaking timing data within age groups reveals that
respondents aged 65 or older experienced the longest average administration times among all age
groups, with an overall mean of more than 82 minutes. The difference between the timings for
the 65 or older group and the younger age groups was greater than other differences between
groups, indicating that the burden increased for that age group.
Respondents aged 50 to 64 also had a mean administration time that was considerably
higher than the mean for all DR English-speaking respondents. Mean interview timings for
respondents aged 12 to 17 were similar to the overall mean for DR respondents, while the
average times for respondents aged 18 to 25 and those aged 26 to 49 were lower than the overall
mean for DR respondents. The overall timing patterns across age groups for DR respondents
were rather consistent with the patterns for the 2012 comparison data interviews and the 2013
quarters 3 and 4 comparison interviews, as shown in Tables 4.10b through 4.10e.
Tables 4.13f through 4.13h provide timing results for the tobacco module for respondents
who answered the question LEADCIG in the DR interviews, the 2012 comparison interviews,
and the 2013 quarters 3 and 4 comparison interviews. One difference between the DR
questionnaire and the 2011 and 2012 quarters 3 and 4 questionnaire was that questions about
chewing tobacco and snuff were combined in the tobacco module for the QFT questionnaire.
This change was intended to increase efficiency in collecting age at first use, recency, and
frequency of smokeless tobacco use. In addition, this section in the DR questionnaire no longer
collected data on the brand of smokeless tobacco that the respondent has used. As expected, the
efficiencies produced by these changes to the DR questionnaire resulted in a slightly lower mean
timing for this module among DR respondents (1.77 minutes) compared with the 2012
comparison respondents (1.98 minutes) and the 2013 quarters 3 and 4 comparison respondents
19

To aid in their readability, Tables 4.13a through 4.13x appear together at the end of this discussion in
Section 4.5.3.

169

(1.93 minutes). Because of the limitations with small sample sizes, these module-specific timing
data are only presented for non-Hispanic respondents who opted to take the interview in English.
Discussions of individual module timings in the next paragraphs are limited to this Englishspeaking non-Hispanic subsample.
As Tables 4.13i through 4.13l indicate, older respondents generally took more time than
younger respondents to complete the four prescription drug module screeners—pain relievers,
tranquilizers, stimulants, and sedatives. The screeners included in the DR questionnaire asked
respondents to report any past year use of prescription pain relievers, tranquilizers, stimulants,
and sedatives. These screener questions then asked respondents to report all use of drugs in each
category, both those that were prescribed and those that were misused. The mean pain relievers
screener administration time was nearly 2½ minutes, which was the longest of the four screeners.
Because the prescription drug screeners were new in the redesign, timing data for these sections
cannot be compared with the 2012 and 2013 quarters 3 and 4 comparison interviews.
In the DR instrument, the four prescription drug main modules followed the screeners
and asked, for each drug used in the past year, whether respondents misused any of them.
Respondents who reported never using a particular class of drug in the past year skipped the
main module and were excluded from the timing data for the four prescription drug main
modules presented in Tables 4.13m through 4.13x. These tables provide timing results for the
prescription drug main modules for the DR interviews, 2012 comparison interviews, and 2013
quarters 3 and 4 comparison interviews. Among DR respondents who answered questions in the
pain reliever, tranquilizer, and stimulant main modules, those aged 50 to 64 had the longest mean
administration times (Tables 4.13m, 4.13n, and 4.13o), although this timing was almost identical
to the timing for the 65 or older age group. This finding was similar to the timing of the 2012 and
2013 quarters 3 and 4 comparison samples, where respondents aged 65 or older generally had the
longest mean administration times for these prescription drug modules among all age groups
(Tables 4.13p through 4.13u), followed by those in the 50 to 64 age group. For the sedatives
main module, respondents aged 65 or older had the longest mean administration times among all
age groups for the DR interviews, 2012 comparison interviews, and 2013 quarters 3 and 4
comparison interviews (Tables 4.13v, 4.13w, and 4.13x). Overall, excluding the new prescription
drug screeners, the mean timings for each of the four prescription drug main modules were lower
for the DR respondents than for the 2012 and 2013 quarters 3 and 4 comparison respondents.

4.6

Other Data Quality Indicators
4.6.1

Overview of Other Data Quality Indicators

Examination of other data quality indicators for the DR focused on the following:
•

choosing "other" responses for which respondents subsequently were asked to specify
a written response (i.e., "OTHER, Specify" data), such as other sources of
prescription psychotherapeutic drugs; and

•

potential patterned responses in answers to the screening questions for past year
prescription drug use or to the questions for past year misuse.

Identification and handling of potential patterned responses in the 2011 and 2012 comparison
data also are discussed in this section.
170

Table 4.13a Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes, in Total
and by Age Groups: All Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
2,020
55

12-17
498
6

18-25
494
30

26-49
713
16

50-64
184
3

65+
131
0

61.77
494.57
22.24

60.00
254.52
15.95

54.06
277.58
16.66

61.96
538.11
23.20

68.42
522.26
22.85

87.25
1017.78
31.90

221.80
71.18
57.33
46.69
30.00
51.05
191.80

137.67
68.77
57.41
48.73
31.98
48.47
105.68

160.83
62.98
51.17
41.75
30.00
34.92
130.83

175.50
70.57
56.85
45.70
30.05
46.07
145.45

142.43
81.79
63.33
50.65
30.03
42.30
112.40

221.80
106.10
82.08
62.13
32.82
85.95
188.98

137.67
104.96
90.23
38.88
35.47
31.00

110.55
90.25
79.42
42.30
38.97
33.12

100.67
84.95
77.35
35.90
33.02
30.25

140.92
106.57
90.82
38.35
34.45
31.00

132.88
116.58
101.22
43.52
40.57
31.40

199.07
143.88
124.63
55.42
47.48
38.03

221.80
199.07
175.50
165.22
163.45
30.12
30.10
30.05
30.03
30.00

137.67
137.40
130.57
111.90
110.55
33.12
32.92
32.88
32.60
31.98

160.83
121.25
109.53
107.78
100.67
30.25
30.15
30.12
30.10
30.00

175.50
163.45
162.67
162.07
161.87
30.87
30.75
30.53
30.28
30.05

142.43
132.88
129.38
125.30
124.90
35.67
34.97
34.35
31.40
30.03

221.80
199.07
165.22
160.35
158.48
40.93
40.23
38.05
38.03
32.82

* Extreme records have an interview length of less than 30 minutes or more than 240 minutes. Respondents with 0 seconds for
this section are also excluded.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

171

Table 4.13b Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes, in Total
and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

59.24
400.24
20.01

58.84
258.65
16.08

51.59
226.57
15.05

57.03
339.76
18.43

64.55
368.57
19.20

82.60
781.46
27.95

199.07
67.65
55.60
45.38
30.00
54.42
169.07

137.67
65.70
56.35
48.12
31.98
40.40
105.68

107.78
59.99
48.41
40.68
30.00
44.85
77.78

161.87
65.55
53.92
44.33
30.53
46.07
131.33

129.38
74.60
62.22
49.85
30.03
42.30
99.35

199.07
99.58
77.01
60.63
32.82
85.95
166.25

124.70
98.63
84.03
38.27
34.45
30.82

130.57
90.07
76.22
42.30
40.02
32.92

94.00
80.62
74.55
34.68
31.73
30.12

116.67
91.42
78.62
37.43
33.40
31.00

120.27
101.58
90.15
42.48
40.53
31.40

160.35
124.70
117.30
53.65
47.42
38.03

199.07
161.87
160.35
147.88
140.92
30.15
30.12
30.10
30.03
30.00

137.67
137.40
130.57
111.90
110.55
33.58
33.20
32.92
32.88
31.98

107.78
96.60
94.00
92.03
90.23
30.25
30.15
30.12
30.10
30.00

161.87
140.92
132.30
122.00
116.67
31.00
30.98
30.87
30.75
30.53

129.38
120.27
114.35
109.85
109.47
37.82
34.97
34.35
31.40
30.03

199.07
160.35
147.88
139.78
129.12
40.93
40.23
38.05
38.03
32.82

172

Table 4.13c Overall Interview Timing Data for the Dress Rehearsal Protocol in Minutes, in Total
and by Age Groups: Spanish-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
183
1

12-17
37
0

18-25
17
1

26-49
99
0

50-64
18
0

65+
12
0

83.94
902.41
30.04

65.57
322.38
17.95

77.04
793.00
28.16

86.64
913.84
30.23

96.17
787.87
28.07

109.71
1120.24
33.47

175.50
98.73
79.32
62.12
33.12
72.82
142.38

108.98
75.97
63.40
53.87
33.12

160.83
80.25
72.18
63.83
46.87

142.43
117.30
94.75
72.82
48.92

165.22
139.99
100.19
88.68
61.58

75.87

113.97

175.50
101.67
85.17
63.82
34.60
68.67
140.90

93.52

103.63

165.22
142.43
129.23
50.30
45.17
34.60

108.98
100.08
91.20
43.95
36.73
33.12

160.83
160.83
121.25
46.88
46.87
46.87

175.50
147.68
129.72
50.48
45.07
34.60

142.43
142.43
132.88
61.17
48.92
48.92

165.22
165.22
143.88
65.13
61.58
61.58

175.50
165.22
163.45
162.67
162.07
38.87
38.87
36.73
34.60
33.12

108.98
100.08
97.08
91.20
89.40
46.00
43.95
38.87
36.73
33.12

160.83
121.25
94.02
85.57
80.25
63.83
58.87
51.15
46.88
46.87

175.50
163.45
162.67
162.07
147.68
45.07
42.02
40.55
38.87
34.60

142.43
132.88
125.30
124.90
117.30
72.82
63.80
61.48
61.17
48.92

165.22
143.88
143.85
136.13
131.80
91.20
90.50
86.87
65.13
61.58

173

Table 4.13d Overall Interview Timing Data for the 2012 Comparison Protocol in Minutes, in Total
and by Age Groups: 2012 Comparison English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,052
212

12-17
17,555
46

18-25
17,932
100

26-49
12,638
48

50-64
4,212
9

65+
2,715
9

60.69
292.19
17.09

59.78
229.39
15.15

58.24
252.37
15.89

60.88
303.67
17.43

66.53
398.87
19.97

72.79
498.31
22.32

237.43
68.95
57.87
48.93
30.00
54.55
207.43

237.43
67.60
57.73
49.35
30.12
50.87
207.32

229.95
66.13
55.58
47.25
30.00
49.10
199.95

234.93
69.53
57.93
48.73
30.03
57.40
204.90

228.65
75.84
63.13
53.07
30.17
62.38
198.48

219.15
84.90
68.97
56.98
31.50
59.20
187.65

115.67
91.53
81.98
42.50
39.10
34.07

105.75
86.23
78.68
43.25
39.80
34.62

108.75
87.17
78.22
41.13
37.87
33.32

115.90
93.07
82.78
42.23
38.88
34.02

130.78
102.45
90.83
45.58
41.93
35.90

140.08
111.77
100.98
48.53
44.02
36.85

237.43
234.93
229.95
228.65
227.67
30.07
30.05
30.03
30.02
30.00

237.43
225.62
221.42
215.20
197.30
30.43
30.43
30.28
30.12
30.12

229.95
216.07
187.40
186.87
186.50
30.12
30.07
30.07
30.02
30.00

234.93
227.67
214.20
204.18
195.47
30.33
30.28
30.20
30.05
30.03

228.65
222.73
215.97
213.45
203.93
32.05
30.80
30.75
30.27
30.17

219.15
218.47
218.40
217.73
209.65
33.07
32.43
31.97
31.67
31.50

174

Table 4.13e Overall Interview Timing Data for the 2013 Comparison Protocol in Minutes, in Total
and by Age Groups: 2013 Comparison English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
28,610
122

12-17
9,532
28

18-25
9,240
55

26-49
6,403
31

50-64
2,122
6

65+
1,313
2

62.03
352.37
18.77

61.58
290.86
17.05

59.26
311.89
17.66

62.26
372.32
19.30

67.88
454.23
21.31

74.14
565.33
23.78

238.67
70.25
58.75
49.63
30.00
50.27
208.67

238.67
69.43
58.93
50.25
30.03
46.20
208.63

215.98
66.82
56.28
47.58
30.03
50.27
185.95

232.70
70.97
58.68
49.38
30.00
54.78
202.70

233.08
76.87
64.35
53.73
30.15
71.00
202.93

235.87
85.98
69.60
58.48
30.63
70.00
205.23

127.97
94.85
84.05
42.97
39.35
33.87

121.40
90.07
80.90
44.10
40.57
34.62

120.45
90.08
80.47
41.18
37.78
32.98

129.73
96.35
85.40
42.63
38.97
33.78

140.98
107.15
92.02
46.38
42.55
36.57

156.00
116.97
99.70
49.63
45.23
37.10

238.67
235.87
233.08
232.70
227.60
30.08
30.03
30.03
30.03
30.00

238.67
195.57
195.10
193.90
193.20
30.33
30.30
30.23
30.03
30.03

215.98
214.27
213.03
207.17
197.70
30.30
30.27
30.20
30.08
30.03

232.70
227.60
225.95
213.22
211.45
30.37
30.27
30.08
30.08
30.00

233.08
218.60
212.80
206.63
184.50
31.98
31.63
31.47
30.77
30.15

235.87
226.30
215.70
213.32
200.40
34.72
33.63
33.38
31.48
30.63

175

Table 4.13f Overall Interview Timing Data for the Dress Rehearsal Tobacco Module in Minutes,
in Total and by Age Groups: English-Speaking Dress Rehearsal Respondents
Answering LEADCIG
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

1.77
1.77
1.33

1.39
1.07
1.03

1.61
1.31
1.14

1.87
1.74
1.32

1.95
1.54
1.24

2.54
4.23
2.06

13.77
2.38
1.40
0.87
0.17
1.27
13.60

8.03
1.52
1.10
0.82
0.35
0.87
7.68

6.45
2.28
1.38
0.66
0.17
0.47
6.28

10.98
2.50
1.57
0.93
0.22
1.22
10.77

7.25
2.45
1.65
1.07
0.38
0.63
6.87

13.77
3.01
2.22
1.26
0.42
1.23
13.35

6.53
4.17
3.27
0.53
0.42
0.25

5.62
3.70
2.63
0.60
0.50
0.37

5.43
3.62
3.12
0.38
0.30
0.17

6.53
4.13
3.30
0.53
0.40
0.27

6.37
4.55
3.98
0.63
0.55
0.42

13.03
5.55
4.22
0.87
0.73
0.47

13.77
13.03
10.98
8.60
8.52
0.18
0.18
0.17
0.17
0.17

8.03
6.60
5.62
5.18
4.67
0.40
0.40
0.37
0.37
0.35

6.45
6.20
5.43
5.42
5.08
0.18
0.18
0.17
0.17
0.17

10.98
8.52
8.45
6.87
6.53
0.27
0.27
0.25
0.25
0.22

7.25
6.37
5.23
4.87
4.65
0.53
0.47
0.47
0.42
0.38

13.77
13.03
8.60
7.17
6.97
0.68
0.65
0.58
0.47
0.42

176

Table 4.13g Overall Interview Timing Data for the 2012 Tobacco Module in Minutes, in Total and
by Age Groups: 2012 Comparison English-Speaking Respondents Answering
LEADCIG
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,040
212

12-17
17,545
46

18-25
17,931
100

26-49
12,637
48

50-64
4,212
9

65+
2,715
9

1.98
1.98
1.41

1.69
1.23
1.11

1.98
2.07
1.44

2.06
1.95
1.40

2.36
2.90
1.70

2.86
3.45
1.86

22.43
2.57
1.68
0.98
0.10
0.80
22.33

17.28
2.07
1.43
0.95
0.22
0.80
17.07

15.53
2.75
1.68
0.85
0.10
0.43
15.43

21.63
2.68
1.78
1.08
0.10
1.97
21.53

22.43
3.00
1.97
1.27
0.12
1.78
22.32

21.87
3.67
2.35
1.68
0.13
2.10
21.73

6.85
4.63
3.73
0.60
0.47
0.30

5.85
3.82
2.90
0.70
0.60
0.45

6.60
4.67
3.87
0.47
0.37
0.25

6.70
4.58
3.75
0.62
0.47
0.32

8.57
5.48
4.38
0.75
0.57
0.35

8.82
6.20
5.12
1.03
0.77
0.50

22.43
21.87
21.63
20.60
19.77
0.12
0.12
0.12
0.10
0.10

17.28
14.93
13.65
11.67
11.53
0.28
0.27
0.27
0.23
0.22

15.53
13.95
13.18
12.50
12.28
0.13
0.13
0.12
0.12
0.10

21.63
20.60
19.77
16.10
12.77
0.15
0.15
0.13
0.12
0.10

22.43
17.52
16.92
15.38
13.42
0.25
0.23
0.20
0.15
0.12

21.87
19.70
16.95
16.27
15.70
0.37
0.35
0.30
0.28
0.13

177

Table 4.13h Overall Interview Timing Data for the 2013 Tobacco Module in Minutes, in Total and
by Age Groups: 2013 Comparison English-Speaking Respondents Answering
LEADCIG
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
12,090
62

12-17
3,904
12

18-25
3,872
31

26-49
2,698
14

50-64
976
2

65+
640
2

1.93
1.91
1.38

1.65
1.11
1.05

1.94
2.04
1.43

2.04
2.01
1.42

2.24
1.91
1.38

2.78
4.07
2.02

22.93
2.50
1.63
0.97
0.13
0.67
22.80

13.38
2.04
1.39
0.93
0.27
0.80
13.12

18.90
2.72
1.63
0.78
0.15
0.42
18.75

22.93
2.63
1.77
1.07
0.13
1.88
22.80

8.63
2.84
1.93
1.27
0.27
2.03
8.37

21.90
3.53
2.30
1.56
0.25
1.98
21.65

6.45
4.52
3.67
0.60
0.45
0.30

5.62
3.63
2.82
0.70
0.60
0.47

6.28
4.65
3.83
0.43
0.35
0.25

6.57
4.63
3.75
0.63
0.45
0.32

7.23
4.98
4.12
0.75
0.60
0.40

10.42
6.21
5.10
0.85
0.67
0.38

22.93
21.90
19.12
18.90
13.38
0.18
0.18
0.17
0.15
0.13

13.38
11.95
10.00
9.27
9.12
0.35
0.33
0.33
0.30
0.27

18.90
12.73
10.43
9.62
9.52
0.18
0.18
0.18
0.17
0.15

22.93
19.12
9.87
9.50
9.47
0.22
0.20
0.20
0.20
0.13

8.63
8.42
8.37
7.88
7.88
0.32
0.32
0.28
0.27
0.27

21.90
13.17
12.83
12.75
12.23
0.37
0.37
0.33
0.27
0.25

178

Table 4.13i

Overall Interview Timing Data for the Dress Rehearsal Pain Reliever Screener in
Minutes, in Total and by Age Groups: English-Speaking Respondents

Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

2.38
2.52
1.59

2.20
1.02
1.01

1.86
1.04
1.02

2.23
1.58
1.26

2.79
3.44
1.85

4.32
8.30
2.88

16.53
2.70
1.97
1.48
0.60
2.10
15.93

8.88
2.62
1.98
1.53
0.62
1.17
8.27

10.93
2.21
1.60
1.26
0.60
1.27
10.33

9.53
2.55
1.92
1.48
0.75
1.58
8.78

9.67
2.88
2.20
1.83
0.97
2.10
8.70

16.53
5.31
3.31
2.36
0.77
2.28
15.77

9.12
5.33
3.83
1.18
1.03
0.75

6.00
3.93
3.40
1.22
1.10
0.67

5.33
3.52
2.88
1.03
0.88
0.73

8.60
4.68
3.35
1.20
1.03
0.87

9.15
8.48
4.68
1.38
1.25
1.03

13.55
9.80
8.97
1.73
1.38
1.07

16.53
13.55
11.88
10.93
10.73
0.67
0.63
0.62
0.62
0.60

8.88
6.02
6.00
5.73
5.33
0.72
0.72
0.67
0.62
0.62

10.93
6.13
5.33
5.25
4.98
0.75
0.73
0.73
0.63
0.60

9.53
9.00
8.87
8.63
8.60
0.87
0.83
0.80
0.77
0.75

9.67
9.15
9.12
8.85
8.73
1.10
1.10
1.08
1.03
0.97

16.53
13.55
11.88
10.73
10.25
1.33
1.28
1.27
1.07
0.77

179

Table 4.13j Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Screener in
Minutes, in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

0.90
0.65
0.81

0.72
0.20
0.45

0.65
0.15
0.39

0.84
0.34
0.58

1.14
0.95
0.97

1.95
2.65
1.63

8.05
0.98
0.68
0.50
0.18
0.48
7.87

4.93
0.85
0.62
0.43
0.18
0.48
4.75

3.22
0.77
0.55
0.42
0.18
0.38
3.03

5.10
0.97
0.68
0.52
0.23
0.53
4.87

5.28
1.17
0.87
0.65
0.38
0.87
4.90

8.05
2.42
1.26
0.88
0.40
0.68
7.65

4.95
2.15
1.43
0.38
0.33
0.25

2.32
1.42
1.17
0.35
0.32
0.27

2.38
1.27
1.05
0.33
0.30
0.23

3.65
1.73
1.28
0.42
0.37
0.27

5.03
3.98
1.77
0.52
0.48
0.42

7.73
5.28
4.93
0.68
0.65
0.50

8.05
7.73
5.92
5.45
5.42
0.23
0.22
0.20
0.18
0.60

4.93
2.47
2.32
2.22
2.02
0.28
0.28
0.27
0.20
0.62

3.22
2.68
2.38
2.28
2.22
0.23
0.23
0.23
0.22
0.60

5.10
4.95
4.83
4.00
3.65
0.27
0.27
0.25
0.23
0.75

5.28
5.03
4.92
4.90
4.83
0.45
0.45
0.43
0.42
0.97

8.05
7.73
5.92
5.45
5.42
0.63
0.62
0.62
0.50
0.77

180

Table 4.13k Overall Interview Timing Data for the Dress Rehearsal Stimulants Screener in
Minutes, in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

0.93
0.69
0.83

0.77
0.22
0.47

0.71
0.18
0.42

0.85
0.36
0.60

1.21
1.06
1.03

1.89
2.91
1.70

10.18
1.03
0.73
0.53
0.15
0.63
10.03

5.28
0.97
0.67
0.48
0.18
0.63
5.10

4.15
0.84
0.60
0.45
0.15
0.52
4.00

5.65
0.95
0.72
0.57
0.22
0.73
5.43

5.73
1.33
0.93
0.68
0.37
0.62
5.37

10.18
2.09
1.29
0.92
0.40
1.00
9.78

5.33
2.03
1.47
0.40
0.33
0.23

2.15
1.48
1.27
0.35
0.32
0.23

2.25
1.48
1.20
0.33
0.28
0.18

4.08
1.58
1.30
0.40
0.35
0.25

5.40
2.95
1.87
0.53
0.48
0.43

6.78
5.48
5.30
0.62
0.50
0.42

10.18
6.78
6.22
5.85
5.73
0.18
0.18
0.18
0.17
0.15

5.28
2.55
2.15
2.07
2.00
0.23
0.23
0.23
0.22
0.18

4.15
2.35
2.25
2.03
2.00
0.23
0.18
0.18
0.17
0.15

5.65
5.38
5.32
5.23
4.08
0.25
0.25
0.23
0.23
0.22

5.73
5.40
5.23
5.23
5.22
0.47
0.45
0.45
0.43
0.37

10.18
6.78
6.22
5.85
5.62
0.50
0.48
0.47
0.42
0.40

181

Table 4.13l

Overall Interview Timing Data for the Dress Rehearsal Sedatives Screener in
Minutes, in Total and by Age Groups: English-Speaking Respondents

5 Lowest

(Lowest)

Overall
1,275
40

12-17
283
6

18-25
284
19

26-49
455
12

50-64
145
3

65+
108
0

0.81
0.76
0.87

0.64
0.19
0.43

0.56
0.10
0.32

0.76
0.68
0.82

1.06
1.01
1.01

1.78
2.80
1.67

15.22
0.87
0.62
0.43
0.10
0.38
15.12

4.72
0.80
0.53
0.38
0.15
0.43
4.57

2.28
0.65
0.48
0.37
0.10
0.38
2.18

15.22
0.82
0.63
0.47
0.15
0.72
15.07

7.48
1.07
0.77
0.57
0.23
0.93
7.25

12.53
1.93
1.21
0.82
0.40
0.93
12.13

4.72
1.87
1.37
0.32
0.27
0.20

1.85
1.28
1.10
0.28
0.25
0.17

2.02
1.12
0.88
0.27
0.22
0.13

3.15
1.47
1.15
0.35
0.30
0.23

5.27
2.77
1.82
0.47
0.43
0.38

5.42
4.90
4.65
0.62
0.57
0.40

15.22
12.53
7.48
5.42
5.27
0.15
0.15
0.13
0.13
0.10

4.72
2.83
1.85
1.82
1.68
0.20
0.20
0.17
0.17
0.15

2.28
2.03
2.02
1.93
1.78
0.18
0.18
0.13
0.13
0.10

15.22
4.65
4.12
3.67
3.15
0.23
0.23
0.22
0.20
0.15

7.48
5.27
4.70
4.67
4.67
0.40
0.40
0.40
0.38
0.23

12.53
5.42
5.27
5.12
5.02
0.50
0.50
0.42
0.40
0.40

182

Table 4.13m Overall Interview Timing Data for the Dress Rehearsal Pain Reliever Module in
Minutes, in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
760
555

12-17
86
6

18-25
152
19

26-49
335
12

50-64
116
3

65+
71
0

0.89
1.16
1.08

0.94
0.88
0.94

0.83
0.98
0.99

0.86
1.44
1.20

1.00
1.18
1.08

0.99
0.52
0.72

11.13
0.98
0.58
0.33
0.07
0.20
11.07

4.32
1.03
0.62
0.38
0.07
0.12
4.25

6.63
0.84
0.49
0.28
0.08
0.20
6.55

11.13
0.90
0.52
0.30
0.10
0.23
11.03

5.48
1.05
0.68
0.43
0.15
0.17
5.33

4.18
1.45
0.80
0.48
0.10
0.77
4.08

5.12
3.08
1.89
0.20
0.15
0.10

4.32
2.77
2.63
0.18
0.12
0.07

5.10
3.17
2.02
0.17
0.15
0.08

6.42
3.08
1.65
0.18
0.15
0.12

5.30
4.03
1.63
0.22
0.17
0.15

4.18
1.77
1.65
0.35
0.22
0.10

11.13
10.38
7.45
6.63
6.42
0.10
0.08
0.08
0.08
0.07

4.32
3.90
3.65
2.97
2.77
0.12
0.12
0.10
0.10
0.07

6.63
5.10
3.78
3.73
3.42
0.10
0.10
0.08
0.08
0.08

11.13
10.38
7.45
6.42
4.75
0.12
0.12
0.12
0.12
0.10

5.48
5.30
5.12
5.03
4.92
0.17
0.17
0.17
0.15
0.15

4.18
4.05
2.02
1.77
1.75
0.23
0.22
0.22
0.18
0.10

183

Table 4.13n Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Module in
Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,052
212

12-17
17,555
46

18-25
17,932
100

26-49
12,638
48

50-64
4,212
9

65+
2,715
9

2.06
1.23
1.11

2.13
1.11
1.05

1.96
1.22
1.10

2.00
1.33
1.15

2.18
1.22
1.10

2.49
1.31
1.15

37.63
2.53
1.87
1.37
0.05
1.63
37.58

24.85
2.62
2.00
1.47
0.07
1.80
24.78

18.62
2.38
1.73
1.25
0.05
1.63
18.57

37.63
2.42
1.78
1.30
0.07
1.68
37.57

17.90
2.63
1.98
1.52
0.12
1.65
17.78

13.87
3.07
2.33
1.75
0.12
2.02
13.75

5.85
3.87
3.27
0.98
0.78
0.42

5.70
3.80
3.25
1.05
0.80
0.40

5.78
3.87
3.22
0.92
0.72
0.37

5.85
3.82
3.17
0.98
0.80
0.48

6.15
3.98
3.32
1.15
0.93
0.58

6.27
4.23
3.63
1.32
1.10
0.62

37.63
24.85
21.67
18.62
18.42
0.08
0.08
0.07
0.07
0.05

24.85
18.42
15.52
14.80
13.68
0.10
0.08
0.08
0.08
0.07

18.62
18.00
17.10
16.48
13.53
0.10
0.10
0.08
0.08
0.05

37.63
21.67
17.82
16.90
15.32
0.10
0.10
0.08
0.08
0.07

17.90
14.32
13.98
13.07
12.85
0.28
0.27
0.20
0.13
0.12

13.87
12.55
12.30
11.50
10.72
0.32
0.20
0.18
0.17
0.12

184

Table 4.13o Overall Interview Timing Data for the Dress Rehearsal Stimulants Module in
Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
28,609
123

12-17
9,532
28

18-25
9,240
55

26-49
6,402
31

50-64
2,122
6

65+
1,313
2

2.06
1.35
1.16

2.15
1.32
1.15

1.92
1.17
1.08

1.99
1.45
1.20

2.17
1.55
1.25

2.51
1.59
1.26

31.78
2.52
1.85
1.33
0.05
1.60
31.73

31.78
2.63
2.00
1.45
0.05
2.17
31.73

17.58
2.33
1.70
1.23
0.05
1.32
17.53

22.97
2.40
1.77
1.28
0.07
1.45
22.90

15.63
2.57
1.93
1.47
0.18
1.60
15.45

13.77
3.03
2.30
1.78
0.15
2.05
13.62

6.00
3.92
3.27
0.97
0.77
0.42

5.77
3.90
3.32
1.03
0.80
0.42

5.88
3.85
3.17
0.88
0.70
0.35

6.03
3.93
3.20
0.95
0.77
0.47

6.65
3.95
3.32
1.10
0.92
0.57

8.20
4.27
3.57
1.37
1.13
0.65

31.78
22.97
20.95
20.52
19.40
0.07
0.07
0.05
0.05
0.05

31.78
20.52
18.30
18.07
14.33
0.12
0.10
0.08
0.08
0.05

17.58
12.63
11.12
10.60
10.47
0.08
0.07
0.07
0.05
0.05

22.97
20.95
19.40
16.98
14.08
0.12
0.10
0.10
0.10
0.07

15.63
14.95
14.38
14.33
13.00
0.37
0.33
0.25
0.20
0.18

13.77
11.18
11.18
11.02
10.77
0.32
0.30
0.28
0.15
0.15

185

Table 4.13p Overall Interview Timing Data for the Dress Rehearsal Sedatives Module in Minutes,
in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
307
1,008

12-17
12
6

18-25
49
19

26-49
156
12

50-64
56
3

65+
34
0

0.68
0.65
0.81

0.72
0.68
0.82

0.93
0.94
0.97

0.61
0.66
0.81

0.57
0.39
0.63

0.85
0.50
0.71

5.07
0.70
0.38
0.20
0.03
0.17
5.03

2.93
0.93
0.37
0.20
0.15
0.22
2.78

3.58
1.63
0.48
0.18
0.03
0.13
3.55

5.07
0.58
0.33
0.18
0.07
0.17
5.00

3.48
0.56
0.38
0.24
0.08
0.20
3.40

3.53
1.15
0.66
0.32
0.15
0.32
3.38

3.58
2.45
1.82
0.13
0.10
0.07

2.93
2.93
1.57
0.17
0.15
0.15

3.58
3.02
2.48
0.13
0.07
0.03

4.08
2.33
1.53
0.12
0.10
0.07

3.48
2.20
1.05
0.15
0.12
0.08

3.53
2.07
1.82
0.27
0.17
0.15

5.07
4.08
4.08
3.58
3.53
0.07
0.07
0.07
0.05
0.05

2.93
1.57
1.15
0.70
0.67
0.22
0.22
0.18
0.17
0.05

3.58
3.25
3.02
2.98
2.48
0.13
0.10
0.07
0.05
0.05

5.07
4.08
4.08
3.28
3.03
0.08
0.08
0.08
0.07
0.07

3.48
2.52
2.20
2.00
1.07
0.13
0.13
0.12
0.12
0.18

3.53
2.07
1.88
1.82
1.55
0.28
0.27
0.20
0.17
0.15

186

Table 4.13q Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Module in
Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,052
212

12-17
17,555
46

18-25
17,932
100

26-49
12,638
48

50-64
4,212
9

65+
2,715
9

1.14
0.56
0.75

1.19
0.48
0.69

1.02
0.52
0.72

1.10
0.59
0.77

1.30
0.62
0.78

1.64
0.81
0.90

30.15
1.47
0.98
0.65
0.03
0.65
30.12

16.67
1.55
1.05
0.70
0.03
0.82
16.63

30.15
1.28
0.87
0.58
0.03
0.65
30.12

27.42
1.37
0.93
0.63
0.03
0.65
27.38

13.92
1.65
1.13
0.78
0.08
0.65
13.83

11.30
2.23
1.48
0.98
0.07
0.83
11.23

3.33
2.48
2.07
0.43
0.33
0.18

3.17
2.43
2.08
0.47
0.35
0.18

3.23
2.25
1.83
0.40
0.30
0.17

3.38
2.37
1.93
0.43
0.35
0.22

3.80
2.62
2.27
0.55
0.43
0.28

4.30
2.93
2.72
0.67
0.52
0.30

30.15
27.42
26.75
19.68
16.75
0.03
0.03
0.03
0.03
0.03

16.67
9.05
8.82
8.05
7.28
0.05
0.05
0.05
0.03
0.03

30.15
19.68
16.75
16.65
10.07
0.05
0.05
0.05
0.05
0.03

27.42
26.75
15.83
8.43
8.35
0.07
0.07
0.07
0.03
0.03

13.92
11.18
7.45
7.28
7.25
0.12
0.10
0.10
0.08
0.08

11.30
8.23
7.95
7.95
7.87
0.13
0.10
0.08
0.08
0.07

187

Table 4.13r Overall Interview Timing Data for the Dress Rehearsal Tranquilizer Module in
Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
28,609
123

12-17
9,532
28

18-25
9,240
55

26-49
6,402
31

50-64
2,122
6

65+
1,313
2

1.14
0.55
0.74

1.20
0.50
0.71

1.02
0.52
0.72

1.09
0.52
0.72

1.28
0.63
0.79

1.62
0.74
0.86

20.73
1.45
0.97
0.63
0.03
0.60
20.70

10.38
1.56
1.05
0.68
0.03
0.60
10.35

20.73
1.27
0.85
0.57
0.03
0.68
20.70

14.60
1.35
0.92
0.62
0.03
0.78
14.57

11.02
1.60
1.11
0.75
0.08
0.72
10.93

7.47
2.15
1.43
1.00
0.10
0.95
7.37

3.47
2.48
2.05
0.43
0.33
0.18

3.33
2.48
2.10
0.47
0.37
0.18

3.33
2.25
1.83
0.38
0.30
0.17

3.53
2.38
1.93
0.43
0.33
0.20

3.97
2.68
2.27
0.55
0.43
0.28

4.63
2.95
2.67
0.68
0.52
0.27

20.73
19.38
14.60
11.02
10.38
0.03
0.03
0.03
0.03
0.03

10.38
8.25
8.07
6.97
6.80
0.07
0.07
0.05
0.03
0.03

20.73
19.38
9.33
8.80
7.72
0.05
0.05
0.03
0.03
0.03

14.60
9.78
8.80
8.60
7.82
0.08
0.07
0.07
0.07
0.03

11.02
8.60
6.92
6.68
5.93
0.12
0.12
0.10
0.10
0.08

7.47
6.07
5.82
5.75
5.27
0.15
0.15
0.13
0.12
0.10

188

Table 4.13s Overall Interview Timing Data for the Dress Rehearsal Stimulants Module in
Minutes, in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
197
1,118

12-17
30
6

18-25
58
19

26-49
80
12

50-64
19
3

65+
10
0

0.60
0.51
0.71

0.61
0.49
0.70

0.91
0.83
0.91

0.44
0.34
0.58

0.37
0.09
0.31

0.50
0.09
0.30

3.42
0.65
0.32
0.15
0.02
0.13
3.40

2.65
0.65
0.37
0.20
0.07
0.07
2.58

3.42
1.50
0.39
0.13
0.02
0.13
3.40

3.12
0.46
0.25
0.15
0.05
0.13
3.07

1.20
0.47
0.27
0.15
0.08
0.13
1.12

1.13
0.78
0.35
0.30
0.27
0.35
0.87

3.18
2.38
1.72
0.10
0.07
0.05

2.65
2.42
1.95
0.08
0.07
0.07

3.42
2.95
2.30
0.10
0.07
0.02

3.12
2.03
0.73
0.10
0.08
0.05

1.20
1.20
0.87
0.12
0.08
0.08

1.13
1.13
0.96
0.28
0.27
0.27

3.42
3.18
3.12
2.95
2.67
0.07
0.07
0.07
0.05
0.02

2.65
2.42
2.38
1.52
0.90
0.13
0.08
0.07
0.07
0.07

3.42
3.18
2.95
2.65
2.60
0.08
0.07
0.07
0.07
0.02

3.12
2.67
2.27
2.07
1.98
0.08
0.07
0.07
0.07
0.05

1.20
0.87
0.82
0.63
0.47
0.15
0.13
0.13
0.12
0.08

1.13
0.78
0.78
0.38
0.35
0.35
0.33
0.30
0.28
0.27

189

Table 4.13t Overall Interview Timing Data for the Dress Rehearsal Stimulants Module in
Minutes, in Total and by Age Groups: 2012 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,052
212

12-17
17,555
46

18-25
17,932
100

26-49
12,638
48

50-64
4,212
9

65+
2,715
9

1.16
0.67
0.82

1.21
0.61
0.78

1.03
0.58
0.76

1.11
0.64
0.80

1.35
0.84
0.92

1.70
1.10
1.05

36.12
1.48
0.97
0.63
0.03
0.57
36.08

21.15
1.58
1.03
0.65
0.03
0.80
21.12

36.12
1.28
0.85
0.55
0.05
0.57
36.07

26.47
1.38
0.93
0.63
0.03
0.75
26.43

24.82
1.70
1.13
0.77
0.07
0.92
24.75

11.72
2.32
1.45
0.92
0.03
0.88
11.68

3.53
2.68
2.15
0.42
0.32
0.17

3.42
2.68
2.23
0.42
0.32
0.17

3.42
2.38
1.88
0.37
0.28
0.15

3.52
2.50
1.97
0.43
0.33
0.20

3.77
2.93
2.40
0.53
0.42
0.27

4.52
3.22
3.05
0.63
0.50
0.27

36.12
26.47
24.82
21.15
19.30
0.05
0.03
0.03
0.03
0.03

21.15
14.08
11.38
10.63
8.37
0.05
0.05
0.05
0.03
0.03

36.12
19.30
9.73
9.68
8.63
0.07
0.05
0.05
0.05
0.05

26.47
17.05
15.12
15.07
12.98
0.08
0.07
0.07
0.05
0.03

24.82
9.73
9.57
9.40
8.92
0.12
0.12
0.10
0.10
0.07

11.72
11.08
9.05
8.97
8.87
0.13
0.10
0.08
0.08
0.03

190

Table 4.13u Overall Interview Timing Data for the Dress Rehearsal Stimulants Module in
Minutes, in Total and by Age Groups: 2013 Comparison English-Speaking
Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
28,609
123

12-17
9,532
28

18-25
9,240
55

26-49
6,402
31

50-64
2,122
6

65+
1,313
2

1.16
0.64
0.80

1.22
0.61
0.78

1.02
0.50
0.71

1.10
0.62
0.79

1.34
0.86
0.93

1.64
0.97
0.98

15.72
1.47
0.95
0.62
0.03
0.73
15.68

13.33
1.60
1.03
0.65
0.05
0.73
13.28

9.95
1.28
0.83
0.55
0.03
0.63
9.92

15.72
1.37
0.92
0.60
0.07
0.78
15.65

12.78
1.67
1.13
0.75
0.07
1.00
12.72

13.40
2.20
1.42
0.93
0.10
1.33
13.30

3.70
2.70
2.17
0.42
0.32
0.17

3.60
2.70
2.27
0.43
0.32
0.17

3.38
2.37
1.88
0.37
0.28
0.15

3.72
2.58
2.00
0.42
0.32
0.20

4.25
3.00
2.45
0.52
0.42
0.25

4.70
3.18
2.98
0.63
0.48
0.22

15.72
13.40
13.33
12.78
12.30
0.05
0.05
0.05
0.05
0.03

13.33
10.10
7.47
6.92
6.07
0.07
0.07
0.07
0.07
0.05

9.95
8.73
7.45
7.23
7.17
0.07
0.05
0.05
0.05
0.03

15.72
11.60
10.98
10.80
9.90
0.10
0.08
0.07
0.07
0.07

12.78
12.30
8.17
7.55
6.85
0.10
0.10
0.10
0.08
0.07

13.40
6.13
5.97
5.77
5.67
0.12
0.12
0.12
0.10
0.10

191

Table 4.13v Overall Interview Timing Data for the Dress Rehearsal Sedatives Module in Minutes,
in Total and by Age Groups: English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
176
1,139

12-17
11
6

18-25
21
19

26-49
90
12

50-64
35
3

65+
19
0

0.47
0.96
0.98

0.58
0.98
0.99

0.24
0.08
0.28

0.38
1.27
1.13

0.49
0.45
0.67

1.03
1.13
1.06

10.63
0.38
0.22
0.13
0.03
0.13
10.60

3.37
0.93
0.15
0.08
0.07
0.08
3.30

1.37
0.25
0.15
0.10
0.07
0.10
1.30

10.63
0.32
0.18
0.13
0.03
0.13
10.60

3.68
0.53
0.28
0.17
0.07
0.22
3.62

4.43
1.15
0.63
0.38
0.20
0.20
4.23

4.43
1.37
0.93
0.08
0.07
0.05

3.37
3.37
1.03
0.08
0.07
0.07

1.37
0.58
0.30
0.08
0.07
0.07

10.63
0.83
0.52
0.09
0.07
0.03

3.68
1.98
0.78
0.12
0.07
0.07

4.43
4.43
2.63
0.20
0.20
0.20

10.63
4.43
3.68
3.37
2.63
0.07
0.05
0.05
0.05
0.03

3.37
1.03
0.93
0.22
0.20
0.12
0.08
0.08
0.08
0.07

1.37
0.58
0.30
0.30
0.28
0.10
0.08
0.08
0.07
0.07

10.63
1.60
1.37
1.37
0.83
0.07
0.05
0.05
0.05
0.03

3.68
1.98
1.22
0.78
0.77
0.12
0.12
0.10
0.07
0.07

4.43
2.63
2.40
1.15
1.15
0.38
0.37
0.25
0.20
0.20

192

Table 4.13w Overall Interview Timing Data for the Dress Rehearsal Sedatives Module in Minutes,
in Total and by Age Groups: 2012 Comparison English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
55,051
213

12-17
17,554
46

18-25
17,932
100

26-49
12,638
48

50-64
4,212
9

65+
2,715
9

0.94
0.46
0.68

1.01
0.46
0.68

0.78
0.31
0.55

0.87
0.37
0.61

1.12
0.61
0.78

1.54
1.07
1.03

16.92
1.17
0.75
0.50
0.02
0.52
16.90

7.30
1.32
0.83
0.52
0.03
0.43
7.27

12.23
0.95
0.63
0.43
0.03
0.52
12.20

16.13
1.07
0.72
0.50
0.02
0.45
16.12

16.92
1.38
0.93
0.63
0.03
0.65
16.88

15.28
2.13
1.28
0.83
0.03
1.00
15.25

3.05
2.32
1.78
0.33
0.25
0.15

3.07
2.40
1.95
0.35
0.27
0.15

2.78
1.80
1.40
0.30
0.23
0.13

2.95
2.02
1.57
0.35
0.27
0.17

3.23
2.58
2.10
0.43
0.33
0.22

3.93
3.02
2.88
0.53
0.40
0.20

16.92
16.13
15.28
13.53
12.73
0.03
0.03
0.03
0.03
0.02

7.30
6.72
6.53
6.48
6.47
0.05
0.05
0.05
0.03
0.03

12.23
10.33
9.98
9.88
9.45
0.05
0.05
0.03
0.03
0.03

16.13
10.18
7.80
7.58
7.27
0.05
0.05
0.05
0.03
0.02

16.92
12.73
8.55
7.80
7.20
0.07
0.07
0.05
0.05
0.03

15.28
13.53
10.62
10.27
8.87
0.10
0.08
0.07
0.07
0.03

193

Table 4.13x Overall Interview Timing Data for the Dress Rehearsal Sedatives Module in Minutes,
in Total and by Age Groups: 2013 Comparison English-Speaking Respondents
Age Group
Sample Used in Analysis
Extreme/Missing Records*
Summary Statistics (Minutes)
Mean
Variance
Standard Deviation
Quartiles
Maximum
Q3
Median
Q1
Minimum
Mode
Range
Percentiles
99%
95%
90%
10%
5%
1%
Extremes
5 Highest (Highest)

5 Lowest

(Lowest)

Overall
28,609
123

12-17
9,532
28

18-25
9,240
55

26-49
6,402
31

50-64
2,122
6

65+
1,313
2

0.94
0.52
0.72

1.02
0.50
0.71

0.79
0.37
0.61

0.87
0.40
0.63

1.13
0.87
0.93

1.50
1.05
1.03

20.37
1.17
0.75
0.50
0.02
0.47
20.35

11.87
1.33
0.83
0.53
0.02
0.57
11.85

16.28
0.95
0.63
0.43
0.05
0.47
16.23

15.18
1.07
0.72
0.48
0.07
0.47
15.12

20.37
1.40
0.92
0.62
0.07
0.70
20.30

12.28
2.03
1.23
0.82
0.12
0.82
12.17

3.20
2.32
1.80
0.33
0.27
0.15

3.25
2.43
1.98
0.35
0.27
0.13

2.93
1.83
1.43
0.30
0.23
0.13

2.97
2.08
1.60
0.33
0.27
0.15

3.63
2.72
2.08
0.42
0.33
0.20

4.45
3.00
2.85
0.53
0.40
0.20

20.37
16.28
15.18
14.48
13.47
0.05
0.03
0.03
0.03
0.02

11.87
7.80
6.47
6.10
5.85
0.05
0.03
0.03
0.03
0.02

16.28
12.48
8.35
7.83
6.42
0.05
0.05
0.05
0.05
0.05

15.18
11.18
6.65
6.17
6.12
0.08
0.08
0.07
0.07
0.07

20.37
14.48
13.47
6.65
5.60
0.13
0.13
0.12
0.10
0.07

12.28
10.68
8.08
7.88
6.65
0.13
0.13
0.13
0.12
0.12

194

Data quality indicators that were discussed in the QFT report for triggering of
inconsistency "flags," triggering of "hard errors" in the DR prescription drug data, and triggering
of consistency checks in the prescription drug initiation data were not examined for the DR. Very
small numbers of respondents in the QFT had triggered flags for inconsistent data in the modules
for smokeless tobacco, methamphetamine, and prescription drugs (i.e., fewer than five
respondents for any given flag that was set). No situations were identified in the audit trail data
for the QFT in which respondents triggered a hard error between the age at first use (AFU)
answers for individual prescription drugs and their current age (Currivan et al., 2013).
4.6.2

Responding to Lead Questions for "OTHER, Specify" Data

As noted in Section 3.3.2 in Chapter 3, only the "OTHER, Specify" data for Hispanic
origin, race, and drugs were coded for use in further data processing or analysis. However, data
for variables or response choices that govern whether respondents were asked "OTHER,
Specify" questions provide an indication of data quality. For example, if predefined categories
for a given question or predefined examples in preceding questions (e.g., specific prescription
drugs) are understandable and encompass the bulk of expected responses, then the rates should
be low for the residual "other" responses (e.g., obtaining pain relievers "some other way").
Estimates are shown in Table 4.14 for the following new, moved, or revised items in the
DR that have associated "OTHER, Specify" data:
•

race (question QD05), including other race;

•

source of the last pain reliever that the respondent misused (PRY42B), including
getting the drug some other way;

•

friend's or relative's source of the pain reliever that the respondent obtained from a
friend or relative for free (PRY42C), including getting the drug some other way;

•

source of the last tranquilizer that the respondent misused (TRY21B);

•

friend's or relative's source of the tranquilizer that the respondent obtained from a
friend or relative for free (TRY21C);

•

source of the last stimulant that the respondent misused (STY26B);

•

friend's or relative's source of the stimulant that the respondent obtained from a friend
or relative for free (STY26C);

•

source of the last sedative that the respondent misused (SVY19B);

•

friend's or relative's source of the sedative that the respondent obtained from a friend
or relative for free (SVY19C);

•

type of cancer (HLTH26), including other cancer;

•

born in the United States (QD14)20; and

Respondents who answer question QD14 as "no" are routed to question QD15, which asks them to
specify the country or territory where they were born.

195

•

immediate family members who are currently in the United States military (QD10D),
including other immediate family members.

Except for question QD14, which does not offer an explicit choice of "other" (i.e., other
country or territory is implied by a response of "no"), and the new question QD10D for
immediate family members serving in the United States military, rates for "other" responses to
these items generally were low in the DR data relative to rates for predefined response
categories. Although 4.8 percent of persons aged 12 or older in the DR data were estimated to be
in the "other" race category, the percentage decreased to 0.6 percent when the estimates did not
include Hispanic and Spanish-language respondents in the DR data. Corresponding rates for the
other race category in the 2012 and 2013 comparison data in Table H-1 in Appendix H that also
did not include Hispanic and Spanish-language respondents were 0.4 and 0.6 percent,
respectively.
Although most of the rates in Table 4.14 were flagged for suppression, the low rates for
"other" responses for most items were consistent with findings from the QFT. Except for other
immediate family members who were reported to be serving in the military, these findings
typically support the conclusion that predefined categories performed adequately in the QFT and
DR.
If DR respondents reported in question QD10D that a member of their immediate family
was serving in the United States military, the most commonly reported response in follow-up
question QD10E was "another member of my immediate family." Specifically, 84 respondents in
the DR chose this relationship, which yielded a weighted percentage of 45.2 percent among
persons with a family member who was serving in the military. The "OTHER, Specify" data for
these 84 respondents indicated that 3 respondents specified a step- parent or stepchild (i.e., which
were included in the explanation of immediate family members in question QD10D). Remaining
respondents specified relationships other than those that were listed in question QD10E or else
had missing data (i.e., don't know, refused, or bad data).21 The other family relationships that
were reported in the DR data were as follows (numbers of respondents in parentheses):
•

a cousin or cousins (n = 33);

•

an uncle or an aunt (n = 11);

•

a nephew, niece, or great nephew or niece (n = 9);

•

an in-law (n = 7); and

•

a grandparent or grandchild, including step relationships (n = 6).

In addition, the "OTHER, Specify" answers for 11 respondents were assigned codes for missing
data, and 4 respondents specified that they had no other family members serving in the military.
On the one hand, the DR data for other family members who were reported to be serving
in the military do not suggest that respondents were unclear about the meanings of the precoded
21

One respondent may have specified a relationship that corresponded to a precoded response option from
question QD10E. However, this respondent's "OTHER, Specify" response was assigned a code for "bad data"
because it was unclear whether the response applied to an immediate family member from question QD10E or
someone else who was related to an immediate family member.

196

response categories in question QD13E for immediate family members. Rather, the "OTHER,
Specify" data suggest that respondents were including family members other than those that were
mentioned in the explanatory text for "immediate family."
Two explanations are possible for the high numbers of DR respondents who specified
family relationships other than those that were listed in question QD13E. First, respondents may
not be attending to the question wording about "immediate family" members and instead are
answering question QD13D as "yes" if they know of any relatives who are currently serving in
the United States military. A second possible explanation is that some respondents could have a
broader interpretation of what constitutes an "immediate family" member based on their cultural
backgrounds, especially for respondents for whom extended family relationships are the norm.
However, the numbers of respondents who answered QD13E as "yes" because they were
thinking of any relatives rather than immediate family members and the number who had
interpreted "immediate family" to include additional relationships cannot be readily determined
from the DR data.
Regardless of the underlying reasons, the higher number of respondents who reported
that another member of their family was serving in the military relative to the precoded
categories in question QD13E suggests that it could be useful to consider ways of revising
questions QD13D and QD13E. Revising these questions could reduce the potential burden on
respondents who otherwise might need to type in an "OTHER, Specify" response and also could
reduce the amount of data coding that otherwise could be required.
However, revising the explanation for "immediate family" to mention relationships that
respondents should not include may not be advisable because (a) these additions would add to
the amount and complexity of the cognitive information that respondents would need to process
in order to answer these questions; and (b) for an undetermined number of respondents, further
revisions could communicate cultural biases about the concept of "immediate family." Rather, it
could be useful for SAMHSA to continue allowing respondents to interpret these questions
according to whatever they mean to the respondents and to consider alternatives for capturing
information on other relationships that could be less burdensome for respondents and that could
require less coding of "OTHER, Specify" data. For example, questions about unmet need for
substance treatment services, unmet need for mental health services among adults, and reasons
why adolescents received mental health services in the past 12 months use an "unfolding"
strategy, in which an initial response choice of "other" leads to a follow-up question.
4.6.3

Patterned Responses in the Core Drug Questions for the Comparison Data

Core modules in the 2012 and 2013 comparison data were reviewed for potential
patterned responses according to the procedures documented in the editing and coding section
(Section 10) of the 2010 methodological resource book (Kroutil et al., 2012a). These checks
were implemented as part of the general editing procedures for editing the full 2012 survey data
and the 2013 survey data from quarters 3 and 4, regardless of whether interviews were within or
outside of the 48 States of the continental United States. However, fewer than 10 cases in the
entire 2012 data were classified as nonrespondents even though they met the usable case criteria
because of patterned responses in their core drug data. Similarly, fewer than 10 cases in the
entire 2012 survey were retained as respondents, but with their original responses in one or more

197

core drug modules being replaced with "bad data" codes. For the 2013 survey in quarters 3 and 4,
fewer than five cases met the usable case criteria but were treated as nonrespondents, and fewer
than five cases were retained as respondents but with their original responses in one or more core
drug modules being replaced with "bad data" codes.
4.6.4

Patterned Responses in the Core Drug Questions for the DR Data

The checks for patterned responses that were used for the comparison data also were
implemented for core DR modules that did not change (or underwent minimal change) relative to
the comparison data. Because the content of the new methamphetamine module for the QFT and
DR was similar to the content of other modules in the comparison data, the relevant checks for
the comparison data were run for the methamphetamine data in the QFT.
Based on the experience in the QFT, particular attention was given to identifying the
occurrence of the following patterns in the prescription drug data and examining the results if
these patterns occurred:
•

keying responses of "1" (and only "1") to all screener questions for a given
prescription drug category;

•

keying responses of "2" (and only "2") to all screener questions for a given
prescription drug category; and

•

No new types of patterned responses were identified in the prescription drug data for the DR.
No cases were dropped from the DR data (i.e., treated as nonrespondents) because of
patterned responses. However, three respondents had a pattern of keying "2" (or mostly "2") in
one or more prescription drug modules. No DR respondents had a pattern of keying only
responses of "1" in the screening questions. No cases were recommended to have their
prescription drug answers set to "bad data" because of high numbers of individual prescription
drugs that were misused.

198

Table 4.14 Estimates and Standard Errors for New, Moved, or Revised Items in the 2013 Dress
Rehearsal among Persons Aged 12 or Older
DR Instrument Item

Race (QD05)
White (QD051)
Black or African American (QD052)
American Indian or Alaska Native
(American Indian Includes North
American, Central American, and
South American Indians)
(QD053)
Native Hawaiian (QD054)
Guamanian or Chamorro (QD055)
Samoan (QD056)
Other Pacific Islander (QD057)
Asian (Including: Asian, Indian,
Chinese, Filipino, Japanese,
Korean, and Vietnamese (QD058)
Other (Specify) (QD059)
Now think about the last time you
used [PRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[PRLASTFILL]? (PRY42B)4
I got a prescription for the
[PRLASTFILL] from just one
Doctor
I got prescriptions for the
[PRLASTFILL] from more
than one doctor
I stole the [PRLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [PRLASTFILL] from a
friend or relative for free
I bought the [PRLASTFILL]
from a friend or relative
I took the [PRLASTFILL] from a
friend or relative without
asking
I bought the [PRLASTFILL]
from a drug dealer or other
stranger
I got the [PRLASTFILL] in some
other way
See notes at end of table.

Type of
Change1

Description of
Change
Added response
categories for
Guamanian or
Chamorro and
Samoan.

DR
Standard Unweighted Unweighted
Estimate2,3 Error
Total
Sample Size

76.8
12.7

2.66
2.03

1,396
270

2,084
2,084

1.8
0.0
0.0
0.0*
0.3

0.33
0.01
0.01
0.00*
0.15

144
3
3
0
19

2,084
2,084
2,084
2,084
2,084

5.3
4.8

1.27
0.79

146
190

2,084
2,084

33.7*

7.15*

126

1.8*

1.82*

126

0.0*

0.00*

126

38.8*

6.01*

126

9.2

2.67

126

2.4*

2.08*

126

7.0*

3.79*

126

7.0

3.04

126
(continued)

Added "fill" and
moved from the
noncore prior
substance use
module.

199

Table 4.14 Estimates and Standard Errors for New, Moved, or Revised Items in the 2013 Dress
Rehearsal among Persons Aged 12 or Older (continued)
DR Instrument Item
How did your friend or relative get
the [PRLASTFILL]? (PRY42C)4
He or she got a prescription for
the [PRLASTFILL] from just
one doctor
He or she got prescriptions for
the [PRLASTFILL] from more
than one doctor
He or she stole the
[PRLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[PRLASTFILL] from a friend
or relative for free
He or she bought the
[PRLASTFILL] from a friend
or relative
He or she took the
[PRLASTFILL] from a friend
or relative without asking
He or she bought the
[PRLASTFILL] from a drug
dealer or other stranger
He or she got the [PRLASTFILL]
in some other Way
Now think about the last time you
used [TRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[TRLASTFILL]? (TRY21B)4
I got a prescription for the
[TRLASTFILL] from just one
doctor
I got prescriptions for the
[TRLASTFILL] from more
than one doctor
I stole the [TRLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [TRLASTFILL] from a
friend or relative for free
I bought the [TRLASTFILL]
from a friend or relative
I took the [TRLASTFILL] from
a friend or relative without
asking
I bought the [TRLASTFILL]
from a drug dealer or other
stranger
I got the [TRLASTFILL] in
some other way
See notes at end of table.

Type of
Change1
R

Description of Change
Added "fill" and moved
from the noncore prior
substance use module.

DR
Standard Unweighted Unweighted
Estimate2,3
Error
Total
Sample Size

87.3*

5.81*

0.4*

0.37*

0.0*

0.00*

4.0*

2.75*

0.1*

0.05*

4.3*

4.17*

0.3*

0.29*

3.7*

2.75*

25.4*

9.88*

0.0*

0.00*

0.0*

0.00*

40.4*

10.03*

13.4*

5.33*

5.7*

3.23*

12.7*

6.64*

2.4*

2.44*

Added "fill" and moved
from the noncore prior
substance use module.

200

54
(continued)

Table 4.14 Estimates and Standard Errors for New, Moved, or Revised Items in the 2013 Dress
Rehearsal among Persons Aged 12 or Older (continued)
DR Instrument Item
How did your friend or relative get
the [TRLASTFILL]? (TRY21C)4
He or she got a prescription for
the [TRLASTFILL] from just
one doctor
He or she got prescriptions for
the[TRLASTFILL] from more
than one doctor
He or she stole the
[TRLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[TRLASTFILL] from a friend
or relative for free
He or she bought the
[TRLASTFILL] from a friend
or relative
He or she took the
[TRLASTFILL] from a friend
or relative without asking
He or she bought the
[TRLASTFILL] from a drug
dealer or other stranger
He or she got the
[TRLASTFILL] in some other
way
How did you get the
[STLASTFILL]? (STY26b)4
I got a prescription for the
[STLASTFILL] from just one
doctor
I got prescriptions for the
[STLASTFILL] from more
than one doctor
I stole the [STLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [STLASTFILL] from a
friend or relative for free
I bought the [STLASTFILL]
from a friend or relative
I took the [STLASTFILL] from a
friend or relative without
asking
I bought the [STLASTFILL]
from a drug dealer or other
stranger
I got the [STLASTFILL] in some
other way
See notes at end of table.

Type of
Change1
R

Description of Change
Added "fill" and moved
from the noncore prior
substance use module.

DR
Standard Unweighted Unweighted
Estimate2,3
Error
Total
Sample Size

97.8*

1.47*

0.0*

0.00*

0.0*

0.00*

0.0*

0.00*

2.2*

1.47*

0.0*

0.00*

0.0*

0.00*

0.0*

0.00*

2.6*

1.74*

6.6*

5.85*

0.0*

0.00*

57.7*

10.70*

25.9*

8.26*

3.8*

3.33*

3.4*

2.72*

0.0*

0.00*

Added "fill" and moved
from the noncore prior
substance use module.

201

46
(continued)

Table 4.14 Estimates and Standard Errors for New, Moved, or Revised Items in the 2013 Dress
Rehearsal among Persons Aged 12 or Older (continued)
DR Instrument Item
How did your friend or relative get
the [STLASTFILL]? (STY26c)4
He or she got a prescription for
the [STLASTFILL] from just
one doctor
He or she got prescriptions for
the [STLASTFILL] from more
than one doctor
He or she stole the
[STLASTFILL] from a
doctor’s office, clinic, hospital,
or pharmacy
He or she got the
[STLASTFILL] from another
friend or relative for free
He or she bought the
[STLASTFILL] from another
friend or relative
He or she took the
[STLASTFILL] from another
friend or relative without
asking
He or she bought the
[STLASTFILL] from a drug
dealer or other stranger
He or she got the
[STLASTFILL] in some other
Way
How did you get the
[SVLASTFILL]? (SVY19B)4
I got a prescription for the
[SVLASTFILL] from just one
doctor
I got prescriptions for the
[SVLASTFILL] from more
than one doctor
I stole the [SVLASTFILL] from
a doctor’s office, clinic,
hospital, or pharmacy
I got the [SVLASTFILL] from a
friend or relative for free
I bought the [SVLASTFILL]
from a friend or relative
I took the [SVLASTFILL] from
a friend or relative without
asking
I bought the [SVLASTFILL]
from a drug dealer or other
stranger
I got the [SVLASTFILL] in
some other way
See notes at end of table.

Type of
Change1

Description of Change

DR
Standard Unweighted Unweighted
Estimate2,3
Error
Total
Sample Size

83.0*

8.52*

0.0*

0.00*

0.0*

0.00*

6.7*

6.19*

1.6*

1.57*

0.0*

0.00*

1.0*

0.86*

7.7*

5.37*

40.5*

20.30*

0.0*

0.00*

0.0*

0.00*

46.3*

18.87*

4.9*

4.98*

0.0*

0.00*

0.0*

0.00*

8.3*

8.27*

Added "fill" and moved
from the noncore prior
substance use module.

202

12
(continued)

Type of
Change1
R

Description of Change
Added "fill" and moved
from the noncore prior
substance use module.

DR
Standard Unweighted Unweighted
Estimate2,3
Error
Total
Sample Size

90.5*

8.67*

8.0*

8.32*

0.0*

0.00*

0.0*

0.00*

1.5*

1.63*

0.0*

0.00*

0.0*

0.00*

0.0*

0.00*

0.0*

0.00*

New questions about
health.

Bladder
Blood

3.5

Bone

0.0*

0.00*

Brain

0.00

5.04

5.38

2.13

3.49

0.0

Breast

10.8

*
*

Cervix (Females Only)

10.3

Colon

2.9

Esophagus

3.5

*
*

3.49

Gallbladder

0.0

0.00

Kidney

0.0*

0.00*

0.00

3.50

0.00

7.01

5.09

7.98

Larynx/Windpipe

0.0

Leukemia

3.7

Liver

0.0

Lung

12.2

*
*

Lymphoma

6.4

Melanoma
See notes at end of table.

19.7

203

41
(continued)

Table 4.14 Estimates and Standard Errors for New, Moved, or Revised Items in the 2013 Dress
Rehearsal among Persons Aged 12 or Older (continued)
DR Instrument Item

Standard Unweighted Unweighted
Type of
DR
Total
Sample Size
Change1 Description of Change Estimate2,3 Error

Mouth/Tongue/Lip
Ovary (Females Only)

3.5*

3.49*

3.58

0.00

4.3

Pancreas

0.0

Prostate (Males Only)

6.7

3.69

Rectum

3.5*

3.49*

8.84

7.44

0.98

3.49

0.00

Skin (Not Melanoma)

31.9

Skin (Don't Know Which Kind)

11.1

Soft Tissue (Muscle or Fat)

1.1

Stomach

3.5

Testis (Males Only)

0.0

Throat/Pharynx

0.0

0.00

Thyroid

6.3*

3.64*

6.25

5.13

Uterus (Females Only)
Other
Were you born in the United States?
(QD14)
Is anyone in your immediate family
currently serving in the United States
military? (QD10d)
Which member or members of your
immediate family are currently in the
United States military? (QD10e)4

11.5
5.3
M

Administered in ACASI
instead of CAPI.
New question about
military service within
immediate family
New question about
military service within
immediate family

84.2

1.83

1,693

2,080

7.6

0.99

184

2,061

My spouse

9.6*

3.71*

179

Unmarried partner

1.1

0.68

179

My mother

1.9

0.96

179

My father

4.9

1.84

179

N
N

My son or sons

16.2
*

My daughter or daughters
My brother or brothers

0.0

0.00

24.4

5.01

179

3.17*

179

5.46

179

6.8*

My sister or sisters
Another member of my immediate
family

4.20

45.2

204

5. Assessments of the Redesigned Protocol
(Research Question 1)
5.1

Overview of DR Protocol Assessment

This chapter presents the results of four efforts to assess the partially redesigned protocol
used for the 2013 Dress Rehearsal (DR). The overall purpose of these assessments was to ensure
that the revised questionnaire and protocol used for the 2013 DR will facilitate continued high
quality and efficiency in the National Survey on Drug Use and Health (NSDUH) data collection
when the partial redesign is implemented in 2015. Section 5.2 presents a description of and
results from a survey that was administered to field interviewers (FIs) at the training session.
Section 5.3 presents findings from a survey on the equipment used by FIs in the DR. Section 5.4
provides selected data compiled from FI debriefing items completed for DR cases. Section 5.5
provides key findings from the debriefing phone calls that were held with select FIs to discuss
their experiences using the redesigned NSDUH interview protocol and tablet computer for
screening. Section 5.6 presents the complete results of field observations of DR FIs. Each section
concludes with recommendations for the 2015 NSDUH.

5.2

Description and Results from DR FI Training Survey

5.2.1

Purpose and Development of the DR FI Training Survey

To gather reactions and suggestions from FIs about the DR FI training program for use in
developing the 2015 NSDUH training programs, a brief electronic feedback survey was
administered at the conclusion of the in-person FI training. FIs were asked about their
satisfaction with the DR training program and materials and about their comfort level with
properly performing various DR tasks. The complete set of DR FI training survey questions and
a summary of the FI responses are provided in Appendix C.
5.2.2

Procedures for Conducting the DR FI Training Survey

The DR FI training survey was administered at the conclusion of the DR training sessions
on August 25 and 27, 2013. FIs completed the survey on their DR laptops during class and were
to transmit the results to RTI at that time.
An introduction screen explained the purpose of the survey and the confidentiality of
individual responses. Results were sent back to RTI via the NSDUH transmission system. Of the
135 DR FIs who successfully completed the DR FI training, 133 FIs completed and transmitted
the survey.
5.2.3

Summary and Discussion of Results from the DR FI Training Survey

A summary of FI feedback on the DR FI training program and materials is in Table 5.1,
which provides the combined counts of FIs who strongly agreed or agreed to each of the
statements. This table also includes counts for the same statements included in the feedback
survey completed by FIs following the Questionnaire Field Test (QFT) FI training in
205

August 2012. Table 5.2 shows how often FIs planned to use the DR handbook while also
providing the counts from the same question asked during the QFT.
•

When responding to the statements, "I feel ready to properly conduct the DR
screening using the tablet" and "I feel ready to properly conduct the DR interview
using the DR laptop," 98 percent of FIs strongly agreed or agreed to each of these
statements.

•

Overall, FIs were highly satisfied with the DR training program, with most FIs
indicating they strongly agreed or agreed to the statements about training.
As indicated in Table 5.1, percentages were above 90 percent for 11 of the
12 statements.

•

The response to the statement, "The overall pace of the DR FI Training Session was
just right for me," was slightly less positive than the others, with 80 percent of the FIs
responding that they strongly agreed or agreed with the statement.

•

As shown in Table 5.1, the percentages from the DR FI training survey are similar to
the percentages from the feedback gathered in the survey conducted at the end of the
QFT training program in August 2012.

Table 5.1 Field Interviewer (FI) Feedback on the FI Training Program

Statement
"Reading the DR/QFT FI Handbook
helped prepare me for training."
"Completing the DR/QFT iLearning
course helped prepare me for training."
"The overall pace of the DR/QFT FI
Training Session was just right for me."
"The paired screening and interview exercises
completed during training were helpful."
"I feel ready to properly conduct DR/QFT screenings
using the tablet."
"I feel ready to properly conduct DR/QFT interviews
using the DR laptop."
"I feel ready to use the email program on the tablet."
"I am comfortable with the process to transmit
wirelessly with the tablet (independent of the laptop)."
"I am comfortable with the process to transmit
wirelessly with the DR laptop."
"Overall, I am satisfied with the training provided on
the DR laptop."
"Overall, the training program has prepared me to
properly complete all DR/QFT tasks."
"I enjoyed attending the DR/QFT FI Training
Session."

DR FI Training Survey
(August 2013)
Agree or Strongly Agree
(n = 133)1
%

QFT FI Survey 1
(August 2012)
Agree or Strongly Agree
(n = 160)
%

123

153

121

151

106

131

122

N/A

130

154

130
121

98
91

152
N/A

95
N/A

125

N/A

125

N/A

128

N/A

127

156

125

149

DR = Dress Rehearsal; FI = field interviewer; N/A = not applicable; QFT = Questionnaire Field Test.
1
See Section 2.3.2 in Chapter 2 for detailed information on the number of FIs hired for the DR.

206

Table 5.2 Field Interviewer (FI) Expectations on Referencing the FI Handbook before Data
Collection
FI Training Survey: "How often do you
think you will reference the DR/QFT FI
Handbook?"
"Each day with DR/QFT work"
"Two to three times a week"
"Rarely, when unusual situations arise"
"Never"

DR FI Training Survey
(August 2013)
(n = 133)1
%
26
62
44
1

20
47
33
1

QFT FI Survey1
(August 2012)
(n = 160)
%
30
65
65
0

19
41
41
0

DR = Dress Rehearsal; FI = field interviewer; QFT = Questionnaire Field Test.
1

See Section 2.3.2 in Chapter 2 for detailed information on the number of FIs hired for the DR.

5.2.4

FI Comments on the DR FI Training Program

The DR FI training survey included one open-ended question that allowed FIs to
comment on any aspect of the training program. Out of 133 DR FIs completing the survey,
69 FIs made comments. Comments were loosely grouped based on their overall content into the
following areas: (a) comments about the training program and/or training staff, (b) comments
about the hotel or other training logistics (such as travel), (c) comments about the DR equipment,
and (d) other comments. Because a number of the comments included thoughts about more than
one topic, exact counts are not presented to prevent counting some comments more than once.
The complete set of raw comments is included in Appendix C. As is appropriate for a training
feedback survey, about three fourths of the comments submitted related to the training program
and/or training staff.
•

Training Program. Overall, these comments stated how thorough and organized the
training program was, describing it as informative and excellent. Several comments
expressed concerns about the pace of the training, with one comment saying it was
too slow, while several others indicated that the pace was too quick. The paired
exercises were mentioned in several comments and were described as excessive, or
irritating, or difficult due to the size of the classroom.

•

Training Staff. Most of the comments about the training staff described trainers as
professional, helpful, prepared, and patient. A few comments stated that the trainers
were "detached" or rude.

•

Hotel Facility. Several comments were complimentary of the hotel facilities, while
others expressed concerns about the classroom size and the fare served at breakfast.

•

Travel. Several comments expressed concerns about the timing of scheduled travel
arrangements and about the use of the metro rail between the airport and the hotel.

•

DR Equipment. Several comments highlighted the positive aspects of the new
equipment, such as its light weight and advanced technology. Two comments
expressed concerns about using wireless technology, while one FI was concerned
about having to repeatedly log on to the tablet while working.

207

5.2.5

Considerations for 2015 NSDUH Training

In planning for the 2015 NSDUH, further consideration will be given to several items
noted in the feedback and comments received from the DR FI training survey, including the
overall pace of training, paired screening and interview exercises, classroom size, and travel
logistics.
Additionally, feedback received from the FI equipment survey (see Section 5.3),
FI debriefing items (see Section 5.4), FI debriefing calls (see Section 5.5), DR field observations
(see Section 5.6), and NSDUH staff serving as trainers at the DR training sessions will be
considered during the development of an FI training plan for the 2015 NSDUH. In particular, DR
trainers noted issues with some FIs being unfamiliar with the computer equipment (particularly
the tablet), lacking knowledge of proper NSDUH procedures, and the unwillingness to follow
protocol. The significance of these concerns warrants further discussion in order to mitigate these
issues at future training sessions. Possible suggestions on ways to address these items within the
training program design include (a) shipping the tablet to FIs prior to the training session to
allow them the opportunity to practice independently on the equipment, (b) extending the
training agenda and content to include extra training time on the equipment and NSDUH
protocols and procedures, and (c) conducting a certification exercise at the conclusion of training
to formally evaluate FIs on adherence to screening and interviewing procedures.

5.3

Description and Results for the FI Equipment Survey
5.3.1

Overview

A survey was developed and administered to FIs during the DR to gather FI feedback
about the new equipment and program modifications that were implemented and tested during
the DR, including the following:
•

A new device for conducting NSDUH's computer-assisted interviewing (CAI)
interviews was selected for field-based evaluation—the Samsung Series 9 laptop.
This laptop was chosen for its small size, light weight, and bright, crisp screen
display, which made it highly portable and easy to see.

•

Smaller carrying cases were purchased to fit the new laptop.

•

The QFT CAI instrument was modified to fit the 13.3-inch display of the Samsung
laptop, and all of the DR FIs used the new laptop and its CAI program to conduct DR
interviews.

In addition to the new laptop, a few changes were made to the DR tablet that was used for
screening, including the following:
•

An email program was configured on the tablet to enable FIs to send and receive
messages to and from their field supervisors (FSs) and other NSDUH staff members.

•

An optional keyboard (the "hacker's" keypad) was added for use during screening to
supplement the default Samsung keyboard because it more closely resembled the
iPAQ keyboard by displaying numbers across the top line of the keyboard.
208

•

Independent wireless transmission functionality was incorporated into the DR
screening program so that FIs could transmit screening data or receive new cases
from the field on any public Wi-Fi network.

The equipment survey included a combination of customized questions used in previous
equipment evaluations and a number of questions adapted from the System Usability Scale
(SUS),22 which is an industry standard scale for measuring the usability of hardware and
software that was first developed and published by engineers at the Digital Equipment
Corporation (DEC) in 1986. The complete set of questions and responses to the equipment
survey are included in Appendix D.
The equipment survey was sent to all of the DR FIs on their DR laptops via the
transmission system on October 7, 2013. The FIs were given 1 week to complete the survey and
transmit results back to RTI. Of the 135 FIs who completed and passed the DR training session,
125 FIs completed the survey. Ten FIs did not complete the survey for one of the following
reasons:
•

they did not work any DR cases,

•

they were unavailable because of travel assignments for their quarter 4 main study
work, or

•

they took a leave of absence (LOA) from their NSDUH work at the time the survey
was released.

5.3.2

Feedback from the FI Equipment Survey

A summary of feedback provided by the FIs on the new laptop and tablet features is
provided in the following paragraphs. Results for all FIs, including both English only and
bilinguals, are presented first in the tables and figures, followed by results for bilingual FIs only.
As shown in Tables 5.3 and 5.4 the FIs were highly satisfied with the laptop as an
interviewing device and would prefer to use it for their NSDUH fieldwork. The vast majority
reported that the laptop was easy to use and that they did not require technical assistance to
become acclimated to it. In fact, they learned to use it quickly, felt confident while using it, and
were satisfied with the training they were given. Also, the FIs were satisfied with the laptop's
physical features and felt that the display was large enough and bright enough for presenting the
NSDUH interview. The FIs were highly satisfied with the weight of the laptop, and many of the
FIs commented in the open-ended question that they enjoyed the light weight of the laptop. Most
of the FIs reported that the touchpad and keyboard were easy to use, although a small number of
the FIs commented in the open-ended question that they felt that the function keys were too
small and were hard to read and that the labels were not close enough to the keys. The FIs were
less satisfied with the carrying case provided for the laptop, and many of the FIs commented that
the diagonal design of the strap was problematic because it interfered with the zippers and made
it difficult to access the case's center pockets. This finding also arose during the FI debriefing
22

For details, see the following online reference: Brooke, J. (n.d.). SUS - A quick and dirty usability scale.
Retrieved from http://hell.meiert.org/core/pdf/sus.pdf.

209

calls (see Section 5.5). A couple of the FIs commented that there were too many zippered
pockets, which made it inconvenient and hard to access stored materials.
The equipment survey also asked the FIs about their experiences using the tablet email
program. As shown in Tables 5.5 and 5.6, the majority of the FIs (79 percent among all FIs and
84 percent among bilingual FIs) used the tablet email program at least a few times a month to
communicate with their supervisors or other NSDUH staff.
Among those FIs who used the email program, the majority were highly satisfied with it,
as shown in Tables 5.7 and 5.8. Most of the FIs reported that the email program was simple and
straightforward as well as easy to use. They were able to send and receive email without needing
technical assistance, learned to use the email program quickly, and felt confident using the
program. Most of the FIs were satisfied with the training provided on the email program and
would like to use the email program on a regular basis. A number of the FIs commented that the
email program was easy to use and that they liked being able to reply to their FS messages,
which "made communicating easy and quick." A couple of the FIs mentioned that they did not
use email as much because of the small DR caseload or because they received only a couple of
FS messages. Other FIs mentioned that they preferred to use the personal email program on their
home computer or smartphone. A couple of the FIs mentioned that they would like a little more
training or refresher training on the email program.
Table 5.3 All Field Interviewers' Satisfaction with the Samsung Laptop (n = 125)

1.
a.
b.
c.
d.
e.
f.
g.
h.
i.
j.
k.
l.

FI Satisfaction with the Samsung Laptop
The laptop was easy to use.
I was able to use the laptop without needing technical assistance.
I learned to use the laptop quickly.
I felt confident using the laptop.
The display size of the laptop screen was large enough for
presenting the NSDUH interview.
The laptop screen was clear and bright enough for displaying the
NSDUH interview.
I was satisfied with the weight of the laptop.
I found the layout of the laptop keyboard easy to use.
The laptop's touchpad was easy to use.
I was satisfied with the carrying case provided for the laptop.
I was satisfied with the training provided on the laptop.
I would prefer to use this laptop for my field work.

DR Equipment Survey
Agree or Strongly Agree
n
%
119
95
113
90
120
96
120
96
119

120
121
115
116
91
119
120

96
97
92
93
73
95
96

DR = Dress Rehearsal; FI = field interviewer; LOA = leave of absence; n = number; NSDUH = National Survey on
Drug Use and Health.
NOTE: A total of 10 FIs were unable to complete the survey because they were on an LOA, they were traveling for
the main study, or they did not work any DR cases.

210

Table 5.4 Bilingual Field Interviewers' Satisfaction with the Samsung Laptop (n = 39)

1.
a.
b.
c.
d.
e.
f.
g.
h.
i.
j.
k.
l.

Bilingual FI Satisfaction with the Samsung Laptop
The laptop was easy to use.
I was able to use the laptop without needing technical assistance.
I learned to use the laptop quickly.
I felt confident using the laptop.
The display size of the laptop screen was large enough for
presenting the NSDUH interview.
The laptop screen was clear and bright enough for displaying the
NSDUH interview.
I was satisfied with the weight of the laptop.
I found the layout of the laptop keyboard easy to use.
The laptop's touchpad was easy to use.
I was satisfied with the carrying case provided for the laptop.
I was satisfied with the training provided on the laptop.
I would prefer to use this laptop for my field work.

DR Equipment Survey
Agree or Strongly Agree
n
%
35
90
34
87
37
95
37
95
37

37
37
34
36
26
37
37

95
95
87
92
67
95
95

DR = Dress Rehearsal; FI = field interviewer; LOA = leave of absence; n = number; NSDUH = National Survey on
Drug Use and Health.

Table 5.5 All Field Interviewers' Frequency of Tablet Email Use (n = 125)
3. During the Dress Rehearsal, how often did you use the email
program on the tablet to communicate with your FS or other
NSDUH staff members?
Each day
Two to three times a week
A few times a month
Never

DR FI Equipment Survey
n
6
39
54
26

%
5
31
43
21

DR = Dress Rehearsal; FI = field interviewer; FS = field supervisor; n = number; NSDUH = National Survey on
Drug Use and Health.

Table 5.6 Bilingual Field Interviewers' Frequency of Tablet Email Use (n = 39)
3. During the Dress Rehearsal, how often did you use the email
program on the tablet to communicate with your FS or other
NSDUH staff members?
Each day
Two to three times a week
A few times a month
Never

DR FI Equipment Survey
n
4
9
20
6

%
10
23
51
15

DR = Dress Rehearsal; FI = field interviewer; FS = field supervisor; n = number; NSDUH = National Survey on
Drug Use and Health.

211

Table 5.7 All Field Interviewers' Satisfaction with the Tablet Email Program (n = 99)

4. FI Satisfaction with the Tablet Email Program
a. I found the email program on the tablet to be simple and
straightforward.
b. The email program was easy to use.
c. I was able to use the email program without needing technical
assistance.
d. I learned to use the email program quickly.
e. I felt confident using the email program on the tablet.
f. I was satisfied with the training provided on the email program.
g. I would like to use the email program on a regular basis.

DR Equipment Survey
Agree or Strongly Agree
n
%
90
92

91
93

91
90
91
88
80

92
91
92
89
81

DR = Dress Rehearsal; FI = field interviewer; n = number.
NOTE: FIs who reported in question 3 of the survey that he or she had never used the tablet email program were not
asked this set of questions about using the tablet email program.

Table 5.8 Bilingual Field Interviewers' Satisfaction with the Tablet Email Program (n = 33)

4. Bilingual FI Satisfaction with the Tablet Email Program
a. I found the email program on the tablet to be simple and
straightforward.
b. The email program was easy to use.
c. I was able to use the email program without needing technical
assistance.
d. I learned to use the email program quickly.
e. I felt confident using the email program on the tablet.
f. I was satisfied with the training provided on the email program.
g. I would like to use the email program on a regular basis.

DR Equipment Survey
Agree or Strongly Agree
n
%
33
33

100
100

31
33
32
30
29

94
100
97
91
88

The DR screening program was configured with two different keyboards—the default
Samsung keypad that is built into the Android operating system and the hacker's keypad that was
installed separately. FIs accessed the hacker's keypad from any screen that required use of the
keypad by tapping the keypad icon displayed at the bottom of the screen and changing the setting
from Samsung keypad to the hacker's keypad. The hacker's keypad was then displayed on that
screen and all subsequent screens in which a keypad was needed to enter data. The same process
was used to switch back to the Samsung keypad.
All DR FIs were shown how to access and switch between both keypads during their inperson training session. For consistency, FIs were instructed to use the Samsung keypad at the
training session and were given the option to select the keypad that best suited their needs once

212

they entered the field. When asked about their preference for the tablet keypad, the vast majority
of the FIs indicated that they preferred the Samsung keypad, as shown in Figures 5.1 and 5.2.
Figure 5.1 All Field Interviewers' Tablet Keypad Preference

Figure 5.2 Bilingual Field Interviewers' Tablet Keypad Preference

Finally, the equipment survey asked about FI satisfaction with the independent wireless
transmission functionality that was incorporated into the DR screening program. The vast
majority of the FIs were highly satisfied with the tablet transmission functionality, as shown in
Figures 5.3 and 5.4. Also, many of the FIs commented that they enjoyed being able to transmit
wirelessly from the tablet, which they found more convenient when they needed to pick up new
cases in the field or at home when they did not complete any interviews.

213

Figure 5.3 All Field Interviewers' Satisfaction with the Tablet's Wireless Transmission

Figure 5.4 Bilingual Field Interviewers' Satisfaction with the Tablet's Wireless Transmission

In addition, the data gathered from the data processing systems revealed that the FIs
made frequent use of the independent tablet transmission functionality. During the DR, 2,330
transmissions (51 percent) were completed using the independent tablet transmission function,
while 2,206 (49 percent) transmissions were completed using the traditional method in which the
tablet is tethered to the laptop.
5.3.3

Summary of Results

Overall, the results of the equipment survey indicate that the FIs were highly satisfied
with the new equipment and programs deployed in the DR although the laptop carrying case was
214

a common area of concern. Most of the FIs indicated that the light weight of the laptop was a
significant advantage, and they also felt that it was easy to learn and easy to use. New tablet
features that included a two-way email program and independent tablet transmission
functionality were also popular among the FIs and were used on a regular basis. The FIs
overwhelmingly preferred the default Samsung keypad over the hacker's keypad.
FIs had lower levels of satisfaction with the laptop carrying case and expressed some
concern with the touchpad and keyboard. The FIs were less than satisfied with the carrying case
provided for the laptop. The FIs consistently reported in their open-ended comments that the
design of the laptop carrying case was problematic, which was primarily due to the diagonal
design of the shoulder strap that connected to the front and back of the bag. This design made it
difficult to open and close zippered pockets and made the bag hard to carry on the shoulder.
Also, the FIs reported that the multiple zippered pockets made it inconvenient and hard to access
items stored in the bag. A couple of the FIs also did not feel there was enough space in the bag
for all of the materials and equipment. In the open-ended question about the laptop, a number of
the FIs commented that the function keys on the laptop were very small and hard to read and that
the standard labels did not match up properly with the keys.
5.3.4

Next Steps for Future Hardware and Software Deployment

Given these equipment survey results, the following steps are being considered with
respect to full-scale equipment deployment and future program implementations:
1. A new method should be explored for labeling the function keys on the laptop to
make them more easily identifiable.
2. A laptop carrying case with a different design should be explored for future use. A
design more suitable for FI use would include a single zippered pocket with multiple
interior pockets and a strap that connects to the sides of the case.
3. The tablet email program should be incorporated into future deployment of the tablet
because it was a popular feature among the FIs and provided an easy way for them to
communicate with their FSs.
4. The independent wireless transmission function built into the screening program was
useful for both field-based and at-home transmissions of case data. This functionality
should be incorporated into future versions of the NSDUH screening program.
5. The default Samsung keypad is suitable for FI use, and an alternate keypad option is
not needed in future versions of the screening program.

5.4

DR FI Debriefing Item Results
5.4.1

Purpose of the Debriefing Items

The FI debriefing items used in the DR can provide valuable insight on the redesigned
protocol in 2015. These FI debriefing questions were administered at the end of each interview
as part of the revised interview protocol. The debriefing items unique to the DR included
questions on respondent reactions to the new laptop computer that was used to administer the

215

survey. Debriefing items that were introduced in the QFT and repeated for the DR included
questions on when the FIs gave respondents the question and answer (Q&A) brochure, whether
respondents expressed any difficulties with or reactions to certain features of the revised protocol
(such as comments about the laptop), the use of on-screen calendars in the audio computerassisted self-interviewing (ACASI) section of the questionnaire, and proxy respondent use of
ACASI. Finally, DR interviewers were asked debriefing questions from the main survey on the
interview location and the privacy of the interview. Although this reporting depends largely on
unprompted information being supplied by interview respondents, these items may provide
valuable information that can be unobtrusively used to identify potential problems with the
redesigned protocol's new features. Specifications for the DR field interviewer debriefing items
are provided in Appendix E.
5.4.2

Results from the Debriefing Items

Tables 5.9, 5.10, and 5.11 present information on FI reports of respondents' comments
about the laptop. First, Table 5.9 shows whether respondents made any comments about the
laptop among three categories: "Yes, Positive Comments," "Yes, Negative Comments," or "No."
Most of the respondents (84.5 percent) did not make any comments about the laptop. As shown
in Table 5.10, among those respondents who provided positive comments, "The laptop was
lightweight" (79.9 percent) was by far the most frequently endorsed positive comment about the
laptop, followed by "The keyboard was easy to use" (10.2 percent), then "The screen was large
or clear or easy to read" (5.5 percent). A review of the "Other" text responses showed that most
of these comments were that the laptop was thin, "nice," "high-tech," modern, or otherwise
attractive.
FIs reported very few negative comments from respondents about the laptop. Only
1.7 percent of respondents provided a negative comment (Table 5.9). As shown in Table 5.11,
among those who provided a negative comment, the most frequently mentioned was that
"The keyboard was hard to use” (47.2 percent), problems specifically with the function keys,
which respondents are instructed to use to back up to a previous screen, view the calendar, or
enter "don't know" or "refused" as a response (33.3 percent), then by the comment that "The
laptop was too hot" (2.8 percent). Among those choosing "Other" as a response, 12 respondents
indicated problems with the function keys, which respondents are instructed to use to back up to
a previous screen, view the calendar, or enter "don't know" or "refused" as a response.
Table 5.9 Comments about the Laptop
DRDBF4 – Did the respondent make any comments about the laptop? Please
include respondent comments about the physical features of the laptop or about
respondent's use of the laptop. Check all that apply.
Yes, Positive Comments
Yes, Negative Comments
No
TOTAL

n
293
36
1,771
2,097

NOTE: Percentages do not sum to 100 percent because of rounding.
NOTE: Interviewer may have indicated that the respondent provided both positive and negative comments.

216

%
14.0
1.7
84.5
100.0

Table 5.10 Positive Comments about the Laptop
DRDBF4a – Which one or more of the following best describes the positive
comments the respondent made about the laptop? Check all that apply.
The screen was large or clear or easy to read
The laptop was lightweight
The keyboard was easy to use
Thin laptop
Nice in general
High-tech, modern, sleek, attractive, etc.
Want to buy/own this
Other
TOTAL

%
5.5
79.9
10.2
6.8
5.8
4.8
4.1
8.9
100.0

n
16
234
30
20
17
14
12
26
293

NOTE: Percentages are based on 293 "Yes, Positive Comments" answers to DRDBF4; more than one response could be chosen.

Table 5.11 Negative Comments about the Laptop
DRDBF4b – Please describe the negative comments the respondent made
about the laptop. Check all that apply.
There were problems reading the screen
The laptop was too hot
The laptop was too heavy
The layout of questions was problematic
The keyboard was hard to use
There were problems with the volume or sound
Problems specifically with function keys
Other
TOTAL

%
13.9
2.8
0.0
0.0
47.2
0.0
33.3
27.8
100.0

n
5
1
0
0
17
0
12
10
36

NOTE: Percentages are based on the 36 "Yes, Negative Comments" answers to DRDBF4; more than one response could be
chosen.

Tables 5.12 and 5.13 provide FI comments about the laptop by respondent age and
education, respectively. In general, very few differences were observed by either age or
education in the proportions of respondents providing positive or negative comments. When
comments were provided about the laptop, they were largely positive across most age groups and
education levels. One exception was that, among persons aged 65 or older, about as many
respondents provided negative comments as did those who provided positive comments.
Table 5.12 Comments about the Laptop, by Interview Respondent Age
DRDBF4 – Did the respondent
make any comments about the
laptop? Please include respondent
comments about the physical
features of the laptop or about
respondent's use of the laptop.
Check all that apply.
Yes, Positive Comments
Yes, Negative Comments
No

12 to 17
(n = 505)

n
62
2
440

%
12.3
0.4
87.1

Interview Respondent Age
18 to 25
26 to 49
50 to 64
(n = 529)
(n = 734)
(n = 187)

n
72
4
452

217

%
13.6
0.8
85.4

n
101
7
627

%
13.8
1.0
85.4

n
37
6
144

%
19.8
3.2
77.0

65 or Older
(n = 132)

n
19
16
99

%
14.4
12.1
75.0

Table 5.13 Respondent Comments on the Laptop, by Interview Respondent Education
Interview Respondent Education
High
< High
School
Some
College
School
Graduate
College
Graduate
(n = 291)
(n = 440)
(n = 530)
(n = 321)
n
%
n
%
n
%
n
%
34 11.7
61 13.9
85
16.0
49
15.3
4
1.4
16
3.6
7
1.3
6
1.9
253 86.9 366 83.2 437 82.5 266
82.9

DRDBF4 – Did the respondent make any comments
about the laptop? Please include respondent comments
about the physical features of the laptop or about
respondent's use of the laptop. Check all that apply.
Yes, Positive Comments
Yes, Negative Comments
No

Table 5.14 shows the timing of providing the Q&A brochure to the respondents. Nearly
80 percent of the FIs (79.8 percent) reported that they provided the Q&A brochure at the end of
the interview, and 19.5 percent of the FIs reported that they provided the Q&A brochure before
the interview. The percentage of FIs who provided the Q&A brochure during the interview was
minimal (0.6 percent). These percentages are relatively similar to those reported for the QFT, in
which 73 percent of the FIs reported providing the brochure at the end of the interview and
25 percent before the interview.
Table 5.14 Timing of Providing Q&A Brochure
DRDBF1 – When did you give the respondent (or parent/guardian of youth
respondent) the Q&A [question and answer] brochure?
Before the interview
During the interview
At the end of the interview
TOTAL

406
13
1,666
2,085

19.5
0.6
79.8
99.9

Table 5.15 shows that FIs reported in 13.2 percent of completed interviews that the
respondent commented that the interview was too long, which is similar to the 12.8 percent
reported in the QFT. Table 5.16 shows that a larger percentage of respondents aged 50 to 64
(21.4 percent) and those aged 65 or older (36.4 percent) made comments about the interview
being too long compared with respondents in other age groups (which ranged from 6.6 to
15.3 percent). This is similar to the finding from the QFT that older respondents were more
likely to comment on the interview being too long. Table 5.17 shows that more than twice as
many respondents with less than a high school education commented that the interview was too
long compared with respondents with higher levels of education. This finding is also similar to
the results observed in the QFT.
Table 5.15 Respondent Comments on the Interview Being Too Long
DRDBF8 – Did the respondent make any comments about the interview being
too long?
Yes
No
TOTAL

218

n
275
1,810
2,085

%
13.2
86.7
99.9

Table 5.16 Respondent Comments on the Interview Being Too Long, by Interview Respondent
Age
DRDBF8 – Did the respondent
make any comments about the
interview being too long?
Yes
No

12 to 17
(n = 505)
n
%
40
7.9
464 92.1

Interview Respondent Age
18 to 25
26 to 49
50 to 64
(n = 529)
(n = 734)
(n = 187)
n
%
n
%
n
%
35
493

6.6
93.4

112
622

15.3
84.7

40
147

65 or Older
(n = 132)
n
%

21.4
78.6

48
84

36.4
63.6

Table 5.17 Respondent Comments on the Interview Being Too Long, by Interview Respondent
Education

DRDBF8 – Did the respondent
make any comments about the
interview being too long?
Yes
No

< High School
(n = 291)
n
%
76
26.1
215
73.9

Interview Respondent Education
High School
Some College
Graduate
(n = 440)
(n = 530)
n
%
n
%
59
13.4
63
11.9
381
86.6
466
88.1

College
Graduate
(n = 321)
n
%
37
11.5
284
88.5

NOTE: Interview Respondent Education is shown only for persons aged 18 or older.

Tables 5.18 and 5.19 show whether respondents had questions or comments on the
on-screen calendars in the ACASI section of the questionnaire and if so the types of comments
that they had. Almost 99 percent (2,057 out 2,085) of the respondents did not have any questions
or comments about the on-screen calendar. The most frequently mentioned comment reported by
FIs on the on-screen calendar was that respondents asked how to close the calendar (41.4 percent
of those who reported any comment).
Table 5.18 Any Interview Respondent Questions or Comments on On-Screen Calendars
DRDBF9 – Did the respondent have any questions or comments about the onscreen calendars in the ACASI [audio computer-assisted self-interviewing] section
of the questionnaire? If the respondent asked how to access the calendar at any
time during the ACASI portion of the interview, select "YES."
Yes
No
TOTAL

n
28
2,057
2,085

%
1.3
98.6
99.9

Table 5.19 Types of Interview Respondent Questions or Comments on On-Screen Calendars
DRDBF9a – What comments did the respondent [R] make about the on-screen
calendars?

The R asked how to access the calendar.
The R asked how to close the calendar.
The R did not see the reference dates on the calendar.
The calendar helped the R answer the question.
The calendar covered the questions or the images on the screen.
Other

24.1

41.4

3.4

6.9

8
27.6
NOTE: Percentages are based on the 28 "Yes" answers to DRDBF9; more than one response could be chosen.

219

Additionally, Table 5.20 shows that 10.3 percent of the FIs reported that the respondent
had trouble understanding any other questions asked during the interview. This is similar to the
9.5 percent of interviews in the QFT in which the FI reported that the respondent had trouble
with any other question.
Table 5.20 Interview Respondents' Troubles with Other Questions
DRDBF10 – Did the respondent have trouble understanding any other questions
asked during the interview?
Yes
No
TOTAL

n
215
1,870
2,085

%
10.3
89.6
99.9

FIs used a text field to provide additional information on other questions that respondents
reported difficulties in understanding. One FI reported that a respondent was not sure how to
enter a response of 8 months on item HLTH30 (age first diagnosed with a heart condition or
heart disease). A value of 1 is the lowest possible value that can be entered for this item. None of
the similar items in the health module asking for the age at which the respondent was first
diagnosed with a health condition (including cancer) provide instructions on how to enter an age
of less than 1 year old. Although this was only reported in one interview, it does raise the
question of whether these items should accommodate medical conditions that can be diagnosed
younger than 1 year old. The conditions asked about in the health module vary as to whether they
can be diagnosed in infancy. For example, cirrhosis of the liver occurs because of the
accumulation of repeated health insults over time, so it is unlikely to be diagnosed in infancy.
On the other hand, infectious diseases such as HIV and hepatitis can arise during infancy.
Table 5.21 shows that 28.8 percent of FIs (600 out of 2,085) reported the use of a proxy
for the income and health insurance questions. Table 5.22 shows that FIs more frequently
reported the use of a proxy for younger respondents than older respondents except among those
in the 65 or older age group. The QFT sample also showed a similar proportion of respondents
using proxies. Similar to the findings from the QFT and the main study, the DR results in Table
5.22 show that about 85 percent of respondents aged 12 to 17 in the DR used a proxy to report on
the income and health insurance questions. Proxy usage was much lower for other age groups.
Table 5.21 Proxy Used for Income and Health Insurance Questions
DRDBF11 – Was a proxy used for the income and health insurance questions?
Yes
No
TOTAL

n
600
1,485
2,085

%
28.8
71.2
100.0

Table 5.22 Proxy Used for Income and Health Insurance Questions, by Interview Respondent Age

DRDBF11 – Was a proxy used for the
income and health insurance questions?
Yes
No

12 to 17
(n = 505)
n
%
430 85.3
74 14.7

220

Respondent Age
18 to 25
26 to 49
50 to 64
(n = 529)
(n = 734)
(n =187)
n
%
n
%
n
%
82 15.5
61 8.3
13 7.0
446 84.5 673 91.7 174 93.0

65 or Older
(n = 132)
n
%
14 10.6
118 89.4

Among the 600 respondents in which the FI reported the use of a proxy, only
5 respondents (0.8 percent) had concerns about revealing their answers to the proxy respondent
(Table 5.23), and only 12 respondents (2.0 percent) had questions or comments about the proxy
interview (Table 5.24). These results are similar to those from the QFT sample, in which
0 percent of FIs reported any respondent concerns about revealing answers to proxy respondents.
Table 5.23 Interview Respondent Concerns about Revealing Answers to Proxy Respondent
DRDBF13 – Did the respondent have any questions or concerns about
his/her answers being revealed to the proxy?
Yes
No
TOTAL
NOTE: Percentages are based on the 600 "Yes" answers to DRDBF11.

n
5
595
600

%
0.8
99.2
100.0

Table 5.24 Interview Respondent Questions or Comments about Proxy Interview
DRDBF14 – Did the respondent have any other questions or comments
about the proxy interview?
Yes
No
TOTAL
NOTE: Percentages are based on the 600 "Yes" answers to DRDBF11.

n
12
588
600

%
2.0
98.0
100.0

In contrast, 7.0 percent of the FIs indicated that there were problems with the proxy use
of ACASI to answer the income and health insurance questions (Table 5.25). Although this is a
low proportion, this appears to be higher than the 3.5 percent of interviews in the QFT in which
an FI reported a problem with a proxy's use of ACASI. Some of this difference may be
attributable to the inclusion of Spanish-language interviews or the oversample of Hispanics in
the DR sample. Among the 600 interviews in which a proxy was reported by an FI, 55 were
completed in Spanish. In 6 out of the 55 interviews conducted in Spanish (10.9 percent),
a problem was reported with the proxy's use of ACASI. Among the interviews conducted in
English, such problems were reported in 6.6 percent of such interviews. Similarly, among
Hispanic respondents, a problem with the proxy's use of ACASI was reported in 8.7 percent of
the interviews as compared with 5.8 percent among non-Hispanics. Although the inclusion of the
Spanish-language interview and the oversample of Hispanics in the DR sample may have
contributed somewhat to the difference of reporting problems with proxy use between the DR
and QFT samples, it was also the case that the English-language interviews and interviews
completed with non-Hispanic respondents alone reported higher percentages of problems with a
proxy's use of ACASI in the DR than in the QFT sample.
Among the types of problems with a proxy's use of ACASI to answer the income and
health insurance questions (Table 5.26), technical difficulty that the proxy did not know how to
enter his or her answer to the question (35.7 percent) had the highest percentage, followed by
"The proxy did not know the answers to the questions" (14.3 percent) and "The proxy did not
know why he/she was asked to answer these questions" (11.9 percent).

221

Table 5.25 Problems with Proxy's Use of ACASI to Answer Income and Health Insurance
Questions
DRDBF12 – Were there any problems with the proxy's use of ACASI
[audio computer-assisted self-interviewing] to answer the income and
health insurance questions?
Yes
No
TOTAL

%
7.0
93.0
100.0

n
42
558
600

NOTE: Percentages are based on the 600 "Yes" answers to DRDBF11.

Table 5.26 Types of Problems with Proxy's Use of ACASI to Answer Income and Health Insurance
Questions
DRDBF12a – Which of the following describes the problems with the
proxy's use of ACASI [audio computer-assisted self-interviewing] in
answering the income and health insurance questions? Check all that apply.
The proxy did not know the answers to the questions.
The proxy did not know how to enter his/her answers to the questions.
The proxy refused to answer some questions.
The proxy did not know why he/she was asked to answer these questions.
Language issues
Interpretation issues
Other

%
14.3
35.7
2.4
11.9
9.5
9.5
33.3

n
6
15
1
5
4
4
14

NOTE: For responses of "OTHER," follow-up information was not collected.
NOTE: Percentages are based on the 42 "Yes" answers to DRDBF12; more than one response could be chosen.

Tables 5.27 and 5.28 present information on the location where the interviews were
conducted in the DR and the comparison samples. In all three samples, 98 percent of the
interviews were conducted in a respondent's home (Table 5.27). The specific locations where the
interviews took place other than in a respondent's home are shown in Table 5.28. In general, the
distribution of locations outside a respondent's home in the 2012 and 2013 quarters 3 and 4
comparison samples were very similar, while those for the DR sample were different. For
example, about 11 percent of the interviews in the 2012 and 2013 comparison samples were
conducted in the home of a friend or relative of the respondent, while almost 29 percent of those
in the DR sample were conducted in such locations. These differences should be interpreted with
caution because only 38 DR interviews total were conducted outside the respondent's home.
Table 5.27 Interviews Conducted at Respondent's Home for the 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal
DRDBF2 – Did you conduct this interview at the
respondent's home, either inside or outside?
Yes
No

2012
Comparison
n
%
65,445
98.4
1,078
1.6

NOTE: Some percentages do not sum to 100 percent because of rounding.

222

2013
Comparison
n
%
31,634
98.4
520
1.6

2013 Dress
Rehearsal
n
%
2,047
98.1
38
1.8

Table 5.28 Interview Location Not at Respondent's Home for the 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal

DRDBF3 – Where did you conduct this interview?

At the respondent's workplace
At the home of the respondent's relative or friend
In some type of conference room in a residence
hall, school or apartment complex
At a library
In some type of common area, such as a lobby,
hallway, stairwell, or laundry room
Some other place

2012
Comparison
n
%

2013
Comparison
n
%

2013 Dress
Rehearsal
n
%

226

21.0

18.1

15.8

118

10.9

10.8

28.9

239

22.2

126

24.2

5.3

204

18.9

18.8

18.4

140

13.0

14.2

5.3

151
14.0
72
13.8
10
26.3
NOTE: Percentages are based on the 1,078, 520, and 38 "No" answers to FIDBF01 (2012), FIDBF01 (2013), and
DRDBF2, respectively.
NOTE: Some percentages do not sum to 100 percent because of rounding.

Using results from the 2012 and 2013 quarters 3 and 4 comparison samples and the DR
sample, Tables 5.29 and 5.30 show the FI evaluation of interview privacy at the respondent's
home and further details on the presence of others during the interview. Table 5.29 shows that
75 percent of the interviews in the DR were rated as completely private, which was lower than
that reported for both the 2012 main study (82.0 percent) and the 2013 quarters 3 and 4
comparison sample (82.6 percent). This lower frequency of having complete privacy appears to
be related to the higher level of minor distractions reported in the DR (16.6 percent) compared
with the 2012 and 2013 comparison samples (13.4 and 13.1 percent, respectively). That is, when
the percentages for the two categories of "completely private" and "minor distractions" are
combined, 91.6 percent of the interviews in the DR were considered as such, which was similar
to that of the QFT, although it was still a bit low compared with that of 2012 and 2013
comparison samples (95.4 and 95.7 percent, respectively).
Table 5.29 Field Interviewer Evaluation of Interview Privacy in Respondent's Home for the 2012
Comparison, 2013 Comparison, and 2013 Dress Rehearsal
DRDBF5 – Please indicate how private the interview
was. Do not count yourself or a project observer as
another person in the room.

Completely private–no one was in the room or
could overhear any part of the interview
Minor distractions–person(s) in the room or
listening less than 1/3 of the time
Person(s) in the room or listening about 1/3 of the
time
Serious interruptions of privacy more than half the
time
Constant presence other person(s)

2012
Comparison
n
%

2013
Comparison
n
%

2013 Dress
Rehearsal
n
%

54,591

82.0

26,565

82.6

1,565

75.0

8,911

13.4

4,207

13.1

347

16.6

1,150

1.7

546

1.7

2.7

283

0.4

121

0.4

0.6

1,588
2.4
NOTE: Some percentages do not sum to 100 percent because of rounding.

715

2.2

104

5.0

223

Table 5.30 Field Interviewer Reports of Others Present during Interview for the 2012
Comparison, 2013 Comparison, and 2013 Dress Rehearsal
DRDBF6 – Not including yourself or project
observers, other people present or listening to the
interview were:

Parent(s)
Spouse
Live-in partner/ boyfriend/ girlfriend
Other adult relative(s)
Other adult(s)
Child(ren) under 15
Other

2012
Comparison
n
%

2013
Comparison
n
%

2013 Dress
Rehearsal
n
%

5,298

44.4

2,493

44.6

170

32.7

1,689

14.2

784

14.0

19.0

725

6.1

325

5.8

7.1

1,470

12.3

636

11.4

14.0

1,148

9.6

491

8.8

7.1

3,893

32.6

1,933

34.6

211

40.6

393
3.3
182
3.3
22
4.2
NOTE: Percentages are based on the 11,932, 5,589, and 520 answers that indicate lack of privacy to FIDBF07
(2012), FIBF07 (2013), and DRDBF5, respectively.

Table 5.30 shows that the DR provided somewhat different patterns in the types of other
persons present during the interview. In both the 2012 and 2013 comparison samples, "parent(s)"
were the most often reported "other people" present during the interview, followed by "child(ren)
under 15" and "spouse." In the DR, FIs reported the presence of parents in 32.7 percent of the
interviews, which was lower than the 44.4 percent in the 2012 comparison sample and the
44.6 percent in the 2013 comparison sample. Similar differences between the 2011 and 2012
main study comparison samples and the QFT were not observed. The oversample of Hispanics
for the DR sample does not appear to account for this difference. For the DR sample, among the
520 interviews in which an FI reported the presence of other persons, 211 were conducted with
Hispanic respondents and 309 among non-Hispanic respondents. Among the non-Hispanic
respondents, a parent was reported as being present in 87 interviews (28.2 percent), while among
the Hispanic respondents, a parent was reported in 83 interviews (39.3 percent).
5.4.3

Summary and Recommendations

5.4.3.1

Summary

The DR debriefing questions provided insight into respondents' reactions to the DR
interview and protocol. Overall, respondents reacted favorably to the new computer. The few
negative comments that were recorded largely concerned difficulty using the function keys.
Respondents' comments about the timing of the interview differed by age. Consistent with the
timing results in Chapter 4, older respondents frequently made comments about the interview
taking a long time.
Data from the debriefing items did not indicate any pervasive problems with respondents'
comprehension of questionnaire items. FIs reported that a proxy was used in 27.8 percent of the
interviews. Among interviews with youths aged 12 to 17, about 85 percent used a proxy. Some
challenges were reported with the process of introducing the proxy respondents to the computer.
These challenges affected a larger proportion of respondents who took the interview in Spanish.
Finally, fewer FIs classified the DR interview as "completely private" compared with
both the 2012 and 2013 comparison interviews and the QFT interviews.
224

5.4.3.2

Recommendations

Analysis of the debriefing items reveals a few recommendations that should be
considered when planning for the 2015 NSDUH:
•

The labeling or assignments of the function keys should be edited to optimize
usability.

•

The debriefing items themselves should be edited for the 2015 NSDUH so as to
assess items of interest in the 2015 survey.

•

An instruction should be considered that indicates to respondents what response
option to choose to report a diagnosis of a health condition (including cancer) at an
age of less than 1 year old.

Plans for the 2015 NSDUH do not call for proxies to complete the health insurance and
income questions using a self-administered mode. This decision was made largely to address
increases in item missing data and shifts in estimates in these modules that occurred in the field
tests. As a result of these changes, any concerns about the process of providing a computer
tutorial to proxy respondents do not need to be addressed.

5.5

Description and Results for the DR FI Debriefing Calls
5.5.1

Purpose of the Debriefing Calls

The purpose of the DR FI debriefing calls was to obtain direct feedback from FIs on their
experiences collecting data using the redesigned NSDUH questionnaire on the new laptop and
screenings completed on the touch screen tablet (i.e., the Samsung Galaxy Tab 7.0") in both
English and Spanish. Information on the complete set of instrumentation, protocol materials, and
equipment changes for the DR is presented in Section 2.4.1. The goal of the debriefing calls was
to gather feedback from FIs (including bilingual FIs) on topics including but not limited to the
following:
•

significant questions or concerns raised by members of sample households about the
redesigned contact materials;

•

challenges encountered using the tablet to conduct household screenings;

•

challenges encountered using the new laptop computer to conduct interviews;

•

challenges encountered in administering the redesigned questionnaire or protocol;

•

significant questions or concerns that respondents raised about aspects of the
redesigned questionnaire or protocol (specifically, the four prescription drug
modules) and the length of the interview (i.e., its overall burden); and

•

significant questions or concerns raised by Spanish-speaking members of sample
households about any of the Spanish-language materials or questionnaires developed
for the DR.

225

The results of the DR FI debriefing calls will be used to inform potential changes to the
preparation, protocol, and procedures for the 2015 NSDUH.
5.5.2

Debriefing Call Procedures

Debriefing calls were held in lieu of in-person focus groups, which had been used to
gather feedback on the 2012 QFT. A total of five debriefing calls were held in order to eliminate
travel costs, expedite the completion of these information-gathering sessions, and allow greater
flexibility in assigning FIs to each call. Table 5.31 provides additional details on the
characteristics of FIs assigned to each debriefing call.
Table 5.31 Debriefing Call Field Interviewer Characteristics, by Call
Call No.
1
2
3
4a
5a

No. of FIs
5
6
6
6
5

No. of QFT FIs
2
3
4
2
1

No. of
Bilingual FIs
0
0
1
6
5

States Represented
FL, MA, ME, NJ, PA
DC, OH, SD, TN, TX, VT
CA, KY, NY,b MI, OK, SC
CA, FL, IL, NM, NV, NY
CA, CA,b IL, TX, TX

QFT = Questionnaire Field test.
a
Bilingual field interviewers (FIs) only.
b
Travelling field interviewer (TFI).

For each debriefing call, 7 FIs were initially assigned (35 in all), with the expectation that
5 or 6 FIs would participate in each call. These 35 FIs were selected from the 133 FIs assigned to
the DR based on their classroom assignment at the DR training, with 1 FI per room chosen for
each call. To the extent possible, other characteristics were balanced across the calls, including
QFT participation, bilingual status, experience level, and the region of the country where they
lived.
A total of 28 FIs participated in the debriefing calls, with each call having at least 5 FIs.
Calls were scheduled for 2 hours in duration, and all five calls lasted at least 90 minutes. Also,
each call included a moderator, an assistant moderator, and a note taker, along with several
observers, including staff from the Substance Abuse and Mental Health Services Administration
(SAMHSA). In addition to using a note taker, each call was audio recorded.
Per SAMHSA request, each of the debriefing calls was moderated by RTI staff from the
training program and field materials team and the field operations team because of the familiarity
of those staff with NSDUH procedures and protocols. Two of the five calls were for bilingual FIs
only, and those calls were also joined by one of the RTI language specialists involved in the
bilingual DR FI training and Spanish-language translation of the redesigned questionnaire.
The moderators referred to and followed a written guide throughout each call. This guide
was developed using the QFT moderator guide as a template and incorporated SAMHSA input;
it also included specific changes made for the DR, such as the new laptop and the inclusion of
Spanish-language materials and instruments. Also, the DR moderator guide incorporated
feedback from the moderators of the QFT in-person debriefings to identify any lessons learned
and problematic sections and questions. The DR moderator's guide is included in Appendix F.

226

A moderator began each debriefing call with an introduction that lasted about 5 minutes
and was intended to set up the discussion rules and familiarize the participants with the group.
The remainder of each call had specific section topics and discussion time limits:
•

Discussion about the redesigned contact materials was allotted 20 minutes and
covered the FIs' views regarding how the DR respondents reacted to the lead letter
and Q&A brochure, as well as the FIs' thoughts about the new laptop bag and
portfolio.

•

The next 30 minutes were devoted to a discussion about the FIs' use of the tablet to
administer household screenings. Topics included the tablet's features, training on the
tablet, respondent reactions to the naming of the "US Department of Health and
Human Services" as the sponsor of the study instead of the "US Public Health
Service," and the use of the new email program installed on the tablets.

•

A total of 25 minutes were allotted to topics surrounding the interview's questionnaire
administration, including respondent comments about the electronic reference date
calendar, whether respondents asked questions about specific modules, and the
experience of the proxy respondents.

•

For the next 10 minutes, the discussion focused on using the DR laptop to administer
the interview, including respondent reaction to the laptop and its overall performance.

•

The next section called for 10 minutes of discussion about the four new prescription
drug modules. The moderator asked questions about the length of administration
time, the electronic pill cards, and the questions designed to capture misuse.

•

A 15-minute section then allowed FIs to share any general comments or concerns
about the redesigned questionnaire, including interview length and burden.

•

A concluding open-ended section gave FIs a final opportunity to make additional
comments on any aspect of the DR.

During the two debriefing calls conducted with only bilingual FIs, the FIs were instructed
to focus their responses on Spanish-language cases to the extent possible in all of the sections
noted above. Also, the discussion section on administering the redesigned questionnaire included
four follow-up questions that asked the bilingual FIs to speak specifically about the DR content
and materials in Spanish.
5.5.3

Feedback from Debriefing Calls, by Topic

5.5.3.1

Reactions to the Redesigned Contact Materials

FIs on all five debriefing calls noted that respondents reacted positively to the changes to
the lead letter and the Q&A brochure. Several FIs commented that many respondents stated that
they had already read the letter prior to the interview and were anticipating the FI's arrival as a
result, and that many more respondents seemed to recall this lead letter than those who received
one as part of main study's data collection. FIs on multiple calls hypothesized that the lead letter's
added color was what led to the improved respondent interest and recall. Similarly, FIs reported
greater respondent interest in the redesigned Q&A brochure than in its main study counterpart,

227

with several stating that the full-color, "professional-looking" design caused more respondents to
open that brochure than the one used during the main study.
Reaction from FIs on the new black leather portfolio for the DR was mixed, with FIs split
fairly evenly among preferring this new portfolio, preferring the main study portfolio, and
preferring using no portfolio at all. Those FIs who did note a preference for using the new black
leather portfolio said that they had no real use for the legal pad of paper inside the portfolio and
would rather see more pockets, both interior and exterior, with some that have zippers. A few FIs
reported that, to appear less intimidating, they prefer to carry as little as possible with them to the
door when approaching respondents for the first time, while others said that they prefer to
carrying a small clipboard with only essential items (such as the lead letter and the study
description).
Many of the FIs offered high praise for the new laptop bag, especially when compared
with their current main study laptop bag, as many appreciated having the extra pockets to store
all of the various items needed to conduct an interview. FIs commented that they felt having the
extra pockets allowed them to set up and break down equipment more efficiently at the
beginning and end of the interview. However, nearly all of the FIs also commented that they did
not like the setup of the large shoulder strap on the laptop bag, specifically, the way it was
attached to the bag itself. FIs felt that this design made it more difficult to unzip several of the
pockets, and several FIs also wished that the strap had more padding for ease of use. A few FIs
also expressed a dislike of the center pocket of the laptop bag that unzipped all the way open
because they found it difficult to prevent items from falling out when that pocket was used.
Given the additional storage in the new laptop bag, many FIs indicated that they would
still prefer to have access to a portfolio in order to have a few items more readily available while
at the door and to keep paper materials from becoming wrinkled.
Also, although most of the FIs stated that they kept their DR FI handbook with them
while working in the field, few reported ever needing to review its contents while conducting
interviews. Those who did reference the handbook reported that they used it for assistance with
transmission (i.e., transmitting data to RTI).
5.5.3.2

Reactions to Administering Household Screenings and Using the Tablet

All of the FIs confirmed that the DR FI training program was effective in preparing them
to use the tablet in the field to conduct screenings. Also, all of them reported feeling comfortable
using the tablet not later than the end of their first day of working DR cases, if not before.
When discussing the tablet, many FIs noted right away that they did not like having to
enter a password each time the tablet was awakened from "sleep" mode. Also, many of the FIs
strongly preferred having more cases available for viewing at one time on the select case
screen—similar to the iPAQ screen, which uses a smaller font size.
Most of the FIs reported using the standard Samsung keyboard (or keypad) throughout
their work on the DR; they found this default keyboard fairly easy to use and did not experience
many problems, if any, with entering record of call (ROC) comments into the tablet. However,
several of the FIs did note their wish that an apostrophe key was easier to access on the default
228

keyboard for use in typing notes. The few FIs who did report switching to use a second keyboard
(i.e., the "hacker's" keypad) stated that their preference was due to the presence of a line of
number keys along the top of that keyboard, rather than having to switch views on the default
keyboard in order to access numbers.
FI preference for using a finger or a stylus when tapping on the tablet was mixed, with
many FIs reporting using both methods in combination depending on the situation. Those FIs
who said that they used only the stylus often did so as a continuation of their use of a stylus with
the iPAQ on the main study, and several of the FIs who used only a finger to tap on the tablet
reported an existing comfort level with that method from previous use with similar electronic
devices.
Also, regarding the canvas case protecting each tablet, FIs had positive comments about
its design, specifically, that it does not feature any Velcro®, unlike its main study counterpart.
The FIs strongly preferred the snaps because using them produces less noise and they do not
become caught on clothing, unlike Velcro®.
The majority of FIs stated that there was little difference in respondent reaction to the use
of "US Department of Health and Human Services" as opposed to "US Public Health Service."
Some of the FIs did say that more respondents recognized the DHHS logo on the various
materials and commented on that. However, several of the FIs also noted an increase in
antigovernment sentiment when contacting respondents during the Federal Government
shutdown, which began on October 1, 2013, in the midst of DR data collection.
With regard to the possible addition of a 30-second video clip on the tablet to play for
respondents, nearly all of the FIs agreed that it could be a useful tool in establishing legitimacy
and providing more information about the study. However, all of the FIs also agreed that any
such video should be added only as an optional tool for their use, in that many respondents are
busy and would not be interested in watching a 30-second video. A few of the FIs also added that
for the more difficult respondents, face-to-face interaction between an FI and a respondent will
always be the best tool for refusal conversion.
All of the FIs admitted to having positive experiences, with little or no problems, using
the email program on the tablet to communicate with their FSs during the DR. They found that
the program was convenient, efficient, and easy to use, and several of the FIs commented that
they appreciated being able to transmit wirelessly from the field to pick up new cases or to send
an email message to their FSs. If given this access on the main study, all of the FIs reported that
it would be a benefit to their work and increase efficiency. Although these DR FIs reported using
the email program a couple of times each week during data collection, several of the FIs noted
that they would use it a great deal more if given a larger assignment with more cases. However, a
few of the FIs commented that it was difficult for them to type out long messages using the tablet
email program because of the keyboard's design.
Overall, FIs reported having few problems with the tablet that required assistance from
technical support staff. Most often, those FIs who did call for help needed assistance with
transmitting data from a hotel or other minor issues. However, several bilingual FIs reported
needing to call technical support staff for an issue with pending cases that disappeared from their

229

tablets without explanation, a problem that occurred with both Spanish- and English-language
cases worked by those bilingual FIs. Technical support staff were not immediately able to
determine why this problem occurred, which resulted in the loss of some screening data for a few
of the FIs.
The only other aspect of the tablet that gave multiple FIs problems during the DR was the
tablet calendar, which some of the FIs found difficult to use. Several of the FIs also reported that
they did not receive notifications for appointments that they felt they had set up correctly.
5.5.3.3

Reactions to Administering the Redesigned Questionnaire and Protocol

Moderators began this section of the debriefing calls by asking about respondent
reactions to and comments about the electronic reference date calendar. Although nearly all of
the FIs reported receiving no comments from respondents about this electronic calendar, several
of the FIs noted that they felt that the respondents used the calendar more during the DR
interview than during the main study interview because it was more readily available on the
screen during the DR interview. Those same FIs said that, in their experience, most main study
respondents disregard the paper calendar. A few of the FIs noted that the electronic calendar also
made it more efficient to conduct interviews, especially outdoors, where it might be difficult for
an FI to properly fill out the paper calendar.
Much like their experience on the main study, most of the DR FIs commented that they
received few or no queries or comments from respondents on specific DR interview questions or
modules. Several of the FIs reported conducting interviews in both English and Spanish that
lasted more than 2 hours, and a few respondents did comment on some fatigue associated with
what they perceived as answering the same question multiple times. FIs said that respondents
thought that the interview was "checking up on their answers" by asking the same question more
than once. A few of the FIs also noted having at least one respondent who asked why the sexual
orientation questions were included in the questionnaire, with some elderly respondents stating
that they felt that those questions were not relevant to them.
With regard to the transition into the second ACASI questions conducted by a proxy, FIs
thought that the transition worked very well except in households with only one resident. FIs
noted that it made for a somewhat awkward transition in those instances, rather than simply
allowing the single-resident respondent to continue answering questions via ACASI.
On the two debriefing calls conducted with just bilingual FIs, moderators asked a few
additional questions to follow up on any issues specific to the Spanish-language interview and its
translation. For the most part, the bilingual FIs reported that the Spanish-language DR interview
worked very well and caused little confusion for respondents. Some of the FIs felt that the
Spanish-language interview was longer than its main study counterpart. Also, a few of the FIs
reported some issues with Spanish-speaking respondents understanding the meaning of
"heterosexual" and "probation" within the DR interview. Also, the bilingual FIs stated that the
INCENT01 language in the Spanish-language interview had not been completely updated to
match the English-language text.

230

5.5.3.4

Reactions to Administering the DR Interview and Using the Laptop

Across all five debriefing calls, FIs confirmed that the DR FI training program was
effective in teaching them how to use the new laptop computer and that from the beginning of
their fieldwork they felt comfortable using the new device. In fact, all of the FIs had
overwhelmingly positive feelings toward the new laptop, with several commenting that they
wished they could keep it for continued use with their main study assignment.
FIs noted that they were impressed by the light weight of the new laptop and its smaller
size when compared with its main study counterpart. The FIs also noted that respondents talked
about how light in weight the laptop was.
Only a few of the FIs reported having minor issues with the laptop that required
assistance from technical support staff, with most of the problems related to entering incorrect
passwords multiple times. However, the majority of the FIs reported that some of their elderly
respondents had problems seeing the function keys along the top row of the laptop keyboard
and/or reading the small lettering on those keys. A few FIs reported emphasizing for those
respondents the larger function key labels just below the laptop screen and showing the
respondents that the function keys they were seeking were directly below the descriptions on the
labels.
5.5.3.5

Reactions to the Redesigned Prescription Drug Modules

Among all of the interviews conducted by these DR FIs, very few comments were about
the new prescription drug modules, with the majority of the FIs noting that respondents had no
questions or made no statements about those modules. As noted previously, a few of the FIs
mentioned the respondents' comments about how repeating a question was a method to check up
on a previous response.
Also, several of the FIs commented that they felt that having the pill images available on
the screen rather than in the showcard booklet resulted in respondents paying more attention to
those images. The FIs' main study experience is that respondents often do not ask to see the pill
cards.
5.5.3.6

Overall Reactions to the Redesigned Questionnaire

When asked to provide their overall reaction to the DR questionnaire changes, including
the recall and length of time, nearly all of the FIs reported that their experience was very similar
to their work on the main study. A few of the FIs did report that a handful of respondents said
that the DR interview took a long time, but that was not a comment unique to the DR, as the FIs
often hear a similar response during main study data collection.
Also, two FIs reported that elderly respondents made a negative comment about the
questions on depression. Those respondents noted that they felt "judged" by the questions and
that answering those questions made them feel depressed.

231

5.5.3.7

Other General Feedback

At the conclusion of each debriefing call, the moderators asked the FIs for any final
comment or question about any topics related to the DR. Included below are the comments or
questions resulting from that final discussion, aside from any general positive comments about
the DR (which were made by the majority of FIs):
•

An FI re-emphasized removing the additional password screen from the tablet.

•

An FI felt that including the income and insurance questions in ACASI allowed the
respondents to give more honest answers.

•

An FI liked completing the debriefing questions on the tablet rather than in the
respondents' home on the laptop.

•

An FI indicated that the new equipment was great and made the interview go more
smoothly, but that it created more idle time for the FIs.

•

An FI said that the headphone cord is too long.

5.5.4

Summary and Recommendations

5.5.4.1

Summary

Despite the mix of characteristics among the 28 FIs who participated in these five DR
debriefing calls, there was a great deal of commonality in the feedback provided by these FIs
across each call. Included below is a list of key points mentioned by FIs across all five of these
DR debriefing calls:
•

FIs noted that respondents seemed to have greater recall of the DR lead letter than its
main study counterpart.

•

There was no dominant opinion among FIs on the black leather portfolio. Some of the
FIs liked it better than the main study portfolio, some liked it less than the main study
version, and some preferred no portfolio at all.

•

Overall, FIs liked having more pockets for storage in the laptop bag, but nearly all of
the FIs strongly disliked the design of the shoulder strap.

•

FIs did not like having to enter a password each time the tablet went into "sleep
mode."

•

Most FIs used the standard Samsung keyboard to make entries in the tablet.

•

FIs were split fairly evenly in their use of the stylus or a finger in entering data into
the tablet.

•

FIs liked that the canvas case for the tablet did not have Velcro® on it.

•

FIs noted very little respondent reaction to the use of "US Department of Health and
Human Services" rather than "US Public Health Service."

•

The possibility of a 30-second video clip on the tablet to show respondents was
deemed a good idea by nearly all of the FIs as long as it is only an optional tool.
232

•

The email program on the tablet was very well received by all of the FIs, and few FIs
had problems with it.

•

Many of the FIs reported having problems using the tablet calendar efficiently.

•

FIs reported greater respondent use of the electronic reference date calendar as
compared with its main study counterpart.

•

There were very few questions or comments from respondents about the content of
the redesigned questionnaire.

•

FIs were very impressed with the light weight and efficient performance of the DR
laptop.

•

Some elderly respondents had problems seeing the function keys clearly.

5.5.4.2

Recommendations

As a result of the feedback provided by the 28 FIs who participated in these DR
debriefing calls, several key recommendations have been selected for consideration for possible
implementation on the NSDUH main study:

5.6

•

Implement the use of the DR lead letter and Q&A brochure on the main study.

•

Investigate the use of a laptop bag different from the bag selected for the DR that still
has lots of pockets for organization, but offers a more functional shoulder strap
design. This recommendation is concordant with that from Section 5.2.

•

Consider removing the extra password from the tablet screen.

•

Allow FIs to have full-time access to an email system that allows them to both send
and receive email messages for improved communication with FSs.

•

Purchase a lightweight laptop similar to the version used on the DR.

•

Revise the Spanish-language questionnaire to address the understanding of the sexual
orientation question and to correct text in INCENT01.

•

Attempt to purchase a laptop with larger function keys or determine a way to make
the function key labels easier to read.

Field Observations of DR FIs

In conjunction with DR data collection, field observations of DR FIs were conducted by
RTI staff (including language methodologists, training and field materials' team members, and
instrument assessment and development team members) as well as other RTI staff and SAMHSA
staff members. Groups of three to four FIs were chosen for field observations in each of seven
metropolitan areas: Miami, Florida; New York City, New York; Los Angeles, California;
San Francisco, California; Dallas, Texas; Houston, Texas; and Chicago, Illinois. RTI staff also
observed FIs locally in North Carolina. SAMHSA staff observed an additional four FIs in
Louisiana, Maryland, Virginia, and the District of Columbia. These observations were of
interviews completed in both English and Spanish. Observations of interviews conducted in
Spanish were completed only by bilingual observers. FIs who were observed committing any
procedural errors were retrained by their FSs using a standardized retraining plan created
233

specifically for the DR field observations. FIs who were observed committing a serious breach of
protocol or committing four or more unrelated errors received disciplinary action in addition to
retraining by their FSs.
All DR field observations were completed between September 3 and September 15, 2013.
During this time period, RTI staff observed a total of 64 screenings with 23 FIs and 27
interviews with 18 FIs in English, and 8 screenings with 5 FIs and 7 interviews with 5 FIs in
Spanish. SAMHSA staff observed 9 screenings with 3 FIs and 6 interviews with 3 FIs. This
section summarizes the DR field observation procedures and errors observed, combining the
results reported by RTI and SAMHSA observers. It also includes comments from observers and
FIs about the materials, procedures, and equipment used for the DR data collection, as well as
potential changes to these items based on field observation results and feedback.
5.6.1

Procedures for Completing Field Observations of DR FIs

Field observation trips were planned during August 2013 to give observers as much time
as possible to prepare for their trip. To keep travel costs to a minimum, FIs were chosen for DR
field observations based on location and proximity to other DR FIs. FIs were observed in
11 States, centered on metropolitan areas, and in the District of Columbia.
Observers used a DR field observation screening checklist and a DR field observation
interviewing checklist to document their observations. A DR field observer reference sheet and a
DR field observer task list were used to help maintain consistency in planning observation
assignments and interacting with FIs and respondents. (The DR field observation materials are
provided in Appendix G.) Observers were asked to ensure that a DR field observation FI
instruction sheet was sent to each FI prior to the observer's arrival in the field. The DR housing
unit (HU) and group quarters unit (GQU) scripts and CAI specifications for the front-end and
back-end computer-assisted personal interviewing (CAPI) questions were provided to observers
for their use during the observations. These materials were developed specifically for the DR
data collection effort based on similar materials used for the QFT and the main study field
observation processes. Field observers were trained to remain neutral during the observations
and were to intercede only if the respondent's rights were being violated or if the equipment was
in jeopardy of being damaged. In all other situations, observers were instructed not to interfere.
Observers were asked to transfer information from paper DR field observation screening
checklists and DR field observation interviewing checklists to spreadsheets designed specifically
for the DR field observations. The DR field observation manager then used the captured
information to process the results of the field observations, which included issuing any
appropriate disciplinary action, creating a retraining plan to address any observed errors, and
sending any comments about the performance of the instrument, equipment, or materials to the
appropriate RTI staff member.
The same standardized retraining process was used for the DR field observations as is
used for the main study field observations. After the DR field observation manager reviewed
each observation form, an FI retraining plan was prepared for each FI who had errors reported on
his or her observation. Using a standardized template, this plan noted the errors that the FI made,
the type of retraining required, and the dates by which the retraining must be completed. The FS

234

used this form to provide standardized feedback and retraining to the FI on each error observed
and issued disciplinary action as directed by the DR field observation manager.
FIs who committed a serious breach of protocol and FIs who were observed committing
four or more unrelated errors were issued disciplinary actions. A serious breach of protocol is
defined as one that could potentially violate a respondent's rights and/or significantly
compromise the accuracy of the data collected. Disciplinary actions issued for the DR field
observations included four verbal warnings. These FIs were added to the NSDUH main study
field observation list for observation in quarter 1 of 2014.
5.6.2

Summary of Results from DR Field Observations

Of 2,025 potential screening errors in the DR field observations (81 completed screenings
multiplied by 25 possible errors on the DR field observation screening checklist), field observers
noted 20 errors, or 0.99 percent of the possible screening errors. Of 560 potential interviewing
errors in the DR field observations (40 completed interviews multiplied by 14 possible errors on
the DR field observation interviewing checklist), field observers noted 16 errors, or 2.86 percent
of the possible interview errors.
The majority of FIs displayed positive behaviors when conducting screenings (see
Table 5.32). Of the 25 items listed on the DR field observation screening checklist, only 1 item
was observed being conducted incorrectly more than 5 percent of the time:
•

Not including name, RTI International, U.S. Department of Health and Human
Services, and the lead letter in introduction.

The relatively high error rate for FIs failing to include all four pieces of required information
(name, RTI International, DHHS, and the lead letter) in the study introduction is concerning. The
eight errors recorded in this category were committed by seven different FIs. Four errors were
caused by the FI using "Research Triangle Institute" instead of "RTI International." Three errors
were due to the FI not including the "Department of Health and Human Services" in the study
introduction. The final error in this category occurred when an FI used "U.S. Public Health
Service" instead of the "Department of Health and Human Services."
To further investigate the potential effects of changes between the DR and main study,
items were added to the DR field observation screening checklist to reflect changes to the
screening procedures, project information, and use of specific DR materials. There were
12 errors recorded for these items (see Table 5.33), including cases where, as noted previously,
the FI did not include all four required pieces of information correctly in the study introduction
and the FI did not correctly answer a respondent's questions with DR-specific information.
Of the eight errors recorded for not including all four required pieces of information in the study
introduction, only five errors (by five different FIs) can be directly attributed to the change in
terminology for the DR. The three cases in which an FI did not include the "Department of
Health and Human Services" could have occurred in a main study observation as well. Based on
observations of these errors, no changes to the equipment or field materials are anticipated.
However, further emphasis will be placed on reading the study introduction screen and providing
the correct information when speaking to a respondent in FI training and project materials for the
2015 NSDUH.
235

Table 5.32 DR Screening Error Rates

Screening Error
Not displaying ID badge prominently when knocking on door
Not being on tablet "Study Introduction" screen when reaching door
Not including name, RTI International, U.S. Department of Health and Human
Services, and lead letter in introduction
If respondent did not recall lead letter, FI did not offer one to respondent
Not confirming that screening respondent was an adult resident of SDU
Not verifying that he/she was at the correct address
Not handing DR study description to respondent
Not reading tablet "informed consent" screen to respondent
Not checking for missed DUs by reading the correct tablet screen verbatim
Not asking all roster questions verbatim
Not recording race based on respondent's answer, but on FI observation instead
Not obtaining all screening information directly from the screening respondent
(by observation or a proxy)
Not confirming accuracy and completeness of roster data with screening
respondent
When no household members were selected for an interview (code 22, 25, 26, or
30), not reading verification instructions verbatim
When one or two household members were selected for an interview (code 31 or
32), not presenting project and interview information accurately
When one or two household members were selected for an interview (code 31 or
32), not demonstrating flexibility in scheduling interview(s)
When one or two household members were selected for an interview (code 31 or
32), not leaving appropriate information about future interview(s)
When one or two household members were selected for an interview (code 31 or
32), not making attempts to begin interview right away
Not providing respondent with correct DR materials
Answer questions correctly and thoroughly, referencing correct DR details (e.g.,
RTI International, DHHS, did not mention DR or field test, sample size, pay or
payment, etc.)
Committing other procedural violation not noted on checklist (sum of individual
errors)
Did not have segment maps readily available for reference while in the field
If this was FI's first time to the DU, did not use segment maps to locate SDU
If this was FI's first time to the DU, did not use segment materials to check for
missed DUs
If a missed DU is found, did not use segment materials to make sure the missed
DU was not already listed
TOTAL

Total Cases = 81
Error
Errors
Rate, %
Observed
1.23
1
0.00
0
9.88
0.00
0.00
0.00
0.00
0.00
1.23
2.47
0.00

8
0
0
0
0
0
1
2
0

0.00

0.00
0.00

0
0

4.94

0.00

0.99

DR = Dress Rehearsal; DU = dwelling unit; FI = field interviewer; SDU = sample dwelling unit.
NOTE: The error rate equals the percentage of observed cases where the error was observed. Bolded items are considered
serious breaches of protocol.

236

Table 5.33 Dress Rehearsal-Specific Errors: Screening

Screening Error
Not including name, RTI International, U.S. Department of Health and Human
Services, and lead letter in introduction
Not providing respondent with correct DR materials
Answer questions correctly and thoroughly, referencing correct DR details (e.g.,
RTI International, U.S. Department of Health and Human Services, did not
mention DR or field test, sample size, or payment)
TOTAL

Total Cases = 81
Error
Errors
Rate, %
Observed
9.88
0.00

8
0

4.94
4.94

4
12

DR = Dress Rehearsal.
NOTE: The error rate equals the percentage of observed cases where the error was observed. A total of 81 screenings were
observed. Note that the total error rate reflects the error rate for DR-specific errors only and is not directly comparable
with the total error rate in Table 5.32.

The majority of FIs also displayed positive behaviors when conducting interviews (see
Table 5.34). Of the 14 items listed on the DR field observation interviewing checklist, only
3 items were observed being conducted incorrectly at least 5 percent of the time. Those items
include the following:
•

not reading all screens verbatim;

•

not following the proper DR quality control form and incentive procedures; and

•

not answering respondent questions correctly and thoroughly, referencing appropriate
DR details.

Verbatim reading errors and FI errors for not following the proper DR quality control
form and incentive procedures were not related specifically to DR procedures and could have
occurred during main study observations. FIs who had difficulty answering respondent questions
used terminology from the main study (i.e., "U.S Public Health Service") instead of the new DR
specific terminology (i.e., "DHHS"). The FI training for the 2015 NSDUH will include
additional emphasis and practice on the importance of using the correct terminology while in the
field.
To further investigate the potential effects of changes between the DR and main study,
items were added to the DR field observation interview checklist to reflect changes to the
interview procedures, project information, and use of specific DR materials. Only five errors
were recorded on these items (see Table 5.35). During DR field observations, FIs generally did
well at following interview procedures during DR field observations, including new procedures
specific to the DR and procedures carried over from the main study.

237

Table 5.34 Dress Rehearsal Interviewing Error Rates

Interviewing Error
If interview respondent was a minor, FI not first obtaining consent from parent or
legal guardian
If interview respondent was not screening respondent, not explaining purpose of
study and visit thoroughly
Not handing DR study description to the respondent
Not reading intro to CAI from DR showcard booklet verbatim to the respondent
Not choosing a private location to conduct interview
Not setting up equipment efficiently
Not explaining headphone usage, offering headphones to respondent, and plugging
them in
Not keeping ACASI portion private (read ACASI) and/or not remaining attentive
Not reading all screens verbatim
Not presenting DR showcards when prompted by the CAI
Not following the proper DR quality control form and incentive procedures
Not answering respondent questions correctly and thoroughly, referencing the
appropriate DR details (e.g., RTI International, DHHS, did not mention DR or
field test, sample size, pay or payment, etc.)
Not providing respondent with correct DR materials
Committing other procedural violations not noted on the checklist (sum of
individual errors)
TOTAL

Total Cases = 40
Error
Errors
Rate, %
Observed
2.50

0.00
2.50
0.00
0.00
0.00

0
1
0
0
0

0.00
0.00
17.50
2.50
5.00

0
0
7
1
2

5.00
2.50

2
1

2.50
2.86

1
16

ACASI = audio computer-assisted self-interviewing; CAI = computer-assisted interviewing; DR = Dress Rehearsal; FI = field
interviewer.
NOTE: The error rate equals the percentage of observed cases where the error was observed. Bolded items are considered
serious breaches of protocol.

Table 5.35 Dress Rehearsal-Specific Errors: Interviewing

Interview Error
Not following the proper DR quality control form and incentive procedures
Not answering respondent questions correctly and thoroughly, referencing the
appropriate DR details (e.g., RTI International, DHHS, did not mention DR or
field test, sample size, or payment)
Not providing respondent with correct DR materials
TOTAL

Total Cases = 40
Error
Errors
Rate, %
Observed
5.00
2

5.00
2.50
4.17

2
1
5

DR = Dress Rehearsal.
NOTE: The error rate equals the percentage of observed cases where the error was observed. A total of 40 interviews were
observed. Note that the total error rate reflects the error rate for DR-specific errors only and is not directly comparable
with the total error rate in Table 5.34.

238

5.6.3

DR Field Observation Comments

Observers were asked to evaluate the performance of the DR equipment and materials
while in the field. No supplementary comments or concerns were received from observers about
the performance of the DR materials during their observations.
Several comments were made about the performance of the tablet in the field:
•

Three FIs mentioned that they had difficulty seeing the screen in direct sunlight.

•

One FI had difficulty tapping and holding items on the tablet instead of just tapping,
as is done on the iPAQ.

•

Two FIs had issues troubleshooting unexpected events with the tablet (e.g., the screen
freezing or the tablet occasionally going blank when first powered on).

Only one comment was received on the performance of the laptop in the field:
•

One FI commented that the function keys were very small and hard to see in low
light.

The DR field observations did not uncover any serious concerns about the DR equipment
or materials. The tablet troubleshooting issues observed could be handled by addressing these
specific items during training and adding documentation to the FI manual on how to resolve
these occurrences. However, these particular technical concerns may not be replicable. The size
of the function keys on the laptop was only mentioned by one FI, but this issue and the
troubleshooting items will be considered further, especially given similar feedback received from
the FI equipment survey and FI debriefing calls.
Observers recorded several comments made by respondents and FIs during the screening
and interview:
•

Seven respondents were confused about the race/ethnicity questions, in both the
screening and interview, and did not know how to classify themselves. These
respondents seemed to think of Hispanic or Latino as their race.

•

Two respondents commented that the interview was time-consuming, and one
respondent commented that the interview did not take very long.

•

One respondent would have preferred if the interview had been read entirely by the FI
rather than having any ACASI portions.

•

Two respondents felt that the interview questions were repetitive.

•

One respondent commented that the ACASI voice read too slowly.

•

One respondent was confused by the phrase "type in your answer." She attempted to
type words for her answer instead of the number corresponding to the appropriate
answer choice.

•

One FI commented that there is not enough space to leave a meaningful comment in
the comment feature on specific questions in the interview.
239

Although these comments occurred during DR field observations, none was clearly
specific to DR screening or interview changes from the main study. These respondents and the
one FI could have expressed similar issues during main study data collection. However, the main
study field observations do not provide comparison data on how many times respondents were
confused or what comments respondents and FIs made on these same issues.
Observers also recorded issues with the proxy portion of the DR interview that do not
apply to the main study. Specifically, the transition from the first ACASI section into the second
is awkward for cases in which only the respondent is included in the household roster. The first
ACASI portion tells the respondent that he or she is "done," then the respondent is almost
immediately handed the computer again to complete the second section. During DR field
observations, two respondents commented that this was confusing. An observer also recorded
that an FI did not know how to switch the language to Spanish from English for the second
ACASI section of the interview.
5.6.4

DR Field Observation Summary and Recommendations

Overall, the DR field observations provided an important opportunity to see firsthand
how the DR instrument, materials, and equipment performed in the field. In general, the DR
components all performed well. Some items went so smoothly that there were no reported issues
or comments from observers, including items regarding the flow of the screening presentation,
the transition between the screening and the interview, and respondent confusion with the laptop.
Based on comments and feedback received during the field observations, the following
changes or enhancements will be considered for the 2015 NSDUH protocols, equipment, and
training materials:
•

Provide additional emphasis and practice with reading the study introduction screen
and using the correct terminology when speaking with respondents.

•

Provide additional training and practice on how to address respondent confusion with
the race/ethnicity questions.

•

Provide additional documentation and training on tablet troubleshooting.

•

Offer research laptop options with larger function keys, given that similar feedback
was received from FIs during the FI equipment survey and FI debriefing calls.

As a result of the increase in item missing data in the health insurance and income questions, and
the differences in estimates in these items between the field tests and the annual NSDUH, the
current plan is to move the health insurance and income questions from ACASI back to CAPI for
the 2015 survey. Furthermore, this change will eliminate the second ACASI portion and the need
to hand the computer back to a respondent. Therefore, no further changes are needed to address
the awkwardness of this transition when a respondent is the only adult in the household.

240

6. Selected Core and Noncore Estimates for
English- and Spanish-Language Dress
Rehearsal Data and Comparison Data
(Research Question 4)
6.1

Overview of Selected Core and Noncore Estimates for English- and
Spanish-Language Dress Rehearsal Data and Comparison Data

This chapter presents findings on selected core and noncore estimates from the 2013
Dress Rehearsal (DR) and comparison data. The following types of analyses are presented:
•

analyses to make decisions for the 2015 survey, presented in Section 6.2;

•

further analyses based on findings from the 2012 Questionnaire Field Test (QFT),
presented in Section 6.3; and

•

analyses to explain anticipated findings in 2015, also presented in Section 6.3.

Most of the analyses in this chapter compare data from the combined English-language
interviews among non-Hispanic respondents only from the QFT and DR with data from Englishlanguage interviews among non-Hispanic respondents in the 2012 and 2103 main study
comparison samples. Selected tables in Section 6.3 provide estimates parallel to the Englishlanguage interviews among non-Hispanic respondents that are based on data from Spanishlanguage interviews in the DR and the 2012 comparison and 2103 comparison samples.

6.2

Analyses to Make Decisions for the 2015 Survey

Analyses for research question 4 that are needed for making decisions for the 2015
survey focus on findings that realistically could affect the content of the 2015 NSDUH
questionnaire. The following sets of analyses fall into this category:
•

analysis of initiation data, particularly for prescription drugs;

•

contributions of certain prescription drugs to estimates of past year use or misuse for
overall prescription drug categories or key subcategories (e.g., benzodiazepine
tranquilizers); and

•

effects on estimates based on changes to the height and weight questions for the DR.

6.2.1

Analysis of Initiation Data

Changes to how initiation of misuse of prescription drugs is measured in the redesigned
questionnaire could affect the following estimates for the initiation of prescription drug misuse:
•

numbers of past year initiates,

•

percentages of past year initiates in the population,
241

•

percentages of past year initiates among persons who are at risk for initiation,

•

percentages of past year initiates among past year misusers of prescription drugs, and

•

mean ages at first use among past year initiates of misuse.

Because the numbers of past year initiates for pain relievers and tranquilizers in 2012 ranked
second and third among illicit drugs (CBHSQ, 2013), changes to the questions for initiation of
misuse of prescription drugs also could affect estimates for initiation of use of illicit drugs other
than marijuana. Therefore, analyses of initiation data that may be affected by changes to the
prescription drug questions are critical for determining whether any modifications to the
prescription drug initiation questions are needed for the 2015 questionnaire.
6.2.1.1

Past Year Initiates in the General Population

Table 6.1A presents estimated numbers of past year initiates aged 12 or older for illicit
drugs, illicit drugs other than marijuana, and specific illicit drugs or categories of illicit drugs
based on the DR data and the comparison data for 2012 and 2013. Table 6.1B presents
corresponding percentages of past year initiates among the persons aged 12 or older in the
general population. Reference is made only to the "comparison data" if similar findings were
observed for both the 2012 and 2013 comparison data relative to the DR.
•

The estimated numbers of past year initiates and percentages of persons aged 12 or
older who were past year initiates of use of illicit drugs did not differ significantly
between the DR and the comparison data. This was the case for the "standard
definition" that included all illicit drugs and alternate definitions that did not include
various groups of illicit drugs.

•

The estimated numbers of past year initiates and percentages of persons aged 12 or
older who were past year initiates of use of illicit drugs other than marijuana did not
differ significantly between the DR and the comparison data for the standard
definition that included all illicit drugs other than marijuana and for various alternate
definitions.

•

Estimated numbers and percentages of persons who were past year initiates for
prescription drugs were similar for the DR and comparison data. For example,
3.0 million persons were estimated to be past year initiates of misuse of prescription
drugs based on data from the DR. The numbers of persons in the comparison data
who were estimated to be past year initiates were 2.6 million for 2012 and 2.3 million
for 2013.

•

For stimulants, an estimated 1.4 million persons based on the DR data were past year
initiates for the "standard definition" that includes methamphetamine as a stimulant,
700,000 were past year initiates based on the 2012 comparison data, and 596,000
were past year initiates based on the 2013 comparison data. An estimated 1.7 million
persons were past year initiates of misuse of prescription stimulants in the DR data
when methamphetamine was not included. (Persons who initiated misuse of
prescription stimulants in the past year but initiated use of methamphetamine more
than 12 months ago were included in the "DR Definition" in Tables 6.1A and 6.1B,
but they were not included as past year initiates in the standard definition.)

242

Table 6.1A Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older: Numbers in
Thousands, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal

Drug Measure
ILLICIT DRUGS, STANDARD
DEFINITION4
Alternate Definition 14
Alternate Definition 24
Alternate Definition 34
Marijuana and Hashish
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
Methamphetamine
Prescription Drug Misuse5
Pain Relievers
Tranquilizers
Stimulants, Standard
Definition6
Stimulants, DR Definition6
Sedatives
ILLICIT DRUGS OTHER THAN
MARIJUANA, STANDARD
DEFINITION7
Alternate Definition 17
Alternate Definition 27

2013
2012
Comparison Comparison
(n = 66,542)1 (n = 32,162)2

2013 DR
(n = 2,087)3

DR vs. 2012
Comparison,
Difference (SE)
(770)
(551)
(580)
(550)
(580)
(385)
(18)
(104)
(218)
(167)
(23)
(163)
(168)
(125)
(817)
(575)
(569)

DR vs. 2013
Comparison,
Difference (SE)

3,170
2,686
2,552
2,684
2,534
676
90a
180
1,142a
434
102a
954a
698
146
2,629
2,031
1,527

3,034
2,631
2,461
2,620
2,496
735
77a
133
1,157a
481
29
758a
578
126
2,275
1,667
1,301

2,868
2,210
2,221
2,210
2,221
857
0*
107
643
360
0*
358
553
155
2,986
1,896
1,332

302
476
331
474
313
-181
90
73
499
74
102
596
145
-9
-357
135
195

166
421
240
410
275
-122
77
26
514
121
29
400
25
-29
-712
-229
-31

(775)
(586)
(597)
(587)
(606)
(384)
(24)
(99)
(238)
(176)
(15)
(154)
(190)
(124)
(810)
(572)
(545)

700
N/A
189

596
N/A
72

1,449
1,654
987

-749 (540)
N/A (N/A)
-798 (552)

-853 (522)
N/A (N/A)
-915 (551)

2,765
1,518
1,513

2,475
1,538
1,531

2,761
1,330
1,301

4 (741)
188 (371)
212 (370)

-286 (744)
208 (392)
230 (393)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine.
NOTE: Data on initiation of substance use were not edited to make them consistent with data on most recent use or vice versa.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
2012 comparison data collected in quarters 1 through 4, 2012. Sample does not include Alaska or Hawaii.
2
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013. Sample does not include Alaska or
Hawaii.
3
DR data collected from September 1 through October 31, 2013. Sample does not include Alaska or Hawaii.
4
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics. Alternate Definition 3 does not include prescription-type psychotherapeutics but includes
methamphetamine.
5
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data but is not included for the DR.
6
The Standard Definition of stimulant misuse for the DR includes methamphetamine. The DR Definition of stimulant misuse
does not include methamphetamine.
7
Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include prescription-type psychotherapeutics.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

243

Table 6.1B Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older: Percentages
of All Persons Aged 12 or Older, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and 2013 Dress Rehearsal

2013
2012
Comparison Comparison
(n = 66,542)1 (n = 32,162)2
1.2
1.0
1.0
1.0
1.0
0.3
0.0a
0.1
0.4a
0.2
0.0a
0.4a
0.3
0.1
1.0
0.8
0.6

1.2
1.0
1.0
1.0
1.0
0.3
0.0a
0.1
0.4a
0.2
0.0
0.3a
0.2
0.0
0.9
0.6
0.5

0.3
N/A
0.1
1.1
0.6
0.6

2013 DR
(n = 2,087)3

DR vs. 2012
Comparison,
Difference (SE)
(0.29)
(0.21)
(0.22)
(0.21)
(0.22)
(0.15)
(0.01)
(0.04)
(0.08)
(0.06)
(0.01)
(0.06)
(0.07)
(0.05)
(0.31)
(0.22)
(0.22)

DR vs. 2013
Comparison,
Difference (SE)

1.1
0.9
0.9
0.9
0.9
0.3
0.0*
0.0
0.2
0.1
0.0*
0.1
0.2
0.1
1.2
0.7
0.5

0.1
0.2
0.1
0.2
0.1
-0.1
0.0
0.0
0.2
0.0
0.0
0.2
0.1
-0.0
-0.1
0.1
0.1

0.1
0.2
0.1
0.2
0.1
-0.0
0.0
0.0
0.2
0.0
0.0
0.2
0.0
-0.0
-0.3
-0.1
-0.0

(0.29)
(0.23)
(0.23)
(0.23)
(0.23)
(0.15)
(0.01)
(0.04)
(0.09)
(0.07)
(0.01)
(0.06)
(0.07)
(0.05)
(0.31)
(0.22)
(0.21)

0.2
N/A
0.0

0.6
0.6
0.4

-0.3 (0.21)
N/A (N/A)
-0.3 (0.21)

-0.3 (0.20)
N/A (N/A)
-0.4 (0.21)

1.0
0.6
0.6

1.1
0.5
0.5

0.0 (0.28)
0.1 (0.14)
0.1 (0.14)

-0.1 (0.29)
0.1 (0.15)
0.1 (0.15)

244

•

For sedatives, an estimated 1.0 million persons based on the DR data were past year
initiates, 189,000 were past year initiates based on the 2012 comparison data, and
72,000 were past year initiates based on the 2013 comparison data. However, these
estimates were not significantly different.

•

Although the content of the initiation questions did not change for the hallucinogens
module in the DR, the numbers and percentages of past year initiates based on the DR
were lower than those based on the 2012 and 2013 comparison data for any
hallucinogen and Ecstasy. Numbers and percentages of past year initiates based on
the DR data also were lower than those in the 2012 comparison data for
phencyclidine (PCP). However, no respondents in the DR reported past year initiation
of PCP use, and these estimates would have been suppressed.

•

The number and percentages of past year initiates for crack cocaine based on the DR
were also lower than those based on the comparison data. However, no respondents in
the DR reported past year initiation of crack cocaine use, and these estimates would
have been suppressed.

6.2.1.2

Initiation among Persons at Risk for Initiation

NSDUH defines persons as being "at risk" for initiation if they did not use a substance (or
substances) in their lifetime or they used the substance for the first time in the past year. Persons
who used a substance in their lifetime but more than 12 months ago are no longer considered to
be at risk for initiation because they already initiated use. However, the focus of the redesigned
prescription drug questions on past year misuse of specific prescription drugs could cause
respondents who misused prescription drugs in a given category more than 12 months ago to
incorrectly report that they never misused prescription drugs in that category. These respondents
would be misclassified as being "at risk" of initiation.
Table 6.2 presents percentages of past year initiates in the DR and comparison data
among persons aged 12 or older who were considered to be at risk for initiation.
•

As for numbers and percentages of all persons aged 12 or older, the percentages of
past year initiates of illicit drugs and illicit drugs other than marijuana among persons
who were at risk for initiation were similar between the DR and comparison data.

•

Percentages of past year initiates of use of crack cocaine, hallucinogens, and Ecstasy
among persons who were at risk for initiation were lower in the DR than in both years
of comparison data. Percentages for PCP also were significantly different for the DR
and 2012 comparison data.

6.2.1.3

Initiation among Past Year Users

Table 6.3 presents percentages of past year initiates in the DR and comparison data
among persons aged 12 or older who were past year users of illicit drugs, marijuana, prescription
drugs, pain relievers, and illicit drugs other than marijuana. A smaller set of drugs is shown in
the table because of smaller numbers of DR respondents who reported past year use or misuse
for specific illicit drugs.

245

Table 6.2 Past Year Initiation of Illicit Drug Use among Persons Aged 12 or Older: Percentages of
Persons at Risk for Initiation of Illicit Drug Use, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal

Drug Measure
ILLICIT DRUGS, STANDARD
DEFINITION5
Alternate Definition 15
Alternate Definition 25
Alternate Definition 35
Marijuana and Hashish
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
Methamphetamine
Prescription Drug Misuse6
Pain Relievers
Tranquilizers
Stimulants, Standard Definition7
Stimulants, DR Definition7
Sedatives
ILLICIT DRUGS OTHER THAN
MARIJUANA, STANDARD
DEFINITION8
Alternate Definition 18
Alternate Definition 28

2012
Comparison
(n = ?c)1,2

2013
Comparison
(n = ?c)1,3

2013 DR
(n = ?c)1,4

DR vs. 2013
DR vs. 2012
Comparison,
Comparison,
Difference
Difference (SE)
(SE)

2.3
1.8
1.7
1.8
1.7
0.3
0.0a
0.1
0.5a
0.2
0.0a
0.4a
0.3
0.1
1.3
0.9
0.6
0.3
N/A
0.1

2.2
1.8
1.7
1.8
1.7
0.3
0.0a
0.1
0.5a
0.2
0.0
0.3a
0.2
0.1
1.1
0.7
0.5
0.2
N/A
0.0

2.0
1.5
1.5
1.5
1.4
0.4
0.0*
0.0
0.3
0.2
0.0*
0.1
0.2
0.1
1.3
0.8
0.5
0.6
0.7
0.4

0.3
0.3
0.2
0.3
0.2
-0.1
0.0
0.0
0.2
0.0
0.0
0.2
0.1
-0.0
-0.1
0.1
0.1
-0.3
N/A
-0.3

(0.52)
(0.37)
(0.37)
(0.37)
(0.37)
(0.18)
(0.01)
(0.04)
(0.10)
(0.07)
(0.01)
(0.07)
(0.07)
(0.05)
(0.36)
(0.24)
(0.23)
(0.23)
(N/A)
(0.22)

1.5
0.7
0.7

1.3
0.8
0.8

1.5
0.7
0.7

-0.0 (0.39)
0.1 (0.18)
0.1 (0.19)

0.2
0.3
0.2
0.3
0.2
-0.1
0.0
0.0
0.2
0.0
0.0
0.2
0.0
-0.0
-0.2
-0.1
0.0
-0.4
N/A
-0.4

(0.52)
(0.39)
(0.39)
(0.39)
(0.39)
(0.18)
(0.01)
(0.04)
(0.11)
(0.08)
(0.01)
(0.06)
(0.08)
(0.05)
(0.36)
(0.24)
(0.22)
(0.22)
(N/A)
(0.22)

-0.2 (0.40)
0.1 (0.20)
0.1 (0.20)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; LSD = lysergic acid diethylamide; N/A = not applicable; PCP = phencyclidine.
NOTE: Data on initiation of substance use were not edited to make them consistent with data on most recent use or vice versa.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample sizes are for all respondents aged 12 or older after this exclusion had been
made. Sample sizes for the specific drugs will vary based on the numbers of persons at risk for initiation.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics. Alternate Definition 3 does not include prescription-type psychotherapeutics but includes
methamphetamine.
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data but is not included for the DR.
7
The Standard Definition of stimulant misuse for the DR includes methamphetamine. The DR Definition of stimulant misuse
does not include methamphetamine.
8
Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include prescription-type psychotherapeutics.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

246

Table 6.3 Past Year Initiation of Use of Selected Illicit Drugs among Persons Aged 12 or Older:
Percentages of Past Year Users, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and 2013 Dress Rehearsal

Drug Measure
ILLICIT DRUGS, STANDARD
DEFINITION5
Alternate Definition 15
Alternate Definition 25
Alternate Definition 35
Marijuana and Hashish
Prescription Drug Misuse6
Pain Relievers
ILLICIT DRUGS OTHER THAN
MARIJUANA, STANDARD
DEFINITION7
Alternate Definition 17
Alternate Definition 27

DR vs. 2012 DR vs. 2013
Comparison, Comparison,
2013
2012
Difference
Difference
2013 DR
Comparison Comparison
(SE)
(SE)
(n = 66,542)1,2 (n = 32,162)1,3 (n = 2,087)1,4
6.8
7.2
7.1
7.1
7.3
14.4
14.4

6.6
6.9
6.7
6.8
7.0
13.3
13.6

5.6
5.2
5.7
5.1
5.8
13.6
12.9

1.2 (1.54)
2.0 (1.35)
1.4 (1.52)
2.0 (1.31)
1.5 (1.56)
0.7 (3.52)
1.5 (3.88)

1.0 (1.54)
1.7 (1.37)
1.0 (1.52)
1.7 (1.33)
1.2 (1.59)
-0.3 (3.52)
0.7 (4.02)

11.7
15.6
5.5

11.2
17.2
7.2

10.4
12.8*
8.2*

1.3 (2.78)
2.7 (5.11)
-2.7 (7.15)

0.8 (2.88)
4.3 (5.30)
-1.0 (7.16)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable;
NOTE: Data on initiation of substance use were not edited to make them consistent with data on most recent use or vice versa.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample sizes are for all respondents aged 12 or older after this exclusion had been
made. Sample sizes for the specific drugs will vary based on the numbers of past year users.
2
2012 comparison data collected in quarters 1 through 4, 2012. Sample does not include Alaska or Hawaii.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013. Sample does not include Alaska or
Hawaii.
4
DR data collected from September 1 through October 31, 2013. Sample does not include Alaska or Hawaii.
5
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics. Alternate Definition 3 does not include prescription-type psychotherapeutics but includes
methamphetamine.
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data but is not included for the DR.
7
Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include prescription-type psychotherapeutics.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

247

Percentages of past year initiates among past year users were not significantly different
between the DR and comparison data for the illicit drug measures shown in Table 6.3. However,
the estimated percentages of past year initiates among past year initiates of illicit drugs other
than marijuana would have been suppressed in the DR for the measures that did not include
prescription drugs.
6.2.1.4

Mean Age at First Use among Past Year Initiates

NSDUH tables and reports present estimates of mean ages at first use among persons
who initiated use of different substances in the past year. As measures of central tendency,
however, means are heavily influenced by the presence of extreme values in the data for persons
aged 12 or older. To reduce the effect of extreme values, values for the mean age at initiation are
calculated for persons aged 12 to 49 and leave out those few respondents who were past year
initiates at age 50 or older. Including data from initiates aged 26 to 49 in this broad age group
also can cause instability of estimates of the mean age at initiation among persons aged 12 to 49,
but this effect is less than that of including data from initiates aged 50 or older.
The DR questionnaire asked respondents to report the age when they first misused each
specific prescription drug for which they reported misuse in the past 12 months. That included up
to 40 possible pain relievers, 19 tranquilizers, 24 stimulants, and 17 sedatives. In comparison,
respondents in the main survey are asked only one age-at-first use question in the tranquilizers
and sedatives modules and are asked only two age-at-first use questions in the modules for pain
relievers (any pain reliever and OxyContin®) and stimulants (any stimulant and
methamphetamine). Even if mean ages at first use in the DR are restricted to persons aged 12 to
49 who were past year initiates, the differences in questions about ages at first use in the DR and
main survey could affect the resulting estimates for means.
Table 6.4 presents mean ages at first use among past year initiates aged 12 to 49 in the
DR and comparison data for illicit drugs, marijuana, prescription drugs, pain relievers, and illicit
drugs other than marijuana.
•

Mean ages at first use among past year initiates aged 12 to 49 were lower in the DR
than in the comparison data for illicit drug definitions that excluded prescription
drugs and for marijuana. However, mean ages for first use of illicit drugs other than
marijuana were similar in the DR and comparison data.

The mean age at first misuse of any prescription psychotherapeutic drug among past year
initiates aged 12 to 49 was 26.0 years in the DR and 22.6 years in both years of comparison data.
The mean age at first misuse of pain relievers was greater for the DR than for the 2013
comparison data (28.8 vs. 20.8 years) but was not significantly different from the mean for the
2012 comparison data (22.1 years).

248

Table 6.4 Mean Age at First Use of Selected Illicit Drugs among Past Year Initiates Aged 12 to 49:
Differences and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and 2013 Dress Rehearsal

Drug Measure
ILLICIT DRUGS, STANDARD
DEFINITION5
Alternate Definition 15
Alternate Definition 25
Alternate Definition 35

2012
Comparison
(n = 59,005)1,2
18.9
17.9

18.2

2013
Comparison
(n = 28,470)1,3
19.5

2013 DR
(n = 1,768)1,4

DR vs. 2012
Comparison,
Difference
(SE)

DR vs. 2013
Comparison,
Difference
(SE)

20.5

-1.6 (3.41)

-1.0 (3.33)

18.5

15.5

2.5 (0.66)

3.0 (0.74)

18.6

15.8

2.4 (0.66)

2.8 (0.72)

18.4

15.5

2.5 (0.66)

3.0 (0.74)

18.7

15.8

2.5 (0.66)

2.9 (0.77)

17.9

Marijuana and Hashish

18.3

Prescription Drug Misuse6

22.6

26.0

-3.4 (3.08)

-3.4 (3.11)

22.1

20.8a

28.8

-6.7 (3.68)

-7.9 (3.75)

20.8

20.5

23.8

-3.0 (3.05)

-3.3 (2.96)

17.9

19.1

18.9

-0.9 (1.93)

0.3 (1.92)

17.9

19.1

18.7

-0.7 (2.00)

0.5 (1.99)

Pain Relievers
ILLICIT DRUGS OTHER
THAN MARIJUANA,
STANDARD DEFINITION7
Alternate Definition 17
Alternate Definition 27

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable;
NOTE: Data on initiation of substance use were not edited to make them consistent with data on most recent use or vice versa.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample sizes are for all respondents aged 12 to 49 after this exclusion had been
made. Sample sizes for the specific drugs will vary based on the numbers of past year initiates.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics. Alternate Definition 3 does not include prescription-type psychotherapeutics but includes
methamphetamine.
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data but is not included for the DR.
7
Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, methamphetamine, or
prescription-type psychotherapeutics that were misused. Alternate Definition 1 does not include methamphetamine or
prescription-type psychotherapeutics. Alternate Definition 2 does not include prescription-type psychotherapeutics.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

249

6.2.1.5

Initiation for Prescription Drugs That Were Misused in the Past Year

Initiation data for prescription drugs in earlier sections focused on comparisons between
the DR and main survey data because QFT respondents who reported only past year initiation for
the prescription drugs that they misused in the past year could have misused other drugs in a
given category (e.g., pain relievers) more than 12 months ago. However, it is possible to compare
some initiation estimates between the QFT and DR based on data from non-Hispanic Englishlanguage respondents in the two datasets. Specifically, both the QFT and DR allow identification
of two groups of persons among those who reported past year misuse: (a) persons who initiated
misuse more than 12 months ago for any of the drugs that they misused in the past year; and
(b) persons who initiated misuse in the past year for all of the drugs that they misused in that
period. Persons in the first group by definition would not be past year initiates. However, the
second group could include past year initiates for the overall category and persons who misused
other drugs in the category that they did not misuse in the past year.
Table 6.5A presents estimated numbers persons aged 12 or older from the QFT and DR
who misused prescription drugs in the past year, including those who reported initiation for some
drugs more than 12 months ago, and those who reported only past year initiation for the drugs
that they misused. Table 6.5B presents corresponding percentages. Although DR respondents
who had missing initiation data for all drugs in a given category that they misused in the past
year were asked whether they ever used any drugs in that category more than 12 months ago,
QFT respondents were not given this opportunity. Therefore, QFT and DR respondents who had
missing data initiation data for all of the individual prescription drugs within a given category
that they misused in the past year were excluded from these analyses.
•

An estimated 16.3 million persons misused prescription drugs in the past year based
on the QFT data, including 11.3 million who initiated misuse of some of these drugs
more than 12 months ago, and 3.9 million who reported past year initiation for all of
the drugs that they misused in that period. For the DR, 17.3 million persons misused
prescription drugs in the past year, including 13.7 million who initiated misuse of
some prescription drugs more than 12 months ago, and 3.1 million who reported past
year initiation for all of the drugs that they misused in that period. (Numbers of
persons who initiated misuse of some drugs more than 12 months ago and those who
reported only past year initiation do not sum to the total number of persons who
misused any prescription drugs because of missing data.)

•

An estimated 12.0 million persons based on QFT data and 11.8 million persons based
on DR data misused pain relievers in the past year. In the QFT, 8.6 million persons
who initiated misuse of pain relievers more than 12 months ago, and 2.3 million
reported past year initiation for all of the pain relievers that they misused in that
period. For the DR, 10.0 million persons initiated misuse of pain relievers more than
12 months ago, and 1.6 million reported past year initiation for all of the pain
relievers that they misused in that period.

250

Table 6.5A Past Year Misuse of Prescription Drugs and Initiation of Misuse among Persons Aged
12 or Older for English-Language Non-Hispanic Interviews: Numbers in Thousands,
Differences, and Standard Error of Differences, 2012 QFT and 2013 Dress Rehearsal
Drug Measure
Prescription Drug Misuse, Past Year3
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Pain Reliever Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Tranquilizer Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Stimulant Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Sedative Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5

2012 QFT
(n = 1,692)1
16,272
11,298

2013 DR
(n = 1,320)2
17,588
13,714

DR vs. 2012 QFT,
Difference (SE)
-1,316 (2,767)
-2,416 (2,472)

3,891
11,963
8,604

3,092
11,793
10,049

799 (1,062)
170 (2,233)
-1,445 (2,103)

2,316
5,305
3,780

1,624
6,269
4,728

691 (800)
-964 (1,551)
-948 (1,303)

1,465
4,085
2,862

617
4,127
2,591

848 (480)
-41 (1,291)
272 (903)

1,102
1,838
1,035

1,469
1,761
787

-367 (676)
77 (867)
248 (562)

803

975

-172 (655)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
QFT = Questionnaire Field Test; DR = Dress Rehearsal; N/A = not applicable.
NOTE: Data on initiation of substance use were not edited to make them consistent with data on most recent use or vice versa.
a
Difference between estimate and 2013 DR is statistically significant at the 0.05 level.
1
QFT data collected from September 1 through November 3, 2012. Sample does not include Alaska or Hawaii and does not
include Spanish-language interviews. Hispanic respondents who completed the QFT interview in English also have been
excluded for these comparisons.
2
DR data collected from September 1 through October 31, 2013. Sample does not include Alaska or Hawaii. Data from Spanishlanguage interviews and Hispanic respondents have been excluded for comparability with the QFT data.
3
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is not included.
4
Estimates are based on edited data and therefore could have missing data for initiation. Respondents who had missing initiation
data for all individual drugs that they misused in the past year were excluded from these analyses.
5
Includes respondents who reported past year initiation of misuse for some drugs and had missing initiation data for any
remaining drugs that were misused in the past year. Estimates for the DR apply only to the individual drugs that were misused
in the past year and do not include data from the new follow-up question for respondents who reported only past year initiation.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

251

Table 6.5B Past Year Misuse of Prescription Drugs and Initiation of Misuse among Persons Aged
12 or Older for English-Language Non-Hispanic Interviews: Percentages of All Persons
Aged 12 or Older, Differences, and Standard Error of Differences, 2012 QFT and 2013
Dress Rehearsal
Drug Measure
Prescription Drug Misuse, Past Year3
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Pain Reliever Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Tranquilizer Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Stimulant Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5
Sedative Misuse, Past Year
Initiation Reported Before the Past Year4
Only Past Year Initiation Reported for Individual
Drugs4,5

2012 QFT
(n = 1,692)1
7.5
5.2

2013 DR
(n = 1,320)2
8.1
6.3

DR vs. 2012 QFT,
Difference (SE)
-0.6 (1.21)
-1.1 (1.11)

1.8
5.5
4.0

1.4
5.4
4.6

0.4 (0.48)
0.1 (1.00)
-0.6 (0.96)

1.1
2.4
1.7

0.7
2.9
2.2

0.3 (0.36)
-0.4 (0.70)
-0.4 (0.60)

0.7
1.9
1.3

0.3
1.9
1.2

0.4 (0.21)
-0.0 (0.60)
0.1 (0.42)

0.5
0.8
0.5

0.7
0.8
0.4

-0.2 (0.31)
0.0 (0.40)
0.1 (0.26)

0.4

-0.1 (0.30)

252

6.2.1.6

Discussion of Past Year Initiation Data

The lack of significant differences between the DR and comparison data for estimated
numbers and percentages of past year initiates for prescription drugs suggests that the structure
of the prescription drug modules may be performing adequately for estimating the prevalence of
initiation of prescription drug misuse. The structural content for the prescription drug modules
includes the addition of follow-up questions to detect misuse more than 12 months ago if
respondents reported only past year initiation for the specific drugs that they misused in the past
year. However, some initiation estimates for prescription drugs that were not significantly
different between the DR and comparison data may become significant in 2015 with a larger
sample size.
Similarly, the mean age at first use for pain relievers approached 30 years in the DR data
and was lower in the 2013 comparison data (Table 6.4). The mean in the 2012 comparison data
also was about 22 years, despite this age not being significantly different from the age in the DR.
One possible explanation for this finding is that extreme values could have had more of an effect
on the mean for the DR than for the comparison data because of the smaller sample size for the
DR. However, this finding also could indicate that the structure of the prescription drug
questions in the DR allows information to be captured about persons who initiate misuse of
prescription drugs at an older age compared with the information that is captured from the
current questionnaire.
As noted in Sections 6.2.1.1 and 6.2.1.2, initiation estimates were significantly different
between the DR and comparison data for crack cocaine, hallucinogens, and Ecstasy and between
the DR and 2012 comparison data for PCP, despite the content of the initiation questions being
the same in the two questionnaires. A possible explanation is that the smaller sample size for the
DR and the sample allocation to obtain Spanish-language interviews in the DR may not have
been optimal for obtaining interviews from past year initiates for these substances, despite the
subsequent weighting.
An alternative explanation that initiation for some of these drugs may be trending
downward is not supported by the significant differences between the DR and 2013 comparison
data. However, the differences between the estimates of PCP initiation in the DR and 2012
comparison data but not between the DR and 2013 comparison data could be attributable to
small numbers of respondents in less than 12 months of survey data who report past year
initiation of PCP use. For example, the 2013 comparison data from quarters 3 and 4 yielded an
estimate of only 29,000 past year initiates for PCP (Table 6.1A).
6.2.2

Contributions of Specific Prescription Drugs to Estimates of Use and Misuse

Including questions about commonly used and misused prescription psychotherapeutic
drugs will be important for accurately measuring the prevalence of misuse based on NSDUH
data. If drugs with specific active ingredients are most commonly prescribed, it also will be
important for the NSDUH questionnaire to include examples of drugs with those active
ingredients that respondents are likely to recognize for reporting use and misuse. Conversely,
questions about low prevalence drugs may not contribute appreciably but could increase
respondent burden. Furthermore, identifying drugs that are still on the market in the

253

United States but have a low prevalence of use or misuse and that appear to contribute little to
prevalence estimates could allow questions about these drugs to be replaced with questions about
other prescription drugs.
Tables 6.6 to 6.9 present the estimated numbers of persons aged 12 or older who used
specific prescription pain relievers, tranquilizers, stimulants, or sedatives in the past year; those
who reported past year misuse; and those who reported past year use but not misuse. To increase
the available sample size and precision of estimates, the estimates presented in these tables were
based on combined QFT and DR data from 3,012 respondents who were not Hispanic and
completed the interview in English.
However, only the data for tranquilizers in Table 6.7 were relevant for resolving the
remaining questions prior to making recommendations to the Substance Abuse and Mental
Health Services Administration (SAMHSA) for the content of the specific prescription drug
questions for 2015. Tables for the prescription drug categories other than tranquilizers provide
supporting documentation for recommendations that can be made to SAMHSA based on
information from other sources. For example, if certain prescription drugs in the QFT and DR
questionnaire have been discontinued in 2013 or earlier, recommendations to drop these drugs
for 2015 can be made without the need to review the results of these analyses. Nevertheless,
discontinued drugs would be expected to have a low prevalence of use and misuse relative to
prevalence estimates for drugs that are still on the market in the United States.
Even with the combined data from the QFT and DR, it also was recognized that the small
sample sizes could limit the kinds of conclusions that could be drawn from these data about the
content of the prescription drug questions for 2015. Therefore, it was assumed that data sources
other than the QFT and DR also would be important for making recommendations to SAMHSA
about the prescription drug questions for 2015. For example, information from IMS Health is
available on rxlist.com for the 200 most commonly dispensed prescription drugs in the United
States in 2012.23 Although the list of most commonly prescribed medications includes drugs that
do not have relevance to NSDUH (e.g., antibiotics), several psychotherapeutic drugs are among
the 200 most commonly prescribed medications, such as hydrocodone-acetaminophen
combinations (e.g., Vicodin®), alprazolam (e.g., Xanax®), amphetamine salts (e.g., Adderall®),
and zolpidem (e.g., Ambien®).
6.2.2.1

Pain Relievers

Table 6.6 presents estimated numbers of persons aged 12 or older who used prescription
pain relievers in the past year, those who reported past year misuse, and those who reported past
year use but not misuse. Estimates are presented for any prescription pain relievers, groups of
pain relievers (e.g., pain relievers containing fentanyl), and specific pain relievers. The latter
group includes estimates for specific pain relievers that were identified as having a low
prevalence of use or misuse in the QFT (Currivan et al., 2013). Particular attention was given to
identifying estimated numbers of persons who reported use or misuse of only these low
prevalence drugs. However, comparative statements in this section refer to relative differences in
23

Retrieved on January 8, 2014, from http://www.rxlist.com/script/main/hp.asp. The list includes some
duplicate mentions of the same active ingredient if manufacturers market the generic equivalent under different
brand names.

254

Table 6.6 Alternate Measures of Past Year Use and Misuse of Specific Pain Relievers among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews: Numbers in
Thousands and Standard Errors, Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal
Pain Reliever
Any Prescription Pain Reliever3
Hydrocodone, Oxycodone, Tramadol,
Codeine, or Morphine Products
Hydrocodone Products
Lorcet® as Only Hydrocodone
Product4
Lorcet® as Only Pain Reliever
Oxycodone Products
Tylox® as Only Oxycodone
Product4
Tylox® as Only Pain Reliever
Tramadol Products
Codeine Products
Morphine Products
Morphine (Generic)5
Avinza® as Only Morphine Product
Kadian® as Only Morphine Product
MS Contin® as Only Morphine
Product4
Avinza®, Kadian, or MS Contin as
Only Pain Relievers
Propoxyphene Products
Propoxyphene Products as
Only Pain Relievers
Fentanyl Products
Fentanyl (Generic)5
Actiq® as Only Fentanyl Product
Duragesic® as Only Fentanyl
Product
Fentora® as Only Fentanyl Product
Fentanyl Products as Only Pain
Relievers
Actiq®, Duragesic®, or Fentora® as
Only Pain Relievers
Buprenorphine Products
Buprenorphine Products as
Only Pain Relievers
Demerol®
Demerol® as Only Pain Reliever
Dilaudid®
Dilaudid® as Only Pain Reliever
Methadone
Methadone as Only Pain Reliever

Any Past Year
Use1
84,034 (4,007)

Past Year Use
But Not Misuse2
71,751 (3,703)

Past Year
Misuse1
12,282
(1,239)

75,231
56,940

(3,800)
(3,208)

64,372
49,799

(3,610)
(2,976)

10,824
7,105

243
103
28,164

(116)
(74)
(1,929)

210
103
23,977

(111)
(74)
(1,745)

0*
0*
4,152

(0*)
(0*)
(651)

193
68
14,655
27,211
7,840
7,223
265
90

(98)
(68)
(1,644)
(2,318)
(1,165)
(1,117)
(164)
(64)

193
68
12,551
24,044
7,106
6,642
265
51

(98)
(68)
(1,596)
(2,258)
(1,139)
(1,106)
(164)
(51)

0*
0*
2,068
3,132
734
580
0*
38

(0*)
(0*)
(449)
(483)
(247)
(215)
(0*)
(38)

263

148

(90)

115

(115)

0* (0*)
5,258 (997)

0*
4,817

(0*)
(983)

0*
442

(0*)
(201)

248
1,337
1,102
132

229
1,252
1,018
132

(183)
(368)
(336)
(132)

19
85
85
0*

(19)
(60)
(60)
(0*)

(51)
(51)

0*
0*

(0*)
(0*)

(0*)

51
51

(146)

(988)
(858)

(184)
(366)
(334)
(132)
(51)
(51)

0* (0*)

51
51
0*

0* (0*)
3,025 (689)

0*
2,155

(0*)
(647)

0*
870

(0*)
(296)

795
1,143
192
1,635
183
1,269
136

524
1,084
192
1,237
183
664
119

(311)
(301)
(192)
(473)
(130)
(209)
(85)

353
60
0*
398
178
605
16

(246)
(45)
(0*)
(127)
(107)
(220)
(16)

See notes at end of table.

(356)
(304)
(192)
(492)
(130)
(307)
(87)

(continued)

255

Any Past Year
Use1
509 (146)
87
114
0*

(62)
(66)

Past Year Use
But Not Misuse2
256 (117)
87
84
0*

Past Year
Misuse1
253
(88)

(62)
(59)

0*
30

(0*)
(30)

(0*)

597

(268)

432

(255)

20,169

(0*)
(1,79
6)

19,542

(0*)
(1,79
9)

6,465

(804)

6,360

(813)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
NOTE: Standard errors are shown in parentheses. Sample does not include Alaska or Hawaii and does not include Spanishlanguage interviews, and Hispanic respondents who completed the interview in English also are excluded for these
comparisons (n = 3,012).
NOTE: Questionnaire Field Test data collected from September 1 through November 3, 2012. Dress Rehearsal data collected
from September 1 through October 31, 2013.
1
Persons with unknown data are excluded.
2
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with
missing data for misuse are excluded.
3
Includes hydrocodone products (Vicodin®, Lortab®, Lorcet®, or generic hydrocodone); oxycodone products (OxyContin®,
Percocet®, Percodan®, Tylox®, or generic oxycodone); propoxyphene products (Darvocet®, Darvon®, or generic propoxyphene);
tramadol products (Ultram®, Ultram® ER, Ultracet®, Ryzolt®, or generic tramadol); codeine products (Tylenol® with codeine 3
or 4 or generic codeine pills); morphine products (Avinza®, Kadian®, MS Contin®, Oramorph® SR, or generic morphine);
fentanyl products (Actiq®, Duragesic®, Fentora®, or generic fentanyl); buprenorphine products (Suboxone®, Subutex®, or
generic buprenorphine); and Demerol®, Dilaudid®, methadone, Opana®, Opana® ER, Talacen®, Talwin®, Talwin® NX, or any
other prescription pain reliever.
4
Includes use/misuse of pain relievers containing other active ingredients.
5
Use/misuse of the generic with or without use/misuse of brand name drugs with the same active ingredient.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

prevalence; testing for statistical significance of differences between estimates was not
conducted.
•

An estimated 84.0 million persons aged 12 or older used any pain reliever in the past
year, including approximately 12.3 million who misused pain relievers and
approximately 71.8 million who used pain relievers but did not misuse them.

•

Of the 84.0 million persons who used any pain reliever in the past year, 75.2 million
used pain relievers that contained hydrocodone, oxycodone, tramadol, codeine, or
morphine; 5.3 million used products that contained propoxyphene; 1.3 million used
products that contained fentanyl; and 3.0 million used products that contained
buprenorphine. (Numbers for the categories of pain relievers are not mutually
exclusive.)

256

•

Numbers of persons who used Demerol®, Dilaudid®, or methadone ranged from
1.1 million to 1.6 million. Numbers of persons who misused these products in the past
year were 60,000 for Demerol®, 398,000 for Dilaudid®, and 605,000 for methadone.

•

An estimated 20.1 million persons reported past year use of pain relievers other than
those that were included in the pain relievers module.

•

Of the 12.3 million persons who misused pain relievers in the past year, most of the
misuse was accounted for by pain relievers that contained hydrocodone, oxycodone,
tramadol, codeine, or morphine (10.8 million persons, or nearly 90 percent of the
persons who misused pain relievers). Among persons who misused pain relievers that
contained any of these five ingredients, the three most commonly misused pain
relievers were those that contained hydrocodone (7.1 million persons), oxycodone
(4.1 million persons), or codeine (3.1 million persons). In addition, 2.1 million
persons misused tramadol products, and 734,000 misused morphine products.

•

Only 243,000 of the estimated 56.9 million persons who used hydrocodone products
in the past year reported that they used Lorcet® but no other hydrocodone products.
No respondents in the combined QFT and DR data reported misuse of Lorcet ® as the
only pain reliever with hydrocodone or the only pain reliever that they misused.

•

An estimated 193,000 persons of the 28.1 million persons who used oxycodone
products in the past year reported that they used Tylox® but no other oxycodone
products. No respondents in the combined QFT and DR data reported misuse of
Tylox ® as the only pain reliever with oxycodone or the only pain reliever that they
misused.

•

Most of the estimated use and misuse of morphine products was captured by reports
of generic morphine. Among the 7.8 million persons who used morphine products in
the past year, 7.2 million (about 90 percent) reported use of the generic. Among the
734,000 persons who misused morphine products in the past year, 580,000 (about
80 percent) reported misuse of the generic. In addition, 115,000 persons who misused
morphine products in the past year were estimated to have misused only MS Contin®,
and 38,000 misused only Kadian®.

•

Similar to morphine, most of the estimated use of fentanyl products was captured by
reports of the generic. Among the 1.3 million persons who used fentanyl products in
the past year, 1.1 million (about 80 percent) reported use of the generic. In addition,
about 10 percent of the past year users of fentanyl products reported that they used
only Actiq® (132,000 persons), and about 8 percent reported that they used only
Duragesic® or only Fentora® (51,000 persons each). No respondents in the combined
QFT and DR data reported misuse of brand-name fentanyl products as the only
fentanyl products that they misused.

•

Among the estimated 1.3 million persons who used methadone in the past year,
nearly half (605,000 persons) misused methadone in that period. Similarly, although
only 509,000 persons were estimated to be past year users of the oxymorphone
products Opana® or Opana® ER, about half of the past year users reported misuse
(253,000 persons).

257

•

Although 20.1 million persons reported past year use of pain relievers other than
those that were included in the pain relievers module, only 597,000 persons reported
past year misuse of other pain relievers (about 3 percent). Among the estimated
12.3 million persons who misused pain relievers in the past year, only 432,000 (about
4 percent) reported that they misused only pain relievers other than those in the
module.

6.2.2.2

Tranquilizers

Table 6.7 presents estimated numbers of persons aged 12 or older who used prescription
tranquilizers in the past year, those who reported past year misuse, and those who reported past
year use but not misuse. Estimates are presented for any prescription tranquilizer, groups of
tranquilizers (e.g., benzodiazepine tranquilizers), and specific tranquilizers.
•

An estimated 39.4 million persons aged 12 or older used any tranquilizer in the past
year, including 5.8 million who misused tranquilizers and 33.6 million who used
tranquilizers but did not misuse them.

•

Of the 39.4 million persons who used any tranquilizer in the past year, 27.1 million
used benzodiazepine tranquilizers; 11.4 million used Flexeril® or Soma®; 3.1 million
used buspirone, hydroxyzine, or meprobamate; and 5.8 million used other
tranquilizers. (Numbers for the categories of tranquilizers are not mutually exclusive.)

•

Of the 27.1 million persons who used benzodiazepine tranquilizers in the past year,
263,000 reported that Librium® was the only benzodiazepine they used, 30,000
reported that Tranxene® was the only benzodiazepine they used, and 387,000 reported
that oxazepam was the only benzodiazepine that they used. In addition, 649,000
persons were estimated to be past year users of Librium®, Tranxene®, or oxazepam
but no other tranquilizers.

•

Of the 5.8 million persons who misused tranquilizers in the past year, 4.7 million
misused benzodiazepines; 1.6 million misused Flexeril® or Soma®; 286,000 misused
buspirone, hydroxyzine, or meprobamate; and 131,000 misused other tranquilizers.
An estimated 206,000 persons who misused tranquilizers misused only buspirone,
hydroxyzine, or meprobamate.

•

The most commonly misused benzodiazepine tranquilizers in the past year were
products containing alprazolam (3.3 million persons), lorazepam (1.0 million
persons), clonazepam (920,000 persons), or diazepam (475,000 persons).
No respondents in the combined QFT and DR data reported misuse of Librium®,
Tranxene®, or oxazepam as the only benzodiazepine tranquilizers or the only
tranquilizers that they misused.

258

Table 6.7 Alternate Measures of Past Year Use and Misuse of Specific Tranquilizers for EnglishLanguage Non-Hispanic Interviews among Persons Aged 12 or Older: Numbers in
Thousands and Standard Errors, Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal
Tranquilizer
Any Prescription Tranquilizer3
Benzodiazepine Tranquilizers
Xanax®, Xanax® XR, Alprazolam, or
Extended-Release Alprazolam
Ativan® or Lorazepam
Klonopin® or Clonazepam
Valium® or Diazepam
Librium® as Only Benzodiazepine
Tranquilizer4
Tranxene® as Only Benzodiazepine
Tranquilizer4
Oxazepam as Only Benzodiazepine
Tranquilizer4
Librium®, Tranxene®, or Oxazepam
as Only Tranquilizers
Flexeril® or Soma®
Flexeril® or Soma ® as Only
Tranquilizers
Buspirone (also known as BuSpar®),
Hydroxyzine (also known as Atarax®
or Vistaril®), or Meprobamate (also
known as Equanil® or Miltown®)
Buspirone, Hydroxyzine, or
Meprobamate as Only
Tranquilizers
Any Other Prescription Tranquilizer
Any Other Prescription Tranquilizer as
Only Tranquilizers

Past Year Use But
Not Misuse1,2
Any Past Year Use1
Past Year Misuse1
39,356 (2,974)
33,569 (2,722)
5,787 (776)
27,113 (2,375)
22,374 (2,168)
4,739 (729)
14,431
6,306
7,262
4,202

(1,501)
(1,051)
(1,072)
(1,057)

11,153
5,291
6,342
3,727

(1,300)
(985)
(1,018)
(1,016)

3,279
1,015
920
475

(548)
(315)
(269)
(284)

263 (263)

0* (0*)

30 (30)

0* (0*)

387 (368)

0* (0*)

649 (452)
11,440 (1,231)

649 (452)
9,879 (1,116)

0* (0*)
1,561 (406)

5,913 (883)

5,399 (818)

712 (252)

3,083 (739)

2,797 (708)

286 (212)

1,215 (506)
5,792 (1,268)

1,009 (462)
5,661 (1,255)

206 (206)
131 (131)

4,044 (1,129)

3,914 (1,115)

131 (131)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
NOTE: Standard errors are shown in parentheses. Sample does not include Alaska or Hawaii and does not include Spanishlanguage interviews, and Hispanic respondents who completed the interview in English also are excluded for these
comparisons (n = 3,012).
NOTE: Questionnaire Field Test data collected from September 1 through November 3, 2012. Dress Rehearsal data collected
from September 1 through October 31, 2013.
1
Persons with unknown data are excluded.
2
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with
missing data for misuse are excluded.
3
Includes benzodiazepine tranquilizers (Xanax®, Xanax® XR, alprazolam, extended-release alprazolam, Ativan®, lorazepam,
Klonopin®, clonazepam, Valium®, diazepam, Librium®, Tranxene®, or oxazepam); Flexeril®, Soma®, buspirone (also known as
BuSpar®), hydroxyzine (also known as Atarax® or Vistaril®), meprobamate (also known as Equanil® or Miltown®), or any other
prescription tranquilizer.
4
Includes use/misuse of other types of tranquilizers or muscle relaxants.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

259

6.2.2.3

Stimulants

Table 6.8 presents estimated numbers of persons aged 12 or older who used prescription
stimulants in the past year, those who reported past year misuse, and those who reported past
year use but not misuse. Estimates are presented for any prescription stimulants, groups of
stimulants (e.g., stimulants containing methylphenidate), and specific stimulants.
•

An estimated 15.6 million persons aged 12 or older used any stimulant in the past
year, including 5.4 million who misused stimulants and 10.3 million who used
stimulants but did not misuse them.

•

Of the 15.6 million persons who used any stimulant in the past year, 10.5 million used
amphetamines24 or methylphenidate products; 2.2 million used stimulants that are
prescribed for weight loss; 444,000 used Provigil®; and 2.4 million used other
stimulants. Of the 10.5 million persons who used amphetamines or methylphenidate
products in the past year, 8.5 million used amphetamines, and 3.7 million used
methylphenidate products.

•

Among the 3.7 million persons who used methylphenidate products in the past year,
about 190,000 reported that they used Metadate® CD (126,000 persons) or Metadate®
ER (66,000 persons) as their only methylphenidate products, or about 5 percent of the
persons who used methylphenidate. No respondents in the combined QFT and DR
data reported that Daytrana®, which delivers methylphenidate through a patch, was
the only form of methylphenidate that they used in the past year.

•

Of the 5.4 million persons who misused stimulants in the past year, most of the
misuse was accounted for by amphetamines or methylphenidate products (4.0 million
persons, or about 75 percent), including 3.8 million persons who misused
amphetamines and 1.1 million who misused methylphenidate. An estimated 84,000
persons misused weight-loss stimulants, and 89,000 persons misused other stimulants.

•

No respondents in the combined QFT and DR data reported that the only stimulants
they misused in the past year were the methylphenidate products Daytrana®,
Metadate® CD or Metadate® ER; the weight-loss stimulants Didrex®, benzphetamine,
Tenuate®, diethylpropion, or phendimetrazine; Provigil®; or other stimulants.

Amphetamines included Adderall®, Adderall® XR, Dexedrine®, dextroamphetamine, or amphetaminedextroamphetamine combinations. Vyvanse® (lisdexamfetamine) was not counted as an amphetamine for these
analyses.

260

Table 6.8 Alternate Measures of Past Year Use and Misuse of Specific Stimulants for EnglishLanguage Non-Hispanic Interviews among Persons Aged 12 or Older: Numbers in
Thousands and Standard Errors, Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal
Past Year Use But
Stimulant
Any Past Year Use1
Not Misuse2
Past Year Misuse1
3
15,612 (1,416)
10,253 (1,269)
5,359 (753)
Any Prescription Stimulant
Amphetamine or Methylphenidate
Products
10,497 (1,159)
6,458 (957)
3,959 (640)
Amphetamine Products
8,492 (981)
4,700 (739)
3,792 (634)
Methylphenidate Products
3,696 (740)
2,563 (687)
1,052 (253)
Daytrana® as Only Methylphenidate
0* (0*)
0* (0*)
0* (0*)
Product4
Metadate® CD as Only
126 (101)
126 (101)
0* (0*)
Methylphenidate Product4
®
Metadate ER as Only
66 (66)
66 (66)
0* (0*)
Methylphenidate Product4
®
®
Daytrana , Metadate CD, or
Metadate® ER, as Only
Stimulants
96 (96)
96 (96)
0* (0*)
Weight-Loss Stimulants
2,274 (559)
2,190 (550)
84 (48)
Didrex® as Only Weight-Loss
0* (0*)
0* (0*)
0* (0*)
Stimulant4
Benzphetamine as Only WeightLoss Stimulant4
0* (0*)
0* (0*)
0* (0*)
Tenuate® as Only Weight-Loss
Stimulant4
0* (0*)
0* (0*)
0* (0*)
Diethylpropion as Only WeightLoss Stimulant4
45 (34)
15 (15)
30 (30)
Phendimetrazine as Only WeightLoss Stimulant4
250 (193)
250 (193)
0* (0*)
Weight-Loss Stimulants as Only
Stimulants
1,542 (470)
1,517 (469)
25 (25)
Didrex®, Benzphetamine,
Tenuate®, Diethylpropion, or
Phendimetrazine as Only
Stimulants
63 (50)
63 (50)
0* (0*)
Provigil®
444 (358)
444 (358)
0* (0*)
Provigil® as Only Stimulant
24 (24)
24 (24)
0* (0*)
Any Other Prescription Stimulant
2,376 (613)
2,287 (606)
89 (89)
Any Other Prescription Stimulant as
Only Stimulants
1,855 (580)
1,855 (580)
0* (0*)
*Low precision; estimate would be suppressed under NSDUH suppression rules.
NOTE: Standard errors are shown in parentheses. Sample does not include Alaska or Hawaii and does not include Spanishlanguage interviews, and Hispanic respondents who completed the interview in English also are excluded for these
comparisons (n = 3,012).
NOTE: Questionnaire Field Test data collected from September 1 through November 3, 2012. Dress Rehearsal data collected
from September 1 through October 31, 2013.
1
Persons with unknown data are excluded.
2
Persons who did not misuse a prescription drug/prescription drugs they reported using in the past year. Past year users with
missing data for misuse are excluded.
3
Includes amphetamine products (Adderall®, Adderall® XR, Dexedrine®, generic dextroamphetamine, or generic amphetaminedextroamphetamine combinations); methylphenidate products (Ritalin®, Ritalin® SR, Ritalin® LA, Concerta®, Daytrana®,
Metadate CD, Metadate ER, Focalin, Focalin XR, generic methylphenidate, or generic dexmethylphenidate); weight-loss
stimulants (Didrex®, benzphetamine, Tenuate®, diethylpropion, phendimetrazine, or phentermine); and Provigil®, Vyvanse®, or
any other prescription stimulant.
4
Includes use/misuse of other types of stimulants.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013.

261

6.2.2.4

Sedatives

Table 6.9 presents estimated numbers of persons aged 12 or older who used prescription
sedatives in the past year, those who reported past year misuse, and those who reported past year
use but not misuse. Estimates are presented for any prescription sedatives, groups of sedatives
(e.g., barbiturates), and specific sedatives.
•

An estimated 18.9 million persons aged 12 or older used any sedative in the past year,
including 1.8 million who misused sedatives and 17.1 million who used sedatives but
did not misuse them.

•

Of the 18.9 million persons who used any sedative in the past year, 13.0 million used
zolpidem products; 1.4 million used Lunesta®; 603,000 used zaleplon products;
2.6 million used benzodiazepines that are indicated for use as sedatives; 511,000 used
barbiturates; and 4.2 million used other sedatives.

•

An estimated 1.6 million persons used only benzodiazepine sedatives in the past year,
511,000 used only barbiturates, and 2.9 million reported that other sedatives were the
only sedatives that they used in the past year.

•

Among the 2.6 million persons who used benzodiazepine sedatives in the past year,
about 1.0 million reported that the only benzodiazepine sedative they used was
Halcion® (586,000 persons), triazolam (the generic equivalent of Halcion®;
301,000 persons), or Restoril® (128,000 persons).

•

Among the 511,000 persons who used barbiturates in the past year, 390,000 (about
76 percent) reported that phenobarbital was the only barbiturate that they used.

•

Most misuse of prescription sedatives in the past year involved misuse of zolpidem
products. Among the 1.8 million persons who misused sedatives in the past year,
1.6 million misused zolpidem products (nearly 90 percent).

•

Although 4.2 million persons reported past year use of sedatives other than those that
were included in the sedatives module, only 24,000 persons reported past year misuse
of other sedatives (about 0.4 percent). Among the estimated 1.8 million persons who
misused sedatives in the past year, only 24,000 (about 1 percent) reported that they
misused only sedatives other than those in the module.

262

Table 6.9 Alternate Measures of Past Year Use and Misuse of Specific Sedatives for EnglishLanguage Non-Hispanic Interviews among Persons Aged 12 or Older: Numbers in
Thousands and Standard Errors, Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal
Sedative
Any Prescription Sedative3
Zolpidem Products
Lunesta®
Zaleplon Products
Benzodiazepine Sedatives
Dalmane® as Only Benzodiazepine
Sedative4
Flurazepam as Only
Benzodiazepine Sedative4
Halcion® as Only Benzodiazepine
Sedative4
Triazolam as Only
Benzodiazepine Sedative4
Restoril® as Only Benzodiazepine
Sedative
Benzodiazepine Sedatives as Only
Sedatives
Barbiturates
Butisol® as Only Barbiturate4
Seconal® as Only Barbiturate4
Phenobarbital as Only Barbiturate4
Barbiturates as Only Sedatives
Any Other Prescription Sedative
Any Other Prescription Sedative as
Only Sedatives

Any Past Year
Use1
18,890 (1,754)
13,000 (1,473)
1,420 (470)
603 (299)
2,640 (670)

Past Year Use
But Not Misuse2
17,091 (1,628)
11,410 (1,332)
1,274 (456)
524 (288)
2,318 (648)

Past Year
Misuse1
1,800
(434)
1,591
(423)
146
(115)
78
(78)
322
(173)

(0*)

0* (0*)

(0*)

0* (0*)

(0*)

586

(346)

450

(319)

136

(136)

301

(154)

233

(138)

(68)

128

(86)

(32)

(80)

1,612
511
34
87
390
290
4,183

(472)
(235)
(34)
(87)
(216)
(185)
(838)

1,503
415
0*
87
328
213
4,160

(466)
(228)
(0*)
(87)
(211)
(177)
(837)

109
96
34
0*
62
77
24

(74)
(58)
(34)
(0*)
(47)
(55)
(24)

2,939

(724)

2,916

(723)

(24)

263

6.2.2.5

Discussion of Specific Prescription Drug Data

Results from Tables 6.6 to 6.9 generally are consistent with available information on
commonly prescribed medications. For example, hydrocodone pain relievers were the most
commonly used and misused prescription pain relievers based on the combined QFT and DR
data; three of the top five most commonly dispensed prescription drugs in the data on rxlist.com
were versions of pain relievers that contain hydrocodone and acetaminophen. Similarly, the four
most commonly misused benzodiazepine tranquilizers in the QFT and DR data (alprazolam,
lorazepam, clonazepam, and diazepam) were among the 200 most commonly dispensed
prescription drugs in 2012. Amphetamine salts (e.g., Adderall®), extended-release
methylphenidate products, and products containing zolpidem also were among the most
commonly dispensed prescription drugs.
Conversely, the benzodiazepines Librium®, Tranxene®, and oxazepam and the nonbenzodiazepine tranquilizers buspirone, hydroxyzine, and meprobamate were not in the top 200.
The absence of these latter tranquilizers from the top 200 list probably reflects a move toward
prescribing newer "generations" of benzodiazepines instead of these non-benzodiazepines and
"older" benzodiazepine tranquilizers. Similarly, the finding that past year users of barbiturates
comprised only about 500,000 of the nearly 19 million persons who were estimated to have used
sedatives in the past year probably reflects changes in prescribing practices for treatment of
insomnia.
Data on misuse of specific subgroups of prescription drugs also are generally consistent
with data from the Drug Enforcement Administration's National Forensic Laboratory
Information System (NFLIS), which reports forensic laboratory results for drugs that were seized
in criminal justice operations. In 2012, for example, pain relievers containing oxycodone or
hydrocodone accounted for nearly 70 percent of the narcotic analgesics that were identified by
NFLIS laboratories (Office of Diversion Control, 2013). These also were the two most
commonly misused subgroups of prescription pain relievers in the QFT and DR data. Similarly,
the 15 most commonly reported tranquilizers and central nervous system depressants in the 2012
NFLIS data included four benzodiazepine tranquilizers (alprazolam, clonazepam, diazepam, and
lorazepam), the muscle relaxants carisoprodol (Soma®) and cyclobenzaprine (Flexeril®),
zolpidem, the benzodiazepine sedative temazepam (Restoril®), the tranquilizer hydroxyzine, and
the barbiturate phenobarbital. Although nearly 85 percent of the forensic laboratory reports for
stimulants in 2012 involved methamphetamine, amphetamines (e.g., Adderall®),
methylphenidate (e.g., Ritalin®), lisdexampfetamine (Vyvanse®), and phentermine were among
the most commonly identified stimulants in 2012 (Office of Diversion Control, 2013).
Consistent with data from these other sources, the QFT and DR data suggest that the
redesigned questionnaire is capturing information about the most commonly used and misused
prescription psychotherapeutic drugs in the United States. Given the constraints placed on
respondent burden, it will be particularly important for redesigning the questionnaire for 2015 to
include not only the subgroups of most commonly used and misused prescription drugs within an
overall category (e.g., pain relievers that contain hydrocodone), but also to include examples of
specific prescription drugs within these subgroups that are most relevant to survey respondents
for reporting use and misuse. For example, findings from Table 6.6 suggest that it would be
reasonable for 2015 to replace the hydrocodone pain reliever Lorcet® and the oxycodone pain
264

reliever Tylox® with other pain relievers that could be more important for estimating the
prevalence of use and misuse of hydrocodone and oxycodone products, respectively.
On the one hand, low estimates for specific prescription drugs in the QFT and DR data—
particularly for past year misuse—can be informative to SAMHSA for identifying prescription
drugs that could be dropped for the 2015 questionnaire without seriously sacrificing the validity
of prevalence estimates. For example, the low numbers of reports (or no reports) of use or misuse
of Librium®, Tranxene®, and oxazepam suggests that these drugs could be dropped for 2015.
However, prevalence is not likely to be the only consideration for decisions to retain or
add prescription drugs for 2015. The following issues also are likely to be relevant even if a drug
is less commonly prescribed and prevalence estimates are low:
•

the potential for serious health consequences if the drug is used outside of medical
supervision or not according to medical directions (e.g., extended-release pain
relievers for which a Risk Evaluation and Mitigation Strategy [REMS] is required);

•

the popularity of a prescription drug to be diverted25 for misuse (e.g., as evidenced in
criminal seizures);

•

the breadth of coverage of prescription drugs within an overall psychotherapeutic
category (e.g., stimulants that are prescribed for weight loss in addition to those that
are prescribed for treatment of attention-deficit/hyperactivity disorder [ADHD]);

•

special characteristics of a drug (e.g., tamper-resistant formulations) that could
warrant prevalence estimation relative to prevalence estimates for other prescription
drugs; and

•

evidence that a recently approved prescription drug has the potential to become more
commonly diverted for misuse.

6.2.3

Height and Weight

Analyses of height and weight data for the QFT included benchmarking analyses that
compared QFT estimates with those from the National Health Interview Survey (NHIS) and the
National Health and Nutrition Examination Survey (NHANES). Comparison of height and
weight data from the DR with data sources external to NSDUH is discussed in Section 7.3.
In addition, analysis of height data from the QFT identified extreme high and low values based
on the allowable ranges for the questions. The same allowable ranges for height were present in
quarter 1 of the 2013 main survey; these ranges for height were adjusted in quarter 2 of the 2013
main survey and for the DR. Therefore, analyses of height and weight data include data from
quarter 1 of 2013 (i.e., prior to the changes for the height questions), in the quarter 3 and 4
comparison data for 2013, and for the DR. Because the height and weight questions were not
added to the main survey until 2013, height and weight data are not available for the 2012
comparison data.
25

Rigg, Kurtz, and Surratt (2012) defined prescription drug diversion as "the transfer of a prescription drug
from a lawful to an unlawful channel of distribution or use." The Centers for Medicare & Medicaid Services (CMS)
referred to the diversion of drugs from legal and medically necessary uses toward uses that are illegal and typically
are not medically authorized or necessary (CMS, 2012).

265

6.2.3.1

Height

Tables 6.10 through 6.15 present summary statistics for height in inches. If respondents
reported their height in metric units (i.e., meters and centimeters or centimeters only), their
heights were converted to inches. Table 6.10 presents summary height statistics for persons aged
12 or older. Tables 6.11 and 6.12 present statistics for males and females aged 12 or older,
respectively. Tables 6.13 through 6.15 present height statistics for persons aged 16 or older,
overall and by gender. In addition, these tables show the number of respondents whose height
data were assigned codes for "bad data" because the heights they reported were below or above
the values based on the revised ranges for height that were fielded in quarter 2 and in the DR.
These cases were treated as having missing data for height.
•

For all persons aged 12 or older (Table 6.10), all persons aged 16 or older
(Table 6.13), and males within these age groups (Tables 6.11 and 6.14), mean
estimates of height were similar between the DR, quarter 1 of 2013, and the quarter 3
and quarter 4 comparison data for 2013. For example, the mean height among all
persons aged 16 or older was 66.9 inches in quarter 1 of 2013 and in the 2013 data
from quarters 3 and 4, and it was 66.6 inches in the DR.

•

Among females, the quarter 1 estimates of mean height were greater than those in the
DR for females aged 12 or older (Table 6.12) and those aged 16 or older
(Table 6.15). The estimated mean height among females aged 16 or older in
quarters 3 and 4 of 2013 also was greater than the corresponding mean for the DR.
Among females aged 16 or older, for example, the mean height was 64.2 inches in
quarter 1 of 2013, 64.1 inches in quarters 3 and 4 of 2013, and 63.8 inches in the DR.

6.2.3.2

Weight

Tables 6.16 through 6.21 present summary statistics for weight in pounds for persons
aged 12 or older (Tables 6.16 through 6.18) and for those aged 16 or older (Tables 6.19 through
6.21). As for the height tables, statistics for weight in pounds also are presented by gender within
these age groups (Tables 6.17 and 6.20 for males; Tables 6.18 and 6.21 for females). Females
aged 12 to 44 who reported that they were currently pregnant were asked to report their weight
before they became pregnant. If respondents reported their weight in kilograms, their weights
were converted to pounds. Unlike the height data, no reported weights were assigned codes for
"bad data."
•

For all persons aged 12 or older (Table 6.16), all persons aged 16 or older
(Table 6.19), and females within these age groups (Tables 6.18 and 6.21), mean
estimates of weight were similar between the DR, quarter 1 of 2013, and the quarter 3
and quarter 4 comparison data for 2013. For example, the mean weight among all
persons aged 16 or older was 177.7 pounds in quarter 1 of 2013, 178.8 pounds in
quarters 3 and 4 of 2013, and 176.3 pounds in the DR.

266

Table 6.10 Summary Statistics for Height in Inches among Persons Aged 12 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 15,368)
14,724
262

2013 Comparison
(n = 32,162)2
31,289
0

2013 DR
(n = 2,087)3
2,027
0

66.8
0.00
0.06
64.0
79.0

66.8
0.00
0.05
64.0
90.2

66.5
0.03
0.17
66.0
76.3

107.0
70.0
66.9
64.0

113.8
70.0
67.0
64.0

107.0
69.0
66.0
63.0

76.0
74.0
72.0
62.0
60.0
57.0

77.0
74.0
72.0
62.0
60.0
56.0

76.0
73.0
72.0
61.0
60.0
57.0

107.0
106.0
105.0
105.0
105.0

113.8
107.9
107.0
107.0
107.0

107.0
105.0
103.0
100.8
87.0

36.0
35.0
29.0
28.3
28.0

23.6
23.6
23.6
23.6
23.6

41.0
41.0
39.0
34.0
30.7

Five Lowest

(Lowest)

DR = Dress Rehearsal.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Answers in metric units (i.e., meters, centimeters) were converted to inches.
2
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
3
DR data collected from September 1 through October 31, 2013. Sample does not include Alaska or Hawaii.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013.

267

Table 6.11 Summary Statistics for Height in Inches among Males Aged 12 or Older: 2013 Quarter
1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 7,324)
7,008
136

2013 Comparison
(n = 15,763)2
15,338
0

2013 DR
(n = 1,002)3
979
0

69.5
0.01
0.07
70.0
78.7

69.7
0.00
0.05
71.0
90.2

69.3
0.04
0.21
71.0
76.3

107.0
72.0
70.0
67.0

113.8
72.0
70.0
68.0

107.0
72.0
69.0
67.0

77.0
75.0
74.0
66.0
64.0
59.0

78.0
75.0
74.0
65.7
64.0
58.0

76.0
75.0
73.0
65.0
64.0
59.0

107.0
105.0
102.0
102.0
102.0

113.8
107.9
107.0
107.0
107.0

107.0
105.0
103.0
100.8
80.7

40.0
40.0
36.0
35.0
28.3

24.0
24.0
23.6
23.6
23.6

53.0
53.0
51.0
43.3
30.7

Five Lowest

(Lowest)

268

Table 6.12 Summary Statistics for Height in Inches among Females Aged 12 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 8,044)
7,716
126

2013 Comparison
(n = 16,399)2
15,951
0

2013 DR
(n = 1,085)3
1,048
0

64.2a
0.00
0.07
64.0
78.0

64.0
0.00
0.05
64.0
83.4

63.7
0.02
0.15
63.0
53.0

106.0
66.0
64.0
62.0

107.0
66.0
64.0
62.0

87.0
66.0
64.0
62.0

72.0
69.0
68.0
61.0
59.1
55.0

72.0
69.0
68.0
60.2
59.0
53.0

71.0
69.0
68.0
60.0
59.0
54.0

106.0
105.0
105.0
105.0
104.0

107.0
105.0
103.0
101.0
101.0

87.0
80.7
79.0
75.0
72.0

40.0
39.0
36.0
29.0
28.0

24.0
24.0
23.6
23.6
23.6

41.3
41.0
41.0
39.0
34.0

Five Lowest

(Lowest)

269

Table 6.13 Summary Statistics for Height in Inches among Persons Aged 16 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 11,978)
11,644
154

2013 Comparison
(n = 24,991)2
24,572
0

2013 DR
(n = 1,743)3
1,704
0

66.9
0.00
0.06
66.0
78.7

66.9
0.00
0.05
66.0
90.2

66.6
0.03
0.18
66.0
76.3

107.0
70.0
67.0
64.0

113.8
70.0
67.0
64.0

107.0
70.0
66.0
63.0

76.0
74.0
72.0
62.0
60.0
58.0

77.0
74.0
73.0
62.0
60.0
57.0

76.0
73.0
72.0
61.0
60.0
58.0

107.0
105.0
105.0
105.0
105.0

113.8
107.0
106.0
105.0
105.0

107.0
103.0
100.8
87.0
80.7

40.2
40.0
40.0
40.0
28.3

24.0
23.6
23.6
23.6
23.6

43.3
41.3
39.0
34.0
30.7

Five Lowest

(Lowest)

270

Table 6.14 Summary Statistics for Height in Inches among Males Aged 16 or Older: 2013 Quarter
1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 5,644)
5,490
80

2013 Comparison
(n = 12,030)2
11,842
0

2013 DR
(n = 809)3
796
0

69.8
0.01
0.07
70.0
78.7

70.0
0.00
0.06
71.0
90.2

69.5
0.05
0.21
71.0
76.3

107.0
72.0
70.0
68.0

113.8
72.0
70.0
68.0

107.0
72.0
69.0
67.0

77.0
75.0
74.0
66.0
65.0
62.0

78.0
75.0
74.0
66.0
65.0
62.0

76.0
75.0
73.0
65.0
64.0
62.0

107.0
105.0
101.0
96.0
95.0

113.8
107.0
106.0
105.0
105.0

107.0
103.0
100.8
80.7
80.0

42.1
40.6
40.2
40.0
28.3

24.0
24.0
24.0
23.6
23.6

58.0
57.0
56.0
43.3
30.7

Five Lowest

(Lowest)

271

Table 6.15 Summary Statistics for Height in Inches among Females Aged 16 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal
Height in Inches1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 6,334)
6,154
74

2013 Comparison
(n = 12,961)2
12,730
0

2013 DR
(n = 934)3
908
0

64.2a
0.01
0.07
64.0
65.0

64.1a
0.00
0.05
64.0
81.4

63.8
0.03
0.16
64.0
53.0

105.0
66.0
64.0
62.0

87.0
66.0
64.0
62.0

72.0
69.0
68.0
61.0
60.0
56.0

72.0
69.0
68.0
61.0
59.1
54.0

71.0
69.0
68.0
60.0
60.0
56.0

105.0
105.0
105.0
101.0
101.0

105.0
103.0
101.0
101.0
98.0

87.0
80.7
72.0
72.0
72.0

43.3
41.3
40.6
40.0
40.0

24.0
24.0
24.0
23.6
23.6

48.0
45.7
41.3
39.0
34.0

Five Lowest

(Lowest)

272

Table 6.16 Summary Statistics for Weight in Pounds among Persons Aged 12 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

Weight in Pounds1
Sample Used in Analysis
Responses Set to "Bad Data" in Editing
Summary Statistics
Mean
Variance
Standard Deviation
Mode
Range
Quartiles
Maximum
75th Percentile
Median
25th Percentile
Percentiles
99th Percentile
95th Percentile
90th Percentile
10th Percentile
5th Percentile
1st Percentile
Five Highest
(Highest)

2013 Q1
(n = 15,368)
15,027
0

2013
Comparison
(n = 32,162)2
31,320
0

2013 DR
(n = 2,087)3
2,027
0

174.7
0.42
0.65
150.0
502.7

175.8
0.24
0.49
150.0
557.8

173.4
2.54
1.59
150.0
423.3

551.2
200.0
170.0
140.0

606.3
200.0
170.0
140.0

463.0
200.0
170.0
140.0

328.0
260.0
235.0
120.0
110.0
90.0

320.0
260.0
235.0
120.0
110.0
92.0

300.0
259.0
235.0
122.0
110.0
90.0

551.2
550.0
540.1
500.0
495.0

606.3
606.3
606.3
551.2
551.2

463.0
436.5
368.2
365.0
360.0

50.0
50.0
48.5
48.5
48.5

50.0
50.0
50.0
50.0
48.5

64.0
58.0
40.0
40.0
39.7

Five Lowest

(Lowest)

DR = Dress Rehearsal.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Answers in metric units (i.e., kilograms) were converted to pounds.
2
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
3
DR data collected from September 1 through October 31, 2013. Sample does not include Alaska or Hawaii.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013.

273

Table 6.17 Summary Statistics for Weight in Pounds among Males Aged 12 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

2013 Q1
(n = 7,324)
7,185
0

2013
Comparison
(n = 15,763)2
15,404
0

190.8a
0.77
0.87
180.0
502.7

192.0a
0.43
0.65
180.0
557.8

185.7
5.25
2.29
150.0
423.3

551.2
215.0
185.0
162.0

606.3
216.0
185.0
163.0

463.0
210.0
180.0
155.0

343.0
275.0
250.0
140.0
125.0
90.0

340.0
275.0
250.0
140.0
125.0
95.0

315.0
255.0
240.0
135.0
124.0
99.0

551.2
550.0
540.1
500.0
495.0

606.3
606.3
606.3
551.2
550.0

463.0
436.5
365.0
360.0
350.0

50.0
50.0
48.5
48.5
48.5

50.0
50.0
50.0
50.0
48.5

75.0
65.0
58.0
40.0
39.7

2013 DR
(n = 1,002)3
980
0

Five Lowest

(Lowest)

274

Table 6.18 Summary Statistics for Weight in Pounds among Females Aged 12 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

2013 Q1
(n = 8,044)
7,842
0

2013
Comparison
(n = 16,399)2
15,916
0

2013 DR
(n = 1,085)3
1,047
0

159.6
0.59
0.77
120.0
430.0

160.4
0.40
0.64
130.0
501.2

161.6
4.84
2.20
130.0
328.2

480.0
180.0
150.0
130.0

551.2
180.0
150.0
130.0

368.2
185.0
150.0
130.0

300.0
240.0
216.0
115.0
107.0
91.0

299.0
245.0
218.3
115.0
106.0
91.0

290.0
259.0
225.0
116.0
105.0
75.0

480.0
450.0
444.0
440.0
424.0

551.2
500.0
460.0
443.1
423.3

368.2
320.0
318.0
316.0
300.0

60.0
58.0
50.0
50.0
50.0

50.0
50.0
50.0
50.0
50.0

67.0
67.0
65.0
64.0
40.0

Five Lowest

(Lowest)

275

Table 6.19 Summary Statistics for Weight in Pounds among Persons Aged 16 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

2013 Q1
(n = 11,978)
11,803
0

2013
Comparison
(n = 24,991)2
24,582
0

2013 DR
(n = 1,743)3
1,707
0

177.7
0.47
0.69
150.0
491.6

178.8
0.27
0.52
150.0
556.3

176.3
2.55
1.60
150.0
423.3

540.1
200.0
172.0
145.0

606.3
202.0
174.0
145.0

463.0
200.0
172.0
141.0

330.0
260.0
235.0
125.0
115.0
100.0

321.9
265.0
240.0
125.0
116.0
100.0

300.0
259.0
236.0
125.0
117.0
95.0

540.1
500.0
495.0
480.0
474.0

606.3
606.3
551.2
551.2
550.0

463.0
436.5
368.2
365.0
360.0

50.0
50.0
50.0
48.5
48.5

50.7
50.0
50.0
50.0
50.0

65.0
64.0
40.0
40.0
39.7

Five Lowest

(Lowest)

276

Table 6.20 Summary Statistics for Weight in Pounds among Males Aged 16 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

2013 Q1
(n = 5,644)
5,586
0

2013
Comparison
(n = 12,030)2
11,877
0

2013 DR
(n = 809)3
799
0

194.7a
0.80
0.89
180.0
491.6

195.9a
0.46
0.68
180.0
556.3

189.0
5.92
2.43
150.0
423.3

540.1
215.0
185.0
165.0

606.3
220.0
190.0
165.0

463.0
215.0
185.0
160.0

343.9
276.0
250.0
147.0
137.0
115.0

340.0
276.0
250.0
146.0
135.0
120.0

315.0
260.0
240.0
140.0
130.0
110.0

540.1
500.0
495.0
474.0
456.0

606.3
606.3
551.2
550.0
501.0

463.0
436.5
365.0
360.0
350.0

50.0
50.0
50.0
48.5
48.5

55.0
50.7
50.7
50.0
50.0

100.0
99.2
94.0
40.0
39.7

Five Lowest

(Lowest)

277

Table 6.21 Summary Statistics for Weight in Pounds among Females Aged 16 or Older: 2013
Quarter 1, 2013 Comparison, and 2013 Dress Rehearsal

2013 Q1
(n = 6,334)
6,217
0

2013
Comparison
(n = 12,961)2
12,705
0

2013 DR
(n = 934)3
908
0

162.0
0.66
0.81
130.0
430.0

162.6
0.46
0.68
130.0
501.2

164.2
4.85
2.20
130.0
328.2

480.0
180.0
154.0
130.0

551.2
185.0
154.3
132.0

368.2
190.0
154.0
130.0

300.0
245.0
220.0
117.0
110.0
98.0

300.0
247.0
220.0
118.0
110.0
96.0

290.0
259.0
230.0
120.0
110.0
78.0

480.0
450.0
444.0
440.0
424.0

551.2
460.0
443.1
423.3
410.0

368.2
320.0
318.0
316.0
300.0

70.0
70.0
65.0
58.0
50.0

60.0
55.0
51.0
50.0
50.0

68.0
67.0
65.0
64.0
40.0

Five Lowest

(Lowest)

278

•

Among males, the DR estimates of weight were lower than those in quarter 1 of 2013
and in quarters 3 and 4 of 2013 for males aged 12 or older (Table 6.17) and for those
aged 16 or older (Table 6.20). The estimated mean weight among males aged 16 or
older was 189.0 pounds in the DR. In comparison, the mean weights among males
aged 16 or older were 194.7 pounds in quarter 1 of 2013 and 195.9 pounds in
quarters 3 and 4 of 2013.

6.2.3.3

Discussion of Height and Weight Data

Except where noted, findings for height and weight were consistent across the different NSDUH
data sources that were examined in this section. Where means differed, these findings could
reflect the influence of extreme high or low values on the resulting means in the DR data because
of the smaller DR sample size. However, further comparison of NSDUH height and weight data
with data from other sources also will be important for (see Section 7.3) increasing SAMHSA's
confidence in the validity of these NSDUH data.

6.3

Further Analyses Based on QFT Findings or Analyses to Explain
Anticipated Findings in 2015

6.3.1

Core Substance Use Items Other Than Methamphetamine and Prescription Drugs

This section presents highlights for core substance use estimates from the 2012 NSDUH
comparison data, the 2013 NSDUH quarters 3 and 4 comparison data, and data from the QFT
and DR for substances other than methamphetamine and prescription drugs. Section 6.3.1.1
discusses highlights for marijuana, cocaine, and heroin. Section 6.3.1.2 discusses highlights for
hallucinogens, and Section 6.3.1.3 discusses highlights for inhalants. Section 6.3.1.4 discusses
multiple definitions of use of "any illicit drug." Sections 6.3.1.5 and 6.3.1.6 present highlights
based on these definitions for any illicit drug use. Sections 6.3.1.7 and 6.3.1.8 present findings
for cigarette use and smokeless tobacco use, respectively. Finally, Section 6.3.1.9 provides
findings on any use of alcohol, and Section 6.3.1.10 presents findings for binge alcohol use, as
defined in that section.
6.3.1.1

Marijuana, Cocaine, and Heroin

This section presents key findings for marijuana, cocaine, and heroin use from the 2012
comparison data and 2013 quarters 3 and 4, as well as data from the QFT and DR. Tables 6.22
through 6.25 provide estimates for lifetime use of these substances for all persons aged 12 or
older, adolescents aged 12 to 17, young adults aged 18 to 25, and adults aged 26 or older.
Likewise, Tables 6.26 through 6.29 provide estimates for past year use of these substances, and
Tables 6.30 through 6.33 provide estimates for past month use of these substances. No changes
were made in the instrument for the QFT and DR for the questions on marijuana, cocaine
(including crack), and heroin use. However, some significant differences between the QFT and
comparison data had been identified that warranted further investigation in the DR analysis
(Currivan et al., 2013). Furthermore, some differences that were not significant in the QFT
analysis could become significant when QFT and DR data are combined for non-Hispanic
English-language respondents.

279

•

In contrast to the finding of no significant differences in the lifetime prevalence of
crack cocaine use between the QFT data and corresponding comparison data for
persons aged 12 or older, the estimate for lifetime crack use among persons aged
12 or older in the combined QFT and DR data (5.2 percent) was greater than the
estimates in the 2012 and 2013 comparison data (3.7 and 3.6 percent, respectively)
(Table 6.22).

•

Among adolescents aged 12 or older, the findings of lower estimates of any cocaine
use in the QFT than in the comparison data continued to be observed in the combined
QFT and DR data for lifetime (Table 6.23), past year (Table 6.27), and past month
use (Table 6.31). However, the past month estimate for the combined QFT and DR
data would have been suppressed in published estimates.

•

Among young adults aged 18 to 25, the rate of past month cocaine use had been
higher in the 2011 comparison data than in the QFT (Currivan et al., 2013). However,
the rates were similar for the combined QFT-DR data and the 2012 and 2013
comparison data (Table 6.32).

•

Among adults aged 26 or older, the lifetime rate of crack use was higher in the
combined QFT-DR data (6.2 percent) compared with the rate of 4.3 percent in both
the 2012 and 2013 comparison data (Table 6.25).

•

For heroin, estimates were lower in the combined QFT-DR data than in the 2012 or
2013 comparison data for past year use among 12 to 17 year olds, past month use
among persons aged 12 or older (Table 6.30), and past month use among adolescents
aged 12 to 17 and adults aged 26 or older (Table 6.33). However, the combined
QFT/DR estimates would have been suppressed for all of these estimates except for
past month use among persons aged 12 or older.

Thus, some differences between field test and comparison data for estimates of cocaine
and heroin use that had been observed in the QFT continued to be observed in the combined
QFT and DR data despite the content of these modules not changing for the QFT and DR.
However, the assumption continues to be that any changes in prevalence for these drugs in 2015
relative to earlier years based on a full sample of approximately 67,000 interviews in 2015 will
reflect an actual change in prevalence in the population. This assumption can be tested by
reviewing trend data that include data from the first 6 months of 2015, which will likely have a
sample size of more than 30,000, or roughly 10 times the sample size of the combined QFT-DR
data. In addition, single-year fluctuations in prevalence would need to be interpreted with
caution. It would be important to examine trends across multiple years—including years beyond
2015—to account for occasional fluctuations in prevalence that may "correct" themselves with
additional years of data.
6.3.1.2

Hallucinogens

This section discusses key findings for use of any hallucinogen, lysergic acid
diethylamide (LSD), phencyclidine (PCP), and Ecstasy in the QFT and DR data and in the 2012
and 2013 comparison data. Although questions about ketamine, tryptamines, and Salvia
divinorum had been moved from the noncore special drugs module to the core hallucinogens

280

module for the QFT and DR, findings from the QFT indicated few differences in lifetime, past
year, or past month prevalence between the QFT and comparison data (Currivan et al., 2013).
•

Among adolescents aged 12 to 17, the estimate of lifetime use of hallucinogens was
greater in the QFT than in the 2011 and 2012 comparison data. Similarly, the
prevalence of lifetime use of hallucinogens among adolescents was greater in the
combined QFT-DR data (4.5 percent) than in the 2013 comparison data (2.5 percent)
(Table 6.23).

•

For Ecstasy, the prevalence of past month use was lower in the combined QFT-DR
data than in the 2013 comparison data for persons aged 12 or older (0.1 vs.
0.3 percent) (Table 6.30). The prevalence of past month use among adults aged 26 or
older also was lower in the combined QFT-DR data (less than 0.05 percent) than in
the 2012 or 2013 comparison data (0.1 and 0.2 percent, respectively) (Table 6.33).

In addition, respondents in the main survey and the QFT were asked about lifetime use of
"any other" hallucinogen besides the ones they had seen in the preceding questions. Respondents
who reported use of other hallucinogens could specify use of up to five other hallucinogens that
they had ever used (subsequently referred to in this section as "OTHER, Specify" data). The
questions about ketamine, tryptamines, and Salvia divinorum had been included in the main
survey since 2006 because of evidence from their "OTHER, Specify" data that these could be
additional important substances for understanding hallucinogen use, especially among
adolescents and young adults aged 18 to 25 (Kroutil, Vorburger, & Aldworth, 2007).
Questions about these hallucinogens were moved from the special drugs module in the main
survey to the core hallucinogens module in the QFT and DR questionnaires. The effect of this
movement had been investigated for the QFT to assess whether it affected the reporting of
(a) use of these three hallucinogens because of their earlier placement in the QFT and DR;
and (b) use of "other" hallucinogens.
Estimates of lifetime use of ketamine, tryptamines, Salvia divinorum, and other
hallucinogens in the QFT-DR data and in the 2012 and 2013 comparison data are shown in
Table 6.34. As was observed for the QFT, estimates of lifetime use of other hallucinogens were
lower in the QFT than in the 2011 or 2012 comparison data for persons aged 12 or older, young
adults aged 18 to 25, and adults aged 26 or older. Among persons aged 12 or older, the estimate
of lifetime use of other hallucinogens was 0.7 percent for the QFT-DR, 1.5 percent for the 2012
comparison data, and 1.9 percent for the 2013 comparison data. In addition, the prevalence of
lifetime use of Salvia divinorum among adolescents aged 12 to 17 in the QFT-DR (1.8 percent)
was greater than that in the 2013 comparison data (0.6 percent) but was not significantly
different from the prevalence in the 2012 comparison data (1.1 percent).
At least for adults, these findings offer further indication that moving the additional
hallucinogen questions from the special drugs module to the core hallucinogens module in the
QFT appears to have affected the reporting for the residual "other" hallucinogen category.
Benefits of this change are that analysts have more information about the specific hallucinogens
that persons have used, whereas the category for other hallucinogens can be a "catchall" for a
wide variety of possible substances. Furthermore, this change could reduce the amount of data
review and coding of "OTHER, Specify" data that is needed for hallucinogens when the
redesigned questionnaire is fielded in 2015. An additional noteworthy finding from these
281

analyses is that moving the questions for these three hallucinogens from the special drugs
module to the core hallucinogens module did not appear to affect most lifetime estimates.
Although one difference between QFT-DR data and comparison data was observed for the
lifetime estimate of Salvia divinorum among adolescents, this was not a consistent pattern across
both sets of comparison data.
6.3.1.3

Inhalants

Questions about lifetime use of felt-tip pens and computer keyboard cleaner (air duster)
were added to the inhalants module for the QFT and DR because review of "OTHER, Specify"
data suggested that these could be other important inhalants that persons used to get high.
Furthermore, prior research has shown that NSDUH respondents are more likely to report use of
a substance if they are asked a direct "yes/no" question about the substance than if they need to
type in its name as part of "OTHER, Specify" questions (Kroutil, Vorburger, Aldworth, &
Colliver, 2010).
Although the only change to the inhalants module for the QFT was the addition of the
questions about lifetime use of these two inhalants, estimates of lifetime use of inhalants were
greater in the QFT than in the 2011 and 2012 comparison data for persons aged 12 or older,
adolescents aged 12 to 17, and adults aged 26 or older (Currivan et al., 2013). In turn, increased
reporting of lifetime use could translate to increased reporting of use in more recent periods.
•

The same finding of a higher prevalence of lifetime use of inhalants was observed for
the combined QFT-DR data relative to the 2012 and 2013 comparison data for
persons aged 12 or older, adolescents aged 12 to 17, and adults aged 26 or older
(Tables 6.22, 6.23, and 6.25). Among adolescents, the estimates of lifetime use of
inhalants were 9.4 percent for the combined QFT-DR data, 6.2 percent for the 2012
comparison data, and 4.9 percent for the 2013 comparison data.

•

As in the QFT, estimates of past year and past month use of inhalants did not differ
significantly between the QFT-DR data and comparison data for persons aged 12 or
older and adults aged 18 to 25 (Tables 6.26 and 6.28, respectively, for the past year
and Tables 6.30 and 6.32 for the past month). Estimates of past year use of inhalants
among adults aged 26 older also were similar in the QFT-DR and comparison data
(Table 6.29).

•

For adolescents aged 12 to 17, a similar pattern of estimates of past year use of
inhalants was observed in the QFT-DR data that was observed in the QFT, with the
QFT-DR estimate (3.5 percent) being higher than the estimate from the 2013
comparison data from quarters 3 and 4 (1.9 percent) but no significant difference
between the QFT-DR estimate and that from all four quarters in the 2012 comparison
data (2.3 percent) (Table 6.27).

•

Among adults aged 26 or older, the estimates of past month use were 0.1 percent in
the QFT-DR data and the 2012 comparison data (Table 6.33). This estimate for the
QFT-DR data was lower than the estimate for the 2013 comparison data (0.2 percent).

As for the hallucinogen data that were described previously, adding the questions to the
QFT about lifetime use of felt-tip pens or computer keyboard cleaner could affect the reporting
282

of the lifetime use of "other" inhalants. Also, computer keyboard cleaner is an aerosol product.
Therefore, asking about lifetime use of computer keyboard cleaner could affect estimates for
lifetime use of other aerosol sprays (i.e., other than spray paint in the main study and other than
spray paint or computer keyboard cleaner in the QFT).
Analysis of QFT data indicated that adolescents aged 12 to 17 who had ever inhaled
felt-tip pens appeared to comprise a substantial portion of the adolescent lifetime inhalant users.
Among young adults aged 18 to 25, those who had ever inhaled felt-tip pens comprised about
half of the lifetime users of inhalants (Currivan et al., 2013).
Estimates of lifetime use of felt-tip pens and computer keyboard cleaner were made for
the combined QFT-DR data. Estimates of lifetime use of other aerosol sprays and other inhalants
also were compared for the QFT-DR data and the data from 2012 and quarters 3 and 4 of 2013.
These estimates are shown in Table 6.35.
•

Consistent with the findings for the QFT, the QFT-DR estimate for lifetime use of
other aerosol sprays among adults aged 18 to 25 was lower than the estimates in the
2012 and 2013 comparison data. The QFT-DR estimate for other inhalants also was
lower than the comparison data, although the QFT-DR estimate was flagged for
suppression.

•

Among adolescents aged 12 to 17, the lifetime estimate of use of other inhalants in
the QFT-DR data than in the 2012 comparison data (0.7 vs. 1.4 percent) but was
similar to the estimate in the 2013 comparison data (0.9 percent). In the QFT, the
estimate of use of other inhalants among adolescents was not significantly different
from the estimates in the comparison data (Currivan et al., 2013).

To further understand the estimates in Table 6.35 and in anticipation of effects on
estimates of inhalant use in 2015, analyses of the QFT data were repeated for the combined
QFT-DR data that categorized users into two groups: (1) lifetime users of felt-tip pens or
computer keyboard cleaner (which could include persons who used other inhalants in addition to
these two); and (2) lifetime users of other inhalants, excluding use of felt-tip pens and computer
keyboard cleaner. Estimates for these two groups of lifetime users were made for persons aged
12 or older and for each age group.
Estimates of persons aged 12 or older and in each age group who reported past year use
also were made for these two groups of lifetime users. These estimates are shown in Table 6.40.
However, the estimates by age group were flagged for suppression. Statistical testing was not
conducted to identify any age group differences in the estimates presented in this table or
differences in the past year estimates. Also, the questions in the QFT and DR did not allow
determination of the specific inhalants that were used in the past year.
•

Percentages of persons who were lifetime users of felt-tip pens or computer keyboard
cleaners were 7.9 percent for 12 to 17 year olds, 7.4 percent for 18 to 25 year olds,
and 2.9 percent for adults aged 26 or older. Percentages of persons who were lifetime
users of other inhalants (but not these two) were 1.4 percent for 12 to 17 year olds,
2.9 percent for 18 to 25 year olds, and 8.5 percent for adults aged 26 or older.

283

•

Among persons aged 12 or older who were lifetime users of felt-tip pens or computer
keyboard cleaners, 12.4 percent used some inhalant in the past year. For lifetime users
of other inhalants excluding these two, 3.3 percent used inhalants in the past year.
These findings from the combined QFT-DR data were consistent with those from the
QFT data alone (Currivan et al., 2013).

•

An estimated 31.5 percent of adolescents who were lifetime users of felt-tip pens or
computer keyboard cleaners and 69.4 percent of lifetime users of other inhalants
excluding these two were past year users. Because of the low precision of these
estimates, however, these findings for adolescents need to be interpreted with caution.

Consistent with the findings from the QFT (Currivan et al., 2013), lifetime use of felt-tip
pens or computer keyboard cleaner appears to be more common among adolescents and young
adults than among adults aged 26 or older. As noted previously, however, age group differences
were not tested.
Taken together, these additional analyses further suggest that adding the questions about
lifetime use of felt-tip pens and computer keyboard cleaner may affect data trends in lifetime use
of inhalants once the new questionnaire is fielded for the 2015 survey, including trends for adults
aged 26 or older. Given that the estimates of past year use of inhalants were not significantly
different between either the QFT data or the combined QFT-DR data relative to the comparison
datasets that included four quarters of data, these findings are inconclusive regarding the
potential that this questionnaire change could affect trends for past year use of inhalants among
adolescents. However, estimates for past month use of inhalants appeared unlikely to be affected
by this change. Because NSDUH national reports tend to focus on estimates of past month use
(i.e., current use), inclusion of these two additional inhalants in the 2015 survey might have a
small impact on trends in the past month use of inhalants. Because long-term trends in lifetime
use and past year use of inhalants are typically included in annual NSDUH detailed tables and
reports of findings, it will be important for SAMHSA to consider how to handle any disruption
in the trends for lifetime use of inhalants in the detailed tables and national reporting for 2015.
6.3.1.4

Definitions of Illicit Drug Summary Measures

This section discusses the definitions for various summary measures of illicit drug use.
The standard definition of any illicit drug use captures use of any of one of nine categories of
illicit drugs: marijuana, cocaine (including crack), heroin, hallucinogens, inhalants, and misuse
of any one of four classes of psychotherapeutics (i.e., pain relievers, tranquilizers, stimulants,
and sedatives). The standard definition of any illicit drug use also includes use of
methamphetamine as reported in the noncore questions that were added in 2005 and 2006.
The standard definition also includes data from the new methamphetamine module in the QFT
and DR. In addition, because marijuana use has historically been the most prevalent form of
illicit drug use, a summary measure of illicit drug use other than marijuana is a standard NSDUH
measure that allows for the detection of trends in any illicit drug use that may be masked by
trends in marijuana use.
Because of extensive changes to questions asking about prescription drug misuse
(including the addition of a new methamphetamine module), the standard definitions of any
illicit drug use (and any illicit drug use other than marijuana) were modified for analyses
284

described in this chapter to exclude the use of methamphetamine and the misuse of any
prescription drugs. Alternate Definition 1 of any illicit drug use covers any use of marijuana,
cocaine (including crack), heroin, hallucinogens, and inhalants. Comparisons between the
combined QFT-DR data and corresponding comparison data for this measure are free of any
measurable differences in the use of methamphetamine and the misuse of psychotherapeutics.
Alternate Definition 3 for any illicit drug use includes use of marijuana, cocaine (including
crack), heroin, hallucinogens, inhalants, and methamphetamine. Similarly, the Alternate
Definition of any illicit drug use other than marijuana covers any use of cocaine (including
crack), heroin, hallucinogens, and inhalants.26
In addition, as noted in Sections 6.3.1.2 and 6.3.1.3, the modules for hallucinogens and
inhalants were modified by explicitly asking respondents about hallucinogens that had
previously been asked about in the special drugs module and asking direct questions about
specific additional inhalants. Thus, Alternate Definition 2 of any illicit drug use is similar to
Alternate Definition 1 except that the use of hallucinogens and inhalants is ignored. Similarly,
ignoring any reported use of hallucinogens and inhalants leads to a measure of any illicit drug
use other than marijuana that only contains two categories of drugs: cocaine (including crack)
and heroin.
Exhibit 6.1 summarizes these measures, which were all were constructed for the lifetime,
past year, and past month reporting periods. These estimates are shown in Tables 6.22 to 6.33
and Tables 6.42 to 6.45. Estimates from Tables 6.22 to 6.33 are discussed in this section and
focus on the effects on summary estimates of illicit drug use that could be attributed to changes
to the hallucinogens and inhalants modules (or other differences); separate from any effects on
these estimates that could be attributed to changes to questions for methamphetamine and
prescription drugs. Estimates from Tables 6.42 to 6.45 are discussed in Section 6.3.2 in the
context of a discussion of the changes to the questions for methamphetamine and prescription
drugs and the effects of these changes on estimates.
Exhibit 6.1

Substances Included in Definitions of Illicit Drugs and Illicit Drugs Other than
Marijuana

Substance
Marijuana
Cocaine (including
Crack)
Heroin
Hallucinogens
Inhalants
Prescription Drug
Misuse
Methamphetamine

Standard
Definition


Illicit Drugs
Alternate
Alternate
Definition
Definition
1
2



Illicit Drugs Other than Marijuana
Alternate
Definition
3


Standard
Definition

Alternate
Definition

Cocaine or
Heroin















 = Use of this substance is included in the summary measure.
26

285

6.3.1.5

Any Illicit Drug

In the QFT summary estimates of lifetime, past year, or past month use of illicit drugs
based on Alternate Definition 1 (i.e., including hallucinogens and inhalants but not
methamphetamine or prescription drugs) and Alternate Definition 2 (i.e., excluding
hallucinogens and inhalants in addition to methamphetamine and prescription drugs) did not
differ between the QFT and comparison data for persons aged 12 or older, adults aged 18 to 25,
or adults aged 26 or older. Estimates of lifetime and past year use among adolescents were
greater in the QFT than in some comparison datasets for Alternate Definition 1, but these
definitions did not remain for Alternate Definition 2 (Currivan et al., 2013).
•

The patterns of similar estimates between QFT and comparison data also were
observed in the QFT-DR and comparison data for lifetime, past year, and past month
use of illicit drugs among persons aged 12 or older (Tables 6.22, 6.26, and 6.30,
respectively) and adults aged 26 or older (Tables 6.25, 6.29, and 6.33, respectively)
for Alternate Definitions 1 and 2.

•

Among adolescents aged 12 to 17, unlike the pattern that was observed with just the
QFT data, the summary estimates of lifetime, past year, and past month use of illicit
drugs based on Alternate Definitions 1 and 2 did not differ between the combined
QFT-DR data and the 2012 or 2013 comparison data (Tables 6.23, 6.27, and 6.31).

•

Among young adults aged 18 to 25, estimates of lifetime use of illicit drugs based on
Alternate Definition 1 in the QFT-DR data (60.3 percent) were greater than the
estimates in the 2012 and 2013 comparison data (55.1 and 53.9 percent, respectively)
(Table 6.24). The estimate of lifetime use based on Alternate Definition 2 also was
greater for the QFT-DR data than for the 2013 comparison data (58.3 vs.
52.7 percent). However, estimates of past year and past month use of illicit drugs
based on Alternate Definitions 1 and 2 were not significantly different between the
QFT-DR and comparison data (Tables 6.28 and 6.32, respectively).

6.3.1.6

Illicit Drugs Other than Marijuana

As noted previously, marijuana historically has been the most commonly used illicit drug.
Consequently, marijuana is likely to drive the estimates of any illicit drug use. Changes to the
QFT and DR questions for hallucinogens and inhalants could have more of an effect on estimates
of use of illicit drugs other than marijuana. Significant differences in rates of use of cocaine,
crack, and heroin in the QFT-DR data relative to the comparison data (Section 6.3.1.1) also
could affect estimates for use of illicit drugs other than marijuana, independent of the changes to
the modules for hallucinogens and inhalants.
In the QFT, rates of lifetime use of illicit drugs other than marijuana based on the
Alternate Definition that included hallucinogens and inhalants but not methamphetamine or
prescription drugs were not significantly different between the QFT and comparison data for
persons aged 12 or older and adults aged 26 or older. However, some of these differences
approached statistical significance. In addition, some rates of lifetime use of illicit drugs other
than marijuana based on the Alternate Definition differed between the QFT and comparison data
(Currivan et al., 2013).

286

•

Unlike the pattern that was observed for the QFT, the prevalence of lifetime use of
illicit drugs other than marijuana among persons aged 12 or older based on the
Alternate Definition was greater in the QFT-DR data than in the 2013 comparison
data (25.1 vs. 22.4 percent) but was not significantly different from the prevalence of
22.7 percent based on the 2012 comparison data (Table 6.22).

•

Consistent with the findings in the QFT, the prevalence of lifetime use of illicit drugs
other than marijuana among adults aged 26 or older did not differ significantly
between the QFT-DR data and the comparison data (Table 6.25).

•

Among adolescents aged 12 to 17, the pattern of a higher rate of lifetime use of illicit
drugs other than marijuana in the QFT than in the comparison data based on the
Alternate Definition (i.e., including hallucinogens and inhalants) also was observed
for the QFT-DR data relative to the 2013 and 2014 comparison data (Table 6.23).
As also was observed for the QFT, the QFT-DR estimate of lifetime use of cocaine or
heroin among adolescents was lower than the corresponding estimates in the
comparison data.

•

Consistent with the findings in the QFT, the lifetime estimate for the Alternate
Definition of any illicit drugs other than marijuana among adults aged 18 to 25 in the
QFT-DR data was higher than that in the 2013 comparison data but was not
significantly different from the estimate in the 2012 comparison data (Table 6.24).
Lifetime estimates of use of cocaine or heroin among 18 to 25 year olds also did not
differ between the QFT-DR data and comparison data.

•

The lack of differences that were observed between the QFT and comparison data for
past year use of illicit drugs other than marijuana based on the Alternate Definition
and for cocaine or heroin also were observed between the QFT-DR and comparison
data for persons aged 12 or older (Table 6.26), adults aged 18 to 25 (Table 6.28), and
adults aged 26 or older (Table 6.29).

•

Among adolescents aged 12 to 17, the QFT-DR estimates of past year use based on
the Alternate Definition did not differ from those in the comparison data (Table 6.27).
In the QFT, the QFT estimate was greater than the estimate for the 2012 comparison
data, but it did not differ from the estimate for the 2011 comparison data (Currivan et
al., 2013).

•

Consistent with the findings in the QFT, estimates of past year use of cocaine or
heroin among adolescents aged 12 to 17 in the QFT-DR data were lower than those
from the 2012 and 2013 comparison data.

•

Unlike the findings in the QFT, the estimate of past month use of illicit drugs other
than marijuana based on the Alternate Definition and the estimate of past month use
of cocaine or heroin were lower in the QFT-DR data than in the comparison data for
persons aged 12 or older (Table 6.30) and adults aged 26 or older (Table 6.33).

Taken together, these findings suggest that changes to the modules for hallucinogens and
inhalants could affect trend data for the use of illicit drugs and illicit drugs other than marijuana
in 2015, especially for adolescents. Although the cocaine and heroin modules did not change for
the QFT and DR, some significant differences for aggregate estimates of use of cocaine or heroin
continued to be observed between the combined QFT-DR data and comparison data. As noted
287

previously, further examination of estimates for cocaine and heroin use in the 6-month data for
2015 will be useful for forecasting the final trends in 2015.
6.3.1.7

Cigarettes

Questions on cigarette use did not change for the QFT and DR relative to the main
survey, both in terms of content or placement as the very first set of substance use questions.
Therefore, the expectation for the QFT analysis was that the QFT estimates would be very
similar to the estimates for the comparison data 2011 comparison data and 2012 quarters 3 and 4
comparison data. Consistent with expectations, the QFT estimates for cigarette use were similar
to the estimates in the comparison data for lifetime, past year, and past month cigarette use
estimates and across all age groups (Currivan et al., 2013).
Lack of significant difference in the prevalence of cigarette use between the QFT-DR and
comparison data for 2012 and 2013 continued to be observed for most estimates in Tables 6.22
to 6.33. However, the prevalence of lifetime use of cigarettes for young adults aged 18 to 25 in
the QFT-DR data was greater than that in the 2013 comparison data (63.7 vs. 57.6 percent)
(Table 6.24). The QFT-DR estimate for lifetime use of cigarettes among 18 to 25 year olds was
not significantly different from the estimate of 60.2 percent in the 2012 comparison data.
Based on these findings, it seems likely that the trend for estimates of cigarette use will
not be disrupted in 2015. The anomalous finding of a difference in lifetime prevalence among
18 to 25 year olds between the QFT-DR and 2013 comparison data (but not between the QFTDR and 2012 comparison data) could be investigated further with the 6-month data from 2015.
6.3.1.8

Smokeless Tobacco

The smokeless tobacco questions underwent some changes for the QFT and DR. In the
main survey, respondents are asked separate sets of questions about their use of snuff and about
their use of chewing tobacco. In the QFT and DR, respondents were asked a single set of
questions about use of any smokeless tobacco product. Smokeless tobacco for the QFT and DR
also was defined somewhat differently than in the main survey and included the use of snuff, dip,
chewing tobacco, or "snus."27 The QFT analysis included investigation of whether these changes
could affect estimates of smokeless tobacco use.
In the QFT, lifetime estimates of smokeless tobacco use did not differ between the QFT
and comparison data for persons aged 12 or older or within any of the three age groups.
However, estimates of past year and past month use were greater in the QFT than in the
comparison data for persons aged 12 or older and adults aged 26 or older. For adolescents aged
12 to 17 and young adults aged 18 to 25, the estimates of past year and past month smokeless
tobacco use did not differ between the QFT and comparison data (Currivan et al., 2013).
•

Unlike the QFT findings, estimates of lifetime smokeless tobacco use were lower in
the QFT-DR data than in the 2012 comparison data for persons aged 12 or older
(Table 6.22) and adults aged 26 or older (Table 6.25). The estimate of lifetime use

"Snus" is a type of Swedish snuff. The question in the QFT is as follows: "The next questions are about
your use of 'smokeless' tobacco such as snuff, dip, chewing tobacco, or 'snus.'"

288

among adults aged 26 or older also was lower in the QFT-DR data than in the 2013
comparison data.
•

As in the QFT, estimates of past year and past month smokeless tobacco use in the
QFT-DR were not significantly different from those in the 2012 or 2013 comparison
data (Tables 6.26 to 6.29 for past year; Tables 6.30 to 6.33 for past month).

As was observed in the QFT, the higher estimates for lifetime use among adults aged
26 or older in the comparison data versus the combined QFT-DR data appear to be driving the
higher lifetime estimates for persons aged 12 or older in the comparison data. The lack of
significant differences for past year and past month estimates suggests that the higher rate of
lifetime use among adults aged 26 or older could be a function of the two opportunities in the
main survey to report lifetime use of smokeless tobacco (i.e., either snuff or chewing tobacco).
In contrast, it may be less of a challenge for some respondents to determine that they used some
type of "smokeless tobacco" in the past year or past month than to determine whether the product
specifically was "snuff" or "chewing tobacco." This explanation is consistent with main survey
data for the brand of snuff or chewing tobacco that respondents reported using most often in the
past 30 days. Specifically, respondents could specify a brand of snuff as some "other" brand of
"chewing tobacco" they used most often, or vice versa (Kroutil et al., 2012a). Although
respondent difficulties in distinguishing between snuff and chewing tobacco in the main survey
can be identified only for the past 30 days, they also are likely to be occurring for reports of these
types of smokeless tobacco use that occurred less recently than the past 30 days but within 12
months of the interview.
These findings also suggest that trends could be disrupted for lifetime use of smokeless
tobacco for all persons aged 12 or older and among adults aged 26 or older in 2015. Estimates of
lifetime, past year, and past month use of smokeless tobacco among persons aged 12 older and
by age group can be monitored in the 6-month data for 2015 to anticipate whether these changes
to the smokeless tobacco questions might affect trends in the final data for 2015.
6.3.1.9

Any Alcohol Use

Because the primary questions for lifetime, past year, and past month alcohol use were
not changed for the QFT instrument, QFT estimates for these items were expected to be similar
to those in the corresponding comparison data. Consistent with expectations, most QFT estimates
for alcohol use were similar to the 2011 comparison estimates and 2012 quarters 3 and 4
comparison estimates (Currivan et al., 2013).
Tables 6.22 through 6.25 provide estimates for lifetime alcohol use for all persons aged
12 or older, adolescents aged 12 to 17, young adults aged 18 to 25, and adults aged 26 or older,
respectively. Likewise, Tables 6.26 through 6.29 provide estimates for past year alcohol use, and
Tables 6.30 through 6.33 provide estimates for any past month alcohol use. In addition,
Tables 6.36 to 6.38 provide estimates of lifetime, past year, and past month alcohol use by
gender.
•

Similar to the QFT, estimates of lifetime and past year use of alcohol were similar
between the QFT-DR and comparison data for persons aged 12 or older and all age

289

groups. Estimates of past month use also were similar between the QFT-DR data and
comparison data for adolescents aged 12 to 17 and adults aged 18 to 25.
•

Unlike the QFT findings, the estimate of past month use of alcohol among persons
aged 12 or older was lower in the QFT-DR data (50.7 percent) than in the 2012 or
2013 comparison data (54.1 and 53.9 percent, respectively) (Table 6.30). Similarly,
the estimate of past month use of alcohol among adults aged 26 or older was lower in
the QFT-DR data (53.5 percent) than in the 2012 or 2013 comparison data (57.5 and
57.6 percent, respectively) (Table 6.33).

•

Among adults aged 26 or older, some estimates of past year and past month alcohol
use among males or females also were lower in the QFT-DR than in the comparison
data (Tables 6.37 and 6.38, respectively). For males aged 26 or older, the estimates of
past year and past month alcohol use were lower in the QFT-DR than in the 2013
comparison data but were not significantly different from the rates in the 2012
comparison data. For females in this age group, the rate of past month alcohol use
was lower in the QFT-DR than in the 2012 comparison data (48.7 vs. 52.9 percent)
but was not significantly different from the rate of 51.6 percent for the 2013
comparison data.

Thus, additional analysis of combined QFT and DR data continued to show a lack of
significant differences in most rates of any alcohol use between the QFT-DR and comparison
data, which suggests that trends in any alcohol use generally will be maintained in 2015.
Furthermore, the lack of significant differences in estimates of past month alcohol use among
adolescents aged 12 to 17 between QFT-DR and comparison data suggests that the significant
difference that was observed between the QFT and 2011 comparison data (Currivan et al., 2013)
may have been an anomaly. However, examination of estimates of past month alcohol use
among adults aged 26 or older in the 6-month data for 2015 could be important for forecasting
trends in this age group and this age group's contribution to trends for the population aged 12 or
older in 2015.
6.3.1.10

Past Month Binge Alcohol Use

One notable change in the QFT and DR instrument involved the definition of binge
alcohol use. In the main survey, binge alcohol use is defined as drinking five or more drinks on
one occasion for both male and female respondents. In the QFT and DR, the definition of binge
alcohol use was changed to drinking four or more drinks on one occasion for female respondents.
This change was investigated for the QFT because it had the potential to increase reports of
binge alcohol use by lowering the threshold for the minimum number of drinks for females.
In the QFT, there were no significant differences in estimates of binge alcohol use in the
past month regardless of gender for persons aged 12 or older or in any of the three age groups.
However, differences approached statistical significance for adults aged 26 or older (Currivan et
al., 2013).
Table 6.39 contains two sets of estimates of binge alcohol use by age group and gender.
The first set of estimates is based only on core data. As noted previously, binge alcohol use in the
comparison data was defined for males and females as drinking five or more drinks on the same

290

occasion on at least 1 day in the past 30 days based on their reports in the core alcohol module.
For the QFT and DR, binge alcohol use was defined for males in the same manner as in the
comparison data. For females, binge alcohol use in the QFT and DR was defined as drinking four
or more drinks on the same occasion based on their reports in the core alcohol module.
Table 6.39 also contains core-plus-noncore (CPN) estimates for the 2012 and 2013
comparison data. In addition to reports of consumption of five or more drinks on a single
occasion on at least 1 day in the past 30 days, these CPN measures took into account females'
reports of usual consumption of four or more drinks on the days that they drank alcohol in the
past 30 days (from the core alcohol module) or their consumption of four or more drinks on the
same occasion on at least 1 day in the past 30 days (from the noncore consumption of alcohol
module). These CPN measures were created to further gauge the potential effects on estimates of
binge alcohol use because of the change to the threshold for females. For males in the
comparison data, the CPN measure was the same as the measure based only on core data.
Estimates for the QFT-DR data based on core alcohol use data (i.e., including the "four or more"
criterion for females) are repeated for comparison with the CPN estimates.
•

Consistent with the findings from the QFT, most rates of binge alcohol use in the past
month based on core data were similar between the QFT-DR and comparison data.
In particular, rates among persons aged 12 or older were similar for all persons and
for males and females.

•

Among 12 to 17 year olds, the QFT-DR estimate for males was lower than the coreonly estimate in the 2012 comparison data (4.4 vs. 6.8 percent) but was not
significantly different from the estimate in the 2013 comparison data (6.3 percent).
The QFT-DR estimates for all adolescents aged 12 to 17 and males in this age group
also were lower than the CPN estimates in the 2012 comparison data. For adolescent
females, however, neither the core-only nor the CPN estimates in the comparison data
were significantly different from the QFT-DR estimates.

•

For females aged 26 or older, the QFT-DR estimate was greater than the core-only
estimate in the 2012 comparison data (17.0 vs. 13.8 percent) but was not significantly
different from the estimate in the 2013 comparison data (14.7 percent). The CPN
estimates for the 2012 and 2013 comparison data that included the lower threshold of
four or more drinks were not significantly different from the QFT-DR estimate.

Consistent with the findings from the QFT, these findings suggest that lowering the
threshold for binge alcohol use among females to consumption of four or more drinks on an
occasion may not affect the trends in binge alcohol use among all persons aged 12 or older or
among all persons within most age groups (i.e., regardless of gender). Females aged 26 or older
may provide an exception to this general conclusion. The higher estimate of binge alcohol use in
the QFT-DR data relative to the core-only estimate for the 2012 comparison data suggests that
the lower threshold for binge alcohol use among females may be more important for estimating
binge alcohol use among females in this age group than it is for females in other age groups.
These findings also suggest that the planned change in the definition of binge alcohol use among
females in 2015 may affect trends for females aged 26 or older.

291

6.3.2

Methamphetamine and Prescription Drug Items

As noted in Chapters 2 and 3, the following changes to the questions for
methamphetamine and prescription drugs were made for the QFT and the DR:
•

A new methamphetamine module was added instead of questions about
methamphetamine use being included as part of the stimulants module.

•

The definition, approach, and terminology for measuring misuse of prescription drugs
were revised.

•

Modules were added that asked respondents about any use of pain relievers,
tranquilizers, stimulants, and sedatives, as opposed to just misuse.

•

The focus of the prescription drug modules was on a 12-month reference period
rather than the lifetime reference period used in the current questionnaire.

•

Electronic images of prescription drugs replaced the current hard-copy pill card
versions, and the images included more than just pills.

•

Questions about discontinued prescription drugs were deleted, and questions were
added for other prescription drugs not previously included in the questionnaire.

•

Questions about prescription drugs that were included in supplemental sections of the
current questionnaire were moved to the appropriate prescription drug module.

These changes are planned for implementation in the redesigned NSDUH questionnaire in 2015
and are likely to affect estimates of methamphetamine use and misuse of prescription drugs
starting in 2015.
This section presents findings on methamphetamine use and prescription drug misuse
from the comparison data for 2012 and quarters 3 and 4 of 2013 and from the combined QFT and
DR data for non-Hispanic English-language respondents. Tables 6.42 to 6.44 present estimates
for these measures for the lifetime, past year, and past month periods.
6.3.2.1

Estimates for Methamphetamine Items

misuse of prescription stimulants. Therefore, the methamphetamine use measures for the
comparison data (i.e., 2012 and quarters 3 and 4 of 2013) were based on reports of
methamphetamine use in the core stimulants module plus reports of use from the supplemental
(or noncore) special drugs module (i.e., core plus noncore, or CPN). However, additional
respondents who reported lifetime use of methamphetamine in the special drugs module were
included in the CPN measures only if their reason for not previously reporting methamphetamine
use was that they did not think of methamphetamine as a prescription drug; respondents who
reported use in the special drugs module were not counted as users if they reported that they did
not previously report methamphetamine use because they "made a mistake" when answering the
methamphetamine questions in the stimulants module or for reasons other than not thinking of
this as a prescription drug (Kroutil, Handley, Bradshaw, Chien, & Felts, 2012b). Consequently,
these CPN measures of methamphetamine use in the comparison data still might underestimate
the prevalence of use.
For the QFT and DR, the methamphetamine use measures were based only on data from
the new methamphetamine module in the core section of the QFT questionnaire. Although QFT
and DR respondents did not have the same multiple opportunities to report methamphetamine
use as in the comparison data, there also was no question (and no need) to check for the reason
that some respondents did not previously report methamphetamine use.
Findings from the QFT analysis indicated that the estimate of lifetime methamphetamine
use among persons aged 12 or older was greater in the QFT than in the 2012 comparison data.
The estimate in the 2011 comparison data was not significantly different from the QFT estimate
but approached statistical significance. Estimates of lifetime use within the age groups were not
significantly different between the QFT and comparison data, although some differences
approached statistical significance. Estimates of methamphetamine use in the past year among
persons aged 12 or older and in each of the three age groups also did not differ significantly
between the QFT and comparison data, although estimates within some age groups approached
significance (Currivan et al., 2013).
•

Consistent with the findings for the QFT, the estimate of lifetime use of
methamphetamine among persons aged 12 or older in the QFT-DR data (7.4 percent)
was greater than the respective estimates of 4.9 and 4.8 percent in the 2012 and 2013
comparison data (Table 6.42).

•

Estimates of past year use of methamphetamine among persons aged 12 or older were
not significantly different between the QFT-DR data and the comparison data
(Table 6.43). This finding was consistent with the QFT findings. The estimates of
past year use were 0.7 percent in the QFT-DR data, 0.4 percent in the 2012
comparison data, and 0.5 percent in the 2013 comparison data.

•

Estimates of methamphetamine use in the past month among persons aged 12 or older
were not significantly different between the QFT-DR data and the comparison data
(Table 6.44). Again, this finding was consistent with the QFT findings. The estimates
of past month use were 0.4 percent in the QFT-DR data, 0.1 percent in the 2012
comparison data, and 0.2 percent in the 2013 comparison data.

293

6.3.2.2

Measurement Issues for Prescription Drug Misuse

The shift in focus of questions about the misuse of specific prescription drugs from the
lifetime reference period in the current questionnaire to a 12-month reference period and the
deletion of questions about discontinued prescription drugs in the QFT and DR could decrease
the estimates of lifetime misuse in these surveys relative to the comparison data. Comparison
data respondents had multiple opportunities to report lifetime misuse of prescription drugs,
including misuse of drugs that currently are no longer available by prescription in the United
States. In contrast, QFT and DR respondents who did not report past year use or misuse of any
prescription drugs in a given category were asked only a single question about misuse of any
prescription drugs in that category in their lifetime. For pain relievers, for example, this question
was worded as follows: "Have you ever, even once, used any prescription pain reliever in any
way a doctor did not direct you to use it?" However, respondents in the QFT and DR were not
given any additional cues or aids to remind them of the types of drugs that qualify as
"prescription pain relievers." Therefore, respondents in the QFT and DR would need to depend
largely on their ability to remember the examples of specific pain relievers that they saw in the
screener section. In light of regular changes in the prescription drug market in the United States,
QFT and DR respondents also would need to consider not only lifetime misuse of prescription
drugs that currently are available, but also any past misuse of prescription drugs that previously
were but no longer are available. Because of the structure and content of the questions in the
QFT and DR, therefore, respondents who last misused prescription drugs more than 12 months
ago might underreport their misuse.
Conversely, the expansion of the number of questions in the QFT and DR about past year
misuse of specific prescription drugs could be expected to increase the estimates of past year
misuse relative to estimates in the main survey. For example, respondents in the QFT and DR
would be classified as having misused prescription pain relievers in the past 12 months if they
reported misuse in that period of any of 40 possible pain relievers, including "any other" pain
reliever. In the main survey, respondents are defined as having misused pain relievers in the past
year principally through their response to the question, "How long has it been since you last used
any prescription pain reliever that was not prescribed for you or that you took only for the
experience or feeling it caused?" Only those respondents in the main survey who reported
lifetime misuse of the pain reliever OxyContin® have an additional opportunity to report past
year misuse through a corresponding question about the last time they used OxyContin® that was
not prescribed for them or that they took only for the experience or feeling the drug caused.
As noted previously, the definition of misuse also was changed for the QFT. The
definition of misuse in the main survey combines a behavior (use of a prescription drug that was
not prescribed for the respondent) and a motivation for misuse (use of a prescription drug only
for the experience or feeling that it caused). In the QFT and DR, the definition of misuse "in any
way a doctor did not direct you to use it" focuses on behaviors. The following examples were
given to QFT and DR respondents for behaviors that constitute misuse:
•

(use) without a prescription of your own;

•

(use) in greater amounts, more often, or longer than you were told to take it; or

•

(use) in any other way a doctor did not direct you to use it.

294

Especially for misuse of prescription pain relievers, alerting respondents in the QFT and DR that
overuse of prescribed medication (e.g., use in greater amounts or more often than prescribed)
constitutes misuse also could increase reporting of misuse.
6.3.2.3

Misuse of Any Prescription Psychotherapeutic Drug

Consistent with the hypotheses noted previously, findings from the QFT analysis
generally indicated that estimates of lifetime misuse of any prescription psychotherapeutic drug
(i.e., pain relievers, tranquilizers, stimulants, or sedatives) were lower in the QFT data than in the
comparison data. In contrast, estimates of past year misuse were greater in the QFT data than in
the comparison data. Estimates of misuse in the past month in the QFT data also were in the
direction of being greater than those in the comparison data; some of these differences
approached but did not attain statistical significance because of the smaller QFT sample size
(Currivan et al., 2013).
•

In accordance with the QFT findings, the estimate of lifetime misuse of any
prescription psychotherapeutic drug among persons aged 12 or older was lower in the
QFT-DR data (14.8 percent) than in the comparison data for 2012 (21.3 percent) or
2013 (20.6 percent) (Table 6.42).

•

The estimate of past year misuse of prescription psychotherapeutic drugs among
persons aged 12 or older (8.0 percent) was greater than the corresponding estimates
in the comparison data for 2012 and 2013 (6.3 and 5.7 percent, respectively)
(Table 6.43). This finding also was consistent with the QFT findings.

•

Estimates of misuse of any psychotherapeutic drugs in the past month among persons
aged 12 or older were not significantly different between the QFT-DR data and the
comparison data (Table 6.44). Again, this finding was consistent with the QFT
findings. The estimates of past month misuse were 2.6 percent in the QFT-DR and
2012 comparison data and 2.3 percent in the 2013 comparison data.
6.3.2.4

Pain Relievers

In the QFT analysis, estimates for misuse of prescription pain relievers followed the same
general pattern as misuse of any prescription drug, with some lower estimates of lifetime misuse
in the QFT than in some comparison datasets and generally higher estimates of past year misuse
in the QFT than in the comparison data (Currivan et al., 2013). Highlights are presented in the
remainder of this section for analyses using combined QFT-DR data and comparisons with data
from 2012 and quarters 3 and 4 of 2013.
•

As for any prescription psychotherapeutic drug, the estimate of lifetime misuse of
pain relievers for persons aged 12 or older in the QFT-DR data (12.0 percent) was
lower than the estimates of 14.4 percent in the 2012 comparison data and 13.8 percent
in the 2013 comparison data (Table 6.42).

•

The estimate of past year misuse of pain relievers among persons aged 12 or older
was greater for the QFT-DR data than for the 2013 comparison data (5.7 vs.
4.1 percent) (Table 6.43). The past year estimate for the QFT-DR data was not
significantly different from the estimate for the 2012 comparison data (4.7 percent).
295

•

Estimates of past year misuse of OxyContin® were available for the QFT-DR data
and the comparison data (Table 6.43). These estimates were similar in these datasets
(0.9 percent in the QFT-DR data and 0.6 percent in both the 2012 and 2013
comparison data).

•

Estimates of misuse of pain relievers in the past month among persons aged 12 or
older were 1.7 percent in the QFT-DR and 2013 comparison data and 1.9 percent in
the 2012 comparison data (Table 6.44).
6.3.2.5

Tranquilizers

•

Consistent with the QFT findings and findings for any psychotherapeutic drug and
pain relievers, the estimate of lifetime misuse of tranquilizers among persons aged
12 or older was lower in the QFT-DR data (5.6 percent) than in the comparison data
for 2012 (9.4 percent) or 2013 (9.3 percent) (Table 6.42).

•

The estimate of past year misuse of tranquilizers among persons aged 12 or older
(2.7 percent) was similar to the corresponding estimates in the comparison data for
2012 and 2013 (2.3 and 2.0 percent, respectively) (Table 6.43). This finding was
consistent with the QFT findings, but it differed from the findings for any
psychotherapeutic drug and pain relievers.

•

Estimates of the misuse of tranquilizers in the past month among persons aged 12 or
older were similar in the QFT-DR data and the comparison data (Table 6.44). Again,
this finding was consistent with the QFT findings. The estimates of past month
misuse were 0.8 percent in the QFT-DR and 2012 comparison data and 0.6 percent in
the 2013 comparison data.
6.3.2.6

Sedatives

Unlike the general pattern that was observed in the QFT analysis for other prescription
drugs, the estimates of lifetime misuse of sedatives generally were similar between the QFT and
the two comparison datasets. Consistent with the findings for other prescription drugs, however,
the estimates of past year sedative misuse in the QFT generally were greater than corresponding
estimates in the comparison data (Currivan et al., 2013).
•

As was observed for the QFT, the estimate of lifetime misuse of sedatives among
persons aged 12 or older in the QFT-DR data (3.4 percent) was similar to the
estimates in the 2012 and 2013 comparison data (3.3 and 3.0 percent, respectively)
(Table 6.42).

•

The estimate of past year sedative misuse in the QFT-DR data for persons aged 12 or
older (0.8 percent) was greater than corresponding estimates in the 2012 and 2013
comparison data (0.2 and 0.1 percent, respectively) (Table 6.43). This finding was
consistent with findings from the QFT and also was consistent with findings from the
QFT-DR data for pain relievers and tranquilizers.

296

•

Estimates of past month sedative misuse among persons aged 12 or older were similar
in the QFT-DR and comparison data (Table 6.44). Estimates ranged from less than
0.05 percent in the 2013 comparison data to 0.2 percent in the QFT-DR data.

However, the estimates for sedative misuse in the comparison data that were described
previously were based only on reports of misuse from the core module. These estimates did not
include data on the misuse of the sedative Ambien® that were in the supplemental (i.e., noncore)
special drugs module. In an analysis of data from the 2006 NSDUH, when questions about
Ambien® were added to the special drugs module, inclusion of these data on Ambien® misuse
had a major impact on estimates of sedative misuse compared with estimates based on core
sedative data alone (Kroutil et al., 2007). Ambien® is one of the specific prescription drugs
included in the core sedatives module for the QFT and DR. Therefore, CPN measures of sedative
misuse that included data on Ambien® misuse also were created for the 2012 and 2013
comparison data. These data are included in Tables 6.48 to 6.50.
Similar analyses had been conducted previously for the QFT to compare estimates of
sedative misuse from the QFT data with CPN estimates from the corresponding comparison
datasets. Inclusion of data for Ambien® raised the CPN estimates of lifetime misuse of sedatives
in the comparison data to the point that most estimates were greater than the QFT estimates.
Ambien® data in the CPN estimates of past year misuse also appeared to erase or reverse the
direction of the differences in prevalence between the QFT and comparison data that were
observed for comparison data estimates based only on core sedatives module data. However,
inclusion of Ambien® data in the CPN estimates had little apparent effect on estimates of past
month sedative misuse or differences between the QFT and comparison data for past month
misuse (Currivan et al., 2013).
•

Consistent with the QFT analysis, CPN estimates of lifetime misuse of sedatives for
the 2012 and 2013 comparison data were greater than the QFT-DR estimates for
persons aged 12 or older and for all age groups (Table 6.48).

•

For past year misuse of sedatives, the CPN estimates that included Ambien® for the
2012 and 2013 comparison data were similar to the QFT-DR estimates for persons
aged 12 or older and for all age groups (Table 6.49). Again, this finding was
consistent with the findings from the QFT analysis.

•

For past month misuse of sedatives, estimates for the QFT-DR and comparison data
were similar for both core-only and CPN estimates in the comparison data
(Table 6.50). Among adults aged 18 to 25, however, the QFT-DR estimate was lower
than the CPN estimate for the 2012 comparison data (0.1 vs. 0.4 percent).

These findings further underscore the conclusion that was reached from the QFT analysis about
the likely importance of including questions about Ambien® for estimating sedative misuse.
6.3.2.7

Stimulants

Because the estimates of methamphetamine use in the 2012 and 2013 comparison data
were based on CPN measures of methamphetamine use (see Section 6.3.2.1), the corresponding
estimates of any stimulant misuse in the comparison data included these CPN methamphetamine

297

use data. These CPN measures are referred to as the "Standard Definition" of stimulant misuse in
Tables 6.42 to 6.44. To produce estimates of stimulant misuse for the combined QFT and DR
data that were as analogous as possible to these estimates in the comparison data, the "standard
definition" estimates of stimulant misuse were based on data from the core methamphetamine
and prescription stimulants modules. A "QFT definition" of stimulant misuse also was created
for the QFT-DR based on data in the core stimulants module but not including data on
methamphetamine use. Because it is not possible to disentangle methamphetamine use from
misuse of other stimulants in the comparison data, however, this QFT definition measure was not
created for the comparison data.
In the QFT analysis, estimates of lifetime stimulant misuse based on the standard
definition including methamphetamine were similar between the QFT and comparison data.
Some estimates of past year and past month use based on the standard definition were greater in
the QFT than in the comparison data (Currivan et al., 2013).
•

Consistent with the QFT findings, estimates of lifetime misuse of stimulants among
persons aged 12 or older based on the standard definition that included
methamphetamine were similar between the QFT-DR data and the comparison data
(Table 6.42). Estimates of lifetime misuse of stimulants for persons aged 12 or older
based on the standard definition were 10.2 percent for the QFT-DR and 8.7 percent in
both the 2012 and 2013 comparison data.

•

The estimate of past year misuse of stimulants for persons aged 12 or older based on
the standard definition was greater in the QFT-DR data (2.5 percent) than in the 2012
or 2013 comparison data (1.3 and 1.5 percent, respectively) (Table 6.43).

•

Estimates of past month misuse of stimulants among persons aged 12 or older based
on the standard definition were 0.7 percent for the QFT-DR data, 0.5 percent for the
2012 comparison data, and 0.5 percent for the 2013 comparison data (Table 6.44).
The estimate for the QFT-DR data was not significantly different from the estimates
in the comparison data.

For the QFT-DR data, statistical tests were not conducted between estimates of stimulant
misuse based on the standard definition that included methamphetamine and the QFT definition
that did not include methamphetamine. Nevertheless, these data provide some indication of the
potential effect if methamphetamine use is no longer included in estimates of stimulant misuse in
2015 and beyond.
•

Estimates of lifetime stimulant misuse in the QFT-DR data for persons aged 12 or
older were 10.2 percent for the standard definition that included methamphetamine
and 4.3 percent for the definition that did not include methamphetamine (Table 6.42).

•

Among persons aged 12 or older, the standard definition estimate of past year
stimulant misuse for the QFT-DR data was 2.5 percent, and the estimate without
methamphetamine was 1.9 percent (Table 6.43).

•

The standard definition estimate in the QFT-DR data for past month stimulant misuse
among persons aged 12 or older was 0.7 percent, and estimate based on the definition
that did not include methamphetamine was 0.4 percent (Table 6.44).

298

As was the case for sedatives, the standard definition estimates for stimulant misuse in
the comparison data that were described previously did not include data on the misuse of the
stimulant Adderall® from the special drugs module. The impact of the Adderall® data on
estimates of nonmedical stimulant use in the 2006 NSDUH was particularly notable for
adolescents aged 12 to 17 and young adults aged 18 to 25 (Kroutil et al., 2007). Adderall® is one
of the specific prescription drugs that was included in the core stimulants module for the QFT
and DR. Therefore, measures of stimulant misuse based on the standard definition plus noncore
data on Adderall® misuse were created for the 2012 and 2013 comparison data. These data are
included in Tables 6.45 to 6.47.
Similar analyses had been conducted previously for the QFT to compare estimates of
stimulant misuse from the QFT data with CPN estimates from the corresponding comparison
datasets. Inclusion of data for Adderall® had relatively little effect on whether differences in
lifetime stimulant misuse between the QFT and comparison data were statistically significant.
For persons aged 12 or older and young adults aged 18 to 25, however, inclusion of data for
Adderall® appeared to erase the differences in the prevalence of past year misuse that were
observed between the QFT and comparison data for the standard definition estimates. Among
persons aged 12 or older, the comparison data estimates for 2011 and 2012 that included noncore
Adderall® data also were similar to the standard definition estimate in the QFT (Currivan et al.,
2013).
•

Consistent with the QFT analysis, CPN estimates of lifetime misuse of stimulants
were not significantly different for the 2012 and 2013 comparison data were not
significantly different from the QFT-DR estimate based on the standard definition for
persons aged 12 or older, young adults aged 18 to 25, and adults aged 26 or older
(Table 6.45). Among adolescents aged 12 to 17, the CPN estimates of lifetime
stimulant misuse for the comparison data that included Adderall® were greater than
the standard definition estimate from the QFT-DR data.

•

For past year misuse of stimulants, the CPN estimates that included Adderall® for the
2012 and 2013 comparison data were similar to the QFT-DR estimates for persons
aged 12 or older, adolescents aged 12 to 17, and adults aged 26 or older (Table 6.46).
Among young adults aged 18 to 25, however, the CPN estimates continued to be
lower than the QFT-DR estimate based on the standard definition.

•

For past month misuse of stimulants, estimates for the QFT-DR and comparison data
were similar for both core-only and CPN estimates in the comparison data
(Table 6.47).

In the QFT analysis, including noncore Adderall® data in the CPN estimates of past year
misuse erased the difference between estimates of past year misuse of stimulants in the QFT and
comparison data for young adults aged 18 to 25. In the combined QFT and DR data, however,
the estimate for young adults based on the standard definition (i.e., including methamphetamine)
continued to be greater than the CPN estimates in the comparison data. For persons aged 12 or
older and for those in other age groups, the CPN estimates for the comparison data that included
Adderall® were similar to the standard definition estimates for the QFT-DR data. Although
including noncore Adderall® data did not affect differences in estimates between the QFT-DR
and comparison data in this analysis, the findings for all persons aged 12 or older and those for
299

other age groups continue to underscore the likely importance of including questions about
Adderall® for estimating misuse of prescription stimulants.
6.3.2.8

Effects of Methamphetamine and Prescription Drugs on Illicit Drug
Use Estimates

As noted in Section 6.3.1.4, the measures of use of any illicit drug and illicit drugs other
than marijuana in current published NSDUH estimates include use of methamphetamine and
misuse of prescription drugs. The changes to the methamphetamine and prescription drug
questions that were previously summarized in Section 6.3.2 for the QFT-DR data (and, by
extension, for the redesigned questionnaire in 2015) also could affect estimates for these other
summary measures of illicit drug use.
In addition to the alternate definitions that were described previously in Section 6.3.1.4, a
third alternate definition for any illicit drug use was developed that included methamphetamine
but did not include prescription drugs (subsequently referred to as Alternate Definition 3).
In addition, measures of use of illicit drugs and illicit drugs other than marijuana were created
based on the standard NSDUH definitions that included both methamphetamine and prescription
drugs. Estimates based on Alternate Definition 3 for illicit drug use and the standard definitions
are presented in this section and in Tables 6.42 to 6.45.
In the QFT analysis, estimates of lifetime use were not significantly different between the
QFT and the comparison data for persons aged 12 or older, adults aged 18 to 25, and adults aged
26 or older for the illicit drug Alternate Definition 3 or for the standard definitions of use of
illicit drugs or illicit drugs other than marijuana. As for the lifetime period, estimates of past year
use of illicit drugs based on the standard definition or Alternate Definition 3 were not
significantly different between the QFT and comparison data for persons aged 12 or older.
For adolescents aged 12 to 17, however, the standard definition estimate and the estimate for
Alternate Definition 3 differed between the QFT and 2012 comparison data. The estimates of use
of illicit drugs other than marijuana in the past year based on the standard definition were greater
in the QFT than in the 2011 or 2012 comparison data for persons aged 12 or older and young
adults aged 18 to 25. Most estimates of past month use of illicit drugs or illicit drugs other than
marijuana did not differ significantly between the QFT and comparison data, regardless of the
definitions (Currivan et al., 2013).
•

Consistent with the QFT findings, estimates of lifetime use of illicit drugs and illicit
drugs other than marijuana that included methamphetamine and prescription drugs
were similar between the QFT-DR and comparison data for persons aged 12 or older
(Table 6.42).

•

Estimates of past year use of illicit drugs among persons aged 12 or older were
similar between the QFT-DR and comparison data (Table 6.43). Consistent with the
QFT findings, however, the standard definition estimate of use of illicit drugs other
than marijuana for persons aged 12 or older was greater in the QFT-DR data
(9.8 percent) than in the 2012 or 2013 comparison data (8.1 and 7.4 percent,
respectively).

300

•

As was observed for the QFT, estimates of past month use of illicit drugs and illicit
drugs other than marijuana that included methamphetamine and prescription drugs
were similar between the QFT-DR and comparison data for persons aged 12 or older
(Table 6.44).
6.3.2.9

Methamphetamine, Prescription Drug, and Illicit Drug Estimation
Issues to Consider for the 2015 Redesign

For methamphetamine, the past year and past month estimates among persons aged 12 or
older did not differ significantly between the QFT-DR and comparison data, but the estimate of
lifetime use for persons aged 12 or older was greater in the QFT-DR data than in the comparison
data. If the prevalence of lifetime methamphetamine use in 2015 is higher than in recent years
for persons aged 12 or older or within different age groups because of changes to the
questionnaire in 2015, SAMHSA will need to decide how to handle the reporting of trends in
lifetime use, as noted previously in Section 6.3.1. One option would be not to report trend data
for lifetime methamphetamine use between 2015 and earlier years or to discontinue the reporting
of lifetime trend data for methamphetamine altogether from 2015 onward. Alternatively,
SAMHSA could start a new baseline for lifetime methamphetamine use beginning in 2015.
Other, more sophisticated options could involve statistical procedures to adjust the trend data for
2002 to 2014. Although data on trends in lifetime prevalence may be of interest for examining
historical changes in the popularity of different drugs, data on trends in the prevalence of
methamphetamine use in the past year and past month are likely to be of more importance to
policymakers, the public health sector, the criminal justice sector, and others because of the
demands that methamphetamine users may place on the criminal justice system, the health care
delivery system (including substance abuse treatment), and systems for providing social services
(including services to dependents of adult substance users).
For prescription drugs, the general findings of lower estimates of lifetime misuse of
prescription drugs but higher past year estimates in both the QFT and combined QFT-DR data
relative to corresponding comparison datasets are expected, given the changes to the prescription
drug questions for the QFT. The structure of the current questionnaire provides respondents with
multiple opportunities to report lifetime misuse of specific prescription drugs but less
opportunity to report past year misuse. This situation was reversed for the questionnaire that was
used in the QFT and DR, with respondents having more opportunity to report past year misuse of
specific prescription drugs and limited opportunity to report misuse of any prescription drugs
that occurred more than 12 months prior to the interview—including misuse of prescription
drugs that are no longer available by prescription in the United States.
These findings from both the QFT and combined QFT-DR data for prescription drugs
also support the conclusion to start a new baseline in 2015 for trends in prescription drug misuse.
In addition, it may be useful for SAMHSA to consider whether to discontinue reporting trend
data for lifetime misuse of prescription drugs after 2014 because of questions about the accuracy
of respondent self-reports of misuse of prescription drugs more than 12 months prior to the
interview.
For summary measures of use of illicit drugs, many estimates of the use of illicit drugs or
the use of illicit drugs other than marijuana were not significantly different between the QFT-DR

301

and comparison data when data for methamphetamine or prescription drugs (or both) were
included in the estimates. However, the estimate for past year use of illicit drugs other than
marijuana that included methamphetamine and prescription drugs was affected for persons aged
12 or older. Given that NSDUH data consistently show the prevalence of misuse of prescription
drugs to be second only to marijuana among illicit drugs, the higher estimate of use of illicit
drugs other than marijuana based on the standard definition in the QFT-DR data than in the
comparison data can be explained by the higher estimate of misuse of prescription drugs in the
QFT-DR data. However, changes to the methamphetamine and prescription drug use questions
were not the only changes made to the questionnaire for the QFT and DR. As discussed
previously, changes also were made to the hallucinogens and inhalants modules in the QFT and
DR questionnaires that could affect estimates of the use of illicit drugs and illicit drugs other than
marijuana. Analysis of 6-month data for 2015 is likely to be useful for assisting SAMHSA in
deciding how to create these summary illicit drug use measures in 2015 and how to report trends
for these measures.
6.3.3

Selected Noncore Items

This section presents and discusses estimates for selected noncore items, including
findings from the DR for results flagged from the QFT report as warranting further investigation
or QFT results that were noted as a potential preview of findings in 2015. Results from some
tables might be relevant to both topics. This section also highlights any new noteworthy findings
for selected noncore items from the DR that were not identified in the QFT and briefly notes any
similar findings between the QFT and DR data.
6.3.3.1

Dependence and Abuse

Table 6.51 presents estimates for substance dependence or abuse in the past year among
all persons aged 12 or older. This table provides estimates only for data based on Englishlanguage non-Hispanic interviews from the 2012 comparison data, 2013 quarters 3 and 4
comparison data, and the combined QFT and DR data. Across all three datasets, the estimates for
both dependence and abuse for the selected substances were generally quite similar, with the
following potential exceptions:
•

The estimate for hallucinogens dependence appeared to be significantly higher in the
2012 comparison data (rounded to 0.0 percent) than in the combined QFT and DR
data (also rounded to 0.0 percent). The estimate for hallucinogens dependence in the
2013 comparison data (rounded to 0.0 percent) met the criteria for low precision and,
therefore, would ordinarily be suppressed.

•

Similarly, the estimate for hallucinogens dependence or abuse appeared to be
significantly higher in the 2012 comparison data (0.1 percent) than in the combined
QFT and DR data (rounded to 0.0 percent). However, the estimates for hallucinogens
dependence or abuse for the combined QFT and DR data and the 2013 comparison
data both met the criteria for low precision and, therefore, would ordinarily be
suppressed.

Overall, few differences were observed for substance dependence or abuse for multiple
substances when comparing the combined QFT and DR data with the two comparison datasets.
302

6.3.3.2

Needle Use

Estimates for the use of five substances with a needle in the lifetime, past year, and past
month among persons aged 12 or older are provided in Table 6.52. This table provides estimates
only for data based on English-language non-Hispanic interviews from the 2012 comparison
data, 2013 quarters 3 and 4 comparison data, and the combined QFT and DR data. Across all
three datasets, the estimates for substance use with a needle for the five substances were
generally quite similar, with the following notable differences:
•

The estimate for use of heroin with a needle in the past month appeared to be
significantly higher in the 2012 comparison data (0.1 percent) than in the combined
QFT and DR data (rounded to 0.0 percent). The estimate for heroin use with a needle
in the past month was also 0.0 percent (rounded) for the 2013 comparison data and
not significantly different from the estimate for the combined QFT and DR data.

•

Similarly, the estimates for use of cocaine with a needle in the past year were
significantly higher for the 2012 comparison data (0.1 percent) and the 2013
comparison data (0.1 percent) than for the combined QFT and DR data (rounded to
0.0 percent). However, the estimate for cocaine use with a needle in the past year for
the combined QFT and DR data met the criteria for low precision and, therefore,
would ordinarily be suppressed.

•

The estimate for use of cocaine with a needle in the past month appeared to be
significantly higher in the 2012 comparison data (0.1 percent) than in the combined
QFT and DR data (rounded to 0.0 percent). The estimate for cocaine use with a
needle in the past month was also 0.0 percent (rounded) for the 2013 comparison data
and not significantly different from the estimate for the combined QFT and DR data.

•

For both past year and past month use of prescription stimulants with a needle, the
estimates for the 2012 and the 2013 comparison data both appeared to be higher than
the parallel estimates for the combined QFT and DR data. However, the estimates for
both past year and past month stimulants use with a needle for the combined QFT and
DR data met the criteria for low precision and, therefore, would ordinarily be
suppressed.

No significant differences were observed for any of the three reference periods for use of
methamphetamine with a needle or the combined use of heroin, cocaine, methamphetamine, or
prescription stimulants with a needle. Overall, the findings for the use of various substances were
similar to the QFT results, where the QFT estimates looked similar to the two comparison
datasets for most needle use estimates but some differences were observed (Currivan et al.,
2013).
6.3.3.3

Substance Use Treatment

Table 6.55 presents estimates for received substance use treatment in the lifetime and
past year, as well as types of past year substance use treatment, among all persons aged 12 or
older. For past year treatment, estimates are presented for alcohol use only, drug use only, and
both alcohol and drug use. This table provides estimates only for data based on English-language

303

non-Hispanic interviews from the 2012 comparison data, 2013 quarters 3 and 4 comparison data,
and the combined QFT and DR data.
Across all three datasets, the estimates for received substance use treatment in the
lifetime and past year substance use treatment were generally quite similar. The only significant
difference observed was for past year treatment for alcohol use only. The estimates for past year
treatment for alcohol use were significantly higher for the 2012 comparison data (0.6 percent)
and the 2013 comparison data (0.6 percent) than for the combined QFT and DR data
(0.3 percent).
6.3.3.4

Adult Mental Health

Table 6.56 presents estimates for adult mental health treatment in the past year and the
type of facility where treatment was received among persons aged 18 or older. This table
provides estimates only for data based on English-language non-Hispanic interviews from the
2012 comparison data, 2013 quarters 3 and 4 comparison data, and the combined QFT and DR
data. In all three datasets, the estimate for staying overnight in a hospital for mental health
treatment was close to 1 percent and not significantly different for any comparisons across
datasets. Similarly, no significant differences were observed among the three datasets for any of
the various types of mental health treatment facilities. Overall, the estimates for adult mental
health treatment were quite similar in the combined QFT and DR data and the comparison data.
Estimates for selected mental health measures among persons aged 18 or older are
provided in Table 6.58. This table provides estimates only for data based on English-language
non-Hispanic interviews from the 2012 comparison data, 2013 quarters 3 and 4 comparison data,
and the combined QFT and DR data. Across all three datasets, the estimates for the selected
mental health measures were generally quite similar, with the following potential exceptions:
•

Estimates for serious psychological distress (SPD) in the past month appeared to be
higher in the 2012 comparison data (5.1 percent) and the 2013 comparison data
(5.0 percent) than in the combined QFT and DR data (4.0 percent). Despite the
similarity between the 2012 and 2013 comparison data, only the difference between
the 2012 main study data and the combined QFT and DR data was statistically
significant. This finding was similar to the QFT results, where the QFT estimate for
past month SPD was significantly lower than the same estimates for the two
comparison datasets (Currivan et al., 2013). However, when SPD estimates for adults
were rerun to compare data just for the DR with data from the 2012 and 2013
comparison datasets (i.e., including Hispanic respondents and Spanish-language
interviews), the estimate of past month SPD in the DR (4.7 percent) was no longer
significantly different from the estimates in the 2012 and 2013 comparison data (5.2
and 5.1 percent, respectively).

•

Estimates for SPD in the past year also appeared to be higher in the 2012 comparison
data (10.6 percent) and the 2013 comparison data (10.6 percent) than in the combined
QFT and DR data (9.2 percent), but these differences were not statistically
significant. When estimates for just the DR were compared with corresponding
estimates from the 2012 and 2013 comparison data, the estimate of past year SPD in

304

the DR (9.7 percent) also was not significantly different from estimates in the 2012
and 2013 comparison data (10.8 and 10.5 percent, respectively).
•

Estimates for loss of interest in things that are usually enjoyable for several days or
longer (in the past year) appeared to be higher in the combined QFT and DR data
(5.8 percent) than in both the 2012 comparison data (4.2 percent) and the 2013
comparison data (4.0 percent), but these differences were not statistically significant.

6.3.3.5

Adolescent Mental Health

Estimates for mental health treatment in the past year and number of nights received
treatment among persons aged 12 to 17 are presented in Table 6.57. This table provides
estimates only for data based on English-language non-Hispanic interviews from the 2012
comparison data, 2013 quarters 3 and 4 comparison data, and the combined QFT and DR data.
In all three datasets, the estimate for staying overnight in hospital for mental health treatment
was close to 2 percent and not significantly different for any comparisons across datasets.
Estimates were also similar across the three datasets for the other three mental health items in
Table 6.57: (a) number of nights in a hospital for mental health treatment, (b) staying overnight
in a residential treatment center for mental health treatment, and (c) number of nights in a
residential treatment center for mental health treatment. For both number of nights in a hospital
and number of nights in a residential treatment center, the estimates for the combined QFT and
DR data met the criteria for low precision and, therefore, would ordinarily be suppressed.
Table 6.59 provides estimates for selected adolescent depression characteristics among
persons aged 12 to 17. The three estimates involve feelings experienced for several days or
longer, including feeling sad, empty or depressed; feeling very discouraged most of the day; and
losing interest in things that are usually enjoyable. Estimates for all three of these adolescent
depression characteristics were quite similar across the three datasets, and none of the small
differences observed was statistically significant.
6.3.3.6

Miscellaneous Noncore Items

Perceived Risk. Estimates for perceived great risk of harm associated with using two
types of substances—cigarettes and marijuana—among all persons aged 12 or older are
presented in Table 6.53. This table provides estimates only for data based on English-language
non-Hispanic interviews from the 2012 comparison data, 2013 quarters 3 and 4 comparison data,
and the combined QFT and DR data. For the first item, perceived risk of smoking one or more
packs of cigarettes per day, estimates were quite similar across the three datasets, and none of the
small differences observed was statistically significant.
For marijuana use, estimates were examined for the perceived risk of both (1) smoking
marijuana once per month and (2) smoking once or twice per week. The estimate for perceived
risk of smoking marijuana once per month was significantly lower in the 2013 comparison data
(26.1 percent) than in the combined QFT and DR data (29.1 percent). The estimate for the 2012
comparison data (28.8 percent), however, was quite similar to the estimate for the combined
QFT and DR data. For the second marijuana risk item, perceived risk of smoking marijuana once
or twice per week, estimates were quite similar across the three datasets, and none of the small
differences observed was statistically significant.
305

Years Since Last Use. Estimates for the number of years since last use of six selected
substances among lifetime users aged 12 to 49 are presented in Table 6.54. The six sets of
estimates for years since last use are for cigarettes, alcohol, marijuana, cocaine, hallucinogens,
and inhalants. This table provides estimates only for data based on English-language nonHispanic interviews from the 2012 comparison data, 2013 quarters 3 and 4 comparison data, and
the combined QFT and DR data. The following highlights summarize the results in Table 6.54:
•

The estimates for years since last use of cigarettes were quite similar across the three
datasets, with a range from 10.2 to 10.6 years, and none of the small differences
observed was statistically significant.

•

The estimates for years since last use of alcohol appeared to be lower in the 2012
comparison data (2.5 years) and the 2013 comparison data (2.5 years) than in the
combined QFT and DR data (3.2 years). Despite the similarity between the 2012 and
2013 comparison data, only the difference between the 2012 main study data and the
combined QFT and DR data was statistically significant.

•

The estimates for years since last use of marijuana were higher in the 2012
comparison data (9.9 years) and the 2013 comparison data (also 9.9 years) than in the
combined QFT and DR data (8.6 years). The differences between the two comparison
datasets and the combined QFT and DR dataset were both statistically significant.

•

The estimates for years since last use of cocaine were nearly identical across the three
datasets at about 10.5 years.

•

The estimates for years since last use of hallucinogens were quite similar across the
three datasets, with a range from 11.0 to 11.4 years, and none of the small differences
observed was statistically significant.

•

The estimates for years since last use of inhalants were quite similar across the three
datasets, with a range from 13.7 to 14.8 years, and none of the small differences
observed was statistically significant. The estimate for the 2013 comparison data
(14.4 years) and the combined QFT and DR data were quite similar (14.8 years).
The estimate for the 2012 comparison data (13.7 years) appeared to be somewhat
lower, but the differences between this estimate and the other two estimates were not
statistically significant.

Arrests. Estimates for being arrested and booked in the lifetime and the past year for
breaking the law among all persons 12 or older are presented in Table 6.60. This table provides
estimates only for data based on English-language non-Hispanic interviews from the 2012
comparison data, 2013 quarters 3 and 4 comparison data, and the combined QFT and DR data.
Estimates for lifetime arrest and booking for breaking the law were similar across the three
datasets, ranging from 16.5 to 17.4 percent. None of the small differences observed was
statistically significant. Similarly, for past year arrest and booking, estimates were similar across
the three datasets, ranging from 2.6 to 3.1 percent, and none of these small differences was
statistically significant.

306

Table 6.22 Substance Use Other Than Methamphetamine or Prescription Drugs in the Lifetime
among Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Substance
ILLICIT DRUGS
Alternate Definition 15
Alternate Definition 26
Marijuana and Hashish
Cocaine
Crack
Heroin
Hallucinogens
LSD
PCP
Ecstasy
Inhalants
ILLICIT DRUGS OTHER
THAN MARIJUANA
Alternate Definition5
Cocaine or Heroin7
CIGARETTES
SMOKELESS TOBACCO8
ALCOHOL

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

46.0
44.7
44.5
15.0
3.7a
1.9
15.3
9.8
2.7
6.3
8.6a

47.0
45.9
45.6
14.8
3.6a
1.9
16.2
10.4
2.6
7.1
8.5a

46.3
44.6
44.4
16.0
5.2
2.2
17.2
11.5
3.6
6.6
11.1

-0.3 (1.49)
0.1 (1.53)
0.1 (1.52)
-1.0 (1.09)
-1.5 (0.66)
-0.3 (0.43)
-1.9 (1.12)
-1.6 (0.93)
-1.0 (0.61)
-0.3 (0.64)
-2.5 (0.75)

0.7 (1.50)
1.2 (1.54)
1.2 (1.53)
-1.2 (1.12)
-1.6 (0.69)
-0.3 (0.44)
-0.9 (1.15)
-1.1 (0.93)
-1.0 (0.62)
0.4 (0.64)
-2.5 (0.75)

22.7
15.1
64.2
19.3a
83.9

22.4a
14.8
63.8
19.1
83.5

25.1
16.0
64.1
17.0
83.2

-2.4 (1.21)
-0.9 (1.09)
0.1 (1.41)
2.3 (1.03)
0.7 (0.99)

-2.7 (1.25)
-1.2 (1.12)
-0.3 (1.39)
2.1 (1.08)
0.4 (0.99)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or inhalants but do not
include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit Drugs Other Than Marijuana in this
definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
6
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include hallucinogens,
inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
7
Cocaine use includes crack.
8
Smokeless tobacco refers to snuff, dip, chewing tobacco, or "snus." Estimates are based on responses to questions about use of any
smokeless tobacco product.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

307

Table 6.22sp Substance Use Other Than Methamphetamine or Prescription Drugs in the Lifetime
among Persons Aged 12 or Older for Spanish-Language Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and 2013 Dress Rehearsal

2012
Comparison
(n = 2,061)1,2

2013
Comparison
(n = 998)1,3

2013 DR
(n = 185)1,4

2012
Comparison
vs. DR,
Difference
(SE)

2013
Comparison
vs. DR,
Difference
(SE)

11.6
10.5
9.2
4.9
0.5
0.5
1.5
0.4
0.2
0.4
2.0

9.5
9.2
8.2
3.6
0.6
0.2
0.8
0.1
0.0
0.4
1.2

14.4*
13.6*
9.5
7.2*
0.3
0.1
4.0
1.9
0.0*
0.6
1.0

-2.8 (5.05)
-3.0 (4.80)
-0.4 (3.51)
-2.3 (3.84)
0.2 (0.39)
0.4 (0.29)
-2.6 (2.04)
-1.5 (1.20)
0.2 (0.14)
-0.2 (0.42)
1.0 (0.77)

-4.9 (5.27)
-4.4 (4.96)
-1.3 (3.51)
-3.6 (3.89)
0.3 (0.39)
0.0 (0.19)
-3.3 (2.14)
-1.8 (1.19)
0.0 (0.05)
-0.2 (0.41)
0.2 (0.69)

6.6
4.9
40.8
1.9a
65.1a

4.4
3.6
39.3
1.2a
64.6

11.2*
7.3*
39.4*
0.1
72.9

-4.6 (5.25)
-2.4 (3.85)
1.4 (6.77)
1.8 (0.53)
-7.8 (3.87)

-6.8 (5.35)
-3.7 (3.89)
-0.1 (6.69)
1.0 (0.42)
-8.3 (4.89)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal.
a
Difference between estimate and 2013 DR is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, or inhalants but do not
include methamphetamine or prescription-type psychotherapeutics that were misused. Illicit Drugs Other Than Marijuana in this
definition include cocaine (including crack), heroin, hallucinogens, or inhalants.
6
Illicit Drugs in this definition include marijuana/hashish, cocaine (including crack), or heroin, but do not include hallucinogens,
inhalants, methamphetamine, or prescription-type psychotherapeutics that were misused.
7
Cocaine use includes crack.
8
Smokeless tobacco refers to snuff, dip, chewing tobacco, or "snus." Estimates are based on responses to questions about use of any
smokeless tobacco product.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

308

Table 6.23 Substance Use Other Than Methamphetamine or Prescription Drugs in the Lifetime
among Persons Aged 12 to 17 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
(n = 8,610)1,3

Combined
2012 QFT and
2013 DR
(n = 707)1,4

20.3
16.5
16.4
0.9a
0.1
0.2a
3.1
1.0
0.3
1.7
6.2a

19.1
15.9
15.8
0.8a
0.1
0.1a
2.5a
1.0
0.2
1.3
4.9a

23.0
17.3
17.3
0.2
0.1
0.0*
4.5
0.8
0.5
1.9
9.4

0.3 (0.95)
0.4 (0.93)
0.2 (0.91)
0.4 (0.28)
0.1 (0.11)
0.1 (0.11)
-0.0 (0.27)
0.0 (0.10)
0.0 (0.02)
0.2 (0.14)
-0.1 (0.16)

0.8 (0.95)
0.8 (0.92)
0.7 (0.91)
0.2 (0.25)
-0.0 (0.11)
0.0 (0.11)
-0.0 (0.27)
0.1 (0.11)
0.0 (0.03)
0.3 (0.15)
-0.2 (0.16)

8.4a
1.0a
17.5
7.2
31.9

7.0a
0.8a
15.5
6.8
29.7

11.9
0.2
16.0
7.0
30.5

0.1 (0.45)
0.3 (0.29)
-1.2 (1.41)
-0.7 (0.50)
2.4 (1.39)

-0.2 (0.42)
0.1 (0.27)
-2.4 (1.37)
-0.8 (0.52)
1.8 (1.35)

2012
Comparison
(n = 17,573)1,2

309

Table 6.24 Substance Use Other Than Methamphetamine or Prescription Drugs in the Lifetime
among Persons Aged 18 to 25 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2013
Comparison
(n = 8,532)1,3

Combined
2012 QFT and
2013 DR
(n = 702)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

55.1a
53.9
53.7
12.3
2.0
2.2
18.2
6.7
1.0
13.1
8.7

53.9a
52.7a
52.6a
11.6
1.6
2.1
17.8
6.9
0.7
12.4
7.7

60.3
58.3
58.0
10.8
2.6
2.6
19.9
7.3
0.9
12.2
10.3

-0.2 (0.69)
-0.3 (0.69)
-0.5 (0.67)
0.3 (0.11)
0.1 (0.05)
0.1 (0.03)
0.1 (0.09)
0.0 (0.04)
-0.0 (0.02)
0.1 (0.06)
0.0 (0.05)

0.2 (0.69)
0.1 (0.69)
-0.1 (0.67)
0.3 (0.11)
0.1 (0.05)
0.1 (0.03)
0.2 (0.10)
0.0 (0.04)
-0.0 (0.02)
0.2 (0.06)
0.1 (0.06)

23.7
12.6
60.2
22.0
85.0

22.5a
11.8
57.6a
22.7
84.3

27.2
10.8
63.7
23.2
86.2

0.5 (0.15)
0.4 (0.12)
-1.0 (1.35)
-0.4 (0.45)
3.5 (1.37)

0.5 (0.16)
0.4 (0.12)
-2.2 (1.32)
-0.5 (0.46)
3.2 (1.40)

2012
Comparison
(n = 18,029)1,2

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

310

Table 6.25 Substance Use Other Than Methamphetamine or Prescription Drugs in the Lifetime
among Persons Aged 26 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 19,630)1,2

2013
Comparison
(n = 9,475)1,3

Combined
2012 QFT and
2013 DR
(n = 1,603)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

47.4
46.4
46.2
17.0
4.3a
2.0
16.2
11.3
3.2
5.7
8.8a

49.0
48.1
47.9
16.8
4.3a
2.1
17.5
12.0
3.2
6.9
9.1a

46.7
45.5
45.3
18.6
6.2
2.4
18.2
13.3
4.4
6.3
11.4

-2.7 (2.14)
-0.8 (1.87)
-0.9 (1.87)
0.7 (0.16)
-0.0 (0.13)
0.2 (0.04)
-1.4 (0.93)
0.2 (0.33)
-0.2 (0.26)
-0.2 (0.49)
-3.2 (1.22)

-3.9 (2.23)
-1.4 (1.95)
-1.5 (1.95)
0.5 (0.17)
-0.0 (0.13)
0.1 (0.02)
-2.0 (0.93)
0.2 (0.34)
-0.3 (0.26)
-0.6 (0.49)
-4.5 (1.21)

24.2
17.1
70.0
20.3a
89.6

24.2
16.9
70.2
19.9a
89.5

26.3
18.6
69.6
17.2
88.6

-3.5 (1.43)
0.8 (0.16)
1.5 (1.99)
0.2 (1.10)
1.5 (2.37)

-5.0 (1.42)
0.6 (0.17)
-0.5 (1.99)
-0.2 (1.14)
-0.7 (2.40)

311

Table 6.26 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

13.0
12.6
12.2
1.8
0.4
0.3
1.6
0.4
0.1
1.0
0.6

13.5
13.0
12.7
1.6
0.3
0.3
1.7
0.4
0.0
1.0
0.6

12.7
12.2
12.0
1.4
0.3
0.2
1.7
0.4
0.0
0.7
0.7

-0.5 (1.63)
0.3 (1.49)
0.2 (1.49)
0.5 (0.10)
0.1 (0.03)
0.1 (0.03)
-0.3 (0.72)
0.3 (0.20)
-0.0 (0.14)
0.1 (0.37)
-1.2 (0.74)

-1.2 (1.68)
-0.4 (1.55)
-0.4 (1.55)
0.4 (0.10)
0.0 (0.02)
0.0 (0.02)
-0.6 (0.72)
0.4 (0.20)
-0.1 (0.14)
-0.1 (0.36)
-1.6 (0.73)

3.3
1.9
26.9
5.1
68.3

3.1
1.7
25.6
5.0
67.7

3.2
1.6
28.0
5.8
65.9

-1.1 (0.95)
0.6 (0.10)
1.1 (1.55)
-0.8 (0.96)
3.1 (1.95)

-1.7 (0.95)
0.4 (0.10)
-0.9 (1.57)
-0.7 (0.99)
1.2 (2.02)

312

Table 6.26sp Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 12 or Older for Spanish-Language Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and 2013 Dress Rehearsal

2012
Comparison
(n = 2,061)1,2

2013
Comparison
(n = 998)1,3

2013 DR
(n = 185)1,4

2012
Comparison
vs. DR,
Difference
(SE)

2013
Comparison
vs. DR,
Difference
(SE)

2.1
1.8
1.5
0.4
0.1
0.0
0.1
0.0
0.0*
0.1
0.3

1.3
1.3
1.3
0.3
0.1
0.0*
0.2
0.1
0.0*
0.1
0.1

1.8
1.8
1.2
0.6*
0.0*
0.1
0.2
0.0*
0.0*
0.2
0.1

0.3 (0.83)
-0.0 (0.83)
0.3 (0.65)
-0.2 (0.54)
0.1 (0.04)
-0.1 (0.14)
-0.1 (0.20)
0.0 (0.03)
0.0 (0.00)
-0.1 (0.19)
0.2 (0.19)

-0.4 (0.63)
-0.5 (0.63)
0.1 (0.45)
-0.2 (0.64)
0.1 (0.14)
-0.1 (0.14)
0.0 (0.22)
0.1 (0.07)
0.0 (0.00)
-0.1 (0.20)
-0.1 (0.14)

0.7
0.4
14.7
0.1
45.0a

0.5
0.3
15.8
0.3
41.9a

0.9
0.7
19.0*
0.1
59.4

-0.2 (0.59)
-0.3 (0.55)
-4.3 (5.71)
-0.0 (0.13)
-14.4 (4.64)

-0.4 (0.67)
-0.4 (0.65)
-3.2 (5.82)
0.1 (0.19)
-17.5 (5.27)

313

Table 6.27 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 12 to 17 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 17,573)1,2

2013
Comparison
(n = 8,610)1,3

Combined
2012 QFT and
2013 DR
(n = 707)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

14.7
13.1
13.1
0.6a
0.1a
0.1a
2.1
0.5
0.1
1.0
2.3

14.0
12.5
12.4
0.5a
0.0a
0.0a
1.7
0.6
0.1
0.8
1.9a

15.2
12.8
12.8
0.1
0.0*
0.0*
2.4
0.3
0.2
0.9
3.5

0.2 (1.11)
0.5 (1.05)
0.4 (1.05)
0.1 (0.03)
0.0 (0.00)
0.0 (0.00)
0.1 (0.27)
0.0 (0.09)
-0.1 (0.10)
0.2 (0.10)
-0.1 (0.28)

-0.3 (1.14)
0.1 (1.08)
0.0 (1.08)
0.1 (0.04)
0.0 (0.00)
0.0 (0.00)
-0.1 (0.27)
0.1 (0.11)
-0.1 (0.10)
0.0 (0.10)
-0.3 (0.28)

4.2
0.7a
11.8
4.3
25.3

3.6
0.5a
9.8
4.3
23.5

5.3
0.1
10.7
5.0
22.3

0.4 (0.34)
0.1 (0.03)
1.0 (1.16)
-0.8 (0.82)
2.4 (1.36)

0.1 (0.34)
0.1 (0.04)
-0.3 (1.18)
-1.1 (0.82)
1.1 (1.37)

314

Table 6.28 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 18 to 25 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 18,029)1,2

2013
Comparison
(n = 8,532)1,3

Combined
2012 QFT and
2013 DR
(n = 702)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

33.9
33.3
32.8
4.6
0.5
0.9
6.7
2.0
0.2
4.1
1.5

33.2
32.5
32.1
4.7
0.3
0.9
6.7
2.3
0.1
4.2
1.4

33.4
32.0
31.7
3.2
0.6
0.8
7.6
2.0
0.2
3.4
1.1

-5.2 (2.43)
-4.4 (2.52)
-4.3 (2.50)
1.5 (1.55)
-0.6 (0.90)
-0.4 (0.69)
-1.6 (2.03)
-0.7 (1.58)
0.1 (0.37)
0.9 (1.50)
-1.6 (1.26)

-6.3 (2.47)
-5.6 (2.56)
-5.4 (2.56)
0.8 (1.56)
-1.0 (0.91)
-0.5 (0.68)
-2.1 (1.99)
-0.4 (1.54)
-0.1 (0.38)
0.2 (1.53)
-2.6 (1.30)

9.6
4.9
42.5
10.3
78.4

9.4
5.0
39.8
10.7
77.7

9.5
3.6
43.8
10.3
78.5

-3.5 (2.25)
1.7 (1.55)
-3.5 (2.43)
-1.2 (2.58)
-1.3 (1.74)

-4.8 (2.23)
1.0 (1.56)
-6.1 (2.40)
-0.5 (2.53)
-1.9 (1.77)

315

Table 6.29 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Year
among Persons Aged 26 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 19,630)1,2

2013
Comparison
(n = 9,475)1,3

9.5
9.3
8.9
1.5
0.4
0.2
0.8
0.1
0.0a
0.5
0.3

10.3
10.0
9.7
1.2
0.3
0.2
0.8
0.1
0.0
0.5
0.3

2.2
1.6
26.1
4.3
71.5

2.0
1.3
25.1
4.1
71.1

Combined
2012 QFT and
2013 DR
(n = 1,603)1,4
9.1
9.0
8.8
1.3
0.3
0.2
0.6
0.1
0.0*
0.3
0.3
2.0
1.5
27.5
5.1
68.8

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

0.5 (2.62)
1.3 (2.54)
1.0 (2.54)
1.4 (0.90)
-0.1 (0.38)
0.1 (0.34)
-0.9 (1.51)
-0.0 (0.60)
-0.0 (0.15)
0.7 (0.77)
0.3 (0.44)

-0.3 (2.59)
0.5 (2.50)
0.3 (2.53)
1.5 (0.88)
-0.3 (0.37)
0.1 (0.33)
-0.8 (1.43)
0.2 (0.57)
-0.1 (0.15)
0.7 (0.74)
0.3 (0.46)

0.1 (1.60)
1.3 (0.94)
-1.3 (2.39)
-0.0 (1.50)
-0.1 (2.25)

-0.0 (1.49)
1.5 (0.90)
-4.0 (2.37)
0.4 (1.50)
-0.8 (2.26)

316

Table 6.30 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

7.8
7.6
7.3
0.7a
0.2a
0.1a
0.4
0.1
0.0
0.2
0.2

8.2
8.0
7.7
0.7a
0.1
0.1a
0.5a
0.1
0.0
0.3a
0.2

8.0
7.9
7.8
0.3
0.1
0.0
0.3
0.1
0.0
0.1
0.1

-0.9 (2.11)
-1.1 (2.11)
-1.4 (2.09)
0.4 (0.41)
-0.3 (0.36)
0.1 (0.19)
0.0 (0.57)
0.1 (0.21)
-0.2 (0.15)
0.2 (0.39)
0.1 (0.19)

-1.5 (2.08)
-1.6 (2.08)
-1.9 (2.08)
0.5 (0.42)
-0.3 (0.36)
-0.0 (0.19)
0.3 (0.58)
0.1 (0.21)
-0.2 (0.15)
0.3 (0.39)
0.0 (0.20)

1.2a
0.7a
23.1
3.9
54.1a

1.3a
0.7a
22.0
3.8
53.9a

0.7
0.3
24.1
4.3
50.7

0.3 (0.71)
0.5 (0.44)
0.5 (2.41)
0.1 (1.27)
0.4 (3.05)

0.6 (0.73)
0.5 (0.45)
-1.3 (2.41)
0.8 (1.25)
-0.9 (3.12)

317

Table 6.30sp Substance Use Other Than Methamphetamine or Prescription Drugs in the Past
Month among Persons Aged 12 or Older for Spanish-Language Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and 2013 Dress Rehearsal

2012
Comparison
(n = 2,061)1,2
1.0
0.8
0.7
0.2
0.0
0.0*
0.1
0.0
0.0*
0.0
0.2
0.4
0.2
11.5
0.0
27.1

2013
Comparison
(n = 998)1,3
0.7
0.7
0.7
0.2
0.1
0.0*
0.0
0.0*
0.0*
0.0
0.1
0.3
0.2
14.1
0.1
25.1

2013 DR
(n = 185)1,4

2012
2013
Comparison vs. Comparison vs.
DR, Difference DR, Difference
(SE)
(SE)

0.9
0.9
0.3
0.6*
0.0*
0.0*
0.0*
0.0*
0.0*
0.0*
0.0*

0.1 (0.79)
-0.0 (0.78)
0.4 (0.31)
-0.4 (0.54)
0.0 (0.04)
0.0 (0.00)
0.1 (0.04)
0.0 (0.01)
0.0 (0.00)
0.0 (0.02)
0.2 (0.13)

-0.1 (0.68)
-0.1 (0.68)
0.4 (0.28)
-0.3 (0.57)
0.1 (0.14)
0.0 (0.00)
0.0 (0.03)
0.0 (0.00)
0.0 (0.00)
0.0 (0.03)
0.1 (0.04)

0.6*
0.6*
18.1*
0.1
33.9*

-0.1 (0.55)
-0.4 (0.54)
-6.5 (5.79)
-0.1 (0.13)
-6.8 (6.28)

-0.3 (0.58)
-0.3 (0.57)
-4.0 (5.89)
-0.0 (0.14)
-8.8 (6.58)

318

Table 6.31 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 12 to 17 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 17,573)1,2

2013
Comparison
(n = 8,610)1,3

Combined
2012 QFT and
2013 DR
(n = 707)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

7.7
7.1
7.1
0.1a
0.0*
0.0*
0.6
0.1
0.0
0.3
0.6

7.2
6.8
6.7
0.1a
0.0*
0.0*
0.5
0.2
0.0*
0.1
0.5

7.5
6.7
6.7
0.0*
0.0*
0.0*
0.5
0.1
0.1
0.1
0.7

0.7 (1.73)
0.9 (1.77)
0.9 (1.76)
-1.6 (1.34)
-1.8 (0.82)
-0.4 (0.54)
-1.9 (1.32)
-2.0 (1.13)
-1.2 (0.78)
-0.6 (0.74)
-2.6 (0.92)

2.3 (1.74)
2.6 (1.78)
2.6 (1.77)
-1.8 (1.38)
-1.9 (0.85)
-0.3 (0.55)
-0.6 (1.36)
-1.3 (1.14)
-1.3 (0.79)
0.6 (0.73)
-2.3 (0.92)

1.3
0.1a
7.0
2.5
12.9

1.0
0.1a
5.7
2.2
11.6

0.9
0.0*
6.0
3.3
10.5

-2.1 (1.42)
-1.6 (1.34)
0.4 (1.56)
3.1 (1.18)
0.9 (1.09)

-2.1 (1.46)
-1.8 (1.38)
0.6 (1.55)
2.7 (1.25)
0.8 (1.09)

319

Table 6.32 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 18 to 25 for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 18,029)1,2

2013
Comparison
(n = 8,532)1,3

Combined
2012 QFT and
2013 DR
(n = 702)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

20.3
19.9
19.5
1.2
0.1
0.4
1.8
0.4
0.0
1.1
0.4

19.8
19.4
19.1
1.3
0.1
0.2
2.1
0.3
0.0
1.1
0.3

21.2
21.0
21.0
0.8
0.4
0.3
1.8
0.3
0.2
0.8
0.3

0.4 (0.88)
0.3 (0.86)
0.1 (0.85)
0.2 (0.31)
0.2 (0.12)
0.0 (0.13)
0.1 (0.19)
0.0 (0.08)
0.0 (0.01)
0.2 (0.11)
-0.1 (0.18)

1.2 (0.90)
1.0 (0.88)
0.9 (0.87)
-0.0 (0.29)
0.0 (0.12)
0.0 (0.12)
0.2 (0.20)
0.0 (0.09)
0.0 (0.02)
0.3 (0.13)
-0.1 (0.18)

3.0
1.4
33.6
6.5
62.0

3.3
1.5
31.9
7.2
60.7

2.7
1.0
33.1
6.4
61.6

0.2 (0.40)
0.1 (0.33)
-1.4 (1.69)
-0.8 (0.58)
2.7 (1.59)

0.0 (0.38)
-0.1 (0.31)
-2.4 (1.65)
-1.0 (0.61)
2.3 (1.57)

320

Table 6.33 Substance Use Other Than Methamphetamine or Prescription Drugs in the Past Month
among Persons Aged 26 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

2012
Comparison
(n = 19,630)1,2

2013
Comparison
(n = 9,475)1,3

Combined
2012 QFT and
2013 DR
(n = 1,603)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

5.8
5.8
5.4
0.6a
0.2a
0.1a
0.2a
0.0
0.0
0.1a
0.1

6.5
6.4
6.0
0.6a
0.2a
0.1a
0.3a
0.1
0.0
0.2a
0.2a

6.0
6.0
5.8
0.3
0.0*
0.0*
0.1
0.0
0.0*
0.0
0.1

-0.1 (0.66)
-0.2 (0.66)
-0.4 (0.65)
0.4 (0.13)
0.2 (0.04)
0.1 (0.02)
0.1 (0.06)
-0.0 (0.03)
0.0 (0.00)
0.1 (0.04)
0.0 (0.05)

0.6 (0.68)
0.4 (0.68)
0.2 (0.67)
0.3 (0.13)
0.2 (0.04)
0.1 (0.03)
0.2 (0.07)
0.0 (0.04)
0.0 (0.02)
0.2 (0.05)
0.1 (0.06)

0.9a
0.7a
23.3
3.6
57.5a

1.0a
0.7a
22.2
3.5
57.6a

0.5 (0.14)
0.4 (0.13)
-1.5 (1.60)
-0.4 (0.51)
4.0 (1.55)

0.6 (0.15)
0.4 (0.14)
-2.5 (1.56)
-0.6 (0.53)
4.1 (1.63)

0.4
0.3
24.7
4.1
53.5

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

321

Table 6.34 Specific Hallucinogen Use in the Lifetime, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Hallucinogen/Age Group
Hallucinogens, Aged 12 or
Older
Ketamine5,6
DMT, AMT, or 5-MeODIPT ("Foxy")5
Salvia divinorum5
Other Hallucinogens7
Hallucinogens, Aged 12 to 17
Ketamine5,6
DMT, AMT, or 5-MeODIPT ("Foxy")5
Salvia divinorum5
Other Hallucinogens7
Hallucinogens, Aged 18 to 25
Ketamine5,6
DMT, AMT, or 5-MeODIPT ("Foxy")5
Salvia divinorum5
Other Hallucinogens7
Hallucinogens, Aged 26 or
Older
Ketamine5,6
DMT, AMT, or 5-MeODIPT ("Foxy")5
Salvia divinorum5
Other Hallucinogens7

2012
2013
Combined 2012 Comparison vs. Comparison vs.
QFT and 2013 Combined QFT Combined QFT
DR
and DR,
and DR,
(n = 3,012)1,4
Difference (SE) Difference (SE)
17.2
-1.9 (1.12)
-0.9 (1.15)

2012
Comparison
(n = 55,232)1,2
15.3

2013
Comparison
(n = 26,617)1,3
16.2

1.0
0.6

1.1
0.7

1.4
1.0

-0.3 (0.28)
-0.4 (0.22)

-0.3 (0.28)
-0.3 (0.22)

2.1
1.5a
3.1
0.3
0.4

2.0
1.9a
2.5a
0.2
0.3

2.5
0.7
4.5
0.4
0.4

-0.5 (0.40)
0.8 (0.22)
-1.4 (0.93)
-0.1 (0.22)
0.0 (0.25)

-0.6 (0.41)
1.2 (0.23)
-2.0 (0.93)
-0.2 (0.22)
-0.1 (0.25)

1.1
1.0
18.2
1.4
2.0

0.6a
0.8
17.8
1.9
2.6

1.8
0.5
19.9
1.4
2.1

-0.7 (0.52)
0.5 (0.24)
-1.6 (2.03)
0.0 (0.46)
-0.1 (0.74)

-1.1 (0.53)
0.3 (0.25)
-2.1 (1.99)
0.5 (0.48)
0.5 (0.75)

8.4
3.3a
16.2

7.4
4.1a
17.5

7.6
1.1
18.2

0.8 (1.30)
2.2 (0.44)
-1.9 (1.32)

-0.2 (1.27)
3.0 (0.45)
-0.6 (1.36)

1.1
0.4

1.5
0.9

-0.4 (0.34)
-0.5 (0.26)

1.2
1.3a

1.2
1.7a

1.8
0.7

-0.7 (0.40)
0.7 (0.26)

-0.6 (0.41)
1.0 (0.28)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
AMT = alpha-methyltryptamine; DMT = dimethyltryptamine; DR = Dress Rehearsal; 5-MeO-DIPT = 5-methoxy-diisopropyltryptamine;
QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Asked in the hallucinogens module in the QFT and DR and in the special drugs module in the 2012 and 2013 comparison data.
6
Ketamine is also known as "Special K" or "Super K."
7
For the 2012 and 2013 comparison data, use of any other hallucinogens besides the following: LSD, also called "acid"; PCP, also called
"angel dust" or phencyclidine; peyote; mescaline; psilocybin; or "Ecstasy," also called MDMA. For the QFT and DR, use of any other
hallucinogens besides the ones in the 2012 and 2013 comparison data, plus the following: ketamine; DMT, AMT, or 5-MeO-DIPT
("Foxy"); or Salvia divinorum.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

322

Table 6.35 Specific Inhalant Use in the Lifetime, by Age Group for English-Language NonHispanic Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Inhalant/Age Group
Inhalants, Aged 12 or Older
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays5
Other Inhalants6
Inhalants, Aged 12 to 17
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays5
Other Inhalants6
Inhalants, Aged 18 to 25
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays5
Other Inhalants6
Inhalants, Aged 26 or Older
Felt-Tip Pens
Computer Keyboard
Cleaner
Other Aerosol Sprays5
Other Inhalants6

2012
Comparison
(n = 55,232)1,2
8.6a
N/A

2013
Comparison
(n = 26,617)1,3
8.5a
N/A

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4
11.1
3.0

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-2.5 (0.75)
N/A (N/A)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-2.5 (0.75)
N/A (N/A)

N/A
0.8
0.4
6.2a
N/A

N/A
0.8
0.4
4.9a
N/A

1.3
0.8
0.7
9.4
7.2

N/A (N/A)
-0.1 (0.24)
-0.2 (0.24)
-3.2 (1.22)
N/A (N/A)

N/A (N/A)
-0.0 (0.25)
-0.3 (0.25)
-4.5 (1.21)
N/A (N/A)

N/A
1.1
1.4a
8.7
N/A

N/A
0.9
0.9
7.7
N/A

1.1
0.8
0.7
10.3
5.2

N/A (N/A)
0.3 (0.36)
0.7 (0.33)
-1.6 (1.26)
N/A (N/A)

N/A (N/A)
0.1 (0.36)
0.2 (0.34)
-2.6 (1.30)
N/A (N/A)

N/A
1.5a
0.8a
8.8a
N/A

N/A
1.2a
0.8a
9.1a
N/A

2.9
0.4
0.0*
11.4
2.1

N/A (N/A)
1.1 (0.24)
0.8 (0.06)
-2.6 (0.92)
N/A (N/A)

N/A (N/A)
0.8 (0.25)
0.8 (0.09)
-2.3 (0.92)
N/A (N/A)

N/A
0.6
0.3

N/A
0.7
0.2

1.1
0.9
0.8

N/A (N/A)
-0.3 (0.30)
-0.5 (0.31)

N/A (N/A)
-0.2 (0.31)
-0.5 (0.31)

323

Table 6.36 Alcohol Use in the Lifetime among Persons Aged 12 or Older, by Age Group and
Gender for English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Gender
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2012
Comparison
(n = 55,232)1,2
83.9
86.3
81.7
31.9
31.4
32.5
85.0
84.9
85.0
89.6
93.1
86.4

2013
Comparison
(n = 26,617)1,3
83.5
85.6
81.6
29.7
29.0
30.4
84.3
84.1
84.5
89.5
92.7
86.5

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4
83.2
84.5
81.9
30.5
30.7
30.2
86.2
85.0
87.5
88.6
91.0
86.6

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.7 (0.99)
2.4 (1.39)
3.5 (1.37)
1.7 (1.17)
3.5 (1.83)
2.8 (2.09)
-0.2 (1.44)
1.5 (1.85)
4.1 (1.69)
1.5 (2.37)
3.1 (1.95)
2.4 (1.36)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.4 (0.99)
1.8 (1.35)
3.2 (1.40)
1.1 (1.20)
3.6 (1.90)
3.7 (2.12)
-0.3 (1.46)
0.2 (1.77)
2.8 (1.70)
-0.7 (2.40)
1.2 (2.02)
1.1 (1.37)

324

Table 6.37 Alcohol Use in the Past Year among Persons Aged 12 or Older, by Age Group and
Gender for English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Gender
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2012
Comparison
(n = 55,232)1,2
68.3
70.9
65.9
25.3
24.5
26.2
78.4
79.2
77.6
71.5
75.2
68.3

2013
Comparison
(n = 26,617)1,3
67.7
71.1
64.6
23.5
22.0
24.9
77.7
78.4
77.0
71.1
75.8a
66.8

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4
65.9
67.5
64.4
22.3
22.5
22.1
78.5
78.4
78.7
68.8
71.0
66.8

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.7 (2.38)
2.0 (2.30)
0.7 (1.89)
2.3 (3.79)
4.1 (2.93)
4.2 (2.07)
-1.3 (1.74)
-0.1 (2.25)
0.4 (3.05)
-0.1 (2.54)
0.9 (3.26)
-1.4 (4.25)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-1.6 (2.58)
-0.4 (2.46)
-0.5 (1.98)
0.2 (3.73)
2.8 (2.93)
2.6 (1.99)
-1.9 (1.77)
-0.8 (2.26)
-0.9 (3.12)
-0.9 (2.57)
0.0 (3.31)
-2.7 (4.36)

325

Table 6.38 Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group and
Gender for English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Gender
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2012
Comparison
(n = 55,232)1,2
54.1a
58.1
50.4a
12.9
12.5
13.3a
62.0
64.2
59.8
57.5a
62.6
52.9a

2013
Comparison
(n = 26,617)1,3
53.9a
59.0
49.1
11.6
11.4
11.8
60.7
63.0
58.5
57.6a
64.1a
51.6

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4
50.7
55.3
46.4
10.5
11.9
9.1
61.6
65.6
57.5
53.5
58.8
48.7

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-2.5 (1.92)
-1.1 (2.41)
2.3 (3.17)
0.9 (1.09)
2.7 (1.59)
4.0 (1.55)
2.1 (1.38)
4.1 (2.19)
3.8 (2.47)
-0.2 (1.53)
1.4 (2.21)
4.2 (2.07)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-3.0 (2.04)
-1.7 (2.49)
1.0 (3.20)
0.8 (1.09)
2.3 (1.57)
4.1 (1.63)
1.7 (1.41)
4.7 (2.30)
5.3 (2.56)
-0.0 (1.60)
0.0 (2.14)
2.9 (2.11)

326

Table 6.38sp Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group and
Gender for Spanish-Language Interviews: Percentages, Differences, and Standard
Error of Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress Rehearsal

Age Group/Gender
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2012
Comparison
(n = 2,061)1,2
27.1
39.5
16.3
11.7
11.9a
11.5*
33.9
44.6
20.3
27.3
40.5
16.2

2013
Comparison
(n = 998)1,3
25.1
37.3
14.4
4.8
5.7*
3.9*
33.3
45.8
18.6
25.7
38.8
14.7

2013 DR
(n = 185)1,4
33.9*
39.4*
29.2*
6.6*
2.7*
18.0*
52.7*
72.3*
42.0*
34.9*
42.1*
29.1*

2012
Comparison vs.
DR, Difference
(SE)
-6.8 (6.28)
0.0 (13.06)
-13.0 (6.95)
5.2 (4.36)
9.3 (4.17)
-6.5 (10.97)
-18.8 (13.66)
-27.6 (23.13)
-21.7 (18.59)
-7.6 (6.79)
-1.6 (14.71)
-12.9 (7.21)

2013
Comparison vs.
DR, Difference
(SE)
-8.8 (6.58)
-2.1 (12.23)
-14.8 (7.62)
-1.8 (5.85)
3.0 (5.20)
-14.1 (11.94)
-19.4 (13.45)
-26.5 (22.58)
-23.4 (17.37)
-9.2 (7.18)
-3.3 (13.90)
-14.4 (7.94)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1

Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2012 comparison data collected in quarters 1 through 4, 2012.
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.
2
3

327

Table 6.39 Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender for English-Language Non-Hispanic Interviews: Percentages, Differences,
and Standard Error of Differences, 2012 Comparison, 2013 Comparison, and
Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Gender
BINGE ALCOHOL USE,
CORE ONLY5
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female
BINGE ALCOHOL USE,
CORE PLUS NONCORE6
Aged 12 or Older
Male
Female
Aged 12 to 17
Male
Female
Aged 18 to 25
Male
Female
Aged 26 or Older
Male
Female

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

Combined
2012 QFT and
2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

22.4
29.8
15.7
6.7
6.8a
6.5
39.9
45.7
34.3
21.4
29.9
13.8a

22.4
29.5
15.8
5.6
6.3
4.8
37.7
44.6
30.8
21.9
29.9
14.7

23.2
28.6
18.2
4.9
4.4
5.4
40.9
45.5
36.1
22.5
28.6
17.0

-0.8 (1.19)
1.2 (1.84)
-2.6 (1.31)
1.8 (1.02)
2.4 (1.06)
1.1 (1.47)
-1.0 (2.94)
0.1 (3.98)
-1.8 (3.10)
-1.0 (1.30)
1.3 (2.07)
-3.2 (1.56)

-0.8 (1.17)
1.0 (1.83)
-2.4 (1.30)
0.7 (1.03)
1.9 (1.08)
-0.6 (1.47)
-3.3 (2.91)
-0.9 (4.06)
-5.3 (3.04)
-0.6 (1.31)
1.3 (2.06)
-2.3 (1.56)

25.0
29.8
20.6
7.5a
6.8a
8.2
43.2
45.7
40.7
24.1
29.9
18.8

24.6
29.5
20.1
6.2
6.3
6.0
41.0
44.6
37.4
24.1
29.9
18.9

23.2
28.6
18.2
4.9
4.4
5.4
40.9
45.5
36.1
22.5
28.6
17.0

1.8 (1.19)
1.2 (1.84)
2.4 (1.31)
2.6 (1.02)
2.4 (1.06)
2.8 (1.46)
2.2 (2.94)
0.1 (3.98)
4.6 (3.09)
1.6 (1.30)
1.3 (2.07)
1.9 (1.55)

1.4 (1.17)
1.0 (1.83)
1.9 (1.29)
1.3 (1.02)
1.9 (1.08)
0.6 (1.47)
0.1 (2.91)
-0.9 (4.06)
1.3 (3.03)
1.7 (1.30)
1.3 (2.06)
2.0 (1.56)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Binge Alcohol Use in the 2012 and 2013 comparison data based on only core alcohol module data is defined as drinking five or more
drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. Binge
Alcohol Use in the Combined QFT and DR is defined for males as drinking five or more drinks on the same occasion and for females
as drinking four or more drinks on the same occasion on at least 1 day in the past 30 days.
6
Binge Alcohol Use in the 2012 and 2013 comparison data based on core plus noncore data is defined for males as drinking five or more
drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. The
measure for females in the 2012 and 2013 comparison data is defined as drinking five or more drinks on the same occasion on at least 1
day in the past 30 days or usually having four drinks on those days when respondents drank alcohol in the past 30 days based on the
core alcohol module data, or drinking four or more drinks on the same occasion on at least 1 day in the past 30 days (including the last
occasion of alcohol use) based on the noncore consumption of alcohol module data. Combined QFT and DR data for binge alcohol use
based on the core alcohol module data are repeated in these rows.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

328

Table 6.39sp Binge Alcohol Use in the Past Month among Persons Aged 12 or Older, by Age Group
and Gender for Spanish-Language Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress
Rehearsal

2012
Comparison
(n = 2,061)1,2

2013
Comparison
(n = 998)1,3

2013 DR
(n = 185)1,4

2012
Comparison vs.
DR, Difference
(SE)

2013
Comparison vs.
DR, Difference
(SE)

16.4
26.2
7.9
7.5a
8.5
6.3a
22.7a
33.7
8.8
16.3
26.5
7.9

14.7
23.6
7.0
2.6
2.5*
2.8*
22.1a
34.0
8.1
14.9
24.1
7.1

19.8*
29.4*
11.6*
2.0*
2.7*
0.0*
52.7*
72.3*
42.0*
19.7*
30.9*
10.8*

-3.4 (5.74)
-3.2 (12.24)
-3.8 (4.76)
5.5 (2.38)
5.9 (4.16)
6.3 (2.46)
-30.0 (13.45)
-38.6 (23.06)
-33.2 (18.04)
-3.4 (6.27)
-4.4 (13.77)
-2.9 (4.98)

-5.1 (5.71)
-5.8 (11.63)
-4.6 (4.88)
0.6 (2.31)
-0.2 (3.15)
2.8 (1.82)
-30.6 (13.81)
-38.3 (22.49)
-33.9 (18.18)
-4.9 (6.20)
-6.8 (13.09)
-3.7 (5.02)

17.4
26.2
9.7
7.8a
8.5
7.1a
23.7a
33.7
11.1
17.4
26.5
9.8

15.1
23.6
7.7
2.6
2.5*
2.8*
23.0a
34.0
10.1
15.3
24.1
7.9

19.8*
29.4*
11.6*
2.0*
2.7*
0.0*
52.7*
72.3*
42.0*
19.7*
30.9*
10.8*

-2.4 (5.75)
-3.2 (12.24)
-1.9 (4.78)
5.8 (2.44)
5.9 (4.16)
7.1 (2.53)
-29.0 (13.64)
-38.6 (23.06)
-30.9 (18.69)
-2.4 (6.26)
-4.4 (13.77)
-1.1 (4.98)

-4.7 (5.77)
-5.8 (11.63)
-3.9 (5.16)
0.6 (2.31)
-0.2 (3.15)
2.8 (1.82)
-29.7 (13.82)
-38.3 (22.49)
-31.9 (18.25)
-4.5 (6.27)
-6.8 (13.09)
-3.0 (5.34)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Binge Alcohol Use in the 2012 and 2013 comparison data based on only core alcohol module data is defined as drinking five or more
drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. Binge
Alcohol Use in the DR is defined for males as drinking five or more drinks on the same occasion and for females as drinking four or
more drinks on the same occasion on at least 1 day in the past 30 days.
6
Binge Alcohol Use in the 2012 and 2013 comparison data based on core plus noncore data is defined for males as drinking five or more
drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. The
measure for females in the 2012 and 2013 comparison data is defined as drinking five or more drinks on the same occasion on at least 1
day in the past 30 days or usually having four drinks on those days when respondents drank alcohol in the past 30 days based on the
core alcohol module data, or drinking four or more drinks on the same occasion on at least 1 day in the past 30 days (including the last
occasion of alcohol use) based on the noncore consumption of alcohol module data. DR data for binge alcohol use based on the core
alcohol module data are repeated in these rows.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

329

Table 6.40 Lifetime Use of Felt-Tip Pens, Computer Cleaners or Other Inhalants, by Age Group
and Past Year Use of Inhalants according to Types of Inhalants Used in the Lifetime
among Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Inhalant/Age Group
LIFETIME USE
Felt-Tip Pens or Computer Keyboard
Cleaner1
Other Inhalants, Excluding Felt-Tip Pens or
Computer Keyboard Cleaner2
PAST YEAR USE
Among Lifetime Users of Felt-Tip Pens or
Computer Keyboard Cleaner1
Among Lifetime Users of Other Inhalants,
Excluding Users of Felt-Tip Pens or
Computer Keyboard Cleaner2

Aged 12 or
Older
(n = 3,012)

Aged 12 to 17
(n = 707)

Aged 18 to 25
(n = 702)

Aged 26 or
Older
(n = 1,603)

3.9

7.9

7.4

2.9

7.2

1.4

2.9

8.5

12.4

31.5*

7.4*

8.5*

3.3

69.4*

19.6*

1.1*

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
NOTE: Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who
completed the interview in English also have been excluded for these comparisons.
NOTE: Questionnaire Field Test data collected from September 1 through November 3, 2012. Dress Rehearsal data collected from
September 1 through October 31, 2013.
NOTE: Denominators for lifetime use estimates consist of the combined QFT and DR sample for persons aged 12 or older or within the
specific age groups. Denominators for past year use estimates among persons aged 12 or older consist of lifetime users of
inhalants aged 12 or older who reported use of felt-tip pens or computer keyboard cleaner (n = 182) or who reported lifetime use
of other inhalants but not these two specific inhalants (n = 185).
1
Estimates could include lifetime use of other inhalants in addition to lifetime use of felt-tip pens, felt-tip markers, or magic markers; or
computer cleaner, also known as air duster.
2
Other inhalants in the combined QFT and DR include the following: amyl nitrite, "poppers," locker room odorizers, or "rush";
correction fluid, degreaser, or cleaning fluid; gasoline or lighter fluid; glue, shoe polish, or toluene; halothane, ether, or other
anesthetics; lacquer thinner or other paint solvents; lighter gases, such as butane or propane; nitrous oxide or "whippits"; spray paints;
other aerosol sprays, or other inhalants besides those that were listed.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013.

330

Table 6.41 Use of Hallucinogens in Lifetime among Persons Aged 12 or Older with or without
Noncore Hallucinogen Data, by Age Group for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

2012
Comparison vs.
2013
Combined Combined QFT Comparison vs.
2012 QFT
and DR,
Combined QFT
and 2013 DR
Difference
and DR,
(n = 3,012)1,4
(SE)
Difference (SE)

15.3
15.8

16.2
16.7

17.2
17.2

-1.9 (1.12)
-1.4 (1.12)

-0.9 (1.15)
-0.5 (1.14)

Aged 12 to 17
Core Only (without Noncore Data)5
Core Plus Noncore5

3.1
3.7

2.5a
2.7

4.5
4.5

-1.4 (0.93)
-0.9 (0.93)

-2.0 (0.93)
-1.8 (0.93)

Aged 18 to 25
Core Only (without Noncore Data)5
Core Plus Noncore5

18.2
20.3

17.8
19.4

19.9
19.9

-1.6 (2.03)
0.5 (2.01)

-2.1 (1.99)
-0.5 (2.00)

Aged 26 or Older
Core Only (without Noncore Data) 5
Core Plus Noncore5

16.2
16.4

17.5
17.8

18.2
18.2

-1.9 (1.32)
-1.7 (1.32)

-0.6 (1.36)
-0.4 (1.36)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
For the 2012 and 2013 comparison data, Core-Only estimates are based on use of any of the following: LSD, also called "acid"; PCP,
also called "angel dust" or phencyclidine; peyote; mescaline; psilocybin; "Ecstasy," also called MDMA; or any other hallucinogen.
Core Plus Noncore estimates are based on use of any of the hallucinogens from the core, plus the following: ketamine, also called
"Special K" or "Super K"; DMT, AMT, or 5-MeO-DIPT ("Foxy"); or Salvia divinorum. Combined QFT and DR estimates are based on
use of any of the hallucinogens available in the Core Plus Noncore data for the 2012 and 2013 comparison data. The Core-Only
estimate for the Combined QFT and DR is repeated in the Core Plus Noncore row.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

331

Table 6.42 Misuse of Prescription Drugs or Methamphetamine in the Lifetime among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Drug Measure
Prescription Drug Misuse5,6
Pain Reliever Misuse
Tranquilizer Misuse
Sedative Misuse
Stimulant Misuse, Standard
Definition6,7
Stimulant Misuse, QFT Definition8
Methamphetamine Use5
Illicit Drugs, Standard Definition5,6,9
Alternate Definition 310
Illicit Drugs Other Than Marijuana,
Standard Definition5,6,9

Combined
2012 QFT
and 2013
DR
(n = 3,012)1,4
14.8
12.0
5.6
3.4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
6.5 (0.95)
2.5 (0.81)
3.8 (0.64)
-0.1 (0.49)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
5.8 (0.94)
1.8 (0.82)
3.7 (0.61)
-0.4 (0.49)

2012
Comparison
(n = 55,232)1,2
21.3a
14.4a
9.4a
3.3

2013
Comparison
(n = 26,617)1,3
20.6a
13.8a
9.3a
3.0

8.7
N/A
4.9a

8.7
N/A
4.8a

10.2
4.3
7.4

-1.5 (0.85)
N/A (N/A)
-2.6 (0.74)

-1.5 (0.87)
N/A (N/A)
-2.6 (0.75)

49.3
46.1

50.3
47.0

48.8
46.3

0.5 (1.45)
-0.2 (1.49)

1.4 (1.44)
0.7 (1.50)

30.7

30.4

31.2

-0.5 (1.26)

-0.8 (1.27)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the 2012 and 2013
comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core plus noncore data).
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data, but is not included for the Combined 2012 QFT and 2013
DR.
7
Estimate for the Combined 2012 QFT and 2013 DR includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the Combined 2012 QFT and 2013 DR includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type
psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens,
inhalants, or prescription-type psychotherapeutics that was misused. For the Combined 2012 QFT and 2013 DR, both measures also
included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because methamphetamine is included
as a stimulant in the 2012 and 2013 comparison data, methamphetamine users in these data by definition also are misusers of
stimulants and psychotherapeutics. However, comparison data respondents who misused psychotherapeutics but did not use
methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

332

Table 6.42sp Misuse of Prescription Drugs or Methamphetamine in the Lifetime among Persons
Aged 12 or Older for Spanish-Language Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and 2013 Dress
Rehearsal

2012
Comparison
vs. DR,
Difference
(SE)
0.2 (4.54)
-2.0 (4.43)
2.8 (0.83)
0.1 (0.40)

2013
Comparison
vs. DR,
Difference
(SE)
0.3 (4.35)
-3.4 (4.24)
3.3 (0.92)
-0.0 (0.46)
0.4 (0.85)
N/A (N/A)
0.8 (0.37)
-3.3 (5.86)
-4.9 (5.27)
-4.9 (6.02)

2012
Comparison
(n = 2,061)1,2
9.1
6.8
2.9a
0.5

2013
Comparison
(n = 998)1,3
9.3
5.4
3.4a
0.4

0.6
N/A
0.3a

1.2
N/A
0.8a

0.7*
0.7*
0.0*

18.6
11.7

16.6
9.6

19.9*
14.4*

-0.1 (0.76)
N/A (N/A)
0.3 (0.16)
-1.3 (5.84)
-2.7 (5.05)

13.9

11.7

16.7*

-2.7 (6.08)

2013 DR
(n = 185)1,4
9.0*
8.9*
0.1
0.4*

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the 2012 and 2013
comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core plus noncore data).
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data, but is not included for the 2013 DR.
7
Estimate for the 2013 DR includes data for methamphetamine and misuse of prescription stimulants.
8
Estimate for the 2013 DR includes data only for misuse of prescription stimulants.
9
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type
psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens,
inhalants, or prescription-type psychotherapeutics that was misused. For the 2013 DR, both measures also included methamphetamine.
10
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because methamphetamine is included
as a stimulant in the 2012 and 2013 comparison data, methamphetamine users in these data by definition also are misusers of
stimulants and psychotherapeutics. However, comparison data respondents who misused psychotherapeutics but did not use
methamphetamine are not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

333

Table 6.43 Misuse of Prescription Drugs or Methamphetamine in the Past Year among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Drug Measure
Prescription Drug Misuse5,6
Pain Reliever Misuse
OxyContin® Misuse7
Tranquilizer Misuse
Sedative Misuse
Stimulant Misuse, Standard
Definition5,8
Stimulant Misuse, QFT Definition9
Methamphetamine Use5
Illicit Drugs, Standard Definition5,6,10
Alternate Definition 311
Illicit Drugs Other Than Marijuana,
Standard Definition5,6,10

Combined
2012 QFT
and 2013
DR
(n = 3,012)1,4
8.0
5.7
0.9
2.7
0.8

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-1.7 (0.67)
-1.0 (0.56)
-0.3 (0.24)
-0.3 (0.35)
-0.6 (0.20)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-2.3 (0.67)
-1.6 (0.56)
-0.4 (0.25)
-0.7 (0.36)
-0.7 (0.20)

2012
Comparison
(n = 55,232)1,2
6.3a
4.7
0.6
2.3
0.2a

2013
Comparison
(n = 26,617)1,3
5.7a
4.1a
0.6
2.0
0.1a

1.3a
N/A
0.4
15.9
13.1

1.5a
N/A
0.5
16.0
13.6

2.5
1.9
0.7
16.7
12.9

-1.2 (0.36)
N/A (N/A)
-0.3 (0.17)
-0.8 (1.01)
0.1 (0.96)

-1.0 (0.37)
N/A (N/A)
-0.2 (0.18)
-0.7 (1.02)
0.7 (0.96)

8.1a

7.4a

9.8

-1.7 (0.75)

-2.4 (0.75)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable; QFT = Questionnaire Field Test.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Estimates of Any Prescription Drug Misuse, Stimulant Misuse, Methamphetamine Use, and Illicit Drug Use for the 2012 and 2013
comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core plus noncore data).
6
Prescription Drug Misuse includes pain reliever, tranquilizer, sedative, or stimulant misuse. Methamphetamine is included as a
stimulant and a prescription drug for the 2012 and 2013 comparison data, but is not included for the 2012 QFT and 2013 DR.
7
Lifetime and Past Month misuse of OxyContin® are not shown because these estimates cannot be produced from the 2012 QFT and
2013 DR.
8
Estimate for the Combined 2012 QFT and 2013 DR includes data for methamphetamine and misuse of prescription stimulants.
9
Estimate for the Combined 2012 QFT and 2013 DR includes data only for misuse of prescription stimulants.
10
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type
psychotherapeutics that was misused. Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens,
inhalants, or prescription-type psychotherapeutics that was misused. For the Combined 2012 QFT and 2013 DR, both measures also
included methamphetamine.
11
Illicit drugs in this definition include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or
methamphetamine, but do not include prescription-type psychotherapeutics that were misused. Because methamphetamine is included
as a stimulant in the 2012 and 2013 comparison data, methamphetamine users in these data by definition also are misusers of stimulants
and psychotherapeutics. However, comparison data respondents who misused psychotherapeutics but did not use methamphetamine are
not included.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

334

Table 6.44 Misuse of Prescription Drugs or Methamphetamine in the Past Month among Persons
Aged 12 or Older for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Drug Measure
Prescription Drug Misuse5,6
Pain Reliever Misuse
Tranquilizer Misuse
Sedative Misuse
Stimulant Misuse, Standard
Definition5,7
Stimulant Misuse, QFT Definition8
Methamphetamine Use5
Illicit Drugs, Standard Definition5,6,9
Alternate Definition 310
Illicit Drugs Other Than Marijuana,
Standard Definition5,6,9

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.0 (0.35)
0.2 (0.27)
0.0 (0.20)
-0.1 (0.10)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
-0.3 (0.36)
-0.0 (0.29)
-0.2 (0.20)
-0.2 (0.10)

2012
Comparison
(n = 55,232)1,2
2.6
1.9
0.8
0.1

2013
Comparison
(n = 26,617)1,3
2.3
1.7
0.6
0.0

Combined
2012 QFT
and 2013
DR
(n = 3,012)1,4
2.6
1.7
0.8
0.2

0.5
N/A
0.1

0.5
N/A
0.2

0.7
0.4
0.4

-0.3 (0.17)
N/A (N/A)
-0.2 (0.14)

-0.2 (0.18)
N/A (N/A)
-0.2 (0.15)

9.3
7.9

9.4
8.3

9.7
8.2

-0.4 (0.74)
-0.4 (0.70)

-0.3 (0.76)
0.1 (0.71)

3.5

3.2

3.4

0.1 (0.39)

-0.1 (0.41)

335

Table 6.45 Misuse of Stimulants in the Lifetime among Persons Aged 12 or Older with or without
Noncore Adderall® Data, by Age Group for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 12 to 17
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 18 to 25
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT Definition7
Aged 26 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7

2012
Comparison vs.
2013
Combined 2012 Combined QFT Comparison vs.
QFT and 2013
and DR,
Combined
DR
Difference
QFT and DR,
(n = 3,012)1,4
(SE)
Difference (SE)

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

8.7

10.2

-1.5 (0.85)

-1.5 (0.87)

10.3
N/A

10.2
4.3

0.1 (0.84)
N/A (N/A)

0.1 (0.88)
N/A (N/A)

2.1

-0.1 (0.60)

-0.0 (0.61)

3.7a
N/A

3.6a
N/A

2.1
1.9

1.6 (0.60)
N/A (N/A)

1.4 (0.61)
N/A (N/A)

10.5a

9.6a

15.3

-4.7 (1.79)

-5.6 (1.75)

17.3
N/A

16.7
N/A

15.3
13.0

2.1 (1.82)
N/A (N/A)

1.5 (1.79)
N/A (N/A)

9.2

9.3

10.3

-1.1 (1.02)

-1.1 (1.05)

9.9
N/A

10.1
N/A

10.3
3.2

-0.4 (1.01)
N/A (N/A)

-0.2 (1.05)
N/A (N/A)

336

Table 6.46 Misuse of Stimulants in the Past Year among Persons Aged 12 or Older with or without
Noncore Adderall® Data, by Age Group for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 12 to 17
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 18 to 25
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 26 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

2012
Comparison vs.
2013
Combined 2012 Combined QFT Comparison vs.
QFT and 2013
and DR,
Combined
DR
Difference
QFT and DR,
(n = 3,012)1,4
(SE)
Difference (SE)

1.3a

1.5a

2.5

-1.2 (0.36)

-1.0 (0.37)

2.0
N/A

2.2
N/A

2.5
1.9

-0.4 (0.36)
N/A (N/A)

-0.3 (0.38)
N/A (N/A)

1.5

1.4

1.6

-0.2 (0.54)

2.4
N/A

2.3
N/A

1.6
1.5

0.8 (0.53)
N/A (N/A)

0.7 (0.55)
N/A (N/A)

4.3a

3.9a

11.3

-6.9 (1.69)

-7.3 (1.70)

7.7a
N/A

7.0a
N/A

11.3
10.4

-3.5 (1.74)
N/A (N/A)

-4.2 (1.75)
N/A (N/A)

0.8

1.1

1.2

-0.4 (0.28)

-0.1 (0.30)

1.1
N/A

1.4
N/A

1.2
0.6

-0.1 (0.28)
N/A (N/A)

0.2 (0.31)
N/A (N/A)

337

Table 6.47 Misuse of Stimulants in the Past Month among Persons Aged 12 or Older with or
without Noncore Adderall® Data, by Age Group for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 12 to 17
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 18 to 25
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7
Aged 26 or Older
Standard Definition5
Standard Definition, Plus Noncore
Adderall®6
QFT and DR Definition7

2012
Comparison vs.
2013
Combined 2012 Combined QFT Comparison vs.
QFT and 2013
and DR,
Combined
DR
Difference
QFT and DR,
(n = 3,012)1,4
(SE)
Difference (SE)

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

0.5

0.7

-0.3 (0.17)

-0.2 (0.18)

0.7
N/A

0.8
N/A

0.7
0.4

-0.0 (0.18)
N/A (N/A)

0.0 (0.19)
N/A (N/A)

0.6

0.4

0.5

0.1 (0.27)

-0.1 (0.26)

0.8
N/A

0.5
0.4

0.3 (0.27)
N/A (N/A)

1.3

1.4

2.1

-0.9 (0.58)

-0.8 (0.58)

2.3
N/A

2.2
N/A

2.1
1.9

0.1 (0.59)
N/A (N/A)

0.3

0.4

0.5

-0.2 (0.19)

-0.1 (0.20)

0.4
N/A

0.5
N/A

0.5
0.1

-0.1 (0.19)
N/A (N/A)

0.0 (0.20)
N/A (N/A)

338

Table 6.48 Misuse of Sedatives in the Lifetime among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

2012
Comparison vs.
2013
Combined Combined QFT Comparison vs.
and DR,
Combined QFT
2012 QFT
Difference
and DR,
and 2013 DR
(SE)
Difference (SE)
(n = 3,012)1,4

3.3
5.3a

3.0
5.0a

3.4
3.4

-0.1 (0.49)
1.9 (0.50)

-0.4 (0.49)
1.6 (0.51)

Aged 12 to 17
Core Only (without Noncore Data)5
Core Plus Noncore5

0.7
1.6a

0.5
1.2a

0.6
0.6

0.1 (0.31)
1.0 (0.31)

-0.1 (0.32)
0.6 (0.33)

Aged 18 to 25
Core Only (without Noncore Data)5
Core Plus Noncore5

1.3
4.3a

1.2
4.0a

2.0
2.0

-0.8 (0.69)
2.3 (0.68)

-0.9 (0.69)
2.0 (0.71)

Aged 26 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

3.9
5.9a

3.5
5.6a

3.9
3.9

-0.0 (0.61)
2.0 (0.62)

-0.4 (0.61)
1.7 (0.62)

339

Table 6.49 Misuse of Sedatives in the Past Year among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

2012
Comparison vs.
2013
Combined Combined QFT Comparison vs.
and DR,
Combined QFT
2012 QFT
Difference
and DR,
and 2013 DR
(SE)
Difference (SE)
(n = 3,012)1,4

0.2a
0.9

0.1a
0.7

0.8
0.8

-0.6 (0.20)
0.0 (0.20)

-0.7 (0.20)
-0.2 (0.20)

Aged 12 to 17
Core Only (without Noncore Data)5
Core Plus Noncore5

0.3
0.8

0.2
0.6

0.6
0.6

-0.3 (0.31)
0.2 (0.30)

-0.4 (0.32)
0.0 (0.32)

Aged 18 to 25
Core Only (without Noncore Data)5
Core Plus Noncore5

0.4
1.5

0.4
1.2

1.3
1.3

-0.9 (0.49)
0.2 (0.50)

-0.9 (0.50)
-0.1 (0.51)

Aged 26 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

0.2a
0.8

0.1a
0.6

0.8
0.8

-0.6 (0.23)
-0.0 (0.24)

-0.7 (0.23)
-0.2 (0.23)

340

Table 6.50 Misuse of Sedatives in the Past Year among Persons Aged 12 or Older with or without
Noncore Ambien® Data, by Age Group for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Age Group/Drug Measure
Aged 12 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

2012
Comparison
(n = 55,232)1,2

2013
Comparison
(n = 26,617)1,3

2012
Comparison vs.
2013
Combined Combined QFT Comparison vs.
and DR,
Combined QFT
2012 QFT
Difference
and DR,
and 2013 DR
(SE)
Difference (SE)
(n = 3,012)1,4

0.1
0.3

0.0
0.2

0.2
0.2

-0.1 (0.10)
0.1 (0.10)

-0.2 (0.10)
-0.1 (0.10)

Aged 12 to 17
Core Only (without Noncore Data)5
Core Plus Noncore5

0.1
0.2

0.1
0.1

0.0 (0.10)
0.1 (0.10)

-0.0 (0.10)
0.1 (0.10)

Aged 18 to 25
Core Only (without Noncore Data)5
Core Plus Noncore5

0.2
0.4a

0.0
0.1

0.1
0.1

0.1 (0.09)
0.3 (0.10)

-0.1 (0.09)
0.0 (0.09)

Aged 26 or Older
Core Only (without Noncore Data)5
Core Plus Noncore5

0.1
0.3

0.0
0.2

0.3
0.3

-0.2 (0.13)
0.0 (0.13)

-0.2 (0.13)
-0.1 (0.12)

341

Table 6.51 Substance Dependence or Abuse in the Past Year among Persons Aged 12 or Older for
English-Language Non-Hispanic Interviews, by Survey Protocol: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Dependence or Abuse Measure
DEPENDENCE
Illicit Drugs5
Marijuana
Hallucinogens
Inhalants
Prescription Drugs6
Pain Relievers
Stimulants Among
Methamphetamine Users
Methamphetamine
Illicit Drugs Other Than Marijuana5
Illicit Drugs Excluding Marijuana7
ABUSE
Illicit Drugs5
Marijuana
Hallucinogens
Inhalants
Prescription Drugs6
Pain Relievers
Illicit Drugs Other Than Marijuana5
Illicit Drugs Excluding Marijuana7
DEPENDENCE OR ABUSE
Illicit Drugs5
Marijuana
Hallucinogens
Inhalants
Prescription Drugs6
Pain Relievers
Illicit Drugs Other Than Marijuana5
Illicit Drugs Excluding Marijuana7

2012
Comparison
(n = 55,232)1,2

2013 Quarters
3 and 4
Comparison
(n = 26,617)1,3

Combined
2012 QFT
and 2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2.0
1.0
0.0a
0.0
0.8
0.6

1.8
1.0
0.0*
0.0
0.7
0.6

1.9
1.0
0.0
0.0
0.8
0.6

0.1 (0.25)
0.1 (0.18)
0.0 (0.01)
0.0 (0.01)
0.0 (0.15)
0.1 (0.13)

-0.0 (0.24)
0.0 (0.18)
-0.0 (0.01)
0.0 (0.01)
-0.1 (0.16)
0.1 (0.14)

0.1
N/A
1.2
1.0

0.0*
N/A
0.9
0.8

N/A
0.2
1.0
0.9

N/A (N/A)
N/A (N/A)
0.1 (0.21)
0.1 (0.20)

N/A (N/A)
N/A (N/A)
-0.1 (0.21)
-0.1 (0.21)

0.8
0.6
0.1
0.0
0.3
0.2
0.4
0.3

0.6
0.4
0.0*
0.0
0.2
0.1
0.2
0.2

0.7
0.5
0.1
0.0
0.3
0.2
0.3
0.4

0.1 (0.15)
0.1 (0.12)
0.0 (0.03)
0.0 (0.02)
-0.0 (0.11)
0.0 (0.09)
0.0 (0.12)
-0.1 (0.13)

-0.1 (0.16)
-0.1 (0.12)
-0.1 (0.03)
0.0 (0.02)
-0.1 (0.11)
-0.1 (0.08)
-0.1 (0.11)
-0.2 (0.12)

2.8
1.6
0.1a
0.0
1.0
0.8
1.5
1.3

2.4
1.4
0.0*
0.0
0.8
0.8
1.2
1.1

2.6
1.5
0.0*
0.0
1.1
0.8
1.4
1.3

0.3 (0.30)
0.2 (0.22)
0.1 (0.01)
0.0 (0.02)
-0.0 (0.19)
0.1 (0.16)
0.1 (0.23)
0.0 (0.23)

-0.1 (0.30)
-0.0 (0.23)
0.0 (0.00)
0.0 (0.02)
-0.2 (0.19)
0.0 (0.17)
-0.2 (0.23)
-0.2 (0.23)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; N/A = not applicable; QFT = Questionnaire Field Test.
NOTE: Dependence or abuse is based on definitions found in the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM-IV).
a

Difference between estimate and Combined 2012 QFT and 2013 DR is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Illicit Drugs include marijuana/hashish, cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics.
Illicit Drugs Other Than Marijuana include cocaine (including crack), heroin, hallucinogens, inhalants, or prescription-type psychotherapeutics.
Estimates for the Combined 2012 QFT and 2013 DR include relevant dependence or abuse data for methamphetamine.
6
Estimates for Prescription Drugs include misuse of pain relievers, tranquilizers, stimulants, or sedatives. Estimates for the Combined 2012 QFT
and 2013 DR do not include dependence or abuse data for methamphetamine.
7
Illicit Drugs Excluding Marijuana include dependence or abuse for cocaine, heroin, hallucinogens, inhalants, or prescription-type
psychotherapeutics and require respondents not to have corresponding dependence or abuse for marijuana. Estimates for the Combined 2012
QFT and 2013 DR include relevant dependence or abuse data for methamphetamine.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

342

Table 6.52 Substance Use with a Needle in the Lifetime, Past Year, and Past Month among
Persons Aged 12 or Older for English-Language Non-Hispanic Interviews:
Percentages, Differences, and Standard Error of Differences, 2012 Comparison, 2013
Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Substance Used with a
Needle/Period of Use
USE OF HEROIN WITH A
NEEDLE
Lifetime
Past Year
Past Month
USE OF COCAINE WITH A
NEEDLE
Lifetime
Past Year
Past Month
USE OF METHAMPHETAMINE
WITH A NEEDLE
Lifetime
Past Year
Past Month
USE OF PRESCRIPTION
STIMULANTS WITH A
NEEDLE5
Past Year
Past Month
USE OF HEROIN, COCAINE,
METHAMPHETAMINE, OR
PRESCRIPTION STIMULANTS
WITH A NEEDLE5
Past Year
Past Month

2012
Comparison
(n = 55,232)1,2

2013 Quarters
3 and 4
Comparison
(n = 26,617)1,3

Combined
2012 QFT
and 2013 DR
(n = 3,012)1,4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

0.9
0.2
0.1a

1.0
0.1
0.0

0.7
0.1
0.0

0.1 (0.27)
0.0 (0.06)
0.1 (0.02)

0.3 (0.27)
0.0 (0.07)
0.0 (0.01)

0.9
0.1a
0.0a

1.1
0.1a
0.0

1.0
0.0*
0.0*

-0.2 (0.32)
0.1 (0.01)
0.0 (0.01)

0.0 (0.31)
0.1 (0.02)
0.0 (0.01)

0.7
0.1
0.0

0.8
0.1
0.0

1.0
0.2
0.1

-0.2 (0.31)
-0.1 (0.09)
-0.1 (0.07)

-0.2 (0.30)
-0.1 (0.09)
-0.1 (0.07)

0.1a
0.0a

0.0*
0.0*

0.1 (0.01)
0.0 (0.01)

0.1 (0.02)
0.0 (0.01)

0.2
0.1

0.0 (0.09)
-0.0 (0.07)

343

Table 6.53 Perceived Great Risk of Harm Associated with Substance Use among Persons Aged 12
or Older for English-Language Non-Hispanic Interviews: Percentages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal

Perception of Great Risk1
PERCEPTIONS OF GREAT
RISK – CIGARETTES
Smoke one or more packs per
day
PERCEPTIONS OF GREAT
RISK – MARIJUANA
Smoke once a month
Smoke once or twice a week

2012
Comparison
(n = 55,232)2,3

2013
Comparison
(n = 26,617)2,4

Combined
2012 QFT
and 2013 DR
(n = 3,012)2,5

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

70.6

69.3

69.2

1.4 (1.38)

0.1 (1.36)

28.8
38.5

26.1a
35.1

29.1
37.0

-0.3 (1.35)
1.5 (1.43)

-2.9 (1.38)
-1.9 (1.44)

344

Table 6.54 Number of Years Since Last Use for Selected Substances among Lifetime Users Aged
12 to 49 for English-Language Non-Hispanic Interviews: Averages, Differences, and
Standard Error of Differences, 2012 Comparison, 2013 Comparison, and Combined
2012 Questionnaire Field Test and 2013 Dress Rehearsal

Substance
Cigarettes
Alcohol
Marijuana
Cocaine
Hallucinogens
Inhalants

2012
Comparison
(n = 48,288)1,2
10.5
2.5a
9.9a
10.5
11.3
13.7

2013
Comparison
(n = 23,236)1,3
10.6
2.5
9.9a
10.6
11.4
14.4

Combined
2012 QFT
and 2013 DR
(n = 2,454)1,4
10.2
3.2
8.6
10.6
11.0
14.8

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.3 (0.49)
-0.7 (0.35)
1.3 (0.44)
-0.1 (0.70)
0.3 (0.68)
-1.1 (0.89)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.4 (0.51)
-0.7 (0.37)
1.4 (0.45)
-0.0 (0.72)
0.4 (0.71)
-0.3 (0.88)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
NOTE: If respondents reported last using a substance more than 30 days ago but within the past 12 months, the number of years since
last use was assumed to be zero, regardless of whether they reported last use more than a year ago based on the age, year, or
month when they last used. In addition, the number of years since last use was set to zero for past month substance users, but
they were not asked the questions pertaining to prior substance use.
NOTE: Within each set of data, sample sizes will vary by substance because nonusers of the substance were excluded from the
analysis.
a
Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed
the interview in English also have been excluded for these comparisons. Sample sizes are for all respondents aged 12 to 49 after these
exclusions had been made. Sample sizes for the specific drugs will vary based on the numbers of lifetime users.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

345

Table 6.55 Received Substance Use Treatment in the Lifetime and Past Year and Types of Past
Year Substance Use Treatment among Persons Aged 12 or Older for English-Language
Non-Hispanic Interviews: Percentages, Differences, and Standard Error of Differences,
2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and
2013 Dress Rehearsal

Substance Use Treatment
LIFETIME TREATMENT
PAST YEAR
TREATMENT
Alcohol Use Only
Drug Use Only
Both Alcohol and Drug
Use

2012
Comparison vs.
Combined QFT
and DR,
Difference
(SE)
-0.3 (0.65)

2013
Comparison vs.
Combined QFT
and DR,
Difference (SE)
-0.3 (0.65)

2012
Comparison
(n = 55,232)1,2
6.3

2013
Comparison
(n = 26,617)1,3
6.3

Combined
2012 QFT
and 2013 DR
(n = 3,012)1,4
6.6

1.4
0.6a
0.4

1.3
0.3
0.5

0.2 (0.24)
0.3 (0.10)
-0.0 (0.14)

0.1 (0.23)
0.3 (0.10)
-0.1 (0.14)

0.4

-0.0 (0.14)

-0.1 (0.14)

346

Table 6.56 Adult Mental Health Treatment in the Past Year and Type of Facility Where Received
Treatment among Persons Aged 18 or Older for English-Language Non-Hispanic
Interviews: Percentages, Differences, and Standard Error of Differences, 2012
Comparison, 2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013
Dress Rehearsal

Past Year Mental Health Treatment1
STAYED OVERNIGHT IN HOSPITAL
FOR MENTAL HEALTH
TREATMENT
FACILITY TYPE – OVERNIGHT
MENTAL HEALTH TREATMENT6
Private or Public Psychiatric Hospital
Psychiatric Unit – General Hospital
Medical unit – General Hospital
Another Type of Hospital
Residential Treatment Center
Other Facility

2012
Comparison
(n = 37,659)2,3

2013
Comparison
(n = 18,007)2,4

Combined
2012 QFT and
2013 DR
(n = 2,305)2,5

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

0.8

1.0

-0.2 (0.27)

-0.1 (0.26)

0.2
0.2
0.2
0.1
0.1
0.1

0.2
0.2
0.3
0.1
0.1
0.1

0.2
0.2
0.3
0.1
0.0
0.2

-0.0 (0.18)
0.0 (0.08)
-0.1 (0.12)
-0.0 (0.11)
0.0 (0.05)
-0.1 (0.08)

-0.1 (0.18)
0.0 (0.08)
-0.0 (0.12)
-0.0 (0.11)
0.1 (0.05)
-0.1 (0.07)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a

Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.

Respondents with unknown mental health treatment information were excluded.
2
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
3
2012 comparison data collected in quarters 1 through 4, 2012.
4
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
5
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
6
Respondents could indicate multiple locations for treatment; thus, these response categories are not mutually exclusive.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

347

Table 6.57 Youth Mental Health Treatment in the Past Year and Number of Nights Received
Treatment among Persons Aged 12 to 17 for English-Language Non-Hispanic
Interviews: Percentages, Chi-Square Test Statistic and P Value, 2012 Comparison,
2013 Comparison, and Combined 2012 Questionnaire Field Test and 2013 Dress
Rehearsal

Past Year Mental Health Treatment1
STAYED OVERNIGHT IN HOSPITAL
FOR MENTAL HEALTH TREATMENT
Yes
No
NUMBER OF NIGHTS IN HOSPITAL FOR
MENTAL HEALTH TREATMENT
1 Night
2 to 6 Nights
7 or More Nights
STAYED OVERNIGHT IN RESIDENTIAL
TREATMENT CENTER FOR MENTAL
HEALTH TREATMENT
Yes
No
NUMBER OF NIGHTS IN RESIDENTIAL
TREATMENT CENTER FOR MENTAL
HEALTH TREATMENT
1 Night
2 to 6 Nights
7 or More Nights

2012
Comparison
(n = 17,573)2,3

2013
Comparison
(n = 8,610)2,4

Combined
2012 QFT
and 2013
DR
(n = 707)2,5

1.9
98.1

2.2
97.8

2.1
97.9

49.9
28.4
21.6

50.5
27.5
21.9

54.5*
26.0*
19.5*

0.9
99.1

1.1
98.9

1.4
98.6

26.8
29.8
43.4

39.4
20.7
39.9

27.1*
44.0*
28.9*

2012
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value

2013
Comparison
vs. Combined
QFT and DR
Chi-Square
Statistic,
P Value

0.08, 0.7746

0.85, 0.3564

0.10, 0.9080

0.29, 0.7496

0.27, 0.6032

0.24, 0.6222

0.21, 0.8079

0.41, 0.6630

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
c

Interaction between the characteristic and survey is significant at the 0.05 level.

Respondents with unknown mental health treatment information were excluded.
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
3
2012 comparison data collected in quarters 1 through 4, 2012.
4
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
5
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.
2

348

Table 6.58 Selected Mental Health Measures among Persons Aged 18 or Older for EnglishLanguage Non-Hispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal

Mental Health Measure
Past Month SPD5
Past Year SPD5
Past Year Thoughts of Suicide6
Past Year Suicide Plans6
Past Year Attempted Suicide6
Several Days or Longer Felt Sad, Empty, or
Depressed7
Several Days When Most of the Day Felt Very
Discouraged7
Several Days or Longer Lost Interest in Things
Usually Enjoyable7
Average Past Month Total K6 Score8
Average Past Year Worst K6 Total Score8
Average WHODAS Score (0 to 24)

2012
Comparison
(n = 37,659)1,2
5.1a
10.6
3.9
1.2
0.5

2013
Comparison
(n = 18,007)1,3
5.0
10.6
3.8
1.2
0.5

Combined
2012 QFT
and 2013 DR
(n = 2,305)1,4
4.0
9.2
3.4
0.8
0.4

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
1.1 (0.52)
1.5 (0.78)
0.5 (0.53)
0.3 (0.20)
0.1 (0.13)

31.3

31.2

30.8

0.5 (1.15)

0.4 (1.16)

12.4

12.0

12.2

0.3 (1.20)

-0.1 (1.19)

4.2
3.8
4.9
3.7

4.0
3.8
4.9
3.7

5.8
3.6
4.7
3.4

-1.6 (0.97)
0.2 (0.12)
0.2 (0.16)
0.2 (0.15)

-1.8 (0.97)
0.2 (0.12)
0.2 (0.16)
0.2 (0.15)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
1.0 (0.54)
1.4 (0.77)
0.4 (0.52)
0.4 (0.22)
0.1 (0.14)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; K6 = Kessler 6; QFT = Questionnaire Field Test; SPD = serious psychological distress; WHODAS = World Health
Organization Disability Assessment Schedule.
a

Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.

Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
SPD is defined as having a score of 13 or higher on the K6 scale.
6
Respondents with unknown suicide information were excluded.
7
Respondents with unknown depression information were excluded.
8
The K6 score is derived from 12 questions asking the frequency that a respondent experienced symptoms of psychological distress. Six new
questions were asked for the first time in 2008 to all respondents aged 18 or older about their past 30-day symptoms. Responses to these six
questions are combined to produce the past month score ranging from 0 to 24. The original six questions are then only asked respondents who
reported that there was a month in the past year when they felt more symptoms than they felt in the past 30 days, and a score ranging from 0 to
24 is produced. The maximum of these two scores is taken to create the past year K6 score.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

349

Table 6.59 Adolescent Depression Characteristics among Persons Aged 12 to 17 for EnglishLanguage Non-Hispanic Interviews: Percentages, Differences, and Standard Error of
Differences, 2012 Comparison, 2013 Comparison, and Combined 2012 Questionnaire
Field Test and 2013 Dress Rehearsal

Depression Characteristic1
Several Days or Longer Felt Sad, Empty or
Depressed
Several Days When Most of the Day Felt
Very Discouraged
Several Days or Longer Lost Interest in
Things Usually Enjoyable

2012
Comparison
(n = 17,573)2,3

2013
Comparison
(n = 8,610)2,4

Combined
2012 QFT
and 2013 DR
(n = 707)2,5

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

2013
Comparison
vs. Combined
QFT and DR,
Difference
(SE)

42.9

44.5

41.5

1.4 (2.19)

3.0 (2.19)

8.6

8.1

0.5 (1.85)

0.0 (1.87)

15.3

15.7

-0.3 (2.58)

0.0 (2.52)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a

Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.

Respondents with unknown depression information were excluded.
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
3
2012 comparison data collected in quarters 1 through 4, 2012.
4
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
5
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.
2

Table 6.60 Arrested and Booked in the Lifetime and Past Year for Breaking the Law among
Persons 12 or Older for English-Language Non-Hispanic Interviews: Percentages,
Differences, and Standard Error of Differences, 2012 Comparison, 2013 Comparison,
and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal

Arrested and Booked1
TIME PERIOD
Lifetime
Past Year

2012
Comparison
(n = 55,232)2,3

2013
Comparison
(n = 26,617)2,4

Combined
2012 QFT and
2013 DR
(n = 3,012)2,5

17.4
3.1

16.5
2.6

16.9
2.9

2012
Comparison
vs. Combined
QFT and DR,
Difference
(SE)
0.5 (1.06)
0.2 (0.36)

2013
Comparison
vs. Combined
QFT and DR,
Difference (SE)
-0.4 (1.08)
-0.2 (0.35)

*Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; QFT = Questionnaire Field Test.
a

Difference between estimate and Combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.

Respondents with unknown arrested and booked information were excluded.
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the
interview in English also have been excluded for these comparisons.
3
2012 comparison data collected in quarters 1 through 4, 2012.
4
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
5
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.
2

350

7. Selected Noncore Estimates for DR,
Comparison Data, and External Data
Sources
7.1

Overview of Selected DR Estimates Compared with Comparison Data
and Other Survey Data

This chapter presents comparisons of estimates from the 2013 Dress Rehearsal (DR) with
estimates from the 2012 comparison sample, the 2013 comparison sample and comparable
sources of data other than the National Survey on Drug Use and Health (NSDUH). Comparable
statistics from other surveys can be used as benchmark tools for evaluating the validity of
estimates from the DR. Such comparisons take into consideration that the external data used in
the comparison have their own error properties and influences, such as mode of administration.
For example, survey modes vary by self-administration versus interviewer administration or use
of paper-and-pencil questionnaires versus computer-assisted interviewing.
Section 7.2 presents comparisons between data from the DR with the comparison data
and external data sources for items that were moved from computer-assisted personal
interviewing (CAPI) to audio computer-assisted self-interviewing (ACASI) administration for
both the 2012 Questionnaire Field Test (QFT) and the DR. DR estimates for measures of income
received from government income support assistance programs, health insurance coverage,
employment, and family income are compared with estimates from the 2012 and 2013 NSDUH
comparison samples and from external data sources, including the National Health Interview
Survey (NHIS), the American Community Survey (ACS), and the Current Population Survey
(CPS). In Section 7.3, estimates from the DR for the height and weight items introduced in the
2012 QFT are compared with estimates from the NHIS and National Health and Nutrition
Examination Survey (NHANES). In addition, items on health conditions, disabilities and
English-language ability that were added to the QFT and DR are compared with estimates from
the NHIS and ACS. Section 7.4 presents comparisons for a new question on sexual orientation
that was added to the DR questionnaire with estimates from two other surveys—the 2012
General Social Survey (GSS) and the 2006 to 2010 National Survey of Family Growth (NSFG).

7.2

Comparisons of Estimates for Items Moved from CAPI to ACASI
Administration

For the QFT, questions in the income and health insurance coverage modules were
administered in ACASI instead of CAPI. For some of these items, estimates from the QFT
differed from estimates from the 2011 and 2012 NSDUH comparison samples and estimates
from external data sources. In particular, QFT estimates differed from other data sources with
respect to income received from government income support assistance programs, health
insurance coverage, employment, and family income.

351

•

QFT estimates for the receipt of Supplemental Security Income (SSI), food stamps,
and welfare payments were generally higher than the estimates from the 2011 and
2012 comparison samples. The QFT estimates for SSI and food stamps were also
higher than the estimates from external data sources (i.e., the NHIS and ACS).

•

There were no statistically significant differences between the QFT estimates for
employment status (full-time, part-time, unemployed, and other) and those from the
2011 and 2012 comparison samples. Unemployment rates from the QFT were also
similar to those from the CPS.

•

QFT estimates for private health insurance coverage (among those who were insured)
were lower than the estimates from the 2011 and 2012 comparison samples and from
the 2011 ACS and 2011 NHIS. In the QFT, 62.1 percent of those with insurance
reported coverage through a private plan. Estimates for the NSDUH comparison
samples and the external data sources ranged from 67.1 percent in the 2011
comparison sample to 68.7 percent in the 2011 NHIS.

•

QFT estimates resulted in higher proportions of persons at lower income levels and
lower proportions at higher income levels compared with the 2011 and 2012
comparison samples and the 2011 NHIS.

Although a decision was made to continue with CAPI administration of these modules
for the 2015 survey, it is nevertheless of interest to the project to determine whether results from
the QFT survey were generalizable to other field tests such as the DR. In this section, estimates
from the DR sample are compared with those from the 2012 and 2013 quarters 3 and 4
comparison samples and estimates from external benchmark surveys.
In Table 7.1, DR estimates for four types of received income or participation in
government assistance programs for all persons aged 12 or older are presented with parallel
estimates from the 2012 and 2013 comparison samples, the 2012 ACS, and the 2012 NHIS.
Several notable comparisons can be observed from this table:
•

For social security, the estimate for the DR sample (23.6 percent) was somewhat
lower than that for the 2012 and 2013 comparison samples, the 2012 ACS, and the
2012 NHIS, all of which were about 26 or 27 percent. However, the difference
between the DR estimate and both the 2012 and 2013 comparison sample estimates
was not statistically significant. This is similar to results from the QFT. In the QFT,
estimates for the receipt of social security were very similar across those in the 2011
and 2012 quarters 3 and 4 comparison samples and in the 2011 ACS and 2011 NHIS,
all at about 27 percent, with no statistically significant differences between the QFT
estimate and estimates for the 2011 and 2012 quarters 3 and 4 comparison samples.

•

For SSI, the DR estimate for all persons aged 12 or older (8.0 percent) appeared to be
higher than the estimates from the NHIS (5.4 percent) and the ACS (6.2 percent).
The difference between the estimate of SSI from the DR sample and both the 2012
and 2013 comparison samples was not statistically significant. These results are
similar to the results from comparisons of the QFT with the NSDUH comparison
samples and external data sources. The QFT estimate for all persons aged 12 or older
(9.4 percent) was especially higher than the estimates from the external sources

352

Table 7.1 Received Income and Program Participation among Persons Aged 12 or Older:
Percentages and Standard Errors for 2012 Comparison, 2013 Comparison, 2013 Dress
Rehearsal, and Other Surveys
Income and Program
Participation
Social Security
Supplemental Security
Income (SSI)
Food Stamps
Welfare Payments

2012
Comparison1,2
Percent (SE)
26.7 (0.44)
7.6 (0.24)
16.4a (0.30)
2.5 (0.12)

2013
Comparison1,3
Percent (SE)
26.8 (0.49)
7.7 (0.31)
16.3a (0.49)
2.1a (0.12)

2013 DR1,4
Percent (SE)
23.6 (2.13)

2012 ACS5
Percent (SE)
26.9 (0.05)

2012 NHIS6
Percent (SE)
25.7 (0.33)

8.0 (1.00)
19.9 (1.47)
3.2 (0.52)

6.2 (0.03)
15.1 (0.05)
3.4 (0.02)

5.4 (0.17)
14.4 (0.30)
3.5 (0.14)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; DR = Dress Rehearsal; NHIS = National Health Interview Survey; SE = standard error.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, or persons in institutional group
quarters.
6
Unknown or invalid data were excluded from the analysis.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013; U.S. Census, American Community Survey, 2012; National Health Interview Survey (NHIS), 2012.

(5.0 percent for the 2011 NHIS and 6.0 percent from the 2011 ACS). Estimates for
SSI from the other surveys were 5.0 percent in the 2011 NHIS and 7.6 percent in the
2012 quarters 3 and 4 comparison sample.
•

The DR estimate for the receipt of assistance from food stamp programs28 for all
persons aged 12 or older (19.9 percent) was also generally higher than the estimates
from the other data sources, ranging from 14.4 percent in the NHIS sample to
16.4 percent in the 2012 comparison sample. The difference between the estimate of
food stamp receipt on the DR and both the 2012 and 2013 comparison samples was
statistically significant. This result was similar to the finding from the QFT in which
the estimate of participation in the QFT sample was higher than the estimate from the
2012 comparison sample. The difference between the DR estimate and the 2012 and
2013 comparison samples appeared slightly larger than the differences between the
QFT estimate and the 2011 and 2012 comparison samples. If the comparison samples
can be considered a point of reference, the difference increased from a 2 or
3 percentage point difference from the QFT to about a 3.5 percentage point difference
in the DR sample.

•

For receipt of welfare payments, such as those from Temporary Assistance for Needy
Families (TANF), the DR estimate for all persons aged 12 or older (3.2 percent) was
higher than the estimate from the 2012 comparison sample (2.5 percent) and the 2013
comparison sample (2.1 percent), but was similar to the ACS estimate (3.4 percent)

Food stamp programs are now more commonly known as the Supplemental Nutrition Assistance
Program (SNAP).

353

and the NHIS estimate (3.5 percent). This result was similar to the result from the
QFT in which the QFT estimate was higher than the estimates from the 2011 and
2012 quarters 3 and 4 comparison samples, but similar to those from the ACS and
NHIS. The difference between the DR estimate and the 2013 comparison sample
estimate was statistically significant.
Estimates of participation in two programs—SSI and food stamps—appeared to be higher for the
DR sample than in the 2012 and 2013 comparison samples as well as estimates from the NHIS
and ACS. In addition, the estimate for the receipt of welfare payments in the DR sample was
higher than those from the 2012 and 2013 comparison samples. These findings reinforce the
results from the QFT, which suggested that QFT respondents were either somewhat lower
overall in socioeconomic status or that the respondents in the QFT sample, even after weighting,
were more likely than respondents in CAPI mode to report participation in these programs in
ACASI mode.
In Table 7.2, DR estimates for four employment categories for all persons aged 18 or
older are presented with parallel estimates from the 2012 and 2013 comparison samples and the
2013 monthly samples for July through October from the 2013 CPS. A few comparisons can be
observed from this table:
•

For all persons aged 18 or older, the DR estimate of persons employed full time
(48.6 percent) was slightly lower than both the 2012 comparison estimate
(50.0 percent) and the 2013 comparison estimate (51.0 percent), and the differences
were not statistically significant. The CPS estimate covering quarter 3 and the first
month of quarter 4 (49.4 percent) was similar to the DR estimate as well.

Table 7.2 Levels of Current Employment among Persons Aged 18 or Older: Percentages and
Standard Errors for 2012 Comparison, 2013 Comparison, 2013 Dress Rehearsal, and
CPS Data

Levels of Current Employment
CURRENT EMPLOYMENT
Full-Time
Part-Time
Unemployed
Other5

2012 Comparison1,2
Percent (SE)

2013
Comparison1,3
Percent (SE)

50.0 (0.45)
14.0 (0.28)
5.8 (0.17)
30.2 (0.39)

51.0 (0.69)
14.0 (0.36)
4.5a (0.17)
30.5 (0.58)

2013 DR1,4
Percent (SE)
48.6 (2.17)
14.0 (1.45)
6.2 (0.83)
31.2 (2.06)

2013 CPS
Q3 & Q41
Percent (SE)
49.4 (0.09)
10.9 (0.56)
4.5 (0.04)
35.2 (0.09)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; DR = Dress Rehearsal; SE = standard error.
1
Sample does not include Alaska or Hawaii.
2
2012 comparison data collected in quarters 1 through 4, 2013.
4
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
3
DR data collected from September 1 through October 31, 2013.
5
The "other" employment category includes students, person keeping house or caring for children full time, retired or disabled
persons, or other persons not in the labor force.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013; Bureau of Labor Statistics and U.S. Census Bureau, Current Population Survey, 2013.

354

•

For all persons aged 18 or older, the DR estimate of persons employed part time
(14.0 percent) was the same as both the 2012 and 2013 comparison sample estimates,
but both of these estimates appeared to be greater than the estimate of 10.9 percent
from the CPS.

•

The DR estimate for being unemployed for all persons aged 18 or older (6.2 percent)
was slightly higher than the 2013 quarters 3 and 4 CPS estimate (4.5 percent), but it
was similar to the 2012 comparison estimate (5.8 percent). Furthermore, the
difference between the 2012 comparison sample and the DR estimate was not
statistically significant. The difference between the DR estimate and the 2013
comparison sample estimate (4.5 percent) was statistically significant.

•

For all persons aged 18 or older, the DR estimate of persons with an employment
status of "other" (31.2 percent), such as being retired or otherwise not in the labor
force, was lower than the 2013 quarters 3 and 4 CPS estimate (35.2 percent), but it
was similar to the 2012 comparison estimate (30.2 percent) and 2013 comparison
sample estimate (30.5 percent).

Table 7.3 provides unemployment rate estimates among persons aged 18 or older for
three age groups for the DR sample, the 2012 and 2013 comparison samples, and the 2013
quarters 3 and 4 CPS. DR unemployment rate estimates were similar to the 2012 comparison
sample and the 2013 quarters 3 and 4 CPS for all persons aged 18 or older and for persons aged
18 to 25. Unemployment rate estimates for the DR sample were higher than the other two
estimates for persons aged 26 or older. Overall, comparisons between the DR and the other
sources of survey data on employment status and unemployment rates showed generally similar
estimates for these measures. No statistically significant differences between the DR and the
2012 comparison sample were uncovered in these analyses for current employment status and
the unemployment rate. However, the difference between the unemployment rate estimate for the
DR sample and the 2013 comparison sample estimate (6.4 percent) was statistically significant.
Table 7.3 Unemployment Rates among Persons Aged 18 or Older, by Age Group: Percentages and
Standard Errors for 2012 Comparison, 2013 Comparison, 2013 Dress Rehearsal, and
CPS Data

Age Group
18 or Older
18 to 25
26 or Older

2012 Comparison1,2
Percent (SE)
8.3 (0.24)
15.3 (0.45)
6.9 (0.27)

2013 Comparison1,3
Percent (SE)
6.4a (0.25)
13.9 (0.56)
5.0 (0.28)

2013 DR1,4
Percent (SE)
9.0 (1.21)
15.4 (2.43)
7.8 (1.41)

2013 CPS
Q3 & Q41
Percent (SE)
7.0 (0.06)
13.4 (0.22)
5.9 (0.58)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
CPS = Current Population Survey; DR = Dress Rehearsal; SE = standard error.
a
Difference between estimate and corresponding 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013; Bureau of Labor Statistics and U.S. Census Bureau, Current Population Survey, 2013.

355

In Table 7.4, DR estimates for four types of health insurance coverage for all persons
aged 12 or older are presented with parallel estimates from the 2012 and 2013 comparison
samples, the 2012 ACS, and the 2012 NHIS. A few notable comparisons can be highlighted from
this table:
•

For all persons aged 12 or older, estimates for the first three types of health insurance
coverage—Medicare; Medicaid; and TRICARE, CHAMPUS, CHAMPVA, or other
military health care sources—are generally different across the four sources of
estimates. For Medicare coverage, the DR estimate was the highest estimate, although
the differences between the DR estimate and those from the 2012 and 2013
comparison samples were not statistically significant. Also, the DR estimate for
Medicaid coverage for all persons aged 12 or older (15.4 percent) was slightly higher
than the parallel estimates from both the 2012 and 2013 comparison samples
(12.1 and 12.6 percent, respectively), the NHIS (11.7 percent), and the ACS
(13.5 percent). The differences between the DR estimate for Medicaid coverage and
those from the 2012 and 2013 comparison samples were statistically significant.

•

In addition, the DR estimate for health insurance coverage through TRICARE,
CHAMPUS, or other military health care sources for all persons aged 12 or older
(3.4 percent) was lower than the estimates from the other four data sources. However,
the differences between the DR estimate and the estimates from the 2012 and 2013
comparison samples (4.8 and 4.4 percent, respectively) were not statistically
significant.

Table 7.4 Health Insurance Coverage among Persons Aged 12 or Older: Percentages and Standard
Errors for 2012 Comparison Data, 2013 Comparison Data, 2013 Dress Rehearsal, and
Other Surveys
Health Insurance
Coverage
Medicare
Medicaid
TRICARE, CHAMPUS,
CHAMPVA, VA,
Military Health Care
Private Health Insurance
Uninsured

2012
2013
Comparison1,2 Comparison1,3
Percent (SE)
Percent (SE)
18.2 (0.35)
18.6 (0.52)
12.1a (0.25)
12.6a (0.33)

2013 DR1,4
Percent (SE)
20.0 (2.00)
15.4 (1.28)

2012 ACS5
Percent (SE)
18.3 (0.01)
13.5 (0.03)

2012 NHIS6
Percent (SE)
17.7 (0.23)
11.7 (0.20)

4.8 (0.20)

4.4 (0.25)

3.4 (0.79)

4.7 (0.02)

3.6 (0.12)

64.5a (0.43)
15.2 (0.27)

65.5a (0.51)
14.5 (0.35)

56.4 (2.07)
16.8 (1.55)

65.4 (0.07)
16.3 (0.06)

65.6 (0.38)
16.5 (0.25)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services;
CHAMPVA = Civilian Health and Medical Program of the Veterans Administration; DR = Dress Rehearsal; NHIS = National
Health Interview Survey; SE = standard error; VA = Department of Veterans Affairs.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Sample does not include persons residing in Alaska or Hawaii, active-duty military personnel, or institutional group quarters.
6
Unknown or invalid data were excluded from the analysis.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013; U.S. Census, American Community Survey, 2012; National Health Interview Survey (NHIS), 2012.

356

•

For all persons aged 12 or older, the DR estimate (56.4 percent) for private health
insurance was lower than the estimates from the other four data sources, which were
very similar to each other, ranging from 64.5 percent in the 2012 comparison sample
to 65.6 percent in the NHIS sample. This finding is similar to a result from the QFT,
which found that estimates of private insurance coverage in the QFT were much
lower than those in the comparison samples and the benchmark surveys.

Benchmarking DR estimates for four types of health insurance coverage to both recent NSDUH
data and other national survey data revealed some results that were similar to those found for the
QFT. Across all age groups, the largest and most consistent differences between the QFT
estimates and estimates from the other four data sources were observed for private health
insurance. In the QFT, there were small differences in estimates of Medicare coverage, while the
differences were larger (but not statistically significant) between the DR estimates and those for
the other four data sources. As with results from the QFT, differences in estimates for military
coverage in the DR sample and the NSDUH comparison samples were not statistically
significant. However, the DR estimate (3.4 percent) was very similar to the 2012 NHIS estimate
of 3.6 percent, while the QFT estimate of 5.0 percent appeared higher than the 2011 NHIS
estimate of 3.5 percent. Finally, the estimate for Medicaid coverage in the DR sample was higher
than in the other four data sources, and this was similar to what was found in the QFT.
In Table 7.5, DR estimates for three income categories for all persons aged 12 or older
are presented with parallel estimates from the 2012 and 2013 comparison samples and the 2012
NHIS. For all persons aged 12 or older, the DR estimate for a family income of $49,999 or lower
(61.3 percent) appeared to be considerably higher than the estimates from the 2012 comparison
sample (51.0 percent), the 2013 comparison sample (50.2 percent), and the NHIS (47.7 percent).
Correspondingly, the DR estimate for the percentage of persons aged 12 or older with a family
income of $75,000 or greater was lower than the estimates for the 2012 comparison sample, the
2013 comparison sample, and the NHIS. These differences were much larger than the similar
patterns reported with respect to the QFT. For all persons aged 12 or older, the QFT estimate for
Table 7.5 Income among Persons Aged 12 or Older: Percentages and Standard Errors for 2012
Comparison Data, 2013 Comparison Data, 2013 Dress Rehearsal, and 2013 NHIS

Income
< $49,999
$50,000 to $74,999
$75,000 or More

2012 Comparison1,2
Percent (SE)
51.0a (0.51)
16.4 (0.27)
32.6a (0.50)

2013
Comparison1,3
Percent (SE)
50.2a (0.79)
17.0 (0.46)
32.8a (0.71)

2013 DR1,4
Percent (SE)
61.3 (2.62)
14.9 (1.23)
23.8 (2.10)

2012 NHIS5
Percent (SE)
47.7 (0.48)
17.6 (0.32)
34.7 (0.48)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; NHIS = National Health Interview Survey; SE = standard error.
a
Difference between estimate and DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Unknown or invalid data were excluded from the analysis.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and
2013; National Health Interview Survey (NHIS), 2012.

357

a family income of $49,999 (52.1 percent) or less was only slightly higher than the 2011 and
2012 quarters 3 and 4 comparison estimate and moderately higher than the 2011 NHIS estimate
(46.5 percent).
Overall, the DR estimates resulted in higher proportions of persons at lower income
levels and lower proportions at higher income levels compared with the other sources of survey
data. As with the QFT, this difference may account for some of the observed differences between
the DR estimates and those from the other data sources for those items that were the most highly
correlated with income level.

7.3

Comparisons of Estimates for Items New to the QFT and DR
Instruments

This section reports on comparisons between estimates from the DR sample with
estimates from external data sources for items that were introduced in the QFT and then repeated
for the DR. Included are items on self-reported height and weight, a question that asks
respondents whether a doctor or other health care professional had ever told them whether they
had one or more of nine health conditions, six items on disabilities and physical limitations, and
an item on English-language ability.
Questions on height and weight were introduced in the QFT, and these questions were
repeated for the DR. Comparisons were carried out with two benchmark data sources—the 2012
NHIS and the 2009-2010 and 2011-2012 NHANES). The latter source provides self-reported
height and weight data and physical measurements of both. Comparisons were limited to
estimates from respondents aged 16 or older because the NHANES only has self-reported height
and weight for persons aged 16 or older. For the DR and comparison NSDUH data, height and
weight estimates were produced with and without using NHIS coding rules. Also, at the time of
this analysis, self-reported measures of height and weight on the NHANES were not available
yet for the 2011-2012 NHANES.
In addition, because the coding of NHIS height and weight data includes specific lower
and upper bounds, the DR estimates for height and weight were calculated both unbounded and
bounded in accordance with NHIS criteria. The second calculation provided a more equivalent
comparison between the DR and 2012 NHIS data. The summary statistics for height presented in
Table 7.6 and the summary statistics for weight presented in Table 7.7 provided some sense of
how the DR statistics for these new questionnaire items compared with other national surveys.
•

Both the unbounded DR mean height estimate (66.6 inches) and the NHIS-bounded
DR mean height estimate (66.5 inches) were very similar to the NHIS mean height
estimate (66.9 inches), the NHANES directly measured mean height estimate
(66.4 inches), and the NHANES self-reported mean height estimate (66.9 inches).

•

The unbounded DR mean weight estimate (176.3 pounds) was similar to both the
self-reported and measured weight estimates based on the NHANES, whereas the
NHIS-bounded DR mean weight estimate (176.0 pounds) was similar to the NHIS
measure of mean weight estimate (174.2 pounds).

358

Table 7.6 2012 NHIS, 2009-2010 NHANES, and 2011-2012 NHANES Height Statistics among
Persons Aged 16 or Older for Comparison with the 2013 Dress Rehearsal
2013 DR1,2
Statistic
Sample Size
Mean
Standard Error
Minimum
Maximum
Median

Unbounded
1,704
66.6
0.11
30.7
107.0
66.0

NHIS Bounds3
1,695
66.5
0.11
30.7
76.0
66.0

2012 NHIS4,5
33,465
66.9
0.03
58.0
76.0
66.18

NHANES5
2009-2010
2011-2012
Self-Reported
Measured
6,730
5,839
66.9
66.4
0.07
0.11
41.0
52.95
81.0
80.51
66.19
66.30

DR = Dress Rehearsal; NHANES = National Health and Nutrition Examination Survey; NHIS = National Health Interview Survey.
NOTE: Answers in metric units (i.e., meters, centimeters) were converted to inches.
1

Sample does not include Alaska or Hawaii.

DR data collected from September 1 through October 31, 2013.

Includes values up to 76 inches for men aged 18 or older and 70 inches for women aged 18 or older. For children, the weighted 1½ and 98½
percentiles for height were computed by age/gender. Respondents with values outside of these bounds were excluded from the estimates.

For adults, these include values of 76 inches for men aged 18 or older and 70 inches for women aged 18 or older. For children, the genderspecific height-for-age values of the highest 1½ percent of records and the lowest 1½ percent of records were changed to "96" or "996" ("Not
available"). In cases where extreme values were reported for either current height or current weight, the data for both variables were changed to
"96" or "996" ("Not available") on the public use data file.

Unknown or invalid data were excluded from the analysis.

Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013;
CDC, National Center for Health Statistics, National Health and Nutrition Examination Survey (NHANES), 2009-2010; National
Health Interview Survey (NHIS), 2012.

Table 7.7 2012 NHIS, 2009-2010 NHANES, and 2011-2012 NHANES Weight Statistics among
Persons Aged 16 or Older for Comparison with the 2013 Dress Rehearsal
2013 DR1,2,3
Statistic
Sample Size
Mean
Standard Error
Minimum
Maximum
Median

Unbounded
1,707
176.3
1.12
39.7
463.0
172.0

NHIS Bounds4
1,699
176.0
1.04
95.0
300.0
172.0

2012 NHIS5,6,7
32,686
174.2
0.31
90.0
299.0
169.6

NHANES7
2009-2010
2011-2012
Self-Reported
Measured
6,741
5,838
177.4
178.9
0.89
1.13
76.0
64.15
670.0
476.4
171.1
173.1

DR = Dress Rehearsal; NHANES = National Health and Nutrition Examination Survey; NHIS = National Health Interview Survey.
NOTE: Answers in metric units (i.e., kilograms) were converted to pounds.
1

Sample does not include Alaska or Hawaii.

DR data collected from September 1 through October 31, 2013.

Pregnant women were asked to report their pre-pregnancy weight. Pregnancy status available for women aged 12 to 44.

For persons aged 18 or older, these include values between 126 and 299 pounds for men and 100 and 274 pounds for women. For children, the
weighted 1½ and 98½ percentiles for weight were computed by gender and age. Respondents with values outside of these bounds were excluded
from the estimates.

For persons aged 18 or older, includes values between 126 and 299 pounds for men and between 100 and 274 pounds for women. For children,
the gender-specific weight-for-age values of the highest 1½ percent of records and the lowest 1½ percent of records were changed to "96" or
"996" ("Not available"). In cases where extreme values were reported for either current height or current weight, the data for both variables were
changed to "96" or "996" ("Not available") on the public use data file.

Pregnant women were asked to report their pre-pregnancy weight. Pregnancy status available for women aged 20 to 44.

Unknown or invalid data were excluded from the analysis.

359

Overall, the DR height and weight estimates aligned closely to the estimates from the
2012 NHIS and the 2009-2010 NHANES and 2011-2012 NHANES, both self-reported and
directly measured.
A new series of questions added to the QFT questionnaire and repeated in the DR
questionnaire asked respondents whether a doctor or other health care professional had ever told
them whether they had one or more of nine health conditions, as shown in Table 7.8. The QFT
and 2011 NHIS estimates were generally similar for some conditions, but significant differences
were observed for a few conditions, with the QFT estimates being lower than NHIS estimates.
Estimates from the QFT and 2011 NHIS were similar for any kind of heart condition or heart
disease, diabetes or sugar diabetes, and kidney disease. For most of the other conditions, the QFT
estimates appeared to be lower than the 2011 NHIS estimates.
•

The difference between the QFT and 2011 NHIS estimates for ever having been
diagnosed with hypertension or high blood pressure was the largest absolute
difference, and the comparison between the DR and 2012 NHIS produced a similar
result; 15.8 percent of DR respondents reported ever having been diagnosed with
hypertension or high blood pressure, while 29.7 percent of NHIS respondents
reported ever having been diagnosed.

•

In the comparison between the QFT and the 2011 NHIS, estimates for ever having
been diagnosed with any kind of heart condition or heart disease were very similar
(10.4 percent for the QFT sample and 10.8 percent for the 2011 NHIS). The
difference for the comparison between the DR sample and the 2012 NHIS was larger
(8.4 percent for the DR sample vs. 10.5 percent for the 2012 NHIS).

A key difference between the QFT and DR instruments on the one hand and the NHIS
instrument is that in the NSDUH instruments, the health conditions were treated as response
categories in a "code all that apply" format, whereas in the NHIS instrument the parallel
categories were administered as separate, individual items.
Another new series of questions added to the QFT instrument asked respondents whether
they had any of six types of disabilities or physical limitations. Estimates from the QFT and 2011
NHIS were very similar for being deaf or having serious hearing difficulty, being blind or having
serious difficulty seeing, and having serious difficulty concentrating, remembering, or making
decisions. QFT estimates appeared to be significantly lower than the comparable 2011 NHIS
estimates for the following disabilities or physical limitations: having serious difficulty
walking or climbing stairs, having difficulty dressing or bathing, and having difficulty doing
errands alone, such as visiting a doctor's office or shopping. Comparisons between DR estimates
and the 2012 NHIS produced more similar results than comparisons between the QFT estimates
and 2011 NHIS with one notable exception. The QFT estimate for serious difficulty
concentrating, remembering, or making decisions (6.6 percent) was very similar to the 2011
NHIS estimate of 6.2 percent. As shown in Table 7.9, the DR estimate was 8.9 percent, while the
2012 NHIS estimate was only 4.5 percent. Estimates for these disabilities and physical
limitations from the 2012 ACS are also presented in Table 7.9. For most items, ACS estimates
were lower than those from the DR. The exceptions were for difficulty dressing or bathing and
difficulty doing errands alone, such as visiting a doctor's office or shopping.

360

Table 7.8 Conditions Told to Respondent by Doctor or Other Health Professional among Persons
Aged 12 or Older: Percentages and Standard Errors, 2013 Dress Rehearsal and 2012
National Health Interview Survey
Condition
Any kind of heart condition or heart disease
Diabetes or sugar diabetes
Chronic bronchitis, emphysema, chronic obstructive
pulmonary disease, also called COPD
Cirrhosis of the liver
Hepatitis
Kidney disease, not including bladder infection or incontinence
Asthma
Cancer or a malignancy of any kind
Hypertension, also called high blood pressure

2013 DR1,2
Percent (SE)
8.4 (1.19)
8.9 (1.06)

2012 NHIS3
Percent (SE)
10.5 (0.20)
8.1 (0.16)

4.2 (0.89)
0.2 (0.16)
1.7 (0.50)
2.3 (0.70)
11.1 (1.18)
4.8 (0.88)
15.8 (1.67)

5.5 (0.16)
1.3 (0.08)
3.0 (0.11)
1.9 (0.90)
13.2 (0.23)
8.2 (0.18)
29.7 (0.37)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; NHIS = National Health Interview Survey; SE = standard error.
1
Sample does not include Alaska or Hawaii.
2
DR data collected from September 1 through October 31, 2013.
3
Unknown or invalid data were excluded from the analysis.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013; CDC,
National Center for Health Statistics, National Health Interview Survey, 2012.

Table 7.9 Disabilities or Physical Limitations among Persons Aged 12 or Older: Percentages and
Standard Errors, 2013 Dress Rehearsal, 2012 National Health Interview Survey, and
2012 American Community Survey
Disability or Physical Limitation
Deaf or serious hearing difficulty
Blind or serious difficulty seeing
Serious difficulty concentrating, remembering, or
making decisions
Serious difficulty walking or climbing stairs
Difficulty dressing or bathing
Difficulty doing errands alone, such as visiting a
doctors' office or shopping4

2013 DR1,2,3
Percent (SE)
4.5 (0.72)
4.2 (0.71)

2012 NHIS3
Percent (SE)
4.9 (0.46)
4.0 (0.42)

2012 ACS1
Percent (SE)
3.9 (0.02)
2.4 (0.01)

8.9 (1.03)
10.3 (1.39)
2.8 (0.57)

4.5 (0.42)
10.7 (0.65)
3.3 (0.39)

5.1 (0.02)
7.6 (0.02)
2.8 (0.01)

4.4 (0.75)

5.8 (0.50)

5.6 (0.02)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; NHIS = National Health Interview Survey; ACS = American Community Survey.
1
Sample does not include Alaska or Hawaii.
2
DR data collected from September 1 through October 31, 2013.
3
Unknown or invalid data were excluded from the analysis.
4
Estimates are for persons aged 15 or older.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013;
National Health Interview Survey (NHIS), 2012; American Community Survey (ACS), 2012.

361

Another item added to the QFT questionnaire and repeated in the DR questionnaire was
an item that assessed respondents' capabilities in speaking English. Item QD55 asks, "How well
do you speak English?" The ACS questionnaire contains the same question, but it is only asked if
there is a response of "yes" to the preceding item, which asks if the person speaks a language
other than English at home. Those who do not speak a language other than English at home are
skipped out of the question on English-language ability. The ACS estimates shown in Table 7.10
are based on classifying those who answered "no" to the preceding question as those who speak
English "very well." Because of the skip pattern in the ACS sequence, it would be expected that
at best only a rough approximation can be made; not all of the ACS respondents who were
skipped out of the English proficiency item would have answered "very well" to that question.
Also, a DR respondent who chose "not at all" would most likely have been asked the ACS
question because they would most likely have reported that another language was used in the
home.
Table 7.10 English-Speaking Proficiency among Persons Aged 12 or Older: Percentages and
Standard Errors, 2013 Dress Rehearsal and 2012 ACS
2013
DR1,2,3
Percent (SE)
83.6 (1.40)
10.9 (1.16)
4.5 (0.68)
1.0 (0.25)

How well do you speak English? (QD55)
Very well
Well
Not well
Not at all

2012
ACS1,4
Percent (SE)
91.2 (0.03)
4.0 (0.02)
3.3 (0.02)
1.5 (0.01)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
ACS = American Community Survey; DR = Dress Rehearsal; SE = standard error.
1
Sample does not include Alaska or Hawaii.
2
DR data collected from September 1 through October 31, 2013.
3
Unknown or invalid data were excluded from the analysis.
4
ACS item is only asked if there is a response of "yes" to the preceding question: "Does this person speak a language other than
English at home?" The estimate shown here for the ACS classifies those who answered "no" to the preceding question as those
who speak English "very well."
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013;
American Community Survey (ACS), 2012.

7.4

Comparisons of Estimates from Items New to the DR Instrument
For the DR, two items on sexual attraction and identity were added:
QD62 People are different in their sexual attraction to other people. Which statement best
describes your feelings?
[IF QD01=5] (asked of males)
1 I am only attracted to females
2 I am mostly attracted to females
3 I am equally attracted to females and males
4 I am mostly attracted to males
5 I am only attracted to males
6 I am not sure
362

DK/REF
[IF QD01=9] (asked of females)
1 I am only attracted to males
2 I am mostly attracted to males
3 I am equally attracted to males and females
4 I am mostly attracted to females
5 I am only attracted to females
6 I am not sure
QD63 Which one of the following do you consider yourself to be?
1 Heterosexual, that is, straight
2 [IF QD01=9 THEN "Lesbian or] Gay"
3 Bisexual
DK/REF
There appear to be few public use datasets that can be used to produce estimates
comparable with those from the DR sample. The NHIS began asking a question similar to QD63
in 2013, but its estimates and data are not currently available. Gates (2011) reviewed a number of
survey-based estimates of sexual orientation, including the 2006-2008 NSFG and the 2008 GSS.
Public use data from the NSFG (through 2010) do not appear to include these items. The 2012
GSS item on sexual orientation asks, "Which of the following best describes you," with response
categories of (1) gay, lesbian, or homosexual; (2) bisexual; or (3) heterosexual or straight.
Estimates (and complex sample standard errors) for persons aged 18 or older, as well as by age
groups, were produced using an online data analysis tool.
Table 7.11 presents estimates for the new item on sexual identity (QD63) and those from
the 2012 GSS both overall for persons aged 18 or older and separately for males and females.
In general, these estimates appear similar to each other, suggesting that these items can be used
to obtain valid estimates on sexual orientation.
Table 7.12 presents estimates of sexual identity for persons aged 18 to 44 years old for
the DR and GSS data and published estimates from the 2006 to 2010 NSFG (Chandra, Copen, &
Moser, 2013). The published NSFG estimates reflect the presence of responses of "something
else" and those who did not provide a response whereas the DR and GSS estimates drop cases
with invalid or missing data and compute percentages based only on those who provided valid
responses. In general, estimates for the sexual identity item appear similar to those from the GSS
and NSFG suggesting that the item can be used to obtain valid estimates of this construct.

363

Table 7.11 Sexual Identity among Persons Aged 18 or Older, by Gender: Percentages and
Standard Errors for 2013 Dress Rehearsal and 2012 GSS Data
Gender and Age / Sexual Identity
All 18 or Older
Heterosexual
Lesbian or Gay
Bisexual
Males, 18 or Older
Heterosexual
Gay
Bisexual
Females, 18 or Older
Heterosexual
Lesbian or Gay
Bisexual

2013 DR1,2,3
Percent (SE)

2012 GSS3
Percent (SE)

96.0 (0.67)
1.3 (0.34)
2.7 (0.53)

96.9 (0.28)
1.5 (0.19)
1.6 (0.18)

97.7 (0.74)
1.3 (0.46)
1.0 (0.45)

97.3 (0.35)
1.7 (0.26)
1.0 (0.22)

94.5 (1.08)
1.3 (0.50)
4.2 (0.94)

96.6 (0.35)
1.3 (0.24)
2.1 (0.25)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; GSS = General Social Survey; SE = standard error.
1
Sample does not include Alaska or Hawaii.
2
DR data collected from September 1 through October 31, 2013.
3
Unknown or invalid data were excluded from the analysis.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013;
NORC, General Social Survey, 2012.

Table 7.12 Sexual Identity among Persons Aged 18 to 44, by Gender: Percentages and Standard
Errors for 2013 Dress Rehearsal and 2012 GSS Data
Gender and Age / Sexual Identity
All, 18 to 44
Heterosexual
Lesbian or Gay
Bisexual
Males, 18 to 44
Heterosexual
Gay
Bisexual
Females, 18 to 44
Heterosexual
Lesbian or Gay
Bisexual

2013 DR1,2,3
Percent (SE)

2012 GSS3
Percent (SE)

2006 – 2010 NSFG4
Percent (SE)

94.8 (0.88)
1.9 (0.57)
3.4 (0.58)

95.7 (0.45)
1.8 (0.27)
2.5 (0.34)

96.6 (1.22)
1.9 (0.80)
1.5 (0.74)

96.6 (0.57)
1.9 (0.41)
1.5 (0.42)

95.6 (0.4)
1.8 (0.2)
1.2 (0.2)

92.9 (1.33)
1.8 (0.84)
5.3 (0.92)

94.9 (0.60)
1.7 (0.40)
3.4 (0.45)

93.6 (0.4)
1.2 (0.2)
3.9 (0.3)

* Low precision; estimate would be suppressed under NSDUH suppression rules.
DR = Dress Rehearsal; GSS = General Social Survey; SE = standard error.
1
Sample does not include Alaska or Hawaii.
2
DR data collected from September 1 through October 31, 2013.
3
Unknown or invalid data were excluded from the analysis.
4
Estimates do not account for unknown or invalid data.
Sources: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2013;
NORC, General Social Survey, 2012; Chandra, Copen and Mosher (2013).

364

7.5

Summary of Comparisons between DR, Comparison Data, and External
Data

Overall, two key differences in estimates between the QFT sample and the comparison
samples remained in place based on a preliminary analysis of data from the DR sample. One
group of differences suggests that the DR sample members were more associated with lower
socioeconomic status than respondents in other surveys, including the NSDUH-based
comparison datasets. It appears that changes introduced in the DR questionnaire for some of
these items did not have any impact on the results of these comparisons with external data
sources.
•

DR estimates for those participating in means-tested government programs, such as
food stamps and SSI, were significantly higher than those for the 2012 and 2013
comparison data and for the external benchmark data sources. This is similar to
findings from the QFT. Similarly, the DR estimate of the percentage of persons with
private insurance remained well below that of the other data sources in these
comparisons, as was observed in the QFT.

•

The DR sample yielded a higher percentage of persons with family incomes of less
than $50,000 when compared with either the 2012 and 2013 comparison data files or
the NHIS benchmark external source. This finding is similar to what was reported for
the QFT, but the magnitude of the difference was larger in the DR than in the QFT.

Another key finding from the QFT was that the NSDUH item on ever having been
diagnosed with specific health conditions produced estimates that were generally lower than
those from the NHIS, and this finding was repeated in the DR field test. This may be due to the
NSDUH items being assessed using a single question with a "code all that apply" format,
whereas the NHIS asks about each condition in separate questions.
Comparisons between DR estimates and external data sources also provided information
to further develop items added to both the QFT and the DR instruments.
•

Data on self-reported height and weight gathered from the DR generally appeared to
be consistent with data from the NHIS and NHANES.

•

DR estimates on six disability items were generally consistent with estimates from the
NHIS and ACS. A notable exception was that the DR estimate for serious difficulty
concentrating, remembering, or making decisions was higher than estimates from the
benchmark data sources and from the QFT. The DR estimates on sexual identity
(heterosexual, gay/lesbian, or bisexual) appeared to be consistent with data from the
2012 GSS and the 2006 to 2010 NSFG.

365

366

8. Summary and Implications
As noted in Chapter 1, the primary goal of the 2013 Dress Rehearsal (DR) was to
measure, using multiple indicators and data sources, the total effect on National Survey on Drug
Use and Health (NSDUH) estimates from the full set of changes to the protocol planned for the
2015 partial redesign. Following the 2012 Questionnaire Field Test (QFT), the DR aimed to
further test the revisions made to the questionnaires, materials, and equipment. Further protocol
revisions that were made following the QFT were tested during the DR, most notably the
addition of Spanish-language interviews and a utilization of new lightweight laptop computers.
Based primarily on the results presented in Chapters 4 to 7, this chapter summarizes the key
findings from the DR with respect to each of the five main research questions and the
implications of these results for implementing a partially redesigned protocol in 2015. Where
appropriate, recommendations for the 2015 NSDUH that are based on results and insights gained
from the DR are noted.
Section 8.1 highlights key outcomes of the DR data collection related to data quality
(Chapter 4), including screening and interview response rates, variable imputation rates and item
missingness rates, interview timing results, and other data quality indicators. Conclusions from
specific assessments of the redesigned protocol in Chapter 5—including field observations,
responses to field interviewer (FI) debriefing surveys, new equipment surveys, FI training
surveys, debriefing calls with FIs, and field observations—are summarized in Section 8.2.
Section 8.3 discusses key findings from comparing DR estimates with main study estimates for
specific substance use items and other core and selected noncore estimates (Chapter 6).
Section 8.4 focuses on key findings from comparing estimates for items that were revised or
moved for the DR and QFT with estimates from the comparison data, as well as those items that
are new, and those items that were moved from computer-assisted personal interviewing (CAPI)
to audio computer-assisted self-interviewing (ACASI) administration, as described in Chapter 7.
Section 8.5 describes the implications for the 2015 partially redesigned instrument and protocol,
based on the results of the DR and QFT.

8.1

Data Collection Outcomes and Data Quality Assessment (Research
Questions 2 and 3)

In most respects, the DR data met the standards of data quality similar to those applied to
the 2013 quarters 3 and 4 and 2012 NSDUH main study comparison data, as detailed in
Chapter 4. Data quality indicators were examined to assess the impact of the redesigned
protocol, including changes to the equipment and questionnaire, on data collection outcomes
achieved by the DR. Data collection outcomes and data quality are assessed by item missingness
and variable imputation rates, interview timing results, screening and interview response rates,
and other data quality indicators, described in the sections below.
8.1.1

Item Missingness Rates and Variable Imputation Rates

Overall, item missingness rates and variable imputation rates examined in the DR results
were similar to the QFT results. Item missingness rates were examined especially for those items

367

that were moved from CAPI to ACASI administration for the QFT and DR, as a number of these
items had higher missingness rates than the parallel items administered via CAPI in the main
study comparison data. Items that were introduced in the QFT then revised between the QFT and
DR and those that were new to the DR questionnaire were also examined.
As detailed in Section 4.4 and Tables B.1 to B.4 in Appendix B, the following items that
will be moved from CAPI to ACASI administration in the 2015 questionnaire had relatively
higher item missingness rates in the QFT and produced similar results in the DR:
•

marital status (QD07);

•

number of home moves in the past year (QD13);

•

full- or part-time student status (QD19);

•

work at a job or business at any time in the past week (QD21); and

•

most of the items that ask about recent employment history, missing workdays, size
of employing organization, and related issues (QD31, QD33, QD36, QD38, QD39a,
QD40, QD41, and QD42).

These items are planned to be administered in ACASI in the partially redesigned 2015
questionnaire, so these items will be examined closely in the 2015 Early Data Review (EDR) to
see whether item missingness rates remain as a potential data quality issue. Although a number
of items asking about health insurance coverage, receipt of various sources of income via
government assistance programs, and income also had relatively higher missingness rates based
on the ACASI data in the QFT and DR, these items will be administered via CAPI in the
partially redesigned 2015 questionnaire.
Another indicator of the quality of the DR data is the proportion of cases for which
imputation was required prior to using specific variables for analysis. Section 4.3 and
Tables 4.9a through 4.9f provided rates of imputation and logical assignment for selected
variables. With a few exceptions, the weighted percentages of cases that were either imputed or
logically assigned for the following variables were similar across the DR, 2012 comparison, and
2013 quarters 3 and 4 comparison datasets:
•

recency of substance use,

•

past year initiation status,

•

health insurance, and

•

income.

The majority of the age at first use variables in the DR data required no logical assignment or
imputation.
8.1.2

Interview Timing Results

Overall interview times were lower or similar for DR respondents compared with the
2012 respondents, the 2013 quarters 3 and 4 respondents, and the QFT respondents for most age
groups. Spanish-language DR respondents aged 12 or older, however, had higher overall

368

interview times when compared with the Spanish-language 2012 respondents and the Spanishlanguage 2013 quarters 3 and 4 respondents. Spanish-language DR timing data were
considerably higher than either the Spanish-language 2012 main study or the Spanish-language
2013 quarters 3 and 4 for respondents aged 65 or older; however, despite this larger difference in
average times for respondents aged 65 or older, the overall timing differences for Spanishlanguage interviews were not of a large magnitude.
As expected, the average timing for the total core substance use sections for all
respondents aged 12 or older was higher for the DR respondents than it was for the 2012
respondents and the 2013 quarters 3 and 4 respondents and only slightly lower than it was for the
QFT respondents. Additions and revisions to the hallucinogens, inhalants, and prescription drug
sections in the partially redesigned DR questionnaire contributed to these higher durations for the
core substance use modules when compared with the main study data.
Timings for the redesigned prescription drug modules for the DR respondents aged 12 or
older were higher than they were for the 2012 respondents and 2013 quarters 3 and 4
respondents, but they were lower than the timings for the QFT respondents. Among the
redesigned prescription drug modules, the pain relievers module accounted for the higher
administration times for the DR respondents compared with the 2012 and 2013 quarters 3 and 4
respondents.
For the health insurance section, a higher average administration time was observed for
the DR respondents compared with the 2012 respondents and the 2013 quarters 3 and 4
respondents. However, the average administration times for this section were similar to those
observed in the QFT. The primary change to this section in the DR questionnaire, relative to the
main study instrument, was moving these questions from CAPI to ACASI administration.
Questionnaire changes to accommodate the ACASI mode were also implemented in both the
QFT and the DR.
8.1.3

Screening and Interview Response Rates

The overall response rates obtained during the DR were lower than for the 2012 main
study comparison sample and very similar to those obtained during the QFT, albeit slightly
lower. These differences are illustrated in Table 4.1 in Chapter 4.
Screening rates account for much of the difference in overall response between the DR
and the 2012 main study comparison sample as well as the QFT. As shown in Table 4.1,
weighted screening rates declined steadily from the 2012 comparison sample to the QFT and
then again during the DR, although the decline was less steep in the latter comparison. Overall,
as shown in Table 4.2, the distribution of visits for completed screenings in the DR sample was
similar to the distributions for the 2012 and 2013 quarters 3 and 4 comparison samples.
Interviewing response rates for the DR were lower than for the 2012 main study
comparison group but higher than for the QFT. Similar to the number of visits for completed
screenings (see Table 4.2), the percentage of completed interviews by the number of visits
during the DR followed a similar pattern as the 2013 quarters 3 and 4 and 2012 main study
comparison groups.

369

8.1.4

Other Data Quality Indicators

Additional indicators were employed to assess the quality of the DR data. Outcomes for
the DR are compared with the 2012 main study and the 2013 quarters 3 and 4 main study
comparison data, where appropriate, by the following indicators:

8.2

•

Responding to lead questions for "OTHER, Specify" data. This data quality
indicator examined instances of choosing "other" responses for which respondents
were subsequently asked to specify an open-ended response, focusing on items that
were new, moved within the questionnaire, or revised. Rates for "other" responses to
these items (shown in Table 4.14) were low in the combined QFT and DR data
relative to the rates for predefined response categories. These low rates for "other"
responses were consistent with the findings from the QFT and support the overall
conclusion that predefined categories performed adequately in the QFT and DR.

•

Patterned responses in the core drug questions. This indicator examined patterned
responses in answers to the screening questions for past year prescription drug use or
to the questions for past year misuse in the DR, relative to the comparison data.
As described in Chapter 4, attention was given to identifying cases in the prescription
drug data where responses of only "1" or "2" were recorded for all screener questions
for a given prescription drug category and/or high numbers of individual prescription
drugs that were misused relative to the overall distribution of the number of drugs that
were misused within a given category. No cases were removed from the DR data
because of patterned responses. Although three respondents had a pattern of keying
"2" in one or more prescription drug modules, no respondents had a pattern of keying
only responses of "1" in the screening questions. No cases were recommended to
have their prescription drug answers set to "bad data" because of high numbers of
individual prescription drugs that were recorded as misused.

Assessments of the Redesigned Protocol (Research Question 1)

As described in Chapter 5, five field-related efforts were used to assess the partially
redesigned questionnaire and protocol used in the DR. Overall, these assessments provided some
assurance that revisions made to the questionnaire and protocol following the QFT will facilitate
continued high quality and efficiency in NSDUH data collection when the partial redesign is
implemented in 2015. Based on these assessments and discussions with the Substance Abuse and
Mental Health Services Administration (SAMHSA), RTI recommends changes for 2015 to
address the instrument and screeners, training and materials, and equipment. In addition to being
described in this section of the report, recommendations for 2105 are documented in Tables 8.1
to 8.3 (shown at the end of this chapter with the chapter's other tables), subject to SAMHSA
approval. Key results from the five field-related assessments are described in the sections below.
8.2.1

Field Interviewer Training Survey

FIs provided their reactions to the DR FI training program through an FI training survey.
Insights derived from their reactions to the training program and materials and the FIs' general
comfort level with DR tasks will be useful in developing the 2015 NSDUH training programs
and related materials. Overall, the results of the DR FI training survey were very positive and
similar to the information gathered at the end of the QFT training program in August 2012.
370

Overall, FIs were highly satisfied with the DR training program, as evidenced by most indicating
agreement ("strongly agree" or "agree") to the statements about training. One important finding,
however, concerned the pace of the DR training session. Fewer FIs reported they strongly agreed
or agreed with the statement concerning the pace of training. Although one FI reported that it
was too slow, more reported that it was too quick.
8.2.2

Field Interviewer Equipment Survey

To assess changes to equipment that are planned for implementation in 2015 and utilized
on the DR, FIs were surveyed about their experiences with the equipment. Responses to the
equipment survey indicate that the FIs overall were very satisfied with the equipment and
programs deployed in the DR. Most of the DR FIs indicated that the decreased weight of the
laptop was advantageous. Additionally, they felt that it was easy to learn and easy to use.
The DR FIs were very satisfied with the new tablet email program, which provided two-way
email capabilities for increased communication between the FIs and their field supervisors (FSs).
The FIs overwhelmingly preferred using the default Samsung keypad instead of the hacker's
keypad.
The results of the equipment survey indicate that the FIs found the design of the laptop's
carrying case to be problematic, which supports efforts to improve this piece of equipment.
The FIs also reported difficulty with the function keys on the laptop, pointing out that they were
very small and hard to read. Recommendations for changes to equipment, emanating from the
equipment survey, are provided in Table 8.3.
8.2.3

Field Interviewer Debriefing Items

FI debriefing items were used to gauge how respondents reacted to the partially
redesigned protocol used during the DR. The DR results for the debriefing questions provided
insight into respondents' reactions to the DR interview and protocol. Overall, the FIs reported
that respondents reacted favorably to the new computer equipment, indicating that it should not
greatly influence respondents' experience with the interview. The function keys were reported as
problematic in some instances, however, which indicates a need to optimize them for 2015 in
order to prevent frustration or confusion that could influence the outcome of the interview.
Data from the debriefing items did not illuminate any significant problems with the respondents'
comprehension of questionnaire items, suggesting that lack of comprehension should not be a
significant factor influencing data quality. The FIs reported that a proxy was used in about one
fourth of the interviews, and some challenges were reported with the process of introducing the
proxy respondents to the computer, which affected a larger proportion of respondents who took
the interview in Spanish.
The FIs indicated overall that 7 percent of the time there were problems with the proxy
use of ACASI to answer the income and health insurance questions (see Table 5.25), but for
Spanish-language interviews it was 11 percent. This will have limited implications in 2015 if
these questions are not administered in ACASI.
The FIs indicated that the interviews during the DR were slightly less private than during
the QFT and the 2012 and 2013 comparison interviews. As indicated in Table 5.29, fewer FIs

371

classified the DR interview as "completely private" compared with both the 2012 and 2013
comparison interviews and the QFT interviews.
As shown in Table 5.15, FIs reported in only 13.2 percent of the completed interviews
that the respondent commented that the interview was too long, which differs insubstantially
from the QFT results. Older respondents and those with lower levels of education were more
likely to make comments about the interview being too long.
8.2.4

Debriefing Calls with Field Interviewers

FI debriefing calls were held with FIs who conducted DR interviews in order to obtain
direct feedback from them on their experiences collecting data using the redesigned NSDUH
questionnaire on the new laptop and completing screenings using the touch screen. Altogether,
five debriefing calls were conducted. The goal of these calls was to gather FIs' feedback and
insights on use of the equipment, challenges encountered, and any significant concerns raised by
respondents regarding the questionnaire or protocol in order to inform potential changes to the
preparation, protocol, and procedures for the 2015 NSDUH. Key findings include the following:
•

FIs on all five debriefing calls noted that respondents reacted positively to the
changes to the lead letter and the question and answer (Q&A) brochure, with many
having already read the lead letter. The FIs suggested that the lead letter's added color
was what led to the improved respondent interest and recall. FIs noted that
respondents seemed to have greater recall of the DR lead letter than its main study
counterpart.

•

All FIs confirmed that the DR FI training program was effective in preparing them to
use the tablet in the field to conduct screenings. Also, all of them reported quickly
feeling comfortable using the tablet.

•

Across all five debriefing calls, FIs confirmed that the DR FI training program was
effective in preparing them to use the new laptop computer, and all had
overwhelmingly positive feelings toward the new laptop.

•

Several FIs commented that they felt that having the pill images available on the
screen rather than in the showcard booklet results in greater attention being paid to
the images on the part of the respondent.

•

Recommendations emanating from the calls are documented in Tables 8.1 to 8.3.
8.2.5

Field Observations of Field Interviewers

The majority of the FIs displayed positive behaviors when conducting DR screenings and
interviews. Some of the errors observed among the DR interviewers were not specifically related
to the redesigned protocol and may have been observed on the main study. Overall, the results
from DR field observations suggested that relatively few specific changes to the protocol are

372

needed in advance of the 2015 survey data collection; however, specific items noted are
described below:
•

Nearly 10 percent of the time (see Table 5.33), the FIs failed to provide their name,
RTI International, U.S. Department of Health and Human Services, and the lead letter
during a screening.

•

During the interviewing, the most common problems were not reading all of the
screens verbatim (see Table 5.34), not answering respondent questions thoroughly
and appropriately (see Table 5.35), and not following proper quality control and
incentive procedures (see Table 5.35).

Enhancements to the training that are recommended on the basis of the field observations are
documented in Tables 8.1 to 8.3.

8.3

Selected Core and Noncore Estimates for English- and SpanishLanguage Dress Rehearsal Data and Comparison Data (Research
Question 4)

As detailed in Chapter 6, findings on a large number of selected core and noncore
estimates from the DR and the 2012 and 2013 quarters 3 and 4 comparison data were presented
for the following types of analyses:
•

analyses to make decisions for the 2015 survey;

•

further analyses based on findings from the QFT; and

•

analyses to explain anticipated findings in 2015.

This section highlights key findings from Chapter 6 and suggests implications for the 2015
NSDUH. Where appropriate, implications of these results and recommendations for the 2015
survey year are presented in Tables 8.1 to 8.3 in Section 8.5.
8.3.1

Core Substance Use Estimates Other Than Methamphetamine and Prescription
Drugs (Research Question 4a)

Some differences between field test and comparison data for estimates of cocaine and
heroin use that had been observed in the QFT continued to be observed in the combined QFT
and DR data for non-Hispanic English-language interviews despite the content of these modules
not changing for the QFT and DR. However, the assumption continues to be that any changes in
prevalence for these drugs in 2015 relative to earlier years based on a full sample of
approximately 67,000 interviews in 2015 will reflect an actual change in prevalence in the
population. However, this assumption can be tested by reviewing trend data from the first
6 months of 2015, which will likely have a sample size of more than 30,000, or roughly 10 times
the sample size of the combined QFT-DR data. In addition, single-year fluctuations in
prevalence would need to be interpreted with caution. It would be important to examine trends
across multiple years—including years beyond 2015—to account for occasional fluctuations in
prevalence that may "correct" themselves with additional years of data.

373

8.3.2

Methamphetamine, Prescription Drug, and Illicit Drug Summary Estimates
(Research Question 4b)

For methamphetamine, separating questions about the use of this drug from questions
about the misuse of prescription stimulants generally increased the prevalence of lifetime use
relative to the estimates in the comparison data, even when noncore data were included in the
comparison datasets for persons who reported their use of methamphetamine outside of the
context of questions about prescription drugs. If the prevalence of lifetime methamphetamine use
in 2015 is higher than in recent years for persons aged 12 or older or within different age groups
because of changes to the questionnaire in 2015, SAMHSA will need to decide how to handle
the reporting of trends in lifetime use. One option would be not to report trend data for lifetime
methamphetamine use between 2015 and earlier years or to discontinue the reporting of lifetime
trend data for methamphetamine altogether from 2015 onward. Alternatively, SAMHSA could
start a new baseline for lifetime methamphetamine use beginning in 2015. Other, more
sophisticated options could involve statistical procedures to adjust the trend data for 2002 to
2014. Although data on trends in lifetime prevalence may be of interest for examining historical
changes in the popularity of different drugs, data on trends in the prevalence of
methamphetamine use in the past year and past month are likely to be of more importance to
policymakers, the public health sector, the criminal justice sector, and others because of the
demands that methamphetamine users may place on the criminal justice system, the health care
delivery system (including substance abuse treatment), and systems for providing social services
(including services to dependents of adult substance users).
For prescription drugs, the general findings of lower estimates of lifetime misuse of
prescription drugs but higher past year estimates in both the QFT and combined QFT-DR data
relative to corresponding comparison datasets were expected, given the changes to the
prescription drug questions for the QFT. In the current questionnaire, respondents have multiple
opportunities to report lifetime misuse of specific prescription drugs whereas respondents in the
QFT and DR had multiple opportunities to report past year misuse of specific prescription drugs.
Similarly, respondents in the main survey have limited opportunity to report past year misuse,
and respondents in the QFT and DR had limited opportunity to report misuse of any prescription
drugs that occurred more than 12 months prior to the interview—including misuse of
prescription drugs that are no longer available by prescription in the United States.
These findings from both the QFT and combined QFT-DR data for prescription drugs
also support the conclusion to start a new baseline in 2015 for trends in prescription drug misuse.
In addition, it may be useful for SAMHSA to consider whether to discontinue reporting trend
data for lifetime misuse of prescription drugs after 2014 because of questions about the accuracy
of respondent self-reports of misuse of prescription drugs more than 12 months prior to the
interview.
The general lack of effect of changes to the methamphetamine and prescription drug
modules in the QFT and DR on summary measures of use of any illicit drug is consistent with
marijuana being the most commonly used illicit drug in the United States. The increases in
estimates of past year use of illicit drugs other than marijuana in the QFT and DR relative to
the comparison data when methamphetamine and prescription drugs were included also can be
explained by the prevalence of misuse of prescription drugs consistently being second only to
374

marijuana among illicit drugs. Analysis of 6-month data for 2015 is likely to be useful for
assisting SAMHSA in deciding how to create these summary illicit drug use measures in 2015
and how to report trends for these measures.
8.3.3

Selected Noncore Estimates (Research Question 4d)

Overall, few differences in prevalence were observed between the data for the QFT and
DR and comparison data for the main survey, particularly for substance dependence and abuse,
needle use, and substance use treatment. No changes were made to the substance treatment
module for the QFT and DR, and relatively minor changes were made to the questions for
dependence and abuse and for needle use.
In addition, estimates for some mental health measures differed between the combined
QFT and DR data for adults but not for adolescents aged 12 to 17. However, the mental health
questions did not change in the QFT and DR for adults or adolescents. As for the drug use
measures that were discussed previously, the estimates for adults that differed between the field
test and comparison data could be examined in the 6-month data for 2015 to test the hypothesis
that the observed differences between the field test and comparison data for these mental health
measures were an artifact of the smaller sample sizes for the QFT and DR datasets.

8.4

Selected Noncore Estimates for the Dress Rehearsal, Comparison Data
and External Data Sources (Research Question 5)

As detailed in Chapter 7, comparisons of several sets of key estimates from the DR were
made with estimates from the 2012 comparison sample and data sources other than NSDUH.
Comparisons were made for the following three sets of estimates for use as benchmark tools for
evaluating the validity of the DR's estimates:
•

selected items that were moved from CAPI to ACASI administration in both the QFT
and the DR,

•

selected items new to the QFT instrument and also included in the DR, and

•

two items on sexual attraction and identity that were first used in NSDUH in the DR
instrument.

In addition to these findings being summarized in this section of the report, any implications of
these results and recommendations for the 2015 survey year are presented in Tables 8.1 to 8.3 in
Section 8.5.
8.4.1

Estimates for Selected Items Moved from CAPI to ACASI Administration

Section 7.2 compared DR estimates with 2012 main study comparison data and various
external sources for the following sets of estimates:
•

received income and program participation,

•

current employment,

•

unemployment rates,
375

•

health insurance coverage, and

•

income.

Overall, some of the key differences in estimates observed between the QFT data, main
study comparison data, and external data sources were observed for the DR data. The majority of
these observed differences suggested that the DR sample was comprised of a higher proportion
of respondents with lower socioeconomic status. Key findings supporting this conclusion were as
follows:

8.4.2

•

DR estimates for those participating in means-tested government programs, such as
food stamps and Supplemental Security Income (SSI), were significantly higher than
those for the 2012 comparison data and for the external data sources.

•

The DR estimate of the percentage of persons with private health insurance coverage
was well below that of the other data sources.

•

The DR data produced a higher percentage of persons with incomes of less than
$50,000 when compared with either the 2012 comparison data file or the external
source.

Estimates for Items New to the QFT Questionnaire and Included in the DR

In Section 7.3, DR estimates were compared with two external data sources for two items
that were new to the QFT questionnaire and included in the DR—height and weight. Overall,
these comparisons revealed that the DR height and weight estimates aligned closely to estimates
from the two external sources, both self-reported and directly measured. Key findings were as
follows:
•

Both the unbounded DR mean height estimate and the National Health Interview
Survey (NHIS)-bounded DR mean height estimate were very similar to the NHIS
mean height estimate, the National Health and Nutrition Examination Survey
(NHANES) directly measured mean height estimate, and the NHANES self-reported
mean height estimate.

•

The unbounded DR mean weight estimate was similar to both the self-reported and
measured weight estimates based on the NHANES, whereas the NHIS-bounded DR
mean weight estimate was very similar to the NHIS measure of mean weight
estimate.

These findings suggest that the items for measuring height and weight in the DR appear likely to
produce estimates that are highly comparable with other large national in-person surveys,
including those that use either bounded estimates of self-reported height and weight or direct
measures of height and weight.
DR estimates were also compared with those from the NHIS for the items asking about
ever having been diagnosed with specific health conditions. The DR estimates were generally
lower than those from the NHIS, and these findings replicated the result from the QFT analysis.
Observed differences between the QFT and DR estimates compared with external data sources
like the NHIS could be attributable to differences in question formats. The NSDUH

376

questionnaire asks about these items via a single question using a "code all that apply" format,
whereas the NHIS asks about each health condition via separate questions.
The pattern of DR estimates being generally lower than those from the NHIS for specific
health conditions was consistent with previous comparisons of NSDUH and NHIS data. A report
comparing 2006 chronic health condition estimates derived from NSDUH and other health data
sources found NSDUH estimates to be generally lower than those from the NHIS (Pemberton et
al., 2013). Estimates varied from the other data sources examined in this report, including data
from the Medical Expenditure Panel Survey (MEPS) and the NHANES, likely due to differences
in the sampling frames, data collection methods, and questionnaires used to produce these
estimates. Overall, no clear pattern was observed in the direction of differences between NSDUH
and the other data sources.
Finally, DR estimates on six items on disabilities or physical limitations were compared
with estimates from the 2012 NHIS and 2012 ACS. Overall, differences between the DR and
2012 NHIS estimates were smaller than those between the QFT and 2011 NHIS. An exception
was the item on "serious difficulty concentrating, remembering, or making decisions." For that
item, the DR estimate was considerably higher than the estimate from the 2012 NHIS.
8.4.3

Estimates for Items New to the DR Questionnaire

Section 7.3 attempted to compare two items on sexual attraction and orientation that were
added as new items to the DR instrument. There appear to be few public use datasets that can be
used to produce estimates comparable with those from the DR; however, this section did
compare the new item on sexual orientation to a comparable item from the 2012 General Social
Survey (GSS) and published estimates from the 2006 to 2010 National Survey of Family Growth
(NSFG). The DR estimates for sexual orientation appeared to be consistent with data from the
GSS across gender and age groups.

8.5

Implications for the 2015 Partially Redesigned Instrument and Protocol

This section summarizes key issues resulting from the protocol assessments and analyses
conducted as part of the QFT and DR. It also identifies recommendations on how each problem
could be addressed for the partially redesigned NSDUH protocol. Table 8.1 presents issues and
recommendations for the screener and questionnaire items, Table 8.2 presents issues and
recommendations for the training and materials, and Table 8.3 presents issues and
recommendations for the field equipment. Combined, these three tables summarize the key
implications of the QFT and DR results for the 2015 data collection.
The 2015 EDR and the 2015 NSDUH 6-month tables can be used to examine the initial
results for items or estimates that were determined to be problematic in the QFT or DR.
Examples include items with relatively higher missingness rates or estimates that differed
significantly from the current main study data. Table 8.4 documents items that could be included
in the 2015 EDR to provide an initial preview of the data, such as missingness rates. As the first
footnote in Table 8.4 indicates, the EDR uses data collected within the first 2 weeks of the
survey year and presents unweighted estimates for these data. Given that the EDR relies upon
unweighted data to help identify significant problems at the start of a new data collection year,

377

such as unusual missingness rates or timing results, one appropriate way to use the EDR would
be to examine unweighted missingness rates for those items identified as problematic in the QFT
and DR in Table 8.4. Table 8.5 documents similar items recommended for priority examination
in the 2015 NSDUH 6-month tables. Data presented in the 6-month tables are weighted, so more
robust comparisons between main study NSDUH data and field test data are feasible.

378

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire
Recommendation for 2015
Redesigned Questionnaire
The routing logic for the comparable question in 2014
(HLTH18) was updated to prevent respondents who
reported not smoking in the past year in the tobacco
module from receiving this question.

SP03r

Issues(s) Identified in
QFT and/or DR
Respondents who reported using tobacco
in their lifetime, but not necessarily in the
past year, were later asked if a doctor had
advised them to quit smoking in the past
year in HLTH21.
The routing logic for the sexual attraction
(QD62) and sexual orientation (QD63)
questions was incorrect and was missing
a reference to the age variable that
restricts these questions to adult
respondents.
These items include the wording "Is that
correct?" whereas elsewhere in the
questionnaire, confirmation questions
say, "Is this correct?"
See above.

Calendar screens
(throughout)

Dates appear without subscript on
calendar screens.

QD01 and QD03

Revise calendar screens so that dates include the
superscript (i.e., "23rd" as opposed to "23"). This
would include edits to the calendar tutorial. Revise
audio as necessary.
Move the introductory text, "The first few questions
are for statistical purposes only…," from QD01 to
QD03. This will allow interviewers to avoid the
confusion of reading a statement that is unrelated to
the information they are entering.

Questionnaire Item
HLTH21

QD62, QD63

MILTERM1

379
See notes at end of table.

Decision
This change will be
carried forward.

The routing was corrected at DR FI training with an
instrument patch, and these items functioned properly
during data collection.

This change will be
carried forward.

Consider editing these items to say "Is this correct?"
for consistency.

Edit approved by
SAMHSA. RTI will
implement.

See above.

Edit approved by
SAMHSA. RTI will
implement.
Edit approved by
SAMHSA. RTI will
implement.
Change approved by
SAMHSA. RTI has made
this change in the CAI
specs.
(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item
Pain relievers screener and
main module

Pain relievers screener and
main module

380
PR01 in the pain relievers
screener and the
appropriate location in the
main module

See notes at end of table.

Issue(s) Identified in
QFT and/or DR
Propoxyphene products (Darvocet®,
Darvon®, and generic propoxyphene)
have been withdrawn from the U.S.
market.
These brand name drugs have been
discontinued: Talacen®, Talwin®, and
Talwin® NX, but generic pentazocine is
still available.
New formulations for Vicodin®
Suboxone® will affect their drug images.
In addition, the current image for
OxyContin® shows a form that was
discontinued because it was prone to
tampering, along with the recently
introduced form that is designed to be
more resistant to tampering. Legislation
that affects the amount of acetaminophen
(APAP) in opioid pain relievers also
could affect the images for other pain
relievers that contain APAP (e.g., generic
hydrocodone plus APAP, Percocet®).
FDA shows Lorcet® HD as being
discontinued, and this had the lowest
prevalence of misuse for the
hydrocodone products in the QFT and
DR. Also, most respondents in the QFT
and DR data who reported misuse of
Lorcet® reported misuse of other
hydrocodone products. No information is
available on plans to reformulate Lorcet®
to include no more than 325 mg of APAP
as per legislation that went into effect in
January 2014.

Recommendation for 2015
Redesigned Questionnaire
Drop these drugs from the pain relievers screener and
main module for 2015.

Drop these drugs from the pain relievers screener and
main module for 2015, and do not replace them with
their generic equivalents.
Replace the current images for these drugs with new
images, as appropriate.

Drop Lorcet® from the pain relievers screener and
main module.
Add the hydrocodone products Norco® and Zohydro®
ER to the pain relievers screener and main module.
Norco® was mentioned in "write-in" data in the DR
and also was mentioned more often than Lorcet® in
write-in data for 2012. Zohydro® ER was approved in
late October 2013. It is the first single-entity
hydrocodone product and the first extended-release
hydrocodone product. For these reasons, its approval
has been controversial. Market launch is anticipated
for 2014.

Decision
Change approved by
SAMHSA. RTI has made
this change in the CAI
specs.
Change approved by
SAMHSA. RTI has made
this change in the CAI
specs.
Change approved by
SAMHSA. RTI will
investigate availability of
new images.

Changes approved by
SAMHSA. RTI replaced
Lorcet® with Norco® in
PR01 and PRY03. Added
a response category in
PR01 and added
questions PRY031 and
PRY03a1 for Zohydro®
ER between PRY03a and
PRY04.

(continued)

Recommendation for 2015
Redesigned Questionnaire
®
Drop Tylox from the pain relievers screener and main
module.
Add the oxycodone products Roxicet® and
Roxicodone® to the pain relievers screener and main
module. Roxicet® and Roxicodone® show evidence of
increasing trends in the main survey write-in data.

There is a generic equivalent of Ultram®
ER and Ryzolt®. Also, Ryzolt® has been
discontinued. Dropping Ryzolt® will
make room for this drug. Tramadol
products were one of the most commonly
misused groups of pain relievers in the
QFT.

Add "Extended-release tramadol (generic)" to question
PR04 in the pain relievers screener for 2015 and in the
appropriate location in the main module. Also will
require creating a new drug image. Drop Ryzolt® from
the questionnaire for 2015.

381

Issue(s) Identified in
QFT and/or DR
No information is available on plans to
reformulate Tylox® to include no more
than 325 mg of APAP. Most respondents
in the QFT and DR data who reported
misuse of Tylox® reported misuse of
other oxycodone products. In the 2012
main survey, Tylox® was mentioned less
often than other oxycodone products in
the write-in data.

PR04 in the pain relievers
screener and the
appropriate location in the
main module

See notes at end of table.

Decision
Changes approved by
SAMHSA. Because
propoxyphene drugs
previously had been
included in PR03 but will
be dropped for 2015, RTI
split the screener
questions about
oxycodone products
between PR02 and PR03.
Included Roxicet® and
Roxicodone® in PR03.
Replaced Tylox with
Roxicet® in PRY08 and
PRY08a and added
PRY081 and PRY081a
for Roxicodone® between
PRY08a and PRY09.
Changes approved by
SAMHSA. RTI added
new drug to PR04.
Created PRY171 and
PRY171a between
PRY17a and PRY18.
(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Questionnaire
®

Decision

The response, "Asked whether Tylenol
with codeine was the same as OTC
Tylenol®," came up in one of the
debriefing write-ins. Misreporting of
OTC Tylenol® as Tylenol® with codeine
also came up in cognitive testing. In the
QFT, Tylenol® with codeine 3 or 4 was
reported as one of the more commonly
misused pain relievers.

Consider adding an explanation to the screener
question for Tylenol® with codeine 3 or 4 and codeine
pills that Tylenol® with codeine 3 or 4 is different from
OTC Tylenol®. Making this change could reduce the
potential for false positives if respondents report
"misuse" of Tylenol® with codeine 3 or 4 (e.g., use
without a prescription) when they really used OTC
Tylenol®.

No changes requested by
SAMHSA.

PR06 in the pain relievers
screener and the
appropriate location in the
main module

The pain relievers pill card in the main
survey shows images of extended-release
morphine, but there was not a response
option for this form in the DR. Also,
Oramorph® SR has been discontinued.

Add "Extended-release morphine (generic)" to
question PR06 in the pain relievers screener for 2015
and in the appropriate location in the main module.
Also will require creating a new drug image and could
require changing the existing drug image for
morphine. Drop Oramorph® SR from the questionnaire
for 2015.

Changes approved by
SAMHSA. RTI added
new drug to PR06.
Created PRY241 and
PRY241a between
PRY24a and PRY25.

PR08 in the pain relievers
screener and the
appropriate location in the
main module

Subutex® has been discontinued.

Drop Subutex® from the pain relievers screener and
main module for 2015.

Change approved by
SAMHSA.

382

PR05 in the pain relievers
screener and the
appropriate location in the
main module

See notes at end of table.

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Questionnaire

Decision

383

PR08 in the pain relievers
screener and the
appropriate location in the
main module

There are the generic equivalents of
Opana® and Opana® ER.

Add "Oxymorphone (generic)" and "Extended-release
oxymorphone (generic)" to the pain relievers screener
and main module for 2015. This also will require
creating new drug images.

Change approved by
SAMHSA. RTI added
these items to PR09 and
replaced Talacen®,
Talwin®, and Talwin®
NX with remaining pain
relievers in PR10. Moved
up PRY35 to PRY36a to
follow PRY31a. Added
PRY361, PRY361a,
PRY362, and PRY362a
between PRY36a and
PRY32 (new order).

Pain relievers screener and
main module

The DR questionnaire does not identify
that hydromorphone is the equivalent of
Dilaudid®.

Consider adding "Hydromorphone (generic)" to the
pain relievers screener and main module, or change the
items for Dilaudid® to "Dilaudid® or hydromorphone
(generic)." These changes would require creating a
new pill image for hydromorphone or changing the
image for Dilaudid® to add pictures for the generic.

SAMHSA approved
changing to "Dilaudid® or
hydromorphone." RTI
added SAMHSAapproved wording to all
instances of Dilaudid®.

Pain relievers screener and
main module

Exalgo® was approved in 2010. Its
patents are scheduled to expire in
July 2014. Adding Exalgo® and generic
hydromorphone would allow more
complete coverage of the misuse of
hydromorphone products.

Consider adding Exalgo® (extended-release
hydromorphone) to the pain relievers screener and
main module in the vicinity of items for Dilaudid®.
This also will require creating a new drug image.

Change approved by
SAMHSA. RTI added
"Exalgo or extendedrelease hydromorphone"
to PR10, one item after
"Dilaudid® or
hydromorphone." Added
PRY331 and PRY331a
between PRY33a and
PRY34.

See notes at end of table.

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item

Issue(s) Identified in
QFT and/or DR
®

Recommendation for 2015
Redesigned Questionnaire
®

Decision

384

Pain relievers screener and
main module

MoxDuo IR is a combination of two
powerful opioids, morphine and
oxycodone, that is scheduled for a new
review in May 2014. Product launch is
targeted for the second half of 2014.

Add MoxDuo IR to the pain relievers screener and
main module. Also will require obtaining a new drug
image (depending on potential cost).

Change approved by
SAMHSA. RTI added
MoxDuo® IR to PR10.
Added PRY332 and
PRY332a between
PRY331a and PRY34.

PRL03 (and TRL03,
STL03, and SVL03)

Some DR cases for pain relievers skipped
out of PRL03 when the respondent
reported some initiation within the past
12 months but initiation of other drugs in
the same month as the interview month
but in the previous calendar year (e.g.,
September 2012).

Adjust the criteria in the past year initiation variables
(e.g., PRYRINIT1 for Vicodin®) to set the initiation
variable to 1 if the YFU equals CURRENT YEAR
MINUS 1, the MFU is between 1 and 12, and the MFU
is greater than OR EQUAL TO the current month.

Change approved by
SAMHSA.

Tranquilizers screener and
main module

Librium®, Tranxene®, and oxazepam
(generic equivalent of Serax®) represent
an older "generation" of benzodiazepine
tranquilizers and had a low prevalence in
the QFT and DR data.

Drop Librium®, Tranxene®, and oxazepam from the
tranquilizers screener and main module.

Changes approved by
SAMHSA. RTI has made
these changes in the CAI
specs.

Tranquilizers screener and
main module

Flexeril® has been discontinued, but the
generic is still available.

Drop Flexeril® from the tranquilizers screener and
main module, or replace it with "Cyclobenzaprine, also
known as Flexeril®." Replacing Flexeril® with the
generic will require creating a new drug image.

Change approved by
SAMHSA to replace
"Flexeril" with
"Cyclobenzaprine
(generic), also known as
Flexeril®," and to create
an image for the generic.

See notes at end of table.

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Questionnaire

Decision

Buspirone, hydroxyzine, and
meprobamate had low prevalence
reported in the QFT.

Consider dropping buspirone, hydroxyzine, and
meprobamate from the tranquilizers screener and main
module.

Final recommendation
made to SAMHSA to
retain questions about
these drugs for 2015.
Final recommendation
approved by SAMHSA.

Stimulants screener and
main module

Generic extended-release amphetaminedextroamphetamine pills are available
that are equivalent to Adderall® XR.

Add "Extended-release amphetaminedextroamphetamine pills other than Adderall XR
(generic)" to the stimulants screener and main module.
This also will require creating a new drug image.

Change approved by
SAMHSA. RTI has split
amphetamine products in
the screener to include
brand-name drugs in
ST01 and generic drugs
in ST02. Added STY051
and STY051a between
STY05a and STY06.

Stimulants screener and
main module

Generic extended-release
methylphenidate was approved in 2011.
This is equivalent to Ritalin® SR/LA and
Concerta®.

Add "Extended-release methylphenidate (generic)" to
the stimulants screener and main module. This also
will require creating a new drug image and could
require changing the existing drug image for
methylphenidate.

Change approved by
SAMHSA. RTI has put
Ritalin®, Ritalin® SR/LA,
Concerta®, and Daytrana®
in ST03 in the screener
and has put Metadate®
CD, Metadate® ER,
methylphenidate, and
extended-release
methylphenidate in ST04.
Moved STY10 and
STY10a to follow
STY12a. Added STY101
and STY101a between
STY10a and STY13.

385

Tranquilizers screener and
main module

See notes at end of table.

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Issues(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Questionnaire

Stimulants screener and
main module

Generic extended-release
dexmethylphenidate was approved in
2013. This is equivalent to Focalin® XR.

Add "Extended-release dexmethylphenidate (generic)"
to the stimulants screener and main module. This also
will require creating a new drug image.

Change approved by
SAMHSA. RTI has put
Focalin®, Focalin® XR,
dexmethylphenidate, and
extended-release
dexmethylphenidate in
ST05 in the screener.
Added STY151 and
STY151a between
STY15a and STY16.

Stimulants screener and
main module

Modafinil is the generic equivalent of
Provigil® and was approved in 2012.
Armodafinil (generic equivalent of
Nuvigil®) also was approved in 2012 and
is chemically similar to Modafinil.

Consider adding "Modafinil (generic)," Nuvigil®, and
"Armodafinil (generic)" to the stimulants screener and
main module. This also will require creating new drug
images.

Final recommendation
made to SAMHSA not to
add these drugs for 2015,
based on QFT and DR
data for Provigil® and
write-in data from the
main survey for
modafinil, Nuvigil®, and
armodafinil. Final
recommendation
approved by SAMHSA.

Sedatives screener and
main module

Patents for Lunesta® are scheduled to
expire in February and August 2014.
Tentative approvals have been given to
the generic equivalent (eszopiclone). Use
and misuse of Lunesta®, Sonata®, and the
generic equivalent of Sonata® (zaleplon)
in the QFT and DR data were not as
prevalent as those for zolpidem products
(e.g., Ambien®) or benzodiazepine
sedatives (e.g., Halcion®, Restoril®).

Change "Lunesta" to "Lunesta or eszopiclone" and
change "Sonata" to "Sonata or zaleplon" in the
sedatives screener and main module. Drop the separate
response category for zaleplon in the sedatives
screener and the corresponding questions in the main
module.

Changes approved by
SAMHSA. RTI has made
these changes in the CAI
specs.

Questionnaire Item

386
See notes at end of table.

Decision

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item

Issue(s) Identified in
QFT and/or DR
®

Recommendation for 2015
Redesigned Questionnaire
®

Decision

387

Sedatives screener and
main module

Dalmane has been discontinued, but the
generic is still available.

Drop Dalmane from the sedatives screener and main
module. Consider replacing the items for Flurazepam
with "Flurazepam (generic), also known as Dalmane."

Changes approved by
SAMHSA to drop
Dalmane® and replace
"Flurazepam" with
"Flurazepam (generic),
also known as Dalmane."
RTI has moved Restoril®
and temazepam to SV03
in the sedatives screener.
Moved SVY13 and
SVY13a to appear
between SVY10a and
SVY11.

DRSV11 and DRSV12

The CAI instrument asks whether the
respondent suffers from one or more
symptoms of withdrawal in DRSV11 and
DRSV12, but the DSM-IV criteria call
for two or more symptoms.

Update the criteria to be consistent with the DSM-IV
for 2015.

Change approved by
SAMHSA. RTI will
correct the criteria for
2015.

QD16a

Those who report not having moved in
the last year (at QD13) are being asked if
they have lived in the United States for at
least a year unnecessarily.

Skip those who report not having moved in the past
year (QD13) out of QD16a (lived in the United States
for at least a year).

Change approved by
SAMHSA. RTI to pursue.

QP02

The default value for "sample member" is
problematic.

Brainstorm a new default value for "sample member."

SAMHSA approved. In
progress.

QH113 and QH114

These screens have electronic calendars.
Because they are being moved to CAPI,
respondents will no longer have calendar
access to assist in answering the
questions.

Add "that is from (fill) through today" to the question
text for these items.

Change approved by
SAMHSA. RTI will
pursue.

See notes at end of table.

(continued)

Table 8.1 Issues Identified for Screener and Questionnaire Items from the QFT or DR Analysis and Recommendation for the 2015
Partially Redesigned Questionnaire (continued)
Questionnaire Item
Q108N

Issue(s) Identified in
QFT and/or DR
QI08N asks about cash assistance from a
State or county welfare program, yet
some States do not have counties (e.g.,
Alaska has boroughs, and Louisiana has
parishes).

Recommendation for 2015
Redesigned Questionnaire
Alter the wording of the question based on State
recorded earlier in the interview.

Decision
Change approved by
SAMHSA. RTI will
pursue.

APAP = acetaminophen; CAI = computer-assisted interviewing; CAPI = computer-assisted personal interviewing; DR = Dress Rehearsal; DSM-IV = Diagnostic and Statistical
Manual of Mental Disorders, 4th edition; FDA = U.S. Food and Drug Administration; FI = field interviewer; MFU = month of first use; mg = milligram; OTC = over-the-counter;
QFT = Questionnaire Field Test; RTI = RTI International (a trade name of Research triangle Institute); SAMHSA = Substance Abuse and Mental Health Services Administration;
specs = specifications; YFU = year of first use.

388

Table 8.2 Issues Identified for Training and Materials from the QFT or DR Analysis and Recommendation for the 2015 Partially
Redesigned Protocol
Training/Materials
Element

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Protocol

Decision

Response options for QHI15 are
currently not read to respondents. This
option could unnecessarily be associated
with high respondent burden if a
respondent was trying to differentiate
between 18 months ago and 2 years ago.
Instead, those fall into the same
response category.

Develop showcard for QHI15.

Will not be pursued.

Materials

During a field observation, a respondent
did not notice the instruction to press the
space bar to enter more than one
response. He asked for help, but the FI
did not know about this type of item and
just told him to do his best, which did
not solve the problem.

Develop a job aid to provide a summary of types of
response options or errors to allow FIs to
troubleshoot respondent issues. This could include a
mention of "check all that apply" items, open-ended
items, a discrete list, open-ended numeric, and so on.
It could also include explanations of range checks
and hard errors that respondents may encounter.

Will be changed for
2015.

Training

Some FIs do not seem to be familiar
with the instruction that they can say,
"Is that you?" if obvious during the CAI
HH roster.

FIs should be reminded of this option during training
so as to minimize respondent confusion when
responding "Self."

Will be changed for
2015.

Training

Some trainees reported that the pace of
the training was not right for them.

Revise the training agenda as needed to reflect timing
and pacing issues noted and make adjustments as
feasible.

Will be pursued,
pending SAMHSA
review.

389

Materials

See notes at end of table.

(continued)

Table 8.2 Issues Identified for Training and Materials from the QFT or DR Analysis and Recommendation for the 2015 Partially
Redesigned Protocol (continued)
Training/Materials
Element

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Protocol

Decision

FI laptops will not have CD drives.

Investigate options for Web-based training.

RTI will summarize the
pros and cons and cost
estimate of switching to
Web-based training and
submit it to SAMHSA.

Training

Some FIs were unfamiliar with the
computer equipment (particularly the
tablet), which caused some difficulty
during training.

Ship the tablets to FIs in advance of the training to
allow them opportunities to practice and gain
familiarity prior to classroom learning. Extend the
training agenda to include additional time on the
equipment.

Change approved by
SAMHSA. Equipment
will be purchased earlier
and shipped to FIs prior
to 2015 veteran training.

Training

Some FIs appeared to lack knowledge
of proper NSDUH procedures.

Extend the training agenda to focus additional time
on the application of proper NSDUH interviewing
procedures. Conduct a certification exercise at the
conclusion of training to formally evaluate each FI's
adherence to screening and interviewing procedures.

Pending SAMHSA
approval. Cost estimate
submitted by RTI to
SAMHSA on 12/9/13.

Training

Observers noticed some FIs struggling
with the introduction screen and use of
correct terminology.

Provide additional emphasis on and practice with
reading the introductory screen and using correct
terminology when speaking to respondents.

Will be pursued,
pending SAMHSA
review.

Training

Observers noticed some difficulties with
respondents being confused when asked
the race/ethnicity questions.

Provide additional training and practice on how to
address respondent confusion with the race/ethnicity
questions.

Will be pursued,
pending SAMHSA
review.

Training

Some FIs have problems with basic
troubleshooting of the interviewing
laptop.

Provide FIs with additional documentation and
training on laptop troubleshooting.

Will be pursued,
pending SAMHSA
review.

390

Training

CAI HH = computer-assisted interviewing household; DR = Dress Rehearsal; FI = field interviewer; NSDUH = National Survey on Drug Use and Health; QFT = Questionnaire
Field Test; RTI = RTI International (a trade name of Research triangle Institute); SAMHSA = Substance Abuse and Mental Health Services Administration.

Table 8.3 Issues Identified for Field Equipment from the QFT or DR Analysis and Recommendation for the 2015 Partially Redesigned
Protocol
Equipment Element

Issue(s) Identified in
QFT and/or DR

Recommendation for 2015
Redesigned Protocol

Decision

At the end of the DR data collection
period, a problem was discovered in the
letter-generating system that prevented
refusal and UTC letters from being sent.
Although letters were requested by the
field, the letters were not sent to
respondents from RTI as planned

RTI to address system problem to ensure that this is
not a problem in subsequent data collection.

Pending.

Laptop bag

The laptop bag was not well received
due to the design of its strap and
pockets, based on equipment survey
results and comments.

Identify a bag that has straps connecting to its sides
(rather than on the front and back), more interior
pockets, and fewer zippered pockets. Consider the
Samsonite bag for 2015 that was considered for the
DR.

Pending.

Laptop

Function keys were not easily
identifiable and confusing to some.

Use a different labeling method to identify function
keys and make them more easily identifiable.
Attempt to purchase laptops that have larger function
keys or otherwise make the function keys more easily
read.

Pending.

Screening/tablet

Some FIs reported that the additional
password on the tablet screen was
cumbersome.

Consider removing the extra password from the
tablet screen.

Pending.

391

Letter-generating system

DR = Dress Rehearsal; FI = field interviewer; QFT = Questionnaire Field Test; RTI = RTI International (a trade name of Research triangle Institute); UTC = unable to contact.

Table 8.4 Estimates and Items Identified from the QFT or DR Analysis for Preview in the 2015
Early Data Review

Estimate or Questionnaire Item1
Are you now married, widowed, divorced, or
separated, or have you never married? (QD07)
Is anyone in your immediate family currently serving
in the U.S. military? (QD10d)
How many times in the past 12 months have you
moved? (QD13)
In what State did you live in 1 year ago today?
(QD13a)
Are you a full-time student or a part-time student?
(QD19)
During the past 30 days, how many whole days of
school did you miss because you were sick or
injured? (QD20)
During the past 30 days, how many whole days of
school did you miss because you skipped or "cut" or
just didn't want to be there? (QD21)
Did you work at a job or business at any time last
week? (QD26)
Did you work at a job or business at any time during
the past 12 months? (QD33)
How many different employers have you had in the
past 12 months? (QD36)
In how many weeks during the past 12 months did
you not have at least one job or business? (QD38)
In what year did you last work at a job or business?
(QD39a)
During the past 30 days, how many whole days of
work did you miss because you were sick or injured?
(QD40)
During the past 30 days, how many whole days of
work did you miss because you just didn't want to be
there? (QD41)
How many people work for your employer out of this
office, store, etc.? (QD42)
In [YEAR], did you receive Supplemental Security
Income or SSI? (QI03N)
In [YEAR], did you receive food stamps? (QI07N)
See notes at end of table.

Item Missingness Rate
Was Significantly
Higher than
Comparison Data

Estimate Was
Significantly Different
from Comparison Data

Yes

N/A

Yes

N/A

Yes

N/A

Yes

Yes
(continued)

392

Table 8.4 Estimates and Items Identified from the QFT or DR Analysis for Preview in the 2015
Early Data Review (continued)

Estimate or Questionnaire Item1
At any time during [YEAR], even for 1 month, did you
receive any cash assistance from a State or county
welfare program such as [TANFFILL]? (QI08N)
In [YEAR], because of low income, did you receive
any other kind of nonmonetary welfare or public
assistance? (QI10N)
What is the highest grade or year of school you have
completed? Just give me the number from the card.
(QD11)2
What grade or year of school are you now attending? /
What grade or year of school will you be attending
when your vacation is over? (QD18)
Please answer this question again. [IF QD17 = 1] What
grade or year of school are you now attending? / [IF
QD17b = 1] What grade or year of school will you be
attending when your vacation is over? (QD18CC03)
Please answer this question again. What is the highest
grade or year of school you have completed?
(QD18CC04)

Item Missingness Rate
Was Significantly
Higher than
Comparison Data

Estimate Was
Significantly Different
from Comparison Data

Yes

DR = Dress Rehearsal; QFT = Questionnaire Field Test.
1
The Early Data Review provides unweighted results from data collected in the first 2 weeks of the survey year.
2
The changes to the education categories in QD11 could affect the classification of highest education for adults in
2015.

393

Table 8.5 Estimates and Items Identified from the QFT or DR Analysis for Priority Examination
in the 2015 6-Month Tables

Estimate or Questionnaire Item1
Lifetime Inhalants Use
Past Year Smokeless Tobacco Use
Past Month Smokeless Tobacco Use
Past Month Serious Psychological Distress (SPD)
Past Year Cocaine Use
Past Month Cocaine Use
Past Year Heroin Use
Past Month Heroin Use
Lifetime Use of Any Prescription Drug
Past Year Use of Any Prescription Drug
Cell Phone Status

Item Missingness Rate
Was Significantly
Higher Than
Comparison Data
No
No
No
No
No
No
No
No
No
No
No

Estimate Was
Significantly Different
from Comparison Data
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

DR = Dress Rehearsal; QFT = Questionnaire Field Test.
1
The 6-month tables provide weighted results for data collected from January through June of the survey year.

394

References
Biemer, P., & Wang, K. H. (2006). National Survey on Drug Use and Health: Adjusting for
nonignorable nonresponse in the NSDUH using interview callback models (prepared for the
Substance Abuse and Mental Health Services Administration). Research Triangle Park, NC:
RTI International.
Brewer, K. R. W. (1963). A model of systematic sampling with unequal probabilities.
Australian Journal of Statistics, 5(1), 5-13. doi: 10.1111/j.1467-842X.1963.tb00132.x
Brewer, K. R. W. (1975). A simple procedure for sampling πpswor. Australian Journal of
Statistics, 17(3), 166-172. doi: 10.1111/j.1467-842X.1975.tb00954.x
Brooke, J. (n.d.). SUS - A quick and dirty usability scale. Retrieved from
http://hell.meiert.org/core/pdf/sus.pdf
Center for Behavioral Health Statistics and Quality. (2013). Results from the 2012 National
Survey on Drug Use and Health: Summary of national findings (HHS Publication No. SMA
13-4795, NSDUH Series H-46). Rockville, MD: Substance Abuse and Mental Health Services
Administration.
Chen, P., Dai, L., Cribb, D., Gordek, H., Laufenberg, J., Sathe, N., & Westlake, M. (2014).
Person-level sampling weight calibration. In 2012 National Survey on Drug Use and Health:
Methodological resource book (Section 12, prepared for the Substance Abuse and Mental Health
Services Administration under Contract No. HHSS283201000003C, Phase I, Deliverable No. 39,
RTI/0212800.001.107.004). Research Triangle Park, NC: RTI International.
Chromy, J. R., & Penne, M. A. (2002). Pair sampling in household surveys. In Proceedings of
the 2002 Joint Statistical Meetings, American Statistical Association, Survey Research Methods
Section, New York, NY (pp. 552-554). Alexandria, VA: American Statistical Association.
Colliver, J. D., Kroutil, L. A., Dai, L., & Gfroerer, J. C. (2006). Misuse of prescription drugs:
Data from the 2002, 2003, and 2004 National Surveys on Drug Use and Health (HHS
Publication No. SMA 06-4192, Analytic Series A-28.) Rockville, MD; Substance Abuse and
Mental Health Services Administration, Office of Applied Studies.
Currivan, D., Kroutil, L., Wang, K., Chen, P., Davis, T., Granger, B., Johnson, I., LeBaron, P.,
McHenry, G., McKamey, A., Morton, K., Myers, S., Scott, V., Shook-Sa, B., Terrey, S., &
Zelko, H. (2013). National Survey on Drug Use and Health: 2012 Questionnaire Field Test
report (prepared for the Substance Abuse and Mental Health Services Administration under
Contract No. HHSS283200800004C, RTI 0211838.102.003.008.005). Research Triangle Park,
NC: RTI International.

395

Currivan, D., Peytcheva, E., Rodriguez, S., Stolzenberg, S., Leiman, J., & Obscura, A. (2009).
Report on the contact materials focus group findings and recommendations (prepared for the
Substance Abuse and Mental Health Services Administration). Research Triangle Park, NC:
RTI International.
Folsom, R. E., & Singh, A. C. (2000). The generalized exponential model for sampling weight
calibration for extreme values, nonresponse, and poststratification. In Proceedings of the 2000
Joint Statistical Meetings, American Statistical Association, Survey Research Methods Section,
Indianapolis, IN (pp. 598-603). Alexandria, VA: American Statistical Association.
Gates, G. J. (2011, April). How many people are lesbian, gay, bisexual and transgender?
Retrieved from http://williamsinstitute.law.ucla.edu/research/census-lgbt-demographicsstudies/how-many-people-are-lesbian-gay-bisexual-and-transgender/
Gfroerer, J., Eyerman, J., & Chromy, J. (Eds.). (2002). Redesigning an ongoing national
household survey: Methodological issues (HHS Publication No. SMA 03-3768). Rockville, MD:
Substance Abuse and Mental Health Services Administration, Office of Applied Studies.
Kroutil, L. A., Handley, W., & Bradshaw, M. R. (2012). General principles and procedures for
editing drug use data in the 2010 NSDUH computer-assisted interview. In 2010 National Survey
on Drug Use and Health: Methodological resource book (Section 10, prepared for the Substance
Abuse and Mental Health Services Administration under Contract No. HHSS283200800004C,
Deliverable No. 39, RTI/0211838.107.003). Research Triangle Park, NC: RTI International.
Kroutil, L. A., Handley, W., Bradshaw, M. R., Chien, C., & Felts, B. J. (2012, January).
Procedures for editing supplementary self-administered data in the 2010 NSDUH computerassisted interview. In 2010 National Survey on Drug Use and Health: Methodological resource
book (Section 10, prepared for the Substance Abuse and Mental Health Services Administration
under Contract No. HHSS283200800004C, Deliverable No. 39, RTI/0211838.107.003).
Research Triangle Park, NC: RTI International.
Kroutil, L., Vorburger, M., & Aldworth, J. (2007, October). Evaluation of the effects of new
noncore drug data on prevalence estimates in the National Survey on Drug Use and Health
(NSDUH). In 2006 National Survey on Drug Use and Health: Methodological resource book
(Section 18, prepared for the Substance Abuse and Mental Health Services Administration,
Office of Applied Studies, under Contract No. 283-2004-00022, Deliverable No. 39,
RTI/0209009). Research Triangle Park, NC: RTI International.
Kroutil, L. A., Vorburger, M., Aldworth, J., & Colliver, J. D. (2010). Estimated drug use based
on direct questioning and open-ended questions: Responses in the 2006 National Survey on Drug
Use and Health. International Journal of Methods in Psychiatric Research, 19(2), 74-87. doi:
10.1002/mpr.302
Morton, K. B., Martin, P. C., Shook-Sa, B. E., Chromy, J. R., & Hirsch, E. L. (2013). Sample
design report. In 2012 National Survey on Drug Use and Health: Methodological resource book
(Section 2, prepared for the Substance Abuse and Mental Health Services Administration under
Contract No. HHSS283201000003C, Phase I, Deliverable No. 8, RTI/0212800.001.103).
Research Triangle Park, NC: RTI International.
396

Morton, K. B., & Shook-Sa, B. E. (2012). 2013 National Survey on Drug Use and Health:
Sample design plan (prepared for the Substance Abuse and Mental Health Services
Administration, Center for Behavioral Health Statistics and Quality, under Contract No.
HHSS283201000003C). Research Triangle Park, NC: RTI International.
Pemberton, M. R., Bose, J., Kilmer, G., Kroutil, L. A., Forman-Hoffman, V. L., & Gfroerer, J. C.
(2013, September). CBHSQ Data Review: Comparison of NSDUH health and health care
utilization estimates to other national data sources. Rockville, MD: Substance Abuse and Mental
Health Services Administration.
RTI International. (2008). SUDAAN®, Release 10.0 [computer software]. Research Triangle
Park, NC: Author.

397

398

Appendix A: Screener and Questionnaire Changes Made for
the Questionnaire Field Test (QFT) and Instrument and
Protocol Revisions Made for the Dress Rehearsal (DR)

Table A.1

Changes between the 2012 NSDUH Screener and the 2012 Questionnaire Field Test (QFT) Screener

Screener Item
Study Introduction Screen



Informed Consent Screen



Confirm Roster Screen



FI Debriefing Items



QFT Changes
On study introduction screen, "RTI International" replaced previous
mentions of "RTI" or "Research Triangle Institute," and "US Public Health
Service" was replaced with "US Department of Health and Human
Services."
Replaced mentions of the incentive by saying, "$30 in cash" as opposed to
"$30 cash payment."
Removed "Other" and "Unspecified" from verify roster screen for roster
members who are missing race, ethnicity, or military status.
New FI debriefing questions were added and administered at the end of the
screening for codes 30, 31, 32, and 70 cases.

Notes

This change also was made in
2013.
This deletion was also made in
2013.

A-1

Table A.2

Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire

Module
Core Demographics






Beginning ACASI Section



ACASI Tutorial
Calendar




A-2

Tobacco




Alcohol
Marijuana
Cocaine
Crack
Heroin
Hallucinogens



Inhalants



See notes at end of table.



QFT Changes
Moved the marital status questions out of this module to the back-end
ACASI.
Revised response categories for educational attainment in QD11, adding
additional degree-level detail (from 17 to 20 categories).
Added "Guamanian or Chamorro" and "Samoan" race categories.
Added questions for military veterans to identify era served and whether in
combat zone.
Added question for the FI or R to indicate that the R has already done the
tutorial so he or she does not have to do the full tutorial again.
No Changes
New electronic version was introduced in the tutorial, so calendar-filling
questions were altered to reflect the on-screen calendar.
This module was moved to follow the ACASI tutorial and was administered
via ACASI.
Combined smokeless sections into one section.
Dropped smokeless brands.
Changed binge definition to four or more drinks for females.
No Changes
No Changes
No Changes
No Changes
Moved questions about Ketamine/Special K, DMT/AMT/Foxy, and Salvia
divinorum from the special drugs module to hallucinogens.
Added questions about markers and air duster.

Notes
Race and military changes also
were made in 2013.
Further revised era response
categories for DR.

Added logic to PREVCOM2 in
DR to only ask of adults.

(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
*Methamphetamine
All Prescription Drugs




**Screener and Main
Module

QFT Changes
Created a new methamphetamine module modeled after cocaine.
Adapted each of the former prescription drug modules to two new modules:
a screener and a main module.
– Added the word "prescription" to the introduction for each section.
– Electronic pill cards replaced the showcards.

Pain Relievers
**Screener and Main
Module



Adopted "ensemble format": All screener modules determining any use in
past 12 months are asked before misuse (main) modules.
The main modules measured misuse among those drugs that were used in
the past 12 months, as well as lifetime misuse.
Discontinued prescription drugs were deleted, and newly available drugs
were added.
Drugs currently asked about in noncore modules were moved to the
appropriate prescription drug module.
Edited the definition of misuse to generally refer to use "in any way a
doctor did not direct you to use them."
See changes described under "All Prescription Drugs."

Tranquilizers
**Screener and Main
Module



See changes described under "All Prescription Drugs."






A-3



See notes at end of table.

Notes

(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
Stimulants
**Screener and Main
Module
Sedatives
**Screener and Main
Module
Special Drugs










A-4



Risk/Availability
Blunts
See notes at end of table.



QFT Changes
See changes described under "All Prescription Drugs."
Moved questions about 12-month and 30-day needle use from special drugs
module to stimulants.
See changes described under "All Prescription Drugs."
Removed all methamphetamine questions except lifetime and recency of
methamphetamine needle use.
Removed "Desoxyn, or Methedrine" because they are no longer on the
market.
Removed Ketamine/Special K, DMT/AMT/Foxy, and Salvia divinorum,
Ambien®, and Adderall®, moving them to other modules as appropriate.
Moved stimulant needle use lifetime and recency questions to stimulants
module.
Added an introduction to SD17 (2013 variable SD12), the question about
reusing needles, to remind Rs about needle use reported in the stimulant
module.
Replaced all instances of "not prescribed for you or that you took only for
the experience or feeling it caused" with "not prescribed for you."
No Changes
Added questions on medical marijuana use (MJMM01 and MJMM02).

Notes

Medical marijuana questions also
were added in 2013.
(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
Substance Dependence and
Abuse
Special Topics
Market Information for
Marijuana
Prior Substance Use







Drug Treatment
Health Care



A-5



Adult Mental Health
Service Utilization
Social Environment
Parenting Experiences
Youth Experiences
Mental Health
Adult Depression
Youth Mental Health
Service Utilization
Adolescent Depression
Consumption of Alcohol




See notes at end of table.

QFT Changes
Updated the stimulant questions to reflect the separate methamphetamine
and prescription stimulants modules.
No Changes
Dropped entire module.
Dropped all prescription drug and methamphetamine questions.
Dropped "which (substance use) came first" questions.
No Changes
Extended the module to ask questions about a number of additional topics,
including revising the list of health conditions, adding questions about
cancer diagnosis, and asking about the age at first diagnosis of other
conditions.
Added height and weight questions.
Added questions on whether R has had discussions with a doctor about
substance use in the past year.
No Changes

Notes

Height, weight, and the
discussions one has had with a
doctor about substance use in the
past year also were added in
2013.

Dropped SEN04 - # of times moved in past 5 years.
No Changes
Dropped YE04 - # of times moved in past 5 years.
No Changes
No Changes
No Changes
No Changes
Dropped all prescription drugs (except for methamphetamine) from "used
with alcohol" question (CA09).
Dropped 4+ binge questions for females.
(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
End of ACASI & BackEnd Demographics

QFT Changes





Education






Employment




A-6

Household Roster
Proxy Information



**Beginning Proxy
Tutorial
**Proxy Tutorial







See notes at end of table.

Moved to ACASI.
QD13 (number of moves in the past 12 months) was edited to add the
interviewer note to the question text.
Added six questions about disability.
Added a question about English-speaking ability.
Moved to ACASI.
Revised consistency check questions to be consistent with the revised
categories in QD11 (educational attainment).
Moved questions about marital status to this module.
Added questions about military families to this module.
Moved to ACASI.
Dropped I&O questions (job titles, industries).
Dropped step, foster, adoptive descriptions of children and parents.
Edited the statement that asks Rs about whether they would like to nominate
a proxy (PROXYINT), adding the information that the next questions are
about the family's income as well.
New module to introduce proxy R to CAI program.
Administered via CAPI.
New module to introduce proxy R to CAI program.
Administered via ACASI.

Notes
Sexual orientation questions
were added in the DR.

Military families' questions were
edited in the DR.

This change also was made in
2013.

(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
Health Insurance

QFT Changes









A-7



See notes at end of table.

Moved to ACASI.
Explanations on what CHAMPUS and CHAMPVA stand for were edited.
Edited wording of QHI02 (Medicaid) to include information formerly in an
interviewer note.
Edited wording of QHI02a (CHIP) to include information formerly in an
interviewer note.
Edited wording of QHI03 (CHAMPUS) to include information formerly in
parentheticals and an interviewer note.
Edited wording of QHI06 (private health insurance) to include information
formerly in an interviewer note.
Edited wording of QHI07 (source of private health insurance) to include
information formerly in an interviewer note.
Edited wording of QHI11 (other health insurance) to include information
formerly in an interviewer note.
Edited wording of QHI14 (time without health insurance) to include
information formerly in an interviewer note.

Notes
Explanations on what
CHAMPUS and CHAMPVA
stand for were also edited in
2013.

(continued)

Table A.2 Changes between the 2012 NSDUH Questionnaire and the 2012 Questionnaire Field Test (QFT) Questionnaire (continued)
Module
Income

QFT Changes









A-8




Incentive
Verification
FI Observation Questions





Moved to ACASI.
Edited wording of QI01N (social security or railroad retirement payments)
to include information formerly in an interviewer note and removed
parenthetical statement about checks being mailed on the 3rd of the month
in a gold envelope.
Edited wording of QI03N (SSI) to include information formerly in an
interviewer note.
Edited wording of QI07N (food stamps) to include information formerly in
an interviewer note.
Removed the information formerly contained in an interviewer note for
QI08N (TANF).
Edited wording of QI10N (non-monetary welfare) to include information
formerly in an interviewer note.
Edited INTRTINN for ACASI administration.
Top response category for income was revised to $150,000 or more.
Replaced telephone landline question with two questions about cellular
phones.
Made slight wording changes, such as change "pay" to "hand" or "given."
Changed "mailing" to "current" address.
Moved to tablet and tailored for QFT.

Notes
The parenthetical statement was
also removed from QI01N in
2013.
Further edits to the wording of
INTRTINN were also made in
2013.
Changed the question to genderappropriate pronoun in
INTROINC in 2013.

Change also made in 2013.
Change also made in 2013.

ACASI = audio computer-assisted self-interviewing; AMT = alpha-methyltryptamine; CAI = computer-assisted interviewing; CAPI = computer-assisted
personal interviewing; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of
the Veterans Administration; CHIP = Children's Health Insurance Program; DMT = dimethyltryptamine; DR = Dress Rehearsal; FI = field interviewer; NSDUH
= National Survey on Drug Use and Health; QFT = Questionnaire Field Test; R = respondent; SSI = supplemental security income; TANF = Temporary
Assistance for Needy Families; VA = Department of Veterans Affairs.
NOTE: The QFT interviews were conducted in English only.

Table A.3

Changes between the 2012 Questionnaire Field Test (QFT) Screener and the 2013 Dress Rehearsal (DR) Screener

Screener Item

DR Changes

Notes

SDU Characteristic
Screen



Removed.

This screen also will be removed
for 2014.

Controlled Access Type
Screen



Removed.

This screen also will be removed
for 2014.

Confirm Roster Screen



Removed "Other" and "Unspecified" from verify roster screen for roster
members who are missing race, ethnicity, or military status.

This deletion was also made in
2013.

Administrative Items



Added ability to enter appointments in tablet calendar.
Added the call distribution feature.
Debriefing items were dropped for codes 30, 31, and 32. FIs will answer
debriefing items only for interview Rs (code 70).



FI Debriefing Items



A-9

Table A.4

Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and the 2013 Dress Rehearsal (DR) Questionnaire

CAI Module
Core Demographics



Beginning ACASI
Section



ACASI Tutorial







A-10

Calendar



Tobacco



DR Changes
Updated response options for military service periods.

No Changes

Marijuana

No Changes

Cocaine

No Changes

Crack

No Changes

Heroin

No Changes
Fixed a routing error in three recency variables that affected nine cases in the
QFT.
No Changes

Inhalants
Methamphetamine
See notes at end of table.



Updated response options also
will be used in 2014.

Removed the F7 function, which muted the sound during ACASI.
Added logic so that interviewers were no longer asked whether an adolescent R
had previously served as a proxy.
Edited questions throughout the tutorial for clarity.
The edited response option in
Edited final response option in ALLAPPLY to "something else" (kinds of music ALLAPPLY also will be used in
listened to).
2014.
Added language about how to enter a response after closing the F2 Help box.
Added "Practice Question #" on each screen to clarify that the questions are
unrelated to the survey content.
Edited the calendar screen to indicate how to close the calendar by pressing F1
again.
Changed the name of "Macanudos" to "Macanudo" for accuracy.

Alcohol

Hallucinogens

Notes

No Changes
(continued)

Table A.4 Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and the 2013 Dress Rehearsal (DR) Questionnaire
(continued)
CAI Module

DR Changes

All Prescription Drug
Screeners
All Prescription Drug
Main Modules

No Changes






A-11

Prescription Pain
Reliever Main Module





Tranquilizers Main

Module
Stimulants Main Module

Sedatives Main Module



Special Drugs



See notes at end of table.

Removed Help screens from all prescription drug questions except age at first
use items.
Added logic to identify Rs with unknown recent initiation. A variable was
created to identify and route these Rs to another new variable that asks about
initiation of misuse of prescription pain relievers more than 12 months ago if the
only definite reports of initiation occurred in the past 12 months, or all initiation
data were missing.
Fixed a skip pattern error that skipped source of prescription drug for adolescents
in the QFT.
Edited logic and ranges throughout the modules for accuracy.
Changes listed above.
Edited response options of PRYMOT1 for consistency with the other
prescription drug questions.
Changes listed above.
Changes listed above.
Changes listed above.
No Changes

Risk/Availability
Blunts

Notes

No Changes
Edited routing into MJMM01 and MJMM02 (medical marijuana) to include all
Rs who earlier reported marijuana use in the past 12 months.
Added the words, "or other health care professional" to the medical marijuana
questions.




Edit also will be used in
2014.
Edit will be used in 2013.
(continued)

Table A.4 Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and the 2013 Dress Rehearsal (DR) Questionnaire
(continued)
CAI Module
Substance Dependence
and Abuse



Special Topics
Prior Substance Use
Drug Treatment



Health Care




A-12



DR Changes
Redefined MET12MON (the variable defining 12-month use of
methamphetamine) to include use of methamphetamine with a needle reported in
the special drugs module.
No Changes

No Changes
Changed the data structure on TX10 to allow R to choose all 12 possible options.
This corrected a QFT error.
Added "B or C" to "Hepatitis" in HLTH25 for more precise description of
New ranges for height and weight
condition.
were changes via a patch at the
Edited logic so that Rs who are diagnosed with cancer at their current age are no start of Q2 data collection.
longer asked whether they had cancer in the past 12 months.
Edited ranges for height in feet and inches for accuracy and increased upper limit
for weight.

Adult Mental Health
Service Utilization

No Changes

Social Environment

No Changes

Parenting Experiences

No Changes

Youth Experiences

No Changes

Mental Health

No Changes

Adult Depression

No Changes

Youth Mental Health
Service Utilization

No Changes

Adolescent Depression

No Changes

Consumption of Alcohol

No Changes

See notes at end of table.

Notes

(continued)

Table A.4 Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and the 2013 Dress Rehearsal (DR) Questionnaire
(continued)
CAI Module
End of ACASI & BackEnd Demographics



Education






Employment





A-13

Household Roster
Proxy Information/
Decision



Beginning Proxy Tutorial 



Proxy Tutorial




See notes at end of table.

DR Changes
Added two questions on sexual attraction and identity.
Further revised consistency check questions to be consistent with categories in
QD11 (educational attainment).
Moved definition of "immediate family" from help screen to question text in
military family questions.
Made other minor wording changes to military questions for clarity.
Added an "Other, Specify" item in the military family series.
Edited the Help instructions to several questions.
Deleted question about size of workplace.
Added a new variable and appropriate logic, which instructs Rs to lock the
ACASI portion of the instrument before returning the computer to the
interviewer (PENTER1).
No Changes
Edited logic so if a R lists a parent, grandparent, or in-law as a household
member, these relations will be listed regardless of missing age data.

Notes

This change also was made in
2013.

Added wording noting that the proxy's family member nominated him or her to
take part in the interview.
Added "and headphones" to IntrAcasi1b.
Deleted the F7 instruction for the proxy.
Edited wording of tutorial questions for clarity.
Added language about how to enter a response after closing the F2 box. Added
this language to all questions that include F2 Help boxes.
(continued)

Table A.4 Changes between the 2012 Questionnaire Field Test (QFT) Questionnaire and the 2013 Dress Rehearsal (DR) Questionnaire
(continued)
CAI Module
Health Insurance





Income








A-14

Incentive
Verification



DR Changes
Updated State program names for Medicaid, CHIP, and TANF.
Removed Help instructions in QHI06 (private health insurance) and moved key
terms to the question itself.
Dropped "including Indian Health Insurance" from QHI11.
Edited the wording of INTROINC for improved flow.
Deleted QI05N (income from wages or pay) and added this source to the list in
INTRTINN.
Edited the wording of QI03N (SSI) for accuracy.
Edited the wordings of QI07N (food stamps) for accuracy.
Edited the list of income sources in INTRTINN. Also edited the introductory
text.
Added a new variable and appropriate logic, which instructs Rs to lock the
ACASI portion of the instrument before returning the computer to the FI.
No Changes
Edited the wording of TOALLR3I for clarity.

Notes

INTROINC wording change also
will be used in 2014.

ACASI = audio computer-assisted self-interviewing; CAI = computer-assisted interviewing; CHIP = Children's Health Insurance Program; DR = Dress
Rehearsal; FI = field interviewer; NSDUH = National Survey on Drug Use and Health; QFT = Questionnaire Field Test; R = respondent; SSI = supplemental
security income; TANF = Temporary Assistance for Needy Families.
NOTE: DR interviews were conducted in both English and Spanish.

Appendix B: Item Missingness Tables for English-Language
and Spanish-Language Interviews

Table B.1 Item Missingness Rates for Moved Items for English-Language Non-Hispanic Interviews in 2012 Comparison Data, 2013
Comparison Data, and Combined 2012 Questionnaire Field Test and 2013 Dress Rehearsal Data among All Persons Aged 12 or
Older
2012 Comparison Data1,2

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

Number of
Number of
Percent
Number of Number of
Number of
Cases with
Percent
Number of Cases with
Percent
Missing
Cases Asked Cases with
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Data5
the Question Missing Data5
the Question
Data5
Data5
the Question
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

Instrument Item
Ever used Ketamine (LS01i6)

55,231

0.1

26,617

0.1

3,012

0.1

55,230

106

0.2

26,617

0.1

3,012

0.2

55,230

112

0.1

26,617

0.2

3,012

0.1

How long has it been since you last used
Ketamine? (LS336)

585

0.1a

296

0.3

0.0*

How long has it been since you last used DMT,
AMT, or Foxy? (LS346)

528

0.2

305

0.3

1.1*

How long has it been since you last used Salvia
divinorum? (LS356)

2,000

0.1a

845

0.1

0.0*

Ever used a needle to inject any drug that was not
prescribed for you (SD157)

55,231

0.0

26,617

0.0

3,012

0.0

Are you now married, widowed, divorced or
separated, or have you never married? (QD07)

46,543

0.0a

22,357

0.1a

2,681

0.3

How many times have you been married? (QD08)

18,031

0.1

8,543

0.0

1,352

0.2

How many times in the past 12 months have you
moved? (QD13)

55,229

0.0a

13,958

0.0a

3,012

0.8

In what State did you live in one year ago today?
(QD13a)

16,540

0.0a

7,854

0.0a*

917

Ever used DMT, AMT, or Foxy (LS01j )
6

Ever used Salvia divinorum (LS01k )

B-1

See notes at end of table.

0.6
(continued)

Instrument Item

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
Data5
the Question
Data5
Data5
the Question
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)
55,229

0.0*

26,614

0.0*

3,012

Have you lived in the United States for at least
one year? (QD16a)

3,289

0.0*

1,580

0.0*

How many years have you lived in the United
States? (QD16b)

3,111

0.1

1,502

0.0

0.0a*

Were you born in the United States? (QD14)

B-2

How many months have you lived in the United
States? (QD16c)

178

0.1

279

0.0*

265

0.4

0.0a*

30.0*

Are you now attending or are you currently
enrolled in school? (QD17)

55,229

0.0*

26,614

0.0*

3,012

0.1

What grade or year of school are you now
attending? (QD18)

27,235

0.1

12,977

0.1

1,080

0.4

Are you a full-time student or a part-time
student? (QD19)

27,235

0.0a

12,977

0.1a

1,080

0.8

During the past 30 days how many whole days of
school did you miss because you were sick or
injured? (QD20)

24,829

0.2a

11,853

0.3

924

0.7

During the past 30 days how many whole days of
school did you miss because you skipped or
"cut" or just didn't want to be there? (QD21)

21,257

0.1a

8,462

0.2

808

0.6

Did you work at a job or business at any time last
week? (QD26)

46,539

0.0a

22,354

0.0a

2,681

0.2

See notes at end of table.

(continued)

Instrument Item

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

B-3

Even though you did not work at any time last
week, did you have a job or business? (QD27)

21,008

0.0

9,838

0.1

1,166

0.2

How many hours did you work last week at all
jobs or businesses? (QD28)

25,525

0.1

12,514

0.1

1,504

0.3

Do you usually work 35 hours or more per week
at all jobs or businesses? (QD29)

28,302

0.1

13,853

0.0

1,666

0.1

Which one of these reasons best describes why
you did not work last week? (QD30)

2,777

0.2

1,339

0.3

162

0.1

Which one of these reasons best describes why
you did not have a job or business last week?
(QD31)

18,231

0.0a

8,499

0.0a

0.5

During the past 30 days, did you make specific
efforts to find work? (QD32)

4,493

0.0*

2,029

0.0

Did you work at a job or business at any time
during the past 12 months? (QD33)

18,235

0.0a

8,501

0.1a

1,015

0.8

How many different employers have you had in
the past 12 months? (QD36)

28,329

0.0a

13,804

0.0a

1,582

1.0

During the past 12 months, was there ever a time
when you did not have at least one job or
business? (QD37)

28,302

0.0

13,853

0.0

1,666

0.2

In how many weeks during the past 12 months
did you not have at least one job or business?
(QD38)

6,134

0.9a

3,101

0.6a

341

3.8

See notes at end of table.

1,004
231

(continued)

Instrument Item
In what year did you last work at a job or
business? (QD39a)

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

0.8a

8,499

1.1a

5,911

0.7

2,797

0.3

During the past 30 days, how many whole days of
work did you miss because you were sick or
injured? (QD40)

28,302

0.0a

13,853

During the past 30 days, how many whole days of
work did you miss because you just didn't want
to be there? (QD41)

28,302

0.0a

13,853

At your workplace, is there a written policy about
employee use of alcohol or drugs? (QD43)

28,302

1,532

4.4a

Does this policy cover only alcohol, only drugs,
or both alcohol and drugs? (QD44)

20,439

358

At your workplace, have you ever been given any
educational information regarding the use of
alcohol or drugs? (QD45)

28,302

176

Through your workplace, is there access to any
type of employee assistance program or other
type of counseling program for employees who
have alcohol or drug-related problems? (QD46)

28,302

Does your workplace ever test its employees for
alcohol use? (QD47)

28,302

In what month in did you last work at a job or
business? (QD39b)

B-4

See notes at end of table.

1,004

4.8

271

0.6

0.1a

1,666

0.6

0.1a

1,666

0.4

13,853

809

5.6a

1,666

2.9

1.8a

9,968

172

2.2a

1,227

0.6

13,853

109

0.8a

1,666

0.3

4,048

11.9a

13,853

2,055

12.2a

1,666

117

7.2

1,721

5.7a

13,853

843

6.5a

1,666

3.2
(continued)

Instrument Item

Combined 2012 QFT
2013 Comparison Data1,3
and 2013 DR Data1,4
2012 Comparison Data1,2
Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
Data5
the Question
Data5
Data5
the Question
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

B-5

Does your workplace ever test its employees for
drug use? (QD48)

28,302

1,391

4.8a

13,853

671

4.8a

1,666

3.3

Does your workplace test its employees for drug
or alcohol use as part of the hiring process?
(QD49)

12,654

212

2.0a

6,269

1.9a

749

0.7

Does your workplace test its employees for drug
or alcohol use on a random basis? (QD50)

12,654

759

5.5a

6,269

384

6.3a

749

2.6

According to the policy at your workplace, what
happens to an employee the first time he or she
tests positive for illicit drugs? (QD51)

12,654

1,743

14.1

6,269

915

16.8a

749

11.3

Would you be more or less likely to want to work
for an employer that tests its employees for
drug use as part of the hiring process? (QD52)

28,302

0.2

13,853

0.2

1,666

0.3

Would you be more or less likely to want to work
for an employer that tests its employees for
drug or alcohol use on a random basis? (QD53)

28,302

0.2

13,853

0.1

1,666

0.3

[SAMPLE MEMBER A] covered by Medicare?
(QHI01)

55,224

193

0.2

26,614

104

0.2

3,011

0.4

1,068

0.1

546

0.1

132

55,224

340

0.4a

26,614

188

0.0*

You have indicated that [SAMPLE MEMBER B]
covered by Medicare. Is this correct? (QHI01v)
[SAMPLE MEMBER A] covered by Medicaid?
(QHI02)
You have indicated that [SAMPLE MEMBER B]
covered by Medicaid. Is this correct? (QHI02v)
See notes at end of table.

175

0.4a
0.2

3,011
19

1*
37
0*

0.6*
0.9
0.0*
(continued)

Instrument Item

B-6

[SAMPLE MEMBER A] currently covered by
[CHIPFILL]? (QHI02A)

22,027

495

2.4

10,692

283

2.5

864

3.0

[SAMPLE MEMBER A] currently covered by
TRICARE, or CHAMPUS, CHAMPVA, the
VA, or military health care? (QHI03)

55,224

191

0.2a

26,614

137

0.3

3,011

0.6

[SAMPLE MEMBER A] currently covered by
private health insurance? (QHI06)

55,224

386

0.4

26,614

212

0.4

3,011

0.7

Was [SAMPLE MEMBER] private health
insurance obtained through work? (QHI07)

35,740

143

0.3a

17,128

0.2a

1,756

0.1

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
alcohol abuse or alcoholism? (QHI08)

35,740

16,350

44.5a

17,128

7,864

43.9a

1,756

486

26.0

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
drug abuse? (QHI09)

35,740

16,257

44.7a

17,128

7,823

44.3a

1,756

500

27.1

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
mental or emotional problems? (QHI10)

35,740

9,681

26.8a

17,128

4,674

27.1a

1,756

311

17.5

[SAMPLE MEMBER A] currently covered by
any kind of health insurance including Indian
Health Insurance? (QHI11)

8,112

0.2

3,792

0.1

565

0.1

During the past 12 months, was there any time
when [SAMPLE MEMBER] did not have any
kind of health insurance or coverage? (QHI13)

47,982

109

0.1

23,191

0.2

2,544

See notes at end of table.

0.2
(continued)

Instrument Item

Number of
Cases Asked
the Question
(unweighted)

Percent
Number of
Missing
Cases with
Data5
Missing Data5
(unweighted) (weighted)

2013 Comparison Data1,3

Combined 2012 QFT
and 2013 DR Data1,4

Percent
Number of Number of
Missing
Cases Asked Cases with
Data5
the Question Missing Data5
(unweighted) (unweighted) (weighted)

B-7

During the past 12 months, about how many
months without any kind of health insurance
or coverage? (QHI14)

3,540

0.4

1,594

0.1

211

0.9

About how long has it been since [SAMPLE
MEMBER] last had any kind of health care
coverage? (QHI15)

6,781

0.4a

3,182

0.8a

429

0.1

Which of these reasons is the main reason why
[SAMPLE MEMBER] stopped being covered
by health insurance? (QHI17)

6,259

0.4

2,927

0.2

352

0.6

Which of these reasons describe why [SAMPLE
MEMBER] never had health insurance
coverage? (QHI188)

55,232

0.0

255

0.0*

0.4*

In [YEAR], did you receive Social Security or
Railroad Retirement payments? (QI01N)

55,224

535

0.6

26,614

289

0.7

3,011

1.0

In [YEAR], did you receive Supplemental
Security Income or SSI? (QI03N)

55,224

703

0.8a

26,614

397

1.0

3,011

1.4

In [YEAR], did you receive food stamps?
(QI07N)

55,224

240

0.3

26,614

113

0.3

3,011

0.6

At any time during [YEAR], even for one
month, did you receive any cash assistance
from a state or county welfare program such
as [TANFFILL]? (QI08N)

55,224

354

0.4a

26,614

189

0.4a

3,011

0.7

See notes at end of table.

(continued)

Instrument Item
In [YEAR ], because of low income, did you
receive any other kind of non-monetary
welfare or public assistance? (QI10N)

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

Percent
Number of Number of
Missing
Cases Asked Cases with
Data5
the Question Missing Data5
(unweighted) (unweighted) (weighted)

B-8

55,224

293

0.3a

For how many months in [YEAR]did you or
your [RELATIONSHIP] receive any type of
welfare or public assistance? (QI12AN)

857

For how many months in [YEAR]did you or
your [RELATIONSHIP] receive any type of
welfare or public assistance, not including
food stamps? (QI12BN)

2,883

Before taxes and other deductions, was your
total personal income from all sources during
[YEAR] more or less than 20,000 dollars?
(QI20N)

Number of
Cases Asked
the Question
(unweighted)

Percent
Number of
Missing
Cases with
Data5
Missing Data5
(unweighted) (weighted)

Percent
Number of Number of
Missing
Cases Asked Cases with
Data5
the Question Missing Data5
(unweighted) (unweighted) (weighted)

26,614

139

0.3a

2.3

374

2.5

109

3.8

1,310

2.9

158

7.2

55,223

681

1.9a

26,614

339

2.1a

3,011

116

3.4

Of these income groups, which category best
represents [SAMPLE MEMBER] total
personal income during [YEAR]? (QI21A)

38,838

467

2.3a

18,604

255

3.2

1,727

4.5

Of these income groups, which category best
represents [SAMPLE MEMBER] total
personal income during [YEAR]? (QI21B)

15,704

361

3.5

7,671

164

2.9

1,181

3.7

Before taxes and other deductions, was the total
combined family income during [YEAR]
more or less than 20,000 dollars? (QI22)

34,968

2,044

7.9

16,930

1,032

8.2

1,576

127

8.9

See notes at end of table.

3,011

0.6

1.0*

(continued)

Instrument Item

Combined 2012 QFT
and 2013 DR Data1,4

2013 Comparison Data1,3

Percent
Number of Number of
Missing
Cases Asked Cases with
Data5
the Question Missing Data5
(unweighted) (unweighted) (weighted)

Number of
Cases Asked
the Question
(unweighted)

Percent
Number of
Missing
Cases with
Data5
Missing Data5
(unweighted) (weighted)

Percent
Number of Number of
Missing
Cases Asked Cases with
Data5
the Question Missing Data5
(unweighted) (unweighted) (weighted)

Of these income groups, which category best
represents your total combined family income
during [YEAR]? (QI23A)

7,387

455

5.9

3,530

216

6.2

521

8.8

Of these income groups, which category best
represents your total combined family income
during [YEAR] (QI23B)

37,224

2,358

6.8

18,028

1,139

6.8

1,894

124

6.1

B-9

*Low precision; estimate would be suppressed under NSDUH suppression rules.
AMT = alpha-methyltryptamine; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Veterans
Administration; CHIP = Children's Health Insurance Program; DMT = dimethyltryptamine; DR = Dress Rehearsal; QFT = Questionnaire Field Test; TANF = Temporary Assistance for
Needy Families; VA = Department of Veterans Affairs.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to self-administered.
a
Difference between estimate and combined 2012 QFT and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the interview in English also have been
excluded for these comparisons.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
QFT data collected from September 1 through November 3, 2012. DR data collected from September 1 through October 31, 2013.
5
Missing data include selection of responses of either "don't know" or "refused" for the question. "Missing Data (weighted)" denotes the weighted percentage of missing data.
Denominators for these percentages were based on the total number of cases (i.e., respondents) who were asked the question.
6
For 2012 and 2013 comparison data, these items correspond to items in the special drugs module, but were moved to the hallucinogens module in the QFT and DR.
7
For 2012 and 2013 comparison data, this item corresponds to special drug item SD05.
8
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not asked the question if all response options were coded as
"blank" (e.g., 98 for 2-digit variables).
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table B.2 Item Missingness Rates for Moved Items for Spanish-Language Interviews in 2012 Comparison Data, 2013 Comparison Data,
and 2013 Dress Rehearsal Data among All Persons Aged 12 or Older
2012 Comparison Data1,2

Instrument Item

2013 Comparison Data1,3

2013 DR Data1,4

Ever used Ketamine? (LS01i6)

2,059

0.0

998

0.1

185

0.0*

Ever used DMT, AMT, or Foxy? (LS01j6)

2,059

0.0

998

0.2

185

0.0*

Ever used Salvia divinorum? (LS01k6)

2,059

0.0*

998

0.1

185

0.1

B-10

How long has it been since you last used
Ketamine? (LS336)

0.0*

82.6*

How long has it been since you last used DMT,
AMT, or Foxy? (LS346)

0.0*

How long has it been since you last used Salvia
divinorum? (LS356)

0.0*

Ever used a needle to inject any drug that was not
prescribed for you? (SD157)

2,059

0.0*

998

0.0

185

0.0*

Are you now married, widowed, divorced or
separated, or have you never married? (QD07)

1,822

0.0

867

0.0*

169

0.3

0.0

476

How many times have you been married? (QD08)

979

How many times in the past 12 months have you
moved? (QD13)

2,059

0.0a*

586

In what State did you live in one year ago today?
(QD13a)

633

0.0

272

See notes at end of table.

0.0*

0.0

117

0.0a*

185

2.0

0.0*

1.6*
(continued)

Instrument Item

2013 Comparison Data1,3

2013 DR Data1,4

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
the Question
Data5
Data5
the Question
Data5
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

Were you born in the United States? (QD14)

2,059

0.0*

998

0.0*

185

0.7

Have you lived in the United States for at least
one year? (QD16a)

1,713

0.0*

804

0.0*

136

0.0*

How many years have you lived in the United
States? (QD16b)

1,649

0.2

761

0.0*

132

1.3

0.0*

20.5*

2,059

0.0*

998

0.0*

185

0.3

How many months have you lived in the United
States? (QD16c)

B-11

Are you now attending or are you currently
enrolled in school? (QD17)
What grade or year of school are you now
attending? (QD18)

535

1.0*

282

1.4*

12.8*

Are you a full-time student or a part-time
student? (QD19)

535

0.1

282

0.0*

12.4*

During the past 30 days how many whole days of
school did you miss because you were sick or
injured? (QD20)

474

0.1

254

0.2

0.6*

During the past 30 days how many whole days of
school did you miss because you skipped or
"cut" or just didn't want to be there? (QD21)

422

0.0

207

0.3

1.0*

Did you work at a job or business at any time last
week? (QD26)

1,820

0.2

867

0.0*

169

See notes at end of table.

1.3
(continued)

Instrument Item
Even though you did not work at any time last
week, did you have a job or business? (QD27)

2013 Comparison Data1,3

2013 DR Data1,4

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
the Question
Data5
Data5
the Question
Data5
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)
765

0.0*

355

0.0*

0.1*

How many hours did you work last week at all
jobs or businesses? (QD28)

1,054

0.2

512

0.0*

1.2*

Do you usually work 35 hours or more per week
at all jobs or businesses? (QD29)

1,137

0.1

542

0.0*

B-12

Which one of these reasons best describes why
you did not work last week? (QD30)

0.0*

11.3*

Which one of these reasons best describes why
you did not have a job or business last week?
(QD31)

682

0.0*

325

0.0*

1.4*

During the past 30 days, did you make specific
efforts to find work? (QD32)

140

0.0*

Did you work at a job or business at any time
during the past 12 months? (QD33)

683

0.0*

325

0.0*

1.6*

How many different employers have you had in
the past 12 months? (QD36)

1,040

0.4*

489

0.0*

2.2*

During the past 12 months, was there ever a time
when you did not have at least one job or
business? (QD37)

1,137

0.0*

542

0.1

0.0*

232

0.0a*

106

0.0a*

32.1*

In how many weeks during the past 12 months
did you not have at least one job or business?
(QD38)
See notes at end of table.

(continued)

Instrument Item

2013 Comparison Data1,3

2013 DR Data1,4

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
the Question
Data5
Data5
the Question
Data5
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

In what year did you last work at a job or
business? (QD39a)

682

2.2

325

0.7a

10.0*

In what month in did you last work at a job or
business? (QD39b)

137

0.0*

2.8*

0.0*

B-13

During the past 30 days, how many whole days of
work did you miss because you were sick or
injured? (QD40)

1,137

0.3a

542

0.0a*

6.2*

During the past 30 days, how many whole days of
work did you miss because you just didn't want
to be there? (QD41)

1,137

0.4a

542

0.0a*

5.1*

At your workplace, is there a written policy about
employee use of alcohol or drugs? (QD43)

1,137

1.9

542

2.7

4.4*

665

1.2

325

0.1

0.0*

At your workplace, have you ever been given any
educational information regarding the use of
alcohol or drugs? (QD45)

1,137

0.6a

542

0.2

0.0*

Through your workplace, is there access to any
type of employee assistance program or other
type of counseling program for employees who
have alcohol or drug-related problems? (QD46)

1,137

7.7

542

9.1

4.8*

Does your workplace ever test its employees for
alcohol use? (QD47)

1,137

1.9

542

1.8

2.3*

Does this policy cover only alcohol, only drugs,
or both alcohol and drugs? (QD44)

See notes at end of table.

(continued)

Instrument Item

2013 Comparison Data1,3

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
the Question
Data5
Data5
the Question
Data5
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

B-14

Does your workplace ever test its employees for
drug use? (QD48)

1,137

1.4

542

Does your workplace test its employees for drug
or alcohol use as part of the hiring process?
(QD49)

422

0.0

202

Does your workplace test its employees for drug
or alcohol use on a random basis? (QD50)

422

2.8

202

According to the policy at your workplace, what
happens to an employee the first time he or she
tests positive for illicit drugs? (QD51)

422

4.8

Would you be more or less likely to want to work
for an employer that tests its employees for
drug use as part of the hiring process? (QD52)

1,137

Would you be more or less likely to want to work
for an employer that tests its employees for
drug or alcohol use on a random basis? (QD53)

1,137

[SAMPLE MEMBER A] covered by Medicare?
(QHI01)

2,059

2013 DR Data1,4

29
2,059
40

1.8

2.3*

0.8*

3.3*

3.8

2.3*

202

6.4

6.8*

1.0

542

0.8

1.9*

1.0a

542

0.8

0.0*

0.3

997

0.2

185

0.0*

0.6a

997

0.0

185

0.0*

10
0*

(continued)

Instrument Item

2013 Comparison Data1,3

2013 DR Data1,4

Number of
Number of
Number of
Number of Cases with
Percent
Number of Cases with
Percent
Number of Cases with
Percent
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
Cases Asked
Missing
Missing
the Question
Data5
the Question
Data5
Data5
the Question
Data5
Data5
Data5
(unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted) (unweighted) (unweighted) (weighted)

B-15

[SAMPLE MEMBER A] currently covered by
[CHIPFILL]? (QHI02A)

520

1.4

278

0.4

0.0*

[SAMPLE MEMBER A] currently covered by
TRICARE, or CHAMPUS, CHAMPVA, the
VA, or military health care? (QHI03)

2,059

0.2

997

0.0*

185

0.6

[SAMPLE MEMBER A] currently covered by
private health insurance? (QHI06)

2,059

0.1

997

0.0*

185

0.0*

Was [SAMPLE MEMBER] private health
insurance obtained through work? (QHI07)

345

0.3

172

0.0*

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
alcohol abuse or alcoholism? (QHI08)

345

128

34.9a

172

41.9a

9.1*

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
drug abuse? (QHI09)

345

134

36.1a*

172

43.8a

9.1*

Does [SAMPLE MEMBER] private health
insurance include coverage for treatment for
mental or emotional problems? (QHI10)

345

20.0

172

23.2

26.6*

[SAMPLE MEMBER A] currently covered by
any kind of health insurance including Indian
Health Insurance? (QHI11)
During the past 12 months, was there any time
when [SAMPLE MEMBER] did not have any
kind of health insurance or coverage? (QHI13)
See notes at end of table.

1,114

996

0.0*

517

0.0*

0.5

515

0.0*

101

0.0*
(continued)

Instrument Item
During the past 12 months, about how many
months without any kind of health insurance
or coverage? (QHI14)

2012 Comparison Data1,2

2013 Comparison Data1,3

2013 DR Data1,4

Number of Number of
Percent
Cases Asked Cases with
Missing
the Question Missing Data5
Data5
(unweighted) (unweighted) (weighted)

0.0*

1,052

0.2

479

0.0*

2.4*

Which of these reasons is the main reason why
[SAMPLE MEMBER] stopped being covered
by health insurance? (QHI17)

438

0.6

207

0.8

7.6*

Which of these reasons describe why [SAMPLE
MEMBER] never had health insurance
coverage? (QHI188)

2,061

0.1

272

0.6

8.1*

In [YEAR], did you receive Social Security or
Railroad Retirement payments? (QI01N)

2,059

0.2

997

0.1

185

0.6

In [YEAR], did you receive Supplemental
Security Income or SSI? (QI03N)

2,059

0.8

997

0.0

185

0.6

In [YEAR], did you receive food stamps?
(QI07N)

2,059

0.6

997

0.4

185

1.5*

At any time during [YEAR], even for one
month, did you receive any cash assistance
from a state or county welfare program such
as [TANFFILL]? (QI08N)

2,059

0.4

997

0.0

185

About how long has it been since [SAMPLE
MEMBER] last had any kind of health care
coverage? (QHI15)

B-16

See notes at end of table.

123

0.3
(continued)

Instrument Item
In [YEAR ], because of low income, did you
receive any other kind of non-monetary
welfare or public assistance? (QI10N)

2013 Comparison Data1,3

Number of Number of
Percent
Cases Asked Cases with
Missing
the Question Missing Data5
Data5
(unweighted) (unweighted) (weighted)

2,059

Number of
Cases Asked
the Question
(unweighted)

2013 DR Data1,4

Number of
Percent
Cases with
Missing
Missing Data5
Data5
(unweighted) (weighted)

Number of Number of
Percent
Cases Asked Cases with
Missing
the Question Missing Data5
Data5
(unweighted) (unweighted) (weighted)

0.1

997

0.0*

185

0.3

For how many months in [YEAR]did you or
your [RELATIONSHIP] receive any type of
welfare or public assistance? (QI12AN)

0.4*

0.0*

47.0*

For how many months in [YEAR]did you or
your [RELATIONSHIP] receive any type of
welfare or public assistance, not including
food stamps? (QI12BN)

123

2.3a*

3.4a*

27.7*

B-17

Before taxes and other deductions, was your
total personal income from all sources during
[YEAR] more or less than 20,000 dollars?
(QI20N)

2,059

1.8

997

1.9

185

6.0

Of these income groups, which category best
represents [SAMPLE MEMBER] total
personal income during [YEAR]? (QI21A)

1,644

2.3

815

2.2

132

3.7

Of these income groups, which category best
represents [SAMPLE MEMBER] total
personal income during [YEAR]? (QI21B)

383

0.7

168

0.1

4.0*

1,563

172

12.1

774

9.9

141

Before taxes and other deductions, was the total
combined family income during [YEAR]
more or less than 20,000 dollars? (QI22)
See notes at end of table.

14.9*
(continued)

Instrument Item

2013 Comparison Data1,3

Number of Number of
Percent
Cases Asked Cases with
Missing
the Question Missing Data5
Data5
(unweighted) (unweighted) (weighted)

Number of
Cases Asked
the Question
(unweighted)

2013 DR Data1,4

Number of
Percent
Cases with
Missing
Missing Data5
Data5
(unweighted) (weighted)

Number of Number of
Percent
Cases Asked Cases with
Missing
the Question Missing Data5
Data5
(unweighted) (unweighted) (weighted)

Of these income groups, which category best
represents your total combined family income
during [YEAR]? (QI23A)

715

6.6

371

6.8

11*

10.7*

Of these income groups, which category best
represents your total combined family income
during [YEAR] (QI23B)

1,006

7.6

487

6.1

5.2*

B-18

*Low precision; estimate would be suppressed under NSDUH suppression rules.
AMT = alpha-methyltryptamine; CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Veterans
Administration; DMT = dimethyltryptamine; DR=Dress Rehearsal; VA = Department of Veterans Affairs.
NOTE: Moved items had no changes but moved to another place in the questionnaire or moved from being interviewer-administered to self-administered.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2
2012 comparison data collected in quarters 1 through 4, 2012.
3
2013 comparison data collected in quarter 3 and quarter 4, 2013, through December 5, 2013.
4
DR data collected from September 1 through October 31, 2013.
5
Missing data include selection of responses of either "don't' know" or "refused" for the question. "Missing Data (weighted)" denotes the weighted percentage of missing data.
Denominators for these percentages were based on the total number of cases (i.e., respondents) who were asked the question.
6
For 2012 and 2013 comparison data, these items correspond to items in the special drugs module, but were moved to the hallucinogens module in the DR.
7
For 2012 and 2013 comparison data, this item corresponds to special drug item SD05.
8
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not asked the question if all response options were coded as
"blank" (e.g., 98 for 2-digit variables).
Source: SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

Table B.3 Item Missingness Rates for New and Revised Items for English-Language Non-Hispanic
Interviews in the 2012 Questionnaire Field Test and 2013 Dress Rehearsal among All
Persons Aged 12 or Older
2012 Questionnaire Field Test1,2

Instrument Item
Race (QD056)
Are you currently serving full-time
in a Reserve component? (V2b)
Have you ever served on active duty
in the United States Armed Forces or
Reserve components? (QD10a)
When did you serve on active duty
in the United States Armed Forces or
Reserve components? (QD10b16)
What is the highest grade or year
of school you have completed?
(QD11)
Use of "smokeless" tobacco such as
snuff, dip, chewing tobacco, or
"snus." (CG25)
How old were you the first time you
used "smokeless" tobacco? (CG26)
Is anyone in your immediate family
currently serving in the United States
military? (QD10d)7
Which member or members of your
immediate family are currently in the
United States military? (QD10e)7,8
People are different in their sexual
attraction to other people. Which
statement best describes your
feelings? (QD62)
Which one of the following do you
consider yourself to be? (QD63)

2013 Dress Rehearsal Data1,3

Number of
Number of
Type of
Cases with
Type of
Cases with
Percent
Percent
Change for Missing Data5 Missing Data5 Change for Missing Data5 Missing Data5
QFT4
(weighted)
DR4
(unweighted)
(unweighted) (weighted)
*
*
R
0
0.0
R
1
0.1
N

0.0*

1.0

0.9

0.3

8.9a

2.7*

N/A

0.2

N/A

0.3

DR = Dress Rehearsal; QFT = Questionnaire Field Test.
*Low precision; estimate would be suppressed under NSDUH suppression rules.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who
completed the interview in English also have been excluded for these comparisons.
2
QFT data collected from September 1 through November 3, 2012.
3
DR data collected from September 1 through October 31, 2013.
4
Changes to questionnaire items fall under two categories: N = new item or R = revised item.
5
Missing data include selection of responses of either "don't know" or "refused" for the question. "Missing Data (weighted)"
denotes the weighted percentage of missing data. Denominators for these percentages were based on the total number of cases
(i.e., respondents) who were asked the question.
6
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not
asked the question if all response options were coded as "blank" (e.g., 98 for 2-digit variables).
7
The definition of "immediate family" was moved from the "Help" screen to the question text, minor wording changes were made
to these questions for clarity, and an "Other, Specify" item was added to this series of questions in the DR questionnaire.
8
Estimates are percentages of all persons aged 12 or older, except where noted.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

B-19

Table B.4 Item Missingness Rates for New and Revised Items for Spanish-Language Interviews in
the 2013 Dress Rehearsal among All Persons Aged 12 or Older
2013 Dress Rehearsal Data1,2
Instrument Item
Race (QD055)
Are you currently serving full-time
in a Reserve component? (V2b)
Have you ever served on active duty
in the United States Armed Forces
or Reserve components? (QD10a)
When did you serve on active duty
in the United States Armed Forces
or Reserve components? (QD10b15)
What is the highest grade or year
of school you have completed?
(QD11)
Use of "smokeless" tobacco such as
snuff, dip, chewing tobacco, or
"snus." (CG25)
How old were you the first time you
used "smokeless" tobacco? (CG26)
Is anyone in your immediate family
currently serving in the
United States military? (QD10d)6
Which member or members of your
immediate family are currently
in the United States military?
(QD10e)6,7
People are different in their sexual
attraction to other people. Which
statement best describes your
feelings? (QD62)
Which one of the following do you
consider yourself to be? (QD63)

Type of Change for DR3
R

Number of Cases with
Missing Data4 (unweighted)
1*

Percent Missing Data4
(weighted)
0.3

0.2

0.0*

2.1*

0.0*

1.3

7.8

*Low precision; estimate would be suppressed under NSDUH suppression rules.
1
Sample does not include Alaska or Hawaii. Sample includes Spanish-language interviews only.
2
DR data collected from September 1 through October 31, 2013.
3
Changes to questionnaire items fall under two categories: N = new item or R = revised item.
4
Missing data include selection of responses of either "don't know" or "refused" for the question. "Missing Data (weighted)"
denotes the weighted percentage of missing data. Denominators for these percentages were based on the total number of cases
(i.e., respondents) who were asked the question.
5
"Enter all that apply" question in which available response options were captured as separate variables. Respondents were not
asked the question if all response options were coded as "blank" (e.g., 98 for 2-digit variables).
6
The definition of "immediate family" was moved from the "Help" screen to the question text, minor wording changes were made
to these questions for clarity, and an "Other, Specify" item was added to this series of questions in the DR questionnaire.
7
Estimates are percentages of all persons aged 12 or older, except where noted.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health, 2013.

B-20

Appendix C: Dress Rehearsal
Field Interviewer Training Survey Results

This appendix provides a summary of field interviewer (FI) responses by question for the
2013 Dress Rehearsal (DR) FI training survey completed at the conclusion of the DR FI training
sessions (August 25 and 27, 2013). Of the 135 DR FIs who successfully completed training,
133 FIs completed and transmitted the DR FI training survey.

DR FI Training Survey
Question
1. How long have you worked on
NSDUH?

Less than
1 Year

1 to 2
Years

More
than 2
Years but
Less than
5 Years

5% (7)

17% (23)

26% (34)

DR FI Training Survey Question
2. Were you trained as an FI for the 2012
Questionnaire Field Test (QFT)?

DR FI Training Survey Question
3a. Reading the DR FI Handbook
helped prepare me for training.
3b. Completing the DR iLearning
course helped prepare me for training.
3c. The overall pace of the DR FI
Training Session was just right for me.
3d. The paired screening and interview
exercises completed during training
were helpful.
3e. I feel ready to properly conduct
DR screenings using the tablet.
3f. I feel ready to properly conduct DR
interviews using the DR laptop.
3g. I feel ready to use the email
program on the tablet.
3h. I am comfortable with the process
to transmit wirelessly with the tablet
(independent of the laptop).
3i. I am comfortable with the process
to transmit wirelessly with the DR
laptop.
3j. Overall, I am satisfied with the
training provided on the DR laptop.
3k. Overall, the training program has
prepared me to properly complete all
DR tasks.
3l. I enjoyed attending the DR FI
Training Session.

More
than 5
Years but
Less than
10 Years

More
than 10
Years

Total n

29% (39)

23% (30)

133

Yes

Total n

47% (62)

53% (71)

133

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total
n

56% (74)

37% (49)

5% (6)

1% (1)

2% (3)

133

58% (77)

33% (44)

7% (9)

0% (0)

2% (3)

133

35% (47)

44% (59)

11% (15)

6% (8)

3% (4)

133

52% (69)

40% (53)

4% (5)

2% (2)

3% (4)

133

71% (94)

27% (36)

0% (0)

2% (3)

133

74% (99)

23% (31)

0% (0)

2% (3)

133

59% (79)

32% (42)

7% (9)

0% (0)

2% (3)

133

65% (86)

29% (39)

4% (5)

0% (0)

2% (3)

133

66% (88)

28% (37)

4% (5)

0% (0)

2% (3)

133

66% (88)

30% (40)

2% (2)

0% (0)

2% (3)

133

66% (88)

29% (39)

2% (3)

0% (0)

2% (3)

133

64% (85)

30% (40)

3% (4)

1% (1)

2% (3)

133

C-1

DR FI Training Survey Question
4. During the next month as you
complete your DR work, how often
do you think you will reference the
DR FI Handbook?
DR FI Training Survey Question
5. Please provide any other comments
you have about the DR FI Training
program.
Response
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

Never
1% (1)

Rarely, When
Unusual
Situations Arise
33% (44)

2 to 3
Times a
Week
47% (62)

Each Day
with QFT
Work
20% (26)

Total n
133

[Verbatim responses from FIs, grouped by content, are provided
below.]

General Comments about DR FI Training Program/Trainers
Went well....great trainers
Had great trainers
Enjoyed the training. Everything was very clear.
It was a very nice experience.
enjoyed the trainers, very professonal and helpful.
staff was very helpful and profesional
Trainers were well prepared and knew exactly where to get answers to questions they did not know
Learned a tremendous amount and am looking forward to using the new equipment.
I liked the training and our 3 trainers who conducted the class. I also came for the 2 hour evening
training and thought it was also helpful.
steady strong pace was useful for me-intense but successful
the trainers where great did a great job.
I really think that the training was excellent and that the Trainers did a wonderful job of preparing us
for the DR field work. I learned a lot and it was quite enjoyable.
Trainers were very patient, and professional. Very helpful when needed.
great training and trainers
tremendously thorough, well prepared, focused and amusing trainers. great mix of FIs. great hotel
accomodations and lunches.
Trainers were helpful and knowledgable
good trainers good materials great food good pace with breaks
our trainers were awesome!
Overall good training experience
The program was thorough and delivered at a pace comfortable for everyone in the class
Our trainers were excellent, very difficult hotel and location for those with special needs, I have two
co-workers whom it was very difficult for. The training stayed on time and was very informative.
Our trainers were knowledgeable, helpful, and upbeat. They did a fine job.
trainers were good about keeping the whole class on task and not letting any one person dominate the
conversation.
The training sessions were great. Very informative in explaining information provided in the manuan
and I-Learning. The trainers were extremely patient with some the FI's inability to keep up.
Training was very helpful, trainers were great.
our training staff was great and well prepared. very knowledgable and patient.
Trainers were well prepared and efficient. Love the ezuipment and hope we get to use it on Main
study sooner rather than later.
Thoriugh training! Trainers are excellent and extremely patient. Thank you!
Great group of trainers with experience and talent. Good job
the trainers were well-prepared and seemed really focused. The training flowed more smoothly. Any
technical issues were addressed immediately and efficiently. Love our trainers :) <3

C-2

Response
No.
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Response
No.
52
53
54
55
56
57
58

General Comments about DR FI Training Program/Trainers
I enjoyed the DR training and appreciate all of the effort that went into preparing us for DR. I am
honored to be able to work on this special project.
excellent job at explaining completely at a good pace
as an experienced FI the training was a good lenghth of time to cover the finer points of th new
equipment and to reiterate the changes in some of the wording we will now use
I truly enjoyed the DR FI training, I feel well equipted to jump right into the DR and be successful.
Thank you for the opportunity.
This training was very well organized
staff very helpful with questions and concerns.
In addition to learning how to use the equipment, it helped me see the wording differences in the
screening and interviews. The trainers were professional and patient and answered all our questions
Just the right amount of information and training, including the paired workshops
the trainers were awesome, very helpful.
Overall training was good, What I disagreed and disliked is the arrival time to my home city, 11pm
arrival is too late and tiring.
By having training for the DR, it allows us to complete how our screening/interviewing techniques are
used and also by following
trainers seemed "detached". Tech support personnel very personable.
the trainers were rude with me i was so sad this is the firs time that this happen to me i had been with
rti for 10 years always try to to the best but the female training was not human
Just one of the trainers attitude was a no very nice and tryining to puting us down when we were trying
to perform the task that we were working on it.
The training program was good. At times is went too fast. I was not happy about having to travel on
the metro from the airport to the hotel. Many fis are elderly and it was not easy.
training could have been longer/attendees were familiar with screening and interviewing but the tablet
was a new experience,the training pace was a little quick
the train was done at a comfortable pace for most of the FI but there really was time set up to review
information right after the training section was gone over yes there was a place to write down qu
when paired with another fi the first day,i was instructed to correct them if they did not read verbatim
during the mock interviews,which i did, which angered the male fi i was working with.
The mock trainings seemed excessive: reading the informed consent mulitple times, going through the
detailed process of signing the incentive receipt form in a particular order, and again for homework
should explain the purpose of the field test at the beginning of training and reiterate it when people
make comments about how they "think" respondents will respond. Spent a lot of time on "I think"
sometimes it felt like we went very slow

Comments about Hotel/Training Logistics
love the hotel and staff
breakfast was not very desirable. using the subway to the hotel was alittle nerve wracking going alone.
One of the trainers was very abrupt and short w/ us when doing training.
Class size was good, but the room could of been larger, very noisy when doing screenings and
interviews..
dress rehearsal classroom size was too small and was a distraction when completing the mock
interviews
Friday night check-in was no) without any near-by attractions or activities for the rest of the evening.
Travel coming to and from the training could have been less tedious such as traveling at 4am and 5am
by Metro to get to a flight leaves less time to sleep and feel refreshed for the training
I am not thrilled about having to rush to the aiport to catch flight when having to take public
transportation on the last day of training.

C-3

Response
No.
59
60
61
62
63
64
65
66
Response
No.
67
68
69

Comments about DR Equipment
I LOVE!!!!! the new lap top. The training was great with helpful trainers to teach us.Very
informative.I feel confident that I am ready to be productive on the DR study.Love everything abt the
new tec
Welcome the new equipment, it is as advance as personal gagets, very confident in using it; will make
working on the field more efficient. Great traininers and a wonderful opportunity to meet FI's.
the lighter weight equipment with newer technology is great!!
I am really excited to see us moving forward with more advanced technology and less waste of paper
and postage. WIN-WIN!!
The laptop functionality bridges the gap between those who have less and more experience with
computer technology; very user friendly.
learning to use wireless @ a location other than home; appreciated trainer staying the course, very
good trainers. Took X to ans. ?'s but did not allow distraction; kept class on track.
When equipment failed to transmit by wirless I began to feel like things would not work
it will be difficult to have to log on to the tablet so many times during the day if you have to put it in
sleep mode like we are suppose to. many will be surprised when they go to the door and are no
Other Comments
Like anything else using the training and materials you need to get out and do it.
training staff threw in several things that vet f.i.s had never heard from all over the country-things that
will not really practical in the real field situations-some trainers not very sensitive to p
I suggest using webinars to save money for training. I do understand not everyone is completely
comfortable with new technology but for people are who in the know webinars would be just as
sufficient

C-4

Appendix D: Field Interviewer Equipment
Survey Questions and Results

Table D.1 FI Satisfaction with Laptop – All Field Interviewers
1. Satisfaction with Samsung
Laptop
a. The laptop was easy to use.
b. I was able to use the laptop
without needing technical
assistance.
c. I learned to use the laptop quickly.
d. I felt confident using the laptop.
e. The display size of the laptop
screen was large enough for
presenting the NSDUH interview.
f. The laptop screen was clear and
bright enough for displaying the
NSDUH interview.
g. I was satisfied with the weight of
the laptop.
h. I found the layout of the laptop
keyboard easy to use.
i. The laptop’s touchpad was easy to
use.
j. I was satisfied with the carrying
case provided for the laptop.
k. I was satisfied with the training
provided on the laptop.
l. I would prefer to use this laptop
for my field work.

Strongly
Agree
79% (99)
73% (91)

Agree
16% (20)
18% (22)

Neutral
2% (2)
3% (4)

Disagree
1% (1)
4% (5)

Strongly
Disagree
2% (3)
2% (3)

Total
125
125

81% (101)
83% (104)
78% (98)

15% (19)
13% (16)
17% (21)

1% (1)
1% (1)
2% (2)

1% (1)
1% (1)
1% (1)

2% (3)
2% (3)
2% (3)

125
125
125

82% (102)

14% (18)

1% (1)

2% (3)

125

90% (113)

6% (8)

0% (0)

3% (4)

125

74% (92)

18% (23)

2% (2)

4% (5)

2% (3)

125

78% (97)

15% (19)

2% (3)

125

55% (69)

18% (22)

9% (11)

13% (16)

6% (7)

125

75% (94)

20% (25)

1% (1)

2% (2)

2% (3)

125

90% (113)

6% (7)

1% (1)

0% (0)

3% (4)

125

D-1

Table D.2 FI Satisfaction with Laptop – Bilingual Field Interviewers Only
1. Satisfaction with Samsung
Laptop
a. The laptop was easy to use.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

87% (34)

3% (1)

5% (2)

3% (1)

b. I was able to use the laptop
without needing technical
assistance.
c. I learned to use the laptop quickly.

85% (33)

3% (1)

3% (4)

5% (2)

3% (1)

87% (34)

8% (3)

0% (0)

1% (1)

3% (1)

d. I felt confident using the laptop.

90% (35)

5% (2)

0% (0)

1% (1)

3% (1)

87% (34)

8% (3)

0% (0)

1% (1)

3% (1)

92% (36)

3% (1)

0% (0)

1% (1)

3% (1)

92% (36)

3% (1)

0% (0)

5% (2)

77% (30)

10% (4)

3% (1)

8% (3)

3% (1)

87% (34)

5% (2)

3% (1)

59% (23)

8% (3)

18% (7)

10% (4)

5% (2)

79% (31)

15% (6)

0% (0)

3% (1)

95% (37)

0% (0)

5% (2)

e. The display size of the laptop
screen was large enough for
presenting the NSDUH interview.
f. The laptop screen was clear and
bright enough for displaying the
NSDUH interview.
g. I was satisfied with the weight of
the laptop.
h. I found the layout of the laptop
keyboard easy to use.
i. The laptop’s touchpad was easy to
use.
j. I was satisfied with the carrying
case provided for the laptop.
k. I was satisfied with the training
provided on the laptop.
l. I would prefer to use this laptop
for my field work.

D-2

Table D.3 All FI Comments about Laptop, Training, and Carrying Case
2. Please provide any other comments you have about the Samsung laptop, carrying case or training on the
laptop.
No.
Comments
Comments about the laptop
1
I love it ,my shoulder loves it,my back loves it,please keep it
respondents liked using new laptop , many positive comments on size and weight
using this new laptop was very practical hope we will be able to use in the near future.in fact i worked my
main study this past weekend and it was not has confortable
the laptop is very convenient and well suited for fi needs. The only difficulty I had was with the touchpad.
Very sensitve and hard to move cursor unless you use 1 finger. Tech support helped me.
I Love it,it is so light weight and can't wait to use it.I am just woundering if the case need more coushning
in the Laptop area
I am very excited about this new equipment easier to handle very light weight
Enjoyed the equipment and made admin of interview much smoother
this lightweight laptop long overdue. fabulous for NYC interviews! let's convert to this laptop ASAP!
I like to be able to transmitting either the laptop or tablet only.
I enjoyed doing interviews on this laptop and not to mention how light wait it is. I loved this laptop, can't
wait to use it on the main project.
hard to see f keys in dim/dark room with bright screen, wish keys were backlit, hard to see which f keys
are due small f markings and symbols on keys
Wonderfully light weight! Interview Respondents enjoyed using it as much as I did.
THE SCREEN IS BLINDING TO THE POINT THAT IT OBSCURES THE FUNCTION KEYS FOR
NOT ONLY THE VISUALLY IMPAIRED..... BUT THE FI!
the screen is too bright
dont know and refuse buttons are much too small most people had a hard time seeing them size of the
font needs to be larger for most people since there is plenty of room for larger print on the displ
older respdnts complained that they couldnt see/read the function keys-too small.
Neither R or I can read the function keys easily and white label tape should have been placed directly
above not on the screen. May need to have option to increase size of font for older people to see
The F keys are not as easily identfiable, the F1, F2, ect are very small.
The F3, F4, and F9 keys are very small.
training was enough /night class should be mandortory/ function keys paper/strip needs lower closer to
keyboard hard to see so close to screen /these numbers are small/laptop easy to carry in cont acx
I do not like the flow to start , pressing the tab is ok but don't like then pressing 0k. The the curser does not
attomatically place on the password code line you have to move the curser to it.
Less AC cord would be much better, like a hide away box retractor or something like that.
This laptop is wonderful, light, perfect size and easy to navigate. I also love the case - perfect pair.
Loved working with this laptop. my only negative thing to point out is that I had to made sure I kept the
cursor away from the bottom of the screen or a window bar would showed up.
The laptop is a joy to use. It's wafer thin! The sooner we start using it in our everyday work, the happier
I will be.
it just made it so much easier to carry everything.
the laptop was very easy to use, specially in situations where there was no support (table, chair, etc)
Comments aboutp
Weight os the old laptop and case is a problem for me. This is soooo light compared to the other and easy
to use.
(continued)

D-3

Table D.3 All FI Comments about Laptop, Training, and Carrying Case (continued)
2. Please provide any other comments you have about the Samsung laptop, carrying case or training on the
laptop.
No.
Comments
The touchpad was highly sensitive and easy for respondents to move mistakenly. Older Rs especially
seemed to touch the pad and needed assistance to get back to the correct screen.
the numbers on the function keys are small and at times are difficult to see, especially in low light areas
that we occasionaly encounter in the field.
i had begun to feel strain on my right thumb due to carrying the main study laptop, since working on dr
and using the samsung my thumb has recovered and i no longer feel discomfort
I really like the laptop, but when I transmit to RTI I would get a pop up about not connecting to a router (I
would x it out and transmit again and it was fine. 200 characters is not enough to explain
I love the light weight laptop, So much easier to carry around in the field.
The F keys are hard to read, especially for older people and those not familiar with computer keyboards.
function keys are too small, configuration of the touchpad kind of confusing/limiting,
The size and light weight features makes this laptop more user friendly, especially when dealing with tight
quarters in DU's, counter space, intv outdoors, etc. less intimidating to non computer resp
There were transmission problems with the laptop and my router at home that Tech was unable to resolve.
Consequently I am transmitting from Starbucks.I would prefer that these issues are resolved.
I prefer this new laptop over the main study laptop. It is lightweight, easier to set-up, and a lot faster.
How will ilearning courses be completed, given that there is no CD drive?
The Function Keys were difficult to use for some people because they could not see F3, F4 since the print
was small. The lables did not seem to be effective.
F-1-F-9hard to see in low light conditions, fnt too small
I can not believe the difference carring a lighter laptop made on my workday. My shoulder did not ache at
the end of a day, it was less bulk and much quicker to set up for an interview!
light weight makes work in field much more comfortable
The only negative thing about the laptop is that the writing on the function keys is too small and not
bright white, like the letter and number keys..
The smaller size and lighter weight make transporting and setting up much less cumbersome than current
gateway models
The laptop was very conveniant when having to walk long distancse to DUs and for going up stairs. I
think people enjoy using a nice, modern, and light laptop.
Most interviews are done at a dining room, a longer cord would be helpful. I did not care for the laptop
bag's straps going across the zipper, other than these 2 minor issues I LOVE this equip!
The Samsung laptop gives the study a more professional image, not to mention that is much easier to
handle. The carrying case was practicle and easy to carry. The training was adequate and clear.
The weight of the laptop and smaller carrying case was excellent. It was especially appreciated in multi
story complexes were it is necessary to carry as opposed to roll on wheels,
The laptop's finish can get easily scratched. I suggest getting some sort of protective film for it.
I am very happy with the Samsung laptop and training. I am satisfied with the carrying case but I could
do without a couple of the zippered sections. Overall, I am very satisfied.
Comments about the Carrying Case
Loved the size and pockets of carrying case but the strap placement was awkard and made using zippers
more difficult. It would be better is zipper went straight across the top from the middle sides.
The hand handles in the center of the laptop carrying case get caught as you are trying to zip up the case
occasionally.
I find the laptop case doesn' t have adequate space for supplies,the shoulder strap is in a awkward place
The shoulder strap slips down the shoulder,actually the side with the nylon material
(continued)

D-4

Table D.3 All FI Comments about Laptop, Training, and Carrying Case (continued)
2. Please provide any other comments you have about the Samsung laptop, carrying case or training on the
laptop.
No.
Comments
Not lg. enough for paper supplies, pens/pencils, extension cord; too small.
the carrying case is not practical the inside pockets are too narrow for puting the charger,the
headphones,what about a comp backpack? ;-)
Laptop bag=TOO many comprtmnts so dif find thgs.
Laptop bag...too many compartments. Because of size and weight a backback laptop bag would be
PERFECT.
A rolling bag is much better
The lap=top carrying case because of the big strap crossing over from 1 side to the other made it very
dificult to get things out of it and to put things back, it would get tangled up all the time.
laptop bag is lghtght but cross over strapping is cumbersome to use.also miss the side pocket of the older
version that could easliy slip items into now have to always unzip.need shldr strp padding
carrying case: the shoulder strip crosses the middle compartment, so it is always in the way.
training was conducted in a rapid manor,were shown how to follow instructions without enough time
being provided to become familiar with equipment
case is not balanced and tips over is seat of car.
the strap criscrossed over the zippered area vs on the sides and it got in the way.I had to take it off and just
use the handle which worked fine. I loooooove thethe weight of the new equipment
the long strap to case is at a diaginal not condusive to getting in out of case however it makes it easier to
carry by the strap
The case could be slimed down to take full advantage of the lightweigt laptop, perhaps something similar
to the faux-leather folder that was also supplied, with a shoulder strap.
The laptop case has the handle - I would prefer to have the handle so that you can still close the zipper.
the carrying case is not easy to open/closed; zippers get caught with shoulder strap, need to zip upwards,
stop and move the shoulder strap away from the zipper, then follow through
i found the notebook lacked room for materials,the functions keys should be lit up so that they can be seen
hard to correlate the written tabs down to the exact function key. excellent overall
Most of the problems I had were with identification of the F keys being way too small, even I have
problems seeing them. It would help if the labels were directly above th keys.
The strap connecting to alternate sides of the case made it difficult to get in and out easily. The laptop
itself is great - so light weight and easy to use.
the carry case handles were in the way when trying to get to each section. needs to be designed alittle
different
laptop bag not easy to unzip
case has too many pockets
I found the laptop was able to fit into a sleeker easier to carry case. The one provided seemed too bulky
and didn't provide for a pocket from which to easily pull handouts without unzipping the case.
It's so much faste and easier to user. We are completing interviews, in most cases, in 45 minutes.
The carrying case can be too cumbersome to put away equipment because of the many pockets and what
each pocket can hold
The carrying case is a nightmare, the person who selected this case decided to leave common sense at the
door, , the handleds are in the middle as well as the conections for the shoulder straps, i
I think the laptop back has zippers that get in the way of proper quick use.
The carring case is frustrating to oen and close due to handle placement.

D-5

Table D.4 Frequency of FI Email Use – All Field Interviewers
Frequency of FI Email Use
3. During the dress rehearsal, how
often did you use the email
program on the tablet to
communicate with the FS or other
NSDUH staff members?

Never

A Few Times
a Month

2-3 Times a
Week

Every Day

Total

21% (26)

43% (54)

31% (39)

5% (6)

125

Table D.5 Frequency of FI Email Use – Bilingual Field Interviewers
Frequency of FI Email Use
3. During the dress rehearsal, how
often did you use the email
program on the tablet to
communicate with the FS or other
NSDUH staff members?

Never

A Few Times
a Month

2-3 Times a
Week

Every Day

Total

15% (6)

51% (20)

23% (9)

10% (4)

Table D.6 FI Satisfaction with Tablet Email – All Field Interviewers
4. FI Satisfaction with Tablet Email
Program
a. I found the email program on the
tablet to be simple and
straightforward.
b. The email program was easy to
use.
c. I was able to use the email
program without needing technical
assistance.
d. I learned to use the email program
quickly.
e. I felt confident using the email
program on the tablet.
f. I was satisfied with the training
provided on the email program.
g. I would like to use the email
program on a regular basis.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

57% (56)

34% (34)

6% (6)

3% (3)

0% (0)

56% (55)

37% (37)

5% (5)

2% (2)

0% (0)

60% (59)

32% (32)

2% (2)

6% (6)

0% (0)

62% (61)

29% (29)

6% (6)

3% (3)

0% (0)

65% (64)

27% (27)

6% (6)

2% (2)

0% (0)

59% (58)

30% (30)

7% (7)

4% (4)

0% (0)

64% (63)

17% (17)

2% (2)

0% (0)

D-6

Table D.7 FI Satisfaction with Tablet Email – Bilingual Interviewers
4. FI Satisfaction with Tablet Email
Program
a. I found the email program on the
tablet to be simple and
straightforward.
b. The email program was easy to
use.
c. I was able to use the email
program without needing technical
assistance.
d. I learned to use the email program
quickly.
e. I felt confident using the email
program on the tablet.
f. I was satisfied with the training
provided on the email program.
g. I would like to use the email
program on a regular basis.

Strongly
Agree

Agree

Neutral

Disagree

Strongly
Disagree

Total

67% (22)

33% (11)

0% (0)

67% (22)

33% (11)

0% (0)

73% (24)

21% (7)

0% (0)

6% (2)

0% (0)

76% (25)

24% (8)

0% (0)

76% (25)

21% (7)

3% (1)

0% (0)

70% (23)

21% (7)

6% (2)

3% (1)

0% (0)

73% (24)

15% (5)

12% (4)

0% (0)

D-7

Figure D.1 FI Tablet Keypad Preference – All Field Interviewers

Figure D.2 FI Tablet Keypad Preference – Bilingual Field Interviewers

D-8

Figure D.3 Tablet Wireless Transmission Satisfaction – All Field Interviewers

Figure D.4 Tablet Wireless Transmission Satisfaction – Bilingual Field Interviewers

D-9

Table D.8 FI Comments about Email, Training, or Transmitting on the Tablet
7. Please provide any other comments you have about the email, training or transmitting on the tablet.
No.
Comments
Comments about Email
1
email was easy to use- all work related email should be on a rti instrument
I loved the fact that you get the e-mail on it and thet you can transmit the tablet alone if you did not get any
2
interviews that day.
3
Didn't use email in the field because I only had a few cases. Need more practice
because I became ill with the flu after the DR training,then having,then having my sister pass away. I could
4
use a refresher's course using the email
5
Seems to have a delay in the receiving emails.
6
The draft function and saving drafts are confusing.
I thought the email function was very easy to use. It can be used to increase communication because I
7
could respond immediately after receiving an email message.
8
My FS has only sent me 2 emails
was not used to emailing , again training covered basics of instruction but no time element for
9
familiarization for a level of comfort prior ,emailers would have had no qualms using tablet email
10
the email program is a great help to send info to FS while in the field....cell phone dead or inaccessible.
I really like having 2-way e-mail to communicate w/FS. I don't have access to wifi @ home, so it's a little
11
inconvenient for me & probably for other FIs who live in rural areas to transmit
using the email function on the tablet would be goodl on receiving project communications, however, not
12
as practica lfor immediate or time sensative or longer communications,I prefer personal email
The email program seems to be more of a hinderance than a convience, I would have the mail sent here to
13
my main account anyways, instead of having to transmit on days I didn't work.
The e-mail was easy to use, however I found myself using the emial on my phone more often. I was
14
extremely satisfied with transmitting without connecting to the laptop, more convinient.
15
got no email activity during dr to date, fs filled me in on due dates.
I thought that the email should update automatially when connected to wi-fi, but I always had to refresh it
16
to see if there was any messages or not. I personally thought it should refresh automatic
I did not get into the E-mail program but use an e-mail program on personal tablet. I would love it in
17
routine work
18
it made communicating very easy and quick
I prefer my personal laptop for communicating with my FS because I use it more often and see her emails
19
in a far more timely way.
Comments about the Tablet or Transmitting
20
I love it does not take much time at all to transmit
the tablet is visible to respondents /it has every thing at 1 touch.well lite /i love it /fast /easy to
21
connect/wireless is better faster/wish we didnt have to wait????????
22
My attempts (4) were unsuccessful
had technicle issues linking tablet to laptop but was able to continue working in field due to being able to
23
transmit both independently . Did have to have replacement equipment to resolve issue
24
having the tablet transmit by itself it's very nice and very practical!
25
its great
26
did not like to always start from beginning on each case w tablet
27
Tablet keyboard has backspace key and "enter" key so close it was easy to hit backspace key.
28
have tried to transmitt from the tablet alone...was not successful
29
I like not having to plug into laptop to transmit
30
Only positive comments - a pleasure to use.
31
cant see tabl screen in the sunlight. attachmnt mssg cnfusing-there was no attachmnt in ema
being able to transmit while on the field was a great tool. I had to used it to pick up new cases. its a great
32
time saver, as I didnt have to wait till i get home to transmit
not being able to use quotations, apostrophes very annoying. reponse time to touch/"pen" very slow and
33
unreliable on tablet. wireless not avail when most needed it. problems more due to tablet than em
(continued)

D-10

Table D.8 FI Comments about Email, Training, or Transmitting on the Tablet (continued)
7. Please provide any other comments you have about the email, training or transmitting on the tablet.
No.
Comments
Only transmitted tablet thru the screener function. Would be more proficient if I had opportunity to use
34
more often.
think we should have a myfi or wireless card. i had difficulttly transmitting and one case sat for several
35
weeks
36
the charge cord on the charger is way to short and the cord on the headphones is way to long
37
Being able to transmit with the tablet was so wonderful!
it was very nice to have this feature on the table and that we didn't have to transmit the laptop if we didn't
38
have an interview, Great idea!!!!!
1st this survey: Easy expt this ? as had trble finding these fields Email prog great.Tablet easy but way cases
39
setup dif to view/count case statuses...
40
tablet size is advantageous for SR to read along with FI while doing screener
41
I think it is great the way it is
42
I was very impressed with this equipment and look forward to seeing it implemented in the main study.
transmitting went fast and easy when i just had to transmit thru the tablet. my reg study transmissions are
43
frustrating, take forever and often have to be done over and over
44
I liked the independence of the tablet.
45
I actually transmitted the way I always did except when I checked email.
I had some problems in the field especially when I was at a hotel. Having to log on to their webpage and
46
then to go back to the email was sometimes not making the connection for wi-fi.
47
the equipment is easy to use, easy to carry with you and functions quite well.
48
I enjoyed the ability to email my FS back and not just receive emails
It was great to be able to transmit immediately without having to hook up the computer. Also, I had to
49
transmit in the field once to pick up cases and it was very easy and convenient.
There was a few times were I was unable to transmit unless I was hooked up to laptop but the issue appears
50
to have been resolved
second password time-consuming ...should be unecessary and operate more like ipaq which allows shut off
51
from hse to hse w/o re-enter password
52
Being able to transmit on the tablet without having to transmit on the laptop saved a lot of time.
Very positive experience over all. Looking forward to using this equipment. Training was very though. No
53
questions were unanswered. Transmitting was fast with no difficulties.
The tablet battery was weak. It's useful time was signifisantly less than the iPAQ.Useful time was 3-5
54
hours. It would be e big improvement to have a battery that would last 8 hours.
55
Love how much quicker and easier transmiingt was. I was able to transmit while traveling very easily.
would like for the cursor to automatically go to the password box rather than having to manuever up to it. F
56
keys are a little difficult to see. can print on font keys br brighter?
57
transmitting from tablet is much mor convenient and again, less cumbersome
58
very conveniant
I enjoyed using the tablet very much, it was easy to navigate and I liked seeing all ROCs and comments in
59
one view. Transmitting from the tablet was AMAZING!
I wasn't totally happy with the different keyboards. Each one had different advantages. I had trouble
60
finding hyphens, apostrophes and quotation marks.
Had some difficulty transmitting an email. what about asking about the tablet programming? I miss the
61
function to pull up sorted cases! I do NOT like not being able to rest the device /out a PW entry
62
Transmitting indepedently of the laptop was phenominal and time saving.
very time consuming to log in sometimes 10 times in a day, also keyboard is extremely sensitive, made
63
errors often, which is also frustrating. easy to be on the wrong line.
The scree of the tablet reflects to much light when doing screening an the sun hits on it. Transmitting
64
without connecting the laptob was one of the greatest things about the program and tablet use.
Pretty satisfied with the email program, training and transmitting on the tablet. The tablet and laptop are
65
more efficient on the DR and much faster.

D-11

D-12

Appendix E: Dress Rehearsal
Field Interviewer Debriefing Items

Document Format:
•

Screen names bolded

•

Screen/question/instructional text designated by black and red text and non-italicized
text in parenthesis (Upper-lower black text to be read, red text is instructions to FI)

•

Fills designated by parentheses and italics

•

Logic designated by brackets

•

Text of instructional message boxes provided in bracketed logic

•

Response categories underlined

•

Variable names from QFT Debriefing Items in parentheses following the DR variable
name

INTERVIEW DEBRIEFING QUESTIONS:
THESE QUESTIONS ARE FOR YOU TO ANSWER. DO NOT READ TO THE R.
DRDBF1 (QFTDBF3) [IF INTERVIEW A CALL RECORD OR INTERVIEW B CALL
RECORD = RESULT CODE 70]
When did you give the respondent (or parent/guardian of youth respondent) the Q&A Brochure?
1. BEFORE THE INTERVIEW
2. DURING THE INTERVIEW
3. AT THE END OF THE INTERVIEW
Next [DRDBF2]
DRDBF2 (QFTDBF4)
Did you conduct this interview at the respondent’s home, either inside or outside?
YES
NO
Next [IF DRDBF2=YES, GO TO DRDBF4]
DRDBF3 (QFTDBF5) [IF DRDBF2=NO]
Where did you conduct this interview?
1. AT THE RESPONDENT’S WORKPLACE
2. AT THE HOME OF THE RESPONDENT’S RELATIVE OR FRIEND
3. IN SOME TYPE OF CONFERENCE ROOM IN A RESIDENCE HALL, SCHOOL
OR APARTMENT COMPLEX
4. AT A LIBRARY

E-1

5. IN SOME TYPE OF COMMON AREA, SUCH AS A LOBBY, HALLWAY,
STAIRWELL, OR LAUNDRY ROOM
6. SOME OTHER PLACE
Next [IF DRDBF3=6, GO TO DRDBF3a]
DRDBF3a (QFTDBF5a) [IFDRDBF3=6]
Where did the interview take place?
ALLOW 140 CHARACTERS
Next [DRDBF4]
DRDBF4
Did the respondent make any comments about the laptop? Please include respondent comments
about the physical features of the laptop or about the respondent’s use of the laptop. Check all
that apply.
1. YES, POSITIVE COMMENTS
2. YES, NEGATIVE COMMENTS
3. NO
Next [IF DRDBF4 = YES, NEGATIVE COMMENTS AND DRDBF4 NE YES, POSITIVE
COMMENTS, GO TO DRDBF4b OR IF DRDBF4 = NO, GO TO DRDBF5]
DRDBF4a [IF DRDBF4 = YES, POSITIVE COMMENTS OR YES, POSITIVE COMMENTS
AND YES, NEGATIVE COMMENTS]
Which one or more of the following best describes the positive comments the respondent made
about the laptop. Check all that apply.
1.
2.
3.
4.

THE SCREEN WAS LARGE OR CLEAR OR EASY TO READ
THE LAPTOP WAS LIGHTWEIGHT
THE KEYBOARD WAS EASY TO USE
OTHER

Next [IF DRDBF4a=4, GO TO DRDBF4a1]
DRDBF4a1 [IF DRDBF4a=OTHER]
Please describe the respondent’s other positive comments about the laptop.
ALLOW 140 CHARACTERS
Next [DRDBF5]

E-2

DRDBF4b [IF DRDBF4 = YES, NEGATIVE COMMENTS OR YES, POSITIVE
COMMENTS AND YES, NEGATIVE COMMENTS]
Please describe the negative comments the respondent made about the laptop. Check all that
apply.
1.
2.
3.
4.
5.
6.
7.

THERE WERE PROBLEMS READING THE SCREEN
THE LAPTOP WAS TOO HOT
THE LAPTOP WAS TOO HEAVY
THE LAYOUT OF QUESTIONS WAS PROBLEMATIC
THE KEYBOARD WAS HARD TO USE
THERE WERE PROBLEMS WITH THE VOLUME OR SOUND
OTHER

Next [IF DRDBF4b=7, GO TO DRDBF4b2]
DRDBF4b2 [IF DRDBF4b=OTHER]
Please describe the respondent’s other negative comments about the laptop.
ALLOW 140 CHARACTERS
Next [DRDBF5]
DRDBF5 (QFTDBF6)
Please indicate how private the interview was. Do not count yourself or a project observer as
another person in the room.
1. COMPLETELY PRIVATE – NO ONE WAS IN THE ROOM OR COULD
OVERHEAR ANY PART OF THE INTERVIEW
2. MINOR DISTRACTIONS – PERSON(S) IN THE ROOM OR LISTENING LESS
THAN 1/3 OF THE TIME
3. PERSON(S) IN THE ROOM OR LISTENING ABOUT 1/3 OF THE TIME
4. SERIOUS INTERRUPTIONS OF PRIVACY MORE THAN HALF THE TIME
5. CONSTANT PRESENCE OF OTHER PERSON(S)
Next [IF DRDBF5=1, GO TO DRDBF8; IF DRDBF5 NE1, GO TO DRDBF6]
DRDBF6 (QFTDBF 7) [IF DRDBF6 NE1]
Not including yourself or project observers, who were the other people present or listening to the
interview? Check all that apply
1.
2.
3.
4.

PARENT(S)
SPOUSE
LIVE-IN PARTNER/BOYFRIEND/GIRLFRIEND
OTHER ADULT RELATIVE(S)

E-3

5. OTHER ADULT(S)
6. CHILD(REN) UNDER 15
7. OTHER
Next [IF DRDBF6=1, 2, 3, 4, 5, OR 6, GO TO DRDBF8]
DRDBF7 (DRDBF8) [IF DRDBF6=7]
Please enter a description of the other person(s) present or listening to the interview. This
description may be relationship to the respondent if you have this information, or simply the
gender and estimated age.
ALLOW 140 CHARACTERS
Next [DRDBF8]
DRDBF8 (QFTBDF9) [IF DRDBF5=1; OR IF DRDBF6=1, 2, 3, 4, 5, OR 6; OR IF DRDBF7
NE BLANK]
Did the respondent make any comments about the interview being too long?
YES
NO
Next [DRDBF9]
DRDBF9 (QFTDBF11)
Did the respondent have any questions or comments about the on-screen calendars in the ACASI
section of the questionnaire? If the respondent asked how to access the calendar at any time
during the ACASI portion of the interview, select “YES.”
YES
NO
Next [IF DRDBF9=NO, GO TO DRDBF10]
DRDBF9a (QFTDBF11a) [IF DRDBF9 = YES]
What comments did the respondent make about the on-screen calendars? Check all that apply
1. THE RESPONDENT ASKED HOW TO ACCESS THE CALENDAR.
2. THE RESPONDENT ASKED HOW TO CLOSE THE CALENDAR.
3. THE RESPONDENT DID NOT SEE THE REFERENCE DATES ON THE
CALENDAR.
4. THE CALENDAR HELPED THE RESPONDENT ANSWER THE QUESTION.
5. THE CALENDAR COVERED THE QUESTIONS OR THE IMAGES ON THE
SCREEN.
6. OTHER
E-4

Next [DRDBF10]
DRDBF10 (QFTDBF12)
Did the respondent have trouble understanding any other questions asked during the interview?
YES
NO
Next [IF DRDBF10=NO, GO TO DRDBF11]
DRDBF10a (QFTDBF12a) [IF DRDBF10=YES]
Enter the screen name and a brief description of what the respondent found confusing. If you do
not know the screen name, please provide as much information as possible.
ALLOW 140 CHARACTERS
Next [DRDBF11]
DRDBF11 (QFTDBF13)
Was a proxy used for the income and health insurance questions?
YES
NO
Next [IF DRDBF11=NO, GO TO DRDBF15]
DRDBF12 (QFTDBF17) [IF DRDBF11=1]
Were there any problems with the proxy’s use of ACASI to answer the income and health
insurance questions?
YES
NO
Next [IF DRDBF12= NO, GO TO DRDBF13]
DRDBF12a (QFTDBF17a) [IF DRDBF12=YES]
Which of the following describes the problems with the proxy’s use of ACASI in answering the
income and health insurance questions? Check all that apply.
1. THE PROXY DID NOT KNOW THE ANSWERS TO THE QUESTIONS
2. THE PROXY DID NOT KNOW HOW TO ENTER HIS/HER ANSWERS TO THE
QUESTIONS
3. THE PROXY REFUSED TO ANSWER SOME QUESTIONS

E-5

4. THE PROXY DID NOT KNOW WHY HE/SHE WAS ASKED TO ANSWER
THESE QUESTIONS
5. OTHER
Next [IF DRDBF12a= 1, 2,3, OR 4, GO TO DRDBF13]
DRDBF12b [IF QFTDBF12a=5]
Please describe the proxy’s other problems with ACASI when answering the income and health
insurance questions.
ALLOW 250 CHARACTERS
Next [DRDBF13]
DRDBF13 (QFDBF14) [IF DRDBF11=YES]
Did the respondent have any questions or concerns about his/ her answers being revealed to the
proxy?
YES
NO
Next [DRDBF14]
DRDBF14 (QFTDBF15) [IF DRDBF13 NE BLANK]
Did the respondent have any other questions or comments about the proxy interview?
YES
NO
Next [IF DRDBF14 =NO, GO TO DRDBF15]
DRDBF14a (QFTDBF15a) [IF DRDBF14=YES]
Please describe the other questions or comments the respondent had about the proxy interview.
ALLOW 140 CHARACTERS
Next [DRDBF15]
DRDBF15
Please note anything else you think would be helpful for the interpretation and understanding of
this interview.
ALLOW 250 CHARACTERS
Next [RECORD OF CALLS]

E-6

Appendix F: Moderator's Guide for the Dress Rehearsal
Field Interviewer Debriefing Calls

NSDUH Dress Rehearsal Field Interviewer Debriefing Calls
Moderator’s Guide – UPDATED DRAFT
September 9, 2013
SECTION I: Introduction (5 minutes)
Hello and thank you for attending today’s debriefing call to discuss your experiences during the
2013 Dress Rehearsal. My name is [MODERATOR’S NAME] from RTI. Also on the call today
from RTI are [ASSISTANT MODERATOR’ S NAME] and [NOTE TAKER’S NAME], as well
as [OTHER RTI OBSERVERS]. I will be leading today’s discussion with help from
[ASSISTANT MODERATOR] and [NOTETAKER’S NAME], who will be taking notes.
In addition, on the call with us today from our client, SAMHSA, are [STAFF NAMES].
Before we get started, I want to remind everyone to have your Dress Rehearsal FI Handbook in
front of you as we talk, so you can reference any notes you made in Chapter 6 as you provide
feedback.
This discussion is intended to gather feedback from your experiences collecting data during the
Dress Rehearsal. As you know, changes to the NSDUH questionnaire, procedures, equipment
and materials were tested in the Dress Rehearsal, but we simply cannot gather all of the
information we need just by analyzing survey data.
Therefore, we’re hoping you can share your experiences from the Dress Rehearsal, including
what sorts of feedback you received from respondents, and what types of issues you encountered,
if any.
A summary of the feedback you provide in this discussion will be included in the Dress
Rehearsal report provided to SAMHSA and may help us make changes to the protocol in the
future.
Before we begin, here are a few ground rules for our discussion:
•

We are recording this call and have a note taker so we can capture all of your comments
here today. However, please know your name will never be associated with any
comments you make and will not be included in reports developed to summarize this call.

•

Please be respectful of everyone on this call, so only one person should speak at a time.
Doing so allows the whole group to hear each person and ensures the recording will be
clear.

•

Also, if you have not done so already, move to a location with minimal background
noise. During the call, please stay focused on the conversation as we want to hear from
all of you.

•

If I haven’t heard from you, I may call on you. If I do call on you but you’d rather not
answer a particular question or if you don’t have anything to add, you can just tell me that
you would like to “pass.”

F-1

•

Since we are on the phone, each time you speak, I would like you to begin your
comments by saying your name, such as, “This is Scott and I think…”

•

Please know there is no right or wrong answer to the questions I will be asking.
Everyone’s input is important and helpful.

Now, I’d like you to briefly introduce yourselves to the group. I’ll call on each of you, one at a
time. When I do, please say your name, the state where you live and how long you’ve worked on
NSDUH.
Any general questions before we get started?
SECTION II: Reactions to the Redesigned Contact Materials (20 minutes)
INTRO: For this first section of this call, we are going to discuss the redesigned contact
materials you provided to respondents during the Dress Rehearsal. I will be asking specifically
about the redesigned Lead Letter and Q&A Brochure as well as the new portfolio and laptop bag
you received for your Dress Rehearsal assignment. NOTE: FOR BILINGUAL-ONLY CALLS,
FIs SHOULD ALSO INCLUDE RESPONSES SPECIFIC TO THE SPANISH MATERIALS.
1. During your Dress Rehearsal screenings, on the Study Introduction screen, you asked
respondents if they remembered receiving the lead letter, reading the sentence “You
should have received a letter explaining the study.” Did you notice any differences
from the main study in the way respondents reacted to that sentence? If so, please
explain.
2. Did you receive any feedback from respondents on the lead letter? [FOR THOSE
THAT INDICATE RECEIVING FEEDBACK ON LETTER: How often did
respondents mention they had seen the lead letter? Also, what sorts of comments, if
any, did they make or questions did they ask about the letter and its content?
FOLLOW UP: And, based on that experience, do you feel this lead letter was more
effective than the current main study version?]
3. Did you receive any feedback from respondents about the question and answer
brochure? [FOR THOSE THAT INDICATE RECEIVING RESPONDENT
FEEDBACK: What type of feedback did you receive? Did respondents focus on the
content of the brochure, the appearance or layout of the brochure, or a mix of both?
PROBE: Please provide examples of any comments on the content or appearance of
the brochure that you can recall.]
4. Overall, do you think respondents’ reactions to the lead letter and question and
answer brochure were similar to the reactions you receive to the current main study
contact materials, or were they different somehow? [FOR ANY WHO INDICATE
REACTIONS THEY RECEIVED WERE DIFFERENT FOR DR HOUSEHOLDS,
ASK: What were the main ways that respondents’ reactions to the contact materials
were different than the reactions you receive to the main study letter and brochure?]

F-2

5. Please describe your experience using the new black leather portfolio you received
for the Dress Rehearsal. Please consider your use of the external pocket, pockets on
the inside front cover and the pad of paper as part of your field work. [FOR ANY
WHO INDICATE THE PORTFOLIO DID NOT MEET THEIR NEEDS, ASK: What
features would be more useful to you for organizing your field materials?]
6. Overall, how would you describe your experience using the new laptop bag? Did
having the extra pockets improve your organization of equipment and field materials?
[FOR ANY WHO INDICATE THE BAG DID NOT MEET THEIR NEEDS, ASK:
What do you think would make the laptop bag more useful?]
7. [IF APPLICABLE] Given the additional storage space in the new laptop bag, do you
feel that having a portfolio provided for you is still necessary?
8. How often did you reference your DR FI Handbook while in the field? [FOR THOSE
WHO INDICATE FREQUENT USE, ASK: What issues did you most commonly
need to reference your handbook for? What suggestions, if any, do you have to
improve the handbook to make it more user-friendly for you in the field?
SECTION III: Administering Household Screenings and Using the Tablet (30 minutes)
INTRO: For this next set of questions, I will be asking about your experience using the Samsung
Galaxy Tablet and the NSDUH screening program and the changes specific to the Dress
Rehearsal NOTE: FOR BILINGUAL-ONLY CALLS, FIs SHOULD ALSO INCLUDE
RESPONSES SPECIFIC TO THEIR ADMINISTRATION OF SPANISH SCREENINGS.
1. Do you feel that the Dress Rehearsal training adequately prepared you to use the new
tablet to conduct screenings? [FOR ANY WHO DO NOT THINK THE
INSTRUCTION WAS ADEQUATE: What recommendations do you have for
improving the training on the tablet (for screening)?]
2. How long did it take you to feel fully comfortable using the tablet to conduct
screenings? [FOR THOSE WHO INDICATE NOT QUICKLY FEELING
COMFORTABLE USING THE TABLET, ASK: What were the greatest challenges
you faced in becoming comfortable using the tablet (for screening)?]
3. Did you receive any reaction from respondents on the use of US Department of
Health and Human Services rather than US Public Health Service? [FOR THOSE
THAT RECEIVED RESPONDENT REACTION: Were these reactions positive or
negative? Did this cause any confusion among respondents?]
4. If a video containing a 20-30 second video clip of the annual press conference were
added to the tablet, do you think this would be a useful tool for gaining cooperation
from respondents at the doorstep? Why or why not?
5. Did you experience any difficulties typing in ROC notes or comments using the
keyboard on the tablet? [FOR ANY WHO INDICATE HAVING DIFFICULTY
TYPING ROC NOTES OR COMMENTS, ASK: How often did you encounter
problems typing in ROC notes or comments using the keyboard on the tablet? How
were you able to overcome this challenge?]

F-3

6. Which of the available keyboard choices did you prefer using with the tablet – the
standard Samsung keyboard or the Hacker’s keyboard? What made you prefer the
keyboard you chose?
7. When using the tablet, did you prefer to use your finger or the provided stylus? What
influenced your choice? [FOR THOSE WHO INDICATE USING ONLY THE
STYLUS, ASK: Were you satisfied with the size of the stylus? Would you prefer a
longer version?]
8. Overall, how would you describe your experience using the email program on the
tablet?
9. [IF APPLICABLE] Did you experience any problems using the email program? If so,
what suggestions do you have to resolve those problems?
10. How often did you use the email program to communicate with your FS? [FOLLOW
UP: What types of messages did you send and receive?]
11. Do you feel that having access to email on the tablet benefitted your work? If so,
how?
12. How often did you connect your tablet to a wireless internet connection outside of
your home? If you did this, did you use this most often to send/receive emails, to
transmit screening cases from your tablet, or both? [FOR THOSE THAT
TRANSMITTED TO PICK UP CASES: Do you feel that having this capability to
transmit and pick up screening cases in the field would help you with your main study
assignment?]
13. Were there any features on or capabilities of the tablet you found especially useful or
especially problematic? If so, please describe in detail.
14. Overall, what are your feelings about the canvas case for the tablet? What changes or
improvements would you make to improve its performance in the field?
15. Did you ever call Technical Support for assistance with the tablet at any point during
data collection? [FOR ANY WHO INDICATE CALLING TECHNICAL SUPPORT,
ASK: Can you tell me why you called?]
SECTION IV: Administering the Redesigned Questionnaire and Protocol (25 minutes)
INTRO: Now I am going to ask a series of questions about the experience you and your
respondents had using the Dress Rehearsal questionnaire NOTE: FOR BILINGUAL-ONLY
CALLS, FIs SHOULD INCLUDE RESPONSES SPECIFIC TO THEIR ADMINISTRATION
OF SPANISH INTERVIEWS.
1. Did you receive any feedback from respondents about their use of the computerized
calendar? [FOR THOSE THAT RECEIVED FEEDBACK: What type of feedback or
questions did you receive from respondents about the calendar? Please provide
examples of any comments or questions that you can recall.]

F-4

2. Would you say respondents made comments or asked questions about the calendar as
often as main study respondents using the paper version, less often, or more often?
[PROBES: Tell me more about that. What do others think?]
3. Did any respondents (or proxy respondents) make any comments about not being able
to clearly understand the ACASI voice in the computer? If so, what particular
comments did they make and/or what specific questions/pronunciations did they seem
to not understand?
4. How often did respondents or proxy respondents make comments or ask questions
about specific questions or modules when completing either the items you
administered to them or completing the ACASI portion of the interview protocol
themselves? [PROBES: Tell me more about that. What do others think?]
5. Did respondents make comments or ask questions about the sexual orientation
questions? [IF YES: Tell me more about that. What do others think?]
6. Did respondents make any comments or ask any questions about any other specific
questions or features of the protocol when completing any of the modules (except
for the prescription drug module)? [PROBE: Please provide examples of any
comments or questions on specific questions or features of the protocol that you can
recall.]
7. Did you experience any issues or notice any repeated problems with the use of the
second ACASI section by the proxy respondents? Do you have any suggestions on
how to improve that transition?
FOR BILINGUAL FI DEBRIEFING CALLS ONLY:
8. As you know, the Dress Rehearsal interview contains a lot of new content and
materials in Spanish. The input gathered during this session will help us assess the
performance of the Spanish instrument in the field. Therefore, please think carefully
about your experiences administering the questionnaire and respondents’ reactions to
the interview. We are interested in detecting any issues in the translation of the
Spanish questionnaire. Did respondents indicate they were confused or unsure about
any Spanish text in the questionnaire? [PROBE: Please provide examples of
questions or wording that caused confusion.]
9. Do you personally have any feedback on questions in the Spanish interview where the
Spanish translation may be problematic? [PROBE: Which questions are problematic?
Can you tell me more about that?]
10. Did respondents report any problems with the Spanish questionnaire that were not
related to translation? For instance, were there any problems with screen layout,
entering responses, or hearing the correct audio?
11. Are there any other comments that you would like to make about the Dress Rehearsal
Spanish questionnaire?

F-5

SECTION V: Administering the DR Interview and Using the Laptop (10 minutes)
INTRO: In this next section, we are going to discuss the new Dress Rehearsal laptop, including
how respondents reacted to it.
1. Do you feel the Dress Rehearsal training adequately prepared you to use the new
laptop to conduct interviews? [FOR ANY WHO DO NOT THINK THE
INSTRUCTION WAS ADEQUATE: What recommendations do you have for
improving the training on the laptop?]
2. How long did it take you to feel fully comfortable using the laptop to conduct
interviews? [FOR THOSE WHO INDICATE NOT QUICKLY FEELING
COMFORTABLE USING THE LAPTOP, ASK: What do you think were the greatest
challenges you faced in becoming comfortable using the laptop (to conduct
interviews)?]
3. How did respondents react to the new laptop? Did respondents make any comments
about the performance of the laptop or their ability to read the questions on the
screen? What about the laptop’s size?
4. Did any of your respondents make any comments about the function keys at the top
of the laptop keyboard? [IF YES: What types of comments did they make? Did they
say anything about the size of the buttons?]
5. Overall, how would you assess the performance of the laptop? Did you encounter any
issues using the laptop, the laptop’s touchpad or the keyboard? If so, please describe.
6. Did you ever call Technical Support for assistance with the laptop at any point during
data collection? [FOR ANY WHO INDICATE REQUESTING TECHNICAL
SUPPORT, ASK: Can you tell me why you called?
SECTION VI: Reactions to the Redesigned Prescription Drug Module (10 minutes)
INTRO: Now I am going to ask a few questions about the Prescription Drug Module that was
redesigned for the Dress Rehearsal. This module asked respondents about their use and misuse of
various prescription drugs and was completed as part of the first ACASI section of the interview.
NOTE: FOR BILINGUAL-ONLY CALLS, FIs SHOULD ALSO INCLUDE RESPONSES
SPECIFIC TO THEIR ADMINISTRATION OF SPANISH INTERVIEWS.
1. Did respondents make comments or react specifically to the amount of recall or the
length of time required to answer the questions in the prescription drug module? [IF
YES: Please provide examples of any comments or reactions to the prescription drug
questions you can recall.]
2. Did respondents make comments or react specifically to the length of time required
to complete the prescription drug module? [IF YES: Please provide examples of any
comments or reactions to the length of the prescription drug module you can recall.]

F-6

3. Did respondents make comments or react specifically to the electronic pill images in
the prescription drug module? [IF YES: Please provide examples of any comments or
reactions to the electronic pill images you can recall.]
4. Did respondents make comments or react specifically to the questions designed to
capture misuse of prescription drugs? [IF YES: Please provide examples of any
comments or reactions to the questions on misuse of prescription drugs you can
recall.]
5. Did respondents make any comments or ask any questions about any other aspects of
the prescription drug module? [IF YES: Please provide examples of any comments or
questions about the prescription drug module that you can recall.]
SECTION VII: Overall Reactions to the Redesigned Questionnaire (15 minutes)
INTRO: I am now going to ask a few additional questions to address any issues respondents may
have had with other redesigned portions of the Dress Rehearsal questionnaire. NOTE: FOR
BILINGUAL-ONLY CALLS, FIs SHOULD ALSO INCLUDE RESPONSES SPECIFIC TO
THEIR ADMINISTRATION OF SPANISH INTERVIEWS.
1. Did respondents make any comments or react specifically to the amount of recall or
length of time required to answer any of the other interview questions? Would
you say respondents commented on the time required to answer the interview
questions about as often as main study respondents, less often, or more often?
[PROBES: Tell me more about that. What do others think?]
2. How often did respondents make comments or react specifically to the length of time
required to complete the entire interview? Would you say respondents commented on
the interview length about as often as main study respondents, less often, or more
often? [PROBES: Tell me more about that. What do others think?]
3. [IF APPLICABLE] Did any other members of the household make comments or react
specifically to the length of time to complete the interview? Were the comments or
reactions mostly positive, mostly negative, or a mix of both? [PROBE: Please provide
examples of any comments or reactions that you can recall.]
4. Did respondents raise any other specific concerns when completing the ACASI
portion or the questions you administered? [PROBE: Please provide examples of any
concerns that you can recall.]
SECTION VIII: Conclusion (5 minutes)
Are there any final comments or any questions on any of the topics we discussed, or other Dress
Rehearsal topics?
I want to thank you all again for your participation on this call.
NOTETAKER WILL NOW STOP THE AUDIO RECORDING.

F-7

F-8

Appendix G: Dress Rehearsal Field Observation Materials

Revised 6/17/2013

DR Screening Observation Checklist
Directions: Complete one DR Screening Observation Checklist for each screening you observe that ends in a
code 22, 25, 26, 30, 31, or 32. For each screening procedure and summary item listed below, place a mark in the
Correct, Error, or N/A column. For each Error or N/A response, provide a brief description in the space just
below that item. If you observe an error that does not fit any of the categories below, describe that error in item
21. You should complete this checklist in hard copy using a clipboard or hard binder while at the household
observing a screening. Within 24 hours you should enter this information into the DR Reporting Spreadsheet
and email the spreadsheet to Gretchen McHenry.
Screening Case ID:

Date of Observation:

FI Name: __________________________________________________________

FI ID:

Observer Name: ___________________________________________ Observer Title: _______________

SCREENING PROCEDURES OBSERVED
1.

Displayed ID Badge prominently when knocking on door

On Tablet “Study Introduction” screen when reached door

Included all required information in introduction (Mark each item when
spoken by FI)
FI Name
RTI International
U.S. Department of Health and Human Services
Lead Letter

If R didn’t recall Lead Letter, FI offered one to R (gave DR version of
LL)

G-1

Correct

Error

N/A

SCREENING PROCEDURES OBSERVED (continued)
5.

Confirmed SR was an adult resident of SDU (FI does not need to
confirm age when it is obvious SR is 18 or older)

Verified that he/she was at the correct address

Gave DR Study Description to R [IF NOT, INTERVENE]

Read Tablet “Informed Consent” screen to R [IF NOT, INTERVENE]

Checked for missed DUs by reading the correct Tablet screen verbatim
(This screen should not be read at apartments/condos)

10. Read all roster questions verbatim (Describe each roster question not
read verbatim)

11. Recorded race based on R answer, not FI observation (If the SR refuses
to answer for the householder, the FI can record an answer based on
his/her observation of the race of the SR)

G-2

Correct

Error

N/A

SCREENING PROCEDURES OBSERVED (continued)
12. Obtained all screening information directly from the SR (Not by
observation or a proxy)

13. Confirmed accuracy & completeness of roster data w/ SR

14. For codes 22, 25, 26, or 30, correctly followed verification procedures

15. For code 31 or 32, presented project and interview information
accurately

16. For code 31 or 32, demonstrated flexibility in scheduling interview time

17. For code 31 or 32, left appropriate information about future interview
(If R asks questions or would like more information about the
interview)

18. For code 31 or 32, made attempts to begin interview right away

G-3

Correct

Error

N/A

SCREENING PROCEDURES OBSERVED (continued)

Correct

19. Provided R with the correct DR materials (did not substitute main study
versions)

20. Answered R questions correctly and thoroughly, referencing the
appropriate DR details [e.g. RTI International, DHHS, did not mention
DR or field test, sample size, pay or payment (should use give or
receive), etc.]

21. OTHER PROCEDURAL VIOLATION NOT NOTED ON THIS
CHECKLIST: [IF BIASING, INTERVENE]

SCREENING SUMMARY
22. Did the screening presentation flow well? If NO, describe:

23. Was visibility an issue when using the Tablet? If YES, describe:

24. Were there any issues with the equipment (Tablet, Tablet case)? If YES, describe:

25. Was there any difficulty using the Tablet keyboard? If YES, describe:

G-4

Error

N/A

SCREENING SUMMARY (continued)
26. Was there any respondent confusion due to something the FI said or did? If YES, describe:

27. Was there any respondent confusion due to a procedure OR to the Tablet screening program itself? If
YES, describe:

28. Was there any FI confusion due to the Tablet or screening program itself? If YES, describe:

29. Were there any respondent comments on the contact materials?

30. Did the respondent make any comments about specific screening questions?

ADDITIONAL OBSERVER COMMENTS:

G-5

SEGMENT MAPS AND

Correct

Error

M1. Had segment maps
readily available for
reference while in the
field (Either in the car or
located with screening and
interviewing materials)
NOTE: If you are unsure,
wait until the END of the
observation and then ask
the FI if he/she has the
maps
M2. [IF THIS IS FI’s
FIRST VISIT TO THE
DWELLING UNIT(s)]
Used segment maps to
locate sample dwelling
unit(s)
M3. [IF THIS IS FI’s
FIRST VISIT TO THE
DWELLING UNIT(s)]
Used the segment maps
and either the printed list
of SDUs or the original
list of dwelling units to
check for missed DUs in
the interval between the
SDU and the next listed
dwelling unit
M4. [IF A MISSED DU
IS FOUND] Used
segment map and original
list of dwelling units to
make sure the missed DU
was not already listed

G-6

N/A

Revised 6/17/2013

DR Interviewing Observation Checklist
Directions: Complete one DR Interviewing Observation Checklist for each interview you observe. For each
interview procedure and summary item listed below, place a mark in the Correct, Error, or N/A column. For
each Error or N/A response, provide a brief description in the space below that item. If you observe an error that
does not fit any of the categories below, describe that error in item 14. You should complete this checklist in
hard copy using a clipboard or hard binder while at the household observing an interview. Within 24 hours you
should enter this information into the DR Reporting Spreadsheet and email the spreadsheet to Gretchen
McHenry.
Interview Case ID:

A / B (please circle A or B)

Date of Observation:
FI Name: _________________________________________________________

FI ID:

Observer Name: __________________________________ Observer Title: _________________________
INTERVIEWING PROCEDURES OBSERVED
1.

If IR was a minor, FI first obtained consent from parent or legal guardian
[IF NOT, INTERVENE]

If IR was not SR, explained purpose of study and visit thoroughly
[IF NOT, INTERVENE]

If IR was not SR, handed DR STUDY DESCRIPTION to the respondent
[IF NOT, INTERVENE]

Read INTRO TO CAI from DR Showcard Booklet verbatim to respondent
[IF NOT, INTERVENE]

G-7

Correct

Error

N/A

INTERVIEWING PROCEDURES OBSERVED (continued)
5.

Chose the most private available location

Set up equipment efficiently

Explained HEADPHONE usage, offered headphones to IR, and plugged in

Kept ACASI portion private (did not read ACASI), but remained attentive

Read all screens verbatim (Record the ID number of all questions not read
verbatim below)

10. Presented DR SHOWCARDS when prompted by the CAI

11. Followed the proper DR Quality Control Form and Incentive procedures

G-8

Correct

Error

N/A

INTERVIEWING PROCEDURES OBSERVED (continued)

Correct

Error

N/A

12. Answered IR questions correctly and thoroughly, referencing the
appropriate DR details [e.g., RTI International, DHHS, did not mention DR
or field test, sample size, pay or payment (should use give or receive), etc.]

13. Provided IR with the correct DR materials (did not substitute main study
versions)

14. OTHER PROCEDURAL VIOLATION NOT NOTED ON THIS
CHECKLIST:
[IF BIASING OR DIVULGING CONFIDENTIAL INFORMATION,
INTERVENE]
INTERVIEWING SUMMARY

15. Did the respondent have trouble understanding any questions asked during the interview? If YES, describe:

16. Were there any issues with transition between the screening and the interview? If YES, describe:

17. Were there any issues with transition between the ACASI and CAPI sections of the interview? If YES,
describe:

G-9

INTERVIEWING SUMMARY (continued)

18. Was there any respondent confusion due to something the FI said or did? If YES, describe:

19. Was there any respondent confusion due to a procedure OR to the CAI instrument itself? If YES, describe:

20. Was there any respondent confusion due to the Samsung laptop?

21. Was there any FI confusion due to the CAI instrument or Samsung laptop? If YES, describe:

22. If a proxy was used, was there any confusion regarding their role, the equipment, adjusting the volume,
etc.? If YES, describe:

23. If a proxy was used, was there any difficulty understanding the ACASI tutorial? If YES, describe:

G-10

INTERVIEWING SUMMARY (continued)

24. Were there any problems with the Samsung laptop that disrupted the flow or completion of the interview?

25. Was there any confusion when the FI was completing the debriefing questions on the Tablet?

26. Did the respondent or proxy make any comments about specific interview questions?

27. Did the respondent or FI make any comments about the length of the interview?

ADDITIONAL OBSERVER COMMENTS

G-11

Revised 6/17/2013

NSDUH DR Field Observations: Field Observer Reference Sheet
DR Field Observer Task List (Task number 0212800.001.102.003.006)
Please follow these steps while planning and conducting field observation trips. It is not necessary to actually complete or submit
this form; it is designed as a helpful tool so you do not skip any protocol steps.
Enter a check mark in the space provided as you complete each item.

TRAVEL PREPARATION
____

1. Receive DR Field Observation Assignment.

____

2. Contact the DR FI’s Field Supervisor. Send the FS an email to obtain the FI’s contact
information and other information that will be pertinent to planning your trip. In the email,
request the following information:
_____
_____
_____

a) FI contact information (FI phone numbers can also be found in the FI Lookup form the
General Information link on the CMS)
b) Location of DR segment and distance between FI segments
c) Any other information the FS feels is significant

You should also request the FS send a copy of the DR FI Field Observations Instructions to the
FI and notify him/her you will soon be in contact.
____

3. Contact the DR FI. Call each FI and make plans for the observation. You will need to discuss
the following:
_____
_____
_____
_____

a)
b)
c)
d)

Date most convenient for observation (Must be completed before September 15th)
DR Workload – How long will the FI have a DR assignment?
Segment information – Location of DR segment, type of attire needed
Other information – Suggested hotels, coordinating transportation to segment

You should also confirm the FS has sent a copy of the DR FI instructions and tell the FI you will
be spending the whole workday in the field with him/her. Let him/her know it is necessary to
observe an interview and encourage him/her to set up an appointment in advance of your
arrival.
____

4. Once the date of observation has been determined, email your observation plans to Gretchen
McHenry, copying the managing FS, RS, and your supervisor. In the email, include the dates
you will observe each FI and any trip details associated with the observation (dates you will fly,
drive, return, etc).

____

5. Are flight or hotel arrangements necessary?
 YES (flights)  continue with 6.
 YES (hotels)  continue with 8.


NO  Skip to Field Preparation.

____

6. Make flight and rental car arrangements with Carlson Wagonlit Travel (online or by phone) at
least 14 days prior to scheduled trip. You will need your Bank of America corporate credit card
number and task number (0212800.001.102.003.006) ready when calling. Before booking your
flight, review flight options on Expedia and select the best and most reasonable flight in terms
of costs and time.

____

7. Immediately after booking your flight, send the completed General Travel Information Form to
the NSDUH Secretaries and Gretchen McHenry, copying your supervisor. A copy of the
General Travel Information Form can be found on the Downloadable Project Forms and
Report Shells on the CMS.

G-12

____

8. Determine the government per diem and lodging rates for the area by clicking the ‘US Gov’t
Per Diems’ link on the General Information page of the CMS. Please keep cost in mind when
identifying a hotel and expensing meals.

____

9. Make hotel reservations at or under the given per diem. When looking for a place to stay,
search online for hotels in the area and/or gather FS and FI suggestions. You cannot pay more
than the official government rate. It is imperative that you verify the government rate on the
‘US Gov’t Per Diems’ link after the hotel tells you what their government rate is. You should
also try to find a hotel that includes free parking and internet access. Call the hotel to confirm
these details before booking.

____

10. Update the CMS travel Calendar (with dates of travel, hotel, and contact information), SRD
travel calendar, and your Outlook Calendar.

FIELD PREPARATION
____

1. Print the DR forms from the email sent by the FO Manager:
____

____
____

____

____
C.

DR Field Observation FI Instructions Form: You should hand a copy of this form to the FI
when you meet him/her in the field. It contains the script the FI is to read to the respondent
when introducing you and your role as the observer.
b. DR Field Observer Reference Sheet: This form outlines your role and responsibilities as the
observer.
c. NSDUH DR Screening Scripts: Print and read through this file before going to the field. Use
the script while observing an FI conducting a screening so you can check whether he/she
reads the tablet screens verbatim. Note that there is an HU script and a GQU script within
this file. If you are a bilingual interviewer, please have both the English and Spanish scripts
with you in the field.
d. NSDUH DR CAI Script: Print and read through this file before going to the field. Use the
script to while observing an FI conducting an interview so you can check whether he/she
reads the CAI screens verbatim. If you are a bilingual interviewer, please have both the
English and Spanish scripts with you in the field.
e. DR Screening Observation Checklist: One copy of this form must be completed for each
screening case you observe than ends in a code 22, 25, 26, 30, 31, or 32. You should complete
this checklist in hard copy using a clipboard or hard binder while at the household observing
a screening.
f. DR Interviewing Observation Checklist: One copy of this form must be completed for each
completed interview you observe. You should complete this checklist in hard copy using a
clipboard or hard binder while at the household observing an interview.

2. Make sufficient copies of both the screening and interviewing checklists before going into the
field (we recommend printing 8 screening checklists and 4 interviewing checklists per FI).

AFTER THE OBSERVATION
____

1. Enter data from your checklists into the DR Screening and Interview Report spreadsheets.
Please enter the results of all cases observed for all FIs in one screening and one interview
spreadsheet and e-mail to Gretchen McHenry, within 24 hours of completing all DR FO
assignments.

____

2. Send an e-mail to the FS, copying the RS, RD, and [NSDUH] DR Field Observations ([email protected]), sharing positive feedback about the FI’s performance within 24 hours of
completing your observation.

____

3. As soon as you have completed all of the field observations you will be conducting for the DR,
please ship all completed hardcopy field observation checklists via United States Postal Service
or interoffice mail to Gretchen McHenry at RTI.

G-13

G-14

Appendix H: Estimates and Standard Errors for All New,
Moved, or Revised Items in the 2012 Questionnaire Field
Test and 2013 Dress Rehearsal for English-Language NonHispanic Interviews among Persons Aged 12 or Older

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older

Instrument Item

Race (QD05)
White (QD051)

78.9

(2.15)

1,268

1,692

78.9

(2.72)

972

1,319

Black or African American (QD052)
American Indian or Alaska Native
(American Indian Includes North
American, Central American, and
South American Indians)
(QD053)

14.7

(1.75)

333

1,692

14.6

(2.43)

233

1,319

1.4

(0.42)

1,692

1.3

(0.34)

1,319

Native Hawaiian (QD054)

0.1

(0.07)

1,692

0.0

(0.01)

1,319

Guamanian or Chamorro (QD055)

0.0*

(0.00*)

1,692

0.0

(0.01)

1,319

(0.00 )

1,319

Samoan (QD056)

0.1

(0.10)

1,692

0.0

Other Pacific Islander (QD057)
Asian (Including Asian, Indian,
Chinese, Filipino, Japanese,
Korean, and Vietnamese (QD058)

0.3

(0.12)

1,692

0.3

(0.18)

1,319

6.0

(0.99)

104

1,692

6.2

(1.51)

133

1,319

0.4

(0.21)

1,692

0.6

(0.33)

1,319

0.0*

(0.00*)

1,692

0.0*

(0.00*)

1,320

8.3

(0.98)

1,692

7.1

(1.29)

1,320

10.3*

(2.89*)

15.8*

(5.00*)

17.4*

(4.75*)

11.0*

(4.49*)

(3.39 )

Other (Specify) (QD059)
Are you currently serving full-time in
a Reserve component? (V2b)
Have you ever served on active duty
in the U.S Armed Forces
or Reserve components? (QD10a)
When did you serve on active duty in
the U.S. Armed Forces or
Reserve components? (QD10b1)5
September 2001 or Later
(QD10b11)
August 1990 to August 2001
(Including Persian Gulf War)
(QD10b12)
May 1975 to July 1990 (QD10b13)
Vietnam Era (August 1964 to
April 1975) (QD10b14)
February 1955 to July 1964
(QD10b15)
Korean War (July 1950 to
January 1955) (QD10b16)
January 1947 to June 1950
(QD10b17)
World War II (December 1941 to
December 1946) (QD10b18)
November 1941 or Earlier
(QD10b19)
See notes at end of table.

21.3

(5.32 )

6.8

46.1*

(6.07*)

42.9*

(8.25*)

9.0*

(3.33*)

5.8*

(5.20*)

8.5*

(3.25*)

12.4*

(6.35*)

1.0*

(0.95*)

0.0*

(0.00*)

5.5*

(2.75*)

11.0*

(5.36*)

0.0*

(0.00*)

0.0*

(0.00*)

54
(continued)

H-1

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Did you ever serve on active duty in
the U.S. Armed Forces or Reserve
components in a military combat
zone or an area where you drew
imminent danger pay or hostile fire
pay? (QD10c) 5
What is the highest grade or year of
school you have completed?
(QD11)6
No Schooling
1st Grade
2nd Grade
3rd Grade

35.9*

(6.65*)

40.3*

(10.57*)

0.1

(0.05)

1,692

0.0*

(0.00*)

0.0

1,692

0.0

1,692

0.2

1,692

0.1

(0.00 )
(0.00 )
(0.00 )

1,320

(0.00 )

1,320

(0.15)

1,320

(0.12)

1,320

(0.00 )

1,320

(0.20)

1,320

4th Grade

0.1

(0.09)

1,692

0.0

5th Grade

0.3

(0.13)

1,692

0.4

6th Grade

1.9

(0.31)

1,692

1.5

(0.47)

1,320

7th Grade

2.7a

(0.44)

1,692

1.3

(0.43)

1,320

8th Grade

3.3

(0.47)

1,692

2.9

(0.68)

1,320

9th Grade

2.8

(0.41)

1,692

4.0

(0.66)

1,320

10th Grade

3.0

(0.43)

1,692

3.1

(0.54)

1,320

11th Grade

3.5

(0.44)

100

1,692

4.4

(0.57)

1,320

Regular High School Diploma

20.4

(1.68)

298

1,692

20.6

(1.97)

216

1,320

12th Grade, No Diploma

1.8

(0.45)

1,692

2.8

(0.60)

1,320

GED Certificate

3.8

(0.63)

1,692

5.0

(0.99)

1,320

19.5

(1.27)

325

1,692

20.2

(1.73)

248

1,320

Associate's Degree

9.2

(0.88)

123

1,692

9.9

(1.29)

107

1,320

Bachelor's Degree

17.4

(1.76)

211

1,692

14.1

(1.50)

180

1,320

7.6

(0.96)

1,692

7.7

(1.35)

1,320

1.2

(0.36)

1,692

0.8

(0.47)

1,320

1.5

(0.40)

1,692

1.0

(0.56)

1,320

Yes

8.0

(1.31)

1,087

9.1

(2.15)

638

92.0

(1.32)

1,026

1,087

90.9

(2.15)

594

638

0.0

(0.04)

1,087

0.0

(0.03)

638

0.0*

(0.00*)

0.0*

(0.00*)

2
(continued)

Some College, No Degree

Master's Degree
Doctorate Degree (e.g., PhD)
Professional Degree Beyond
Bachelor's Degree (e.g., MD)
Previously served as a proxy for
another respondent? (PREVCOM)

I am not sure
Previously completed any part of this
interview yourself, including
answering questions on behalf of a
member of your household?
(PREVCOM2)5
See notes at end of table.

H-2

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Use of "smokeless" tobacco such as
snuff, dip, chewing tobacco, or
"snus." (CG25)
How old were you the first time you
used "smokeless" tobacco? (CG26)7
How long has it been since you last
used, have you used "smokeless"
tobacco? (CG27and CG28)

(1.16)

307

1,692

15.5

(1.62)

18.4

(0.71)

N/A

307

18.1

(0.74)

209

1,320

N/A

207

Within the past 30 days
More than 30 days ago but within
the past 12 months

5.4

(0.60)

1,692

3.2

(0.62)

1,318

1.8

(0.36)

1,692

1.2

(0.34)

1,318

More than 12 months ago

1.5

(0.30)

1,692

1.6

(0.46)

1,318

9.9

(0.91)

138

1,692

9.5

(1.28)

112

1,318

23.9

(1.34)

415

1,679

22.3

(1.82)

301

1,309

1.4

(0.33)

1,690

1.3

(0.43)

1,320

0.7

(0.20)

1,689

1.4

(0.39)

1,318

Ever used Salvia divinorum (LS01k)
How long has it been since you last
used Ketamine? (LS33)

2.4

(0.46)

1,689

2.6

(0.65)

1,320

Within the past 30 days
More than 30 days ago but within
the past 12 months

0.1

(0.04)

1,689

0.0*

(0.00*)

1,320

0.3

(0.16)

1,689

0.0*

(0.00*)

1,320

More than 12 months ago
How long has it been since you last
used DMT, AMT, or Foxy? (LS34)

1.1

(0.27)

1,689

1.3

(0.43)

1,320

Within the past 30 days
More than 30 days ago but within
the past 12 months

0.1

(0.05)

1,688

0.0*

(0.00*)

1,318

More than 3 years ago
During the past 30 days, did you have
[Insert #] or more drinks on the
same occasion? (AL08)7,8
Ever used Ketamine (LS01i)
Ever used DMT, AMT, or Foxy
(LS01j)

More than 12 months ago
See notes at end of table.

0.2

(0.12)

1,688

0.1

(0.05)

1,318

0.4a

(0.15)

1,688

1.3

(0.39)

1,318
(continued)

H-3

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How long has it been since you last
used Salvia divinorum? (LS35)
Within the past 30 days
More than 30 days ago but within
the past 12 months
More than 12 months ago
Have you ever, inhaled felt-tip pens,
felt-tip markers, or magic markers for
kicks or to get high? (IN01h1)
Have you ever inhaled computer
keyboard cleaner, also known as air
duster, for kicks or to get high?
(IN01ii)
Have you ever used methamphetamine?
(ME01)
How old were you the first time you
used methamphetamine? (ME02)7
How long has it been since you last used
methamphetamine? (MELAST3)

0.1

(0.09)

1,689

0.0*

(0.00*)

1,320

0.2

(0.10)

1,689

0.1

(0.08)

1,320

2.1

(0.40)

1,689

2.5

(0.61)

1,320

3.1

(0.35)

1,690

2.8

(0.51)

1,320

1.2

(0.27)

1,691

1.3

(0.34)

1,318

6.9

(0.91)

1,691

8.0

(1.15)

1,319

20.7

(0.67)

N/A

21.3

(0.83)

N/A

Within the past 30 days
More than 30 days ago but within
the past 12 months

0.5

(0.18)

1,691

0.3

(0.22)

1,319

0.1

(0.08)

1,691

0.4

(0.16)

1,319

More than 12 months ago
How many days you've used
methamphetamine during the past 12
months? (MEFRAME3, MEYRAVE,
MEMONAVE, MEWKAVE)7
During the past 30 days, on how many
days did you use methamphetamine?
(ME06)7
In the past 12 months, which, if any, of
these pain relievers have you used?
(PR01)

6.2

(0.86)

1,691

7.3

(1.05)

1,319

161.2

(45.87)

N/A

195.6

(39.61)

N/A

17.7*

(4.51*)

N/A

25.9*

(1.34*)

N/A

12.7

(1.23)

201

1,679

10.9

(1.30)

159

1,311

5.8

(0.78)

1,679

4.4

(0.87)

1,311

1.1

(0.28)

1,679

1.7

(0.66)

1,311

14.6a

(1.29)

227

1,679

18.8

(1.45)

181

1,311
(continued)

Vicodin®
®

Lortab

Lorcet

Hydrocodone
See notes at end of table.

H-4

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR02)
OxyContin®
®

2.5

(0.39)

1,675

3.1

(0.67)

1,312

Percocet

6.5

(0.89)

107

1,675

5.4

(1.07)

1,312

Percodan®

0.5

(0.16)

1,675

0.8

(0.36)

1,312

0.3

(0.13)

1,675

0.1

(0.05)

1,312

6.9

(0.94)

112

1,675

7.9

(1.15)

1,312

1.8

(0.46)

1,676

1.6

(0.65)

1,313

0.6

(0.33)

1,676

0.7

(0.29)

1,313

0.3

(0.13)

1,676

0.5

(0.34)

1,313

Ultram®

2.3

(0.61)

1,677

1.9

(0.56)

1,312

0.5

(0.26)

1,677

0.4

(0.27)

1,312

0.4

(0.17)

1,677

0.2

(0.12)

1,312

(0.00 )

1,312

Tylox

Oxycodone
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR03)
Darvocet®
®

Darvon

Propoxyphene
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR04)
Ultram ER
®

Ultracet
®

Ryzolt

0.0

(0.03)

1,677

0.0

Tramadol
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR05)

4.5

(0.57)

1,677

5.7

(1.16)

1,312

Tylenol® with Codeine 3 or 4

11.1

(1.06)

199

1,675

12.7

(1.56)

146

1,309

Codeine Pills
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR06)

1.7

(0.33)

1,675

1.7

(0.50)

1,309

0.1

(0.12)

1,679

0.1

(0.09)

1,312

(0.00 )

1,312

(0.12)

1,312

(0.00 )

1,312

(0.85)

1,312
(continued)

Avinza®
®

Kadian

0.1
®

MS Contin

Oramorph SR
Morphine
See notes at end of table.

0.1
*

(0.06)

1,679

0.0

(0.06)

1,679

0.2

0.0

(0.00 )

1,679

0.0

3.9

(0.61)

1,679

2.8

H-5

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR07)
Actiq®

0.1
®

(0.12)

1,678

0.0*
*

(0.00*)
*

1,311

Duragesic

0.0

(0.05)

1,678

0.0

(0.00 )

1,311

Fentora®

0.0

(0.05)

1,678

0.0*

(0.00*)

1,311

0.7

(0.26)

1,678

0.3

(0.17)

1,311

0.8

(0.26)

1,678

1.3

(0.44)

1,312

0.3

(0.11)

1,678

0.3

(0.16)

1,312

0.0

(0.04)

1,678

0.5

(0.31)

1,312

Demerol®

0.7

(0.16)

1,677

0.4

(0.24)

1,311

Dilaudid®

0.9

(0.25)

1,677

0.6

(0.39)

1,311

Methadone

Fentanyl
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR08)
Suboxone®
®

Subutex

Buprenorphine
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR09)

Opana

0.7

(0.19)

1,677

0.5

(0.22)

1,311

0.2

(0.07)

1,677

0.1

(0.06)

1,311

0.2

(0.08)

1,677

0.2

(0.10)

1,311

0.0*

(0.00*)

1,677

0.0*

(0.00*)

1,312

0.0

(0.03)

1,677

0.0

(0.04)

1,312

(0.00 )

1,312

Opana ER
In the past 12 months, which, if any,
of these pain relievers have you
used? (PR10)
Talacen®
®

Talwin

0.0

(0.04)

1,677

0.0

8.9

(0.87)

150

1,676

9.8

(1.20)

108

1,311

68.5

(1.66)

1,001

1,667

66.2

(1.67)

774

1,312

Xanax®

4.9

(0.75)

1,686

3.9

(0.73)

1,314

Xanax XR

0.4

(0.17)

1,686

0.1

(0.07)

1,314

Alprazolam

1.6

(0.38)

1,686

3.1

(0.69)

1,314

0.4

(0.27)

1,686

0.3

(0.20)

1,314
(continued)

Extended-Release Alprazolam
See notes at end of table.

H-6

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR02)
Ativan®

1.1
®

(0.32)

1,686

0.7

(0.31)

1,314

Klonopin

1.2

(0.29)

1,686

1.8

(0.53)

1,314

Lorazepam

2.0

(0.35)

1,686

2.5

(0.77)

1,314

2.1

(0.45)

1,686

2.5

(0.89)

1,314

Valium®

2.0

(0.46)

1,686

1.4

(0.29)

1,315

Diazepam

1.1

(0.30)

1,686

2.8

(0.92)

1,315

0.1

(0.07)

1,686

0.2

(0.24)

1,315

0.0

(0.04)

1,686

0.1

(0.12)

1,315

0.0

(0.02)

1,686

0.3

(0.34)

1,315

4.5

(0.64)

1,686

3.0

(0.62)

1,314

Soma
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR05)
Buspirone (also known as
BuSpar®)
Hydroxyzine (also known as
Atarax® or Vistaril®)

1.3

(0.36)

1,686

2.1

(0.66)

1,314

0.5

(0.23)

1,686

1.0

(0.37)

1,313

0.6

(0.27)

1,686

0.8

(0.43)

1,313

Meprobamate
In the past 12 months, have you used
any other prescription tranquilizer?
(TR06)
Have you ever, even once, used any
prescription tranquilizer?
(TR07)
See notes at end of table.

0.0

(0.03)

1,686

0.1

(0.04)

1,313

1.8

(0.38)

1,686

3.5

(1.03)

1,316

27.1

(1.72)

369

1,683

30.2

(1.99)

311

1,309
(continued)

Clonazepam
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR03)

Librium

Tranxene
Oxazepam (also known as
Serax®)
In the past 12 months, which, if any,
of these tranquilizers have you
used? (TR04)
Flexeril®`
®`

H-7

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, which, if any,
of these stimulants have you used?
(ST01)

2013 DR
2013 DR
2013 DR
2012 QFT 2012 QFT 2012 QFT
2012 QFT Standard Unweighted Unweighted 2013 DR Standard Unweighted Unweighted
Estimate1,2,3 Error
Total
Sample Size
Total
Sample Size Estimate1,3,4 Error

Adderall®

2.4

Adderall XR
Dexedrine®
Dextroamphetamine
Amphetamine-Dextroamphetamine
Combinations
In the past 12 months, which, if any,
of these stimulants have you used?
(ST02)
Ritalin®
®

Ritalin SR or Ritalin LA
®

Concerta

Daytrana

Methylphenidate
In the past 12 months, which, if any,
of these stimulants have you used?
(ST03)
Metadate® CD
®

(0.41)

1.2

(0.26)

1,687

0.3

(0.13)

1,687

0.2

(0.12)

1,687

0.8

(0.31)

0.5

(0.16)

0.3
0.6

0.0
0.5

0.0

1,687

3.5

(0.68)

1,314

1.3

(0.45)

1,314

0.2

(0.17)

1,314

0.1

(0.07)

1,314

1,687

0.3

(0.13)

1,314

1,687

0.5

(0.16)

1,314

(0.11)

1,687

0.0

(0.02)

1,314

(0.17)

1,687

1.1

(0.59)

1,314

(0.00 )

1,314

(0.02)

1,687

0.0

(0.15)

1,687

0.1

(0.05)

1,314

(0.03)

1,687

0.1

(0.09)

1,313

Metadate ER

0.1

(0.06)

1,687

0.1

(0.05)

1,313

Focalin®

0.3

(0.12)

1,687

0.1

(0.05)

1,313

0.3

(0.14)

1,687

0.1

(0.05)

1,313

0.2

(0.11)

1,687

0.0

(0.01)

1,313

0.0*

(0.00*)

1,687

0.0*

(0.00*)

Focalin XR
Dexmethylphenidate
In the past 12 months, which, if any,
of these stimulants have you used?
(ST04)
Benzphetamine
®

Didrex

0.0

(0.03)

1,687

0.0

Diethylpropion

0.0

(0.03)

1,687

0.0

Phendimetrazine
Phentermine
In the past 12 months, which, if any,
of these stimulants have you used?
(ST05)
Provigil®
®

Tenuate

Vyvanse
In the past 12 months, have you used
any other prescription stimulant?
(ST06)
Have you ever, even once, used any
prescription stimulant? (ST07)
See notes at end of table.

1,313

(0.00 )

1,313

(0.01)

1,313

0.2

(0.17)

1,687

0.1

(0.05)

1,313

0.8

(0.27)

1,687

1.0

(0.40)

1,313

(0.06)

1,687

0.3

(0.32)

1,313

(0.00 )

1,313

0.1
*

0.0

(0.00 )

1,687

0.0

0.8

(0.26)

1,687

0.6

(0.25)

1,313

1.1

(0.28)

1,686

1.1

(0.48)

1,316

12.1

(1.08)

224

1,684

16.1

(1.77)

195

1,312
(continued)

H-8

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, which, if any,
of these sedatives have you used?
(SV01)
Ambien®
®

Ambien CR
Zolpidem
Extended-Release Zolpidem
In the past 12 months, which, if any,
of these sedatives have you used?
(SV02)
Lunesta®
®

Sonata

4.7

(0.69)

1,686

0.7

(0.25)

1,686

0.8

(0.54)

1,314

1.8

(0.51)

1,686

1.4

(0.48)

1,314

0.1

(0.08)

1,686

0.8

(0.56)

1,314

1.0

(0.32)

1,687

0.4

(0.29)

1,314

(0.27)

1,687

0.0

(0.01)

1,314

0.5
*

(0.00 )

0.0*

(0.00*)

Halcion

0.2

Flurazepam

Zaleplon
In the past 12 months, which, if any,
of these sedatives have you used?
(SV03)
Dalmane
®

Triazolam
In the past 12 months, which, if any,
of these sedatives have you used?
(SV04)
Restoril®
Temazepam
In the past 12 months, which, if any,
of these sedatives have you used?
(SV05)
Butisol®
Seconal®
Phenobarbital
In the past 12 months, have you used
any other prescription sedative?
(SV06)
Have you ever used any prescription
sedative? (SV07)
Have you ever, even once, used any
prescription pain reliever in any
way a doctor did not direct you to
use it? (PRL01 and PRL02)
In the past 12 months, did you use
Vicodin in any way a doctor did
not direct you to use it? (PRY01)
How old were you when you first
used Vicodin in a way a doctor did
not direct you to use it? (PRY01a)7
See notes at end of table.

(0.61)

1,314

1,687

0.0

(0.00 )

1,687

0.0*

(0.00*)

1,314

(0.20)

1,687

0.5

(0.27)

1,314

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.2

(0.12)

1,687

0.1

(0.07)

1,314

0.1

(0.07)

1,687

0.0

(0.03)

1,314

0.7

(0.28)

1,687

0.8

(0.39)

1,314

0.0

(0.03)

1,687

0.0*

(0.00*)

1,314

0.0

3.5

0.1

(0.08)

1,687

0.0

(0.00 )

1,314

0.2

(0.17)

1,687

0.1

(0.11)

1,314

1.2

(0.29)

1,687

2.7

(0.74)

1,313

17.3

(1.45)

213

1,683

16.6

(1.71)

178

1,309

12.2

(1.02)

222

1,663

11.6

(1.15)

168

1,312

2.4a

(0.46)

1,683

1.3

(0.31)

1,315

(2.32)

N/A

27.6

(2.94)

N/A

26
(continued)

24.3

H-9

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
Lortab in a way a doctor did not
direct you to use it? (PRY02)
How old were you when you first
used Lortab in a way a doctor did
not direct you to use it?
(PRY02a)7
In the past 12 months, did you use
Lorcet in any way a doctor did not
direct you to use it? (PRY03)
How old were you when you first
used Lorcet in a way a doctor did
not direct you to use it?
(PRY03a)7
In the past 12 months, did you use
hydrocodone in any way a doctor
did not direct you to use it?
(PRY04)
How old were you when you first
used hydrocodone in a way a
doctor did not direct you to use it?
(PRY04a)7
In the past 12 months, did you use
OxyContin in any way a doctor
did not direct you to use it?
(PRY05)
How old were you when you first
used OxyContin in a way a doctor
did not direct you to use it?
(PRY05a)7
In the past 12 months, did you use
Percocet in any way a doctor did
not direct you to use it? (PRY06)
How old were you when you first
used Percocet in a way a doctor
did not direct you to use
it? (PRY06a)7
In the past 12 months, did you use
Percodan in any way a doctor did
not direct you to use it? (PRY07)
How old were you when you first
used Percodan in a way a doctor
did not direct you to use it?
(PRY07a)7
In the past 12 months, did you use
Tylox in any way a doctor did not
direct you to use it? (PRY08)
How old were you when you first
used Tylox in a way a doctor did
not direct you to use it?
(PRY08a)7
See notes at end of table.

(0.29)

1,682

0.4

(0.18)

1,315

23.7

(2.69)

N/A

27.3*

(4.13*)

N/A

0.2

(0.12)

1,683

0.4

(0.17)

1,315

18.1a*

(0.74*)

N/A

34.8*

(4.60*)

N/A

1.9

(0.37)

1,682

1.8

(0.47)

1,315

25.8a

(2.76)

N/A

38.2

(4.41)

N/A

0.9

(0.23)

1,682

0.7

(0.28)

1,316

20.8a

(1.98)

N/A

35.2*

(4.15*)

N/A

1.0

(0.24)

1,681

0.7

(0.25)

1,316

23.0

(2.54)

N/A

27.1

(3.31)

N/A

0.2

(0.09)

1,682

0.1

(0.12)

1,316

19.6a*

(2.46*)

N/A

30.0*

(0.00*)

N/A

0.0

(0.03)

1,682

0.0*

(0.00*)

1,316

15.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0
(continued)

H-10

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
oxycodone in any way a doctor
did not direct you to use it?
(PRY09)
How old were you when you first
used oxycodone in a way a doctor
did not direct you to use it?
(PRY09a)7
In the past 12 months, did you use
Darvocet in a way a doctor did not
direct you to use it? (PRY10)6
How old were you when you first
used Darvocet in a way a doctor
did not direct you to use it?
(PRY10a)7
In the past 12 months, did you use
Darvon in any way a doctor did
not direct you to use it? (PRY11)
How old were you when you first
used Darvon in a way a doctor
did not direct you to use it?
(PRY11a)7
In the past 12 months, did you use
propoxyphene in any way a doctor
did not direct you to use it?
(PRY12)
How old were you when you first
used propoxyphene in a way a doctor
did not direct you to use it?
(PRY12a)7
In the past 12 months, did you use
Ultram in any way a doctor did
not direct you to use it? (PRY13)
How old were you when you first
used Ultram in a way a doctor did
not direct you to use it?
(PRY13a)7
In the past 12 months, did you use
Ultram ER in any way a doctor
did not direct you to use it?
(PRY14)
How old were you when you first
used Ultram ER in a way a doctor did
not direct you to use it?
(PRY14a)7
In the past 12 months, did you use
Ultracet in any way a doctor did
not direct you to use it? (PRY15)
How old were you when you first
used Ultracet in a way a doctor did
not direct you to use it?
(PRY15a)7
In the past 12 months, did you use
Ryzolt in any way a doctor did not
direct you to use it? (PRY16)
How old were you when you first
used Ryzolt in a way a doctor did
not direct you to use it?
(PRY16a)7
See notes at end of table.

1.3

(0.31)

1,681

0.7

(0.28)

1,316

23.5

(1.77)

N/A

23.8

(3.52)

N/A

0.1

(0.08)

1,683

0.3

(0.17)

1,316

16.2a*

(0.67*)

N/A

26.0*

(2.66*)

N/A

0.0*

(0.00*)

1,683

0.1

(0.12)

1,316

0.0*

(0.00*)

N/A

40.0*

(0.00*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.5

(0.20)

1,682

0.4

(0.26)

1,316

33.3*

(5.80*)

N/A

42.5*

(9.97*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.11)

1,683

0.0*

(0.00*)

1,316

33.6*

(11.61*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0
(continued)

H-11

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
tramadol in any way a doctor did
not direct you to use it? (PRY17)
How old were you when you first
used tramadol in a way a doctor
did not direct you to use it?
(PRY17a)7
In the past 12 months, did you use
Tylenol with codeine 3 or 4 in any
way a doctor did not direct you to
use it? (PRY18)
How old were you when you first
used Tylenol with codeine 3 or 4
in a way a doctor did not direct
you to use it? (PRY18a)7
In the past 12 months, did you use
codeine pills in any way a doctor
did not direct you to use them?
(PRY19)
How old were you when you first
used codeine pills in a way a
doctor did not direct you to use
them? (PRY19a)7
In the past 12 months, did you use
Avinza in any way a doctor did
not direct you to use it? (PRY20)
How old were you when you first
used Avinza in a way a
doctor did not direct you to use
them? (PRY20a)7
In the past 12 months, did you use
Kadian in any way a doctor did
not direct you to use it? (PRY21)
How old were you when you first
used Kadian in a way a doctor did
not direct you to use it?
(PRY21a)7
In the past 12 months, did you use
MS Contin in any way a doctor
did not direct you to use it?
(PRY22)
How old were you when you first
used MS Contin in a way a doctor did
not direct you to use it?
(PRY22a)7
In the past 12 months, did you use
morphine in any way a doctor did
not direct you to use it? (PRY24)
How old were you when you first
used morphine in a way a doctor
did not direct you to use it?
(PRY24a)7
In the past 12 months, did you use
Actiq in any way a doctor did not
direct you to use it? (PRY25)
How old were you when you first
used Actiq in a way a doctor
did not direct you to use it?
(PRY25a)7
See notes at end of table.

(0.18)

1,683

0.5

(0.16)

1,315

26.5

(3.32)

N/A

25.5*

(3.21*)

N/A

1.4

(0.29)

1,677

1.1

(0.28)

1,316

27.5

(5.24)

N/A

25.3

(2.38)

N/A

0.3

(0.13)

1,680

0.3

(0.17)

1,316

17.2*

(0.77*)

N/A

18.7*

(0.65*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.04)

1,683

0.0*

(0.00*)

1,316

17.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,683

0.1

(0.11)

1,316

0.0*

(0.00*)

N/A

22.0*

(0.00*)

N/A

0.4

(0.16)

1,683

0.1

(0.12)

1,316

17.5a*

(1.55*)

N/A

47.8*

(1.58*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0
(continued)

H-12

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
Duragesic in any way a doctor did
not direct you to use it? (PRY26)
How old were you when you first
used Duragesic in a way a doctor
did not direct you to use it?
(PRY26a)7
In the past 12 months, did you use
Fentora in any way a doctor did
not direct you to use it? (PRY27)
How old were you when you first
used Fentora in a way a doctor
did not direct you to use it?
(PRY27a)7
In the past 12 months, did you use
fentanyl in any way a doctor did
not direct you to use it? (PRY28)
How old were you when you first
used fentanyl in a way a doctor
did not direct you to use it?
(PRY28a)7
In the past 12 months, did you use
Suboxone in any way a doctor did
not direct you to use it? (PRY29)
How old were you when you first
used Suboxone in a way a doctor
did not direct you to use it?
(PRY29a)7
In the past 12 months, did you use
Subutex in any way a doctor did
not direct you to use it? (PRY30)
How old were you when you first
used Subutex in a way a doctor
did not direct you to use it?
(PRY30a)7
In the past 12 months, did you use
buprenorphine in any way a doctor
did not direct you to use it?
(PRY31)
How old were you when you first
used buprenorphine in a way a
doctor did not direct you to use it?
(PRY31a)7
In the past 12 months, did you use
Demerol in any way a doctor did
not direct you to use it? (PRY32)
How old were you when you first
used Demerol in a way a doctor
did not direct you to use it?
(PRY32a)7
In the past 12 months, did you use
Dilaudid in any way a doctor did
not direct you to use it? (PRY33)
How old were you when you first
used Dilaudid in a way a doctor
did not direct you to use it?
(PRY33a)7
See notes at end of table.

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.1

(0.06)

1,683

0.0*

(0.00*)

1,316

22.1*

(2.83*)

N/A

0.0*

(0.00*)

N/A

0.3

(0.11)

1,683

0.3

(0.23)

1,316

24.2*

(2.03*)

N/A

33.2*

(4.93*)

N/A

0.1

(0.08)

1,683

0.2

(0.14)

1,316

23.5*

(0.65*)

N/A

20.8*

(3.42*)

N/A

0.0

(0.04)

1,683

0.0*

(0.00*)

1,316

17.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.1

(0.04)

1,683

0.0*

(0.00*)

1,316

18.6*

(0.61*)

N/A

0.0*

(0.00*)

N/A

0.3

(0.09)

1,683

0.1

(0.08)

1,316

21.5a*

(2.42*)

N/A

38.0*

(0.00*)

N/A

1
(continued)

H-13

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
methadone in any way a doctor
did not direct you to use it?
(PRY34)
How old were you when you first
used methadone in a way a doctor
did not direct you to use it?
(PRY34a)7
In the past 12 months, did you use
Opana in any way a doctor did not
direct you to use it? (PRY35)
How old were you when you first
used Opana in a way a doctor did
not direct you to use it?
(PRY35a)7
In the past 12 months, did you use
Opana ER in any way a doctor did
not direct you to use it? (PRY36)
How old were you when you first
used Opana ER in a way a doctor
did not direct you to use it?
(PRY36a)7
In the past 12 months, did you use
Talwin in any way a doctor did
not direct you to use it? (PRY38)
How old were you when you first
used Talwin in a way a doctor did
not direct you to use it?
(PRY38a)7
In the past 12 months, did you use
Talwin NX in any way a doctor
Did not direct you to use it?
(PRY39)
How old were you when you first
used Talwin NX in a way a doctor did
not direct you to use it?
(PRY39a)7
In the past 12 months, did you use
any other prescription pain
reliever in a way a doctor did not
direct you to use it? (PRY40)
How old were you when you first
used any other prescription pain
reliever in a way a doctor did not
direct you to use it? (PRY40a)7
In the past 30 days, did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it?
(PRM01)
See notes at end of table.

0.3

(0.12)

1,683

0.3

(0.17)

1,316

20.9*

(2.42*)

N/A

25.0*

(4.01*)

N/A

0.1

(0.06)

1,683

0.0

(0.01)

1,316

16.2*

(1.16*)

N/A

16.0*

(0.00*)

N/A

0.1

(0.06)

1,683

0.0*

(0.00*)

1,316

17.7*

(0.24*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.03)

1,683

0.0*

(0.00*)

1,316

13.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,683

0.0*

(0.00*)

1,316

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.09)

1,677

0.5

(0.27)

1,314

21.2*

(2.74*)

N/A

17.2*

(2.64*)

N/A

2.0

(0.40)

1,674

1.4

(0.37)

1,314
(continued)

H-14

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 30 days, on how
many days did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it?
(PRM02)7
During the past 30 days, did you use
[PRNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (PRM03)
Which of these statements describe
your use of [PRNAMEFILL] at
any time in the past 12 months?
(PRY41)5
I used [PRNAMEFILL] without
a prescription of my own.
I used [PRNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [PRNAMEFILL] more
often than it was/they were
prescribed.
I used [PRNAMEFILL] for
longer than it was/they were
prescribed.
I used [PRNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
What were the reasons you used
[PRLASTFILL2] that time?
(PRYMOTIV)5
To relieve physical pain
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
See notes at end of table.

8.7

(1.56)

N/A

10.1

(2.93)

N/A

0.8a

(0.26)

1,482

0.2

(0.12)

1,161

70.0a

(4.93)

120

51.0*

(7.76*)

25.8

(5.13)

120

18.1*

(4.72*)

22.6

(4.96)

120

24.0*

(6.25*)

12.6

(3.68)

120

21.4*

(6.94*)

21.5

(4.37)

120

19.1*

(5.85*)

73.3

(4.25)

117

68.6*

(6.58*)

(5.79 )

(1.80*)

22.5
8.9*
22.8

(4.27)

117

(3.50*)

117

3.1*

117

15.6

11.1

(4.26)

21.1

(3.97 )

(5.26 )

14.7

(3.48)

117

8.1

(2.64)

117

6.3*

(3.35*)

2.3

(1.48)

117

1.1*

(1.13*)

1.9

(1.27)

117

5.5*

(3.02*)

2.4*

(1.77*)

117

3.0*

(1.84*)

80
(continued)

H-15

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Which was the main reason you
used [PRLASTFILL2] that time?
(PRYMOT1)5
To relieve physical pain
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
Now think about the last time you
used [PRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[PRLASTFILL]? (PRY42B)5
I got a prescription for the
[PRLASTFILL] from just one
doctor
I got prescriptions for the
[PRLASTFILL] from more
than one doctor
I stole the [PRLASTFILL] from
a doctor's office, clinic,
hospital, or pharmacy
I got the [PRLASTFILL] from a
friend or relative for free
I bought the [PRLASTFILL]
from a friend or relative
I took the [PRLASTFILL] from a
friend or relative without
asking
I bought the [PRLASTFILL]
from a drug dealer or other
stranger
I got the [PRLASTFILL] in some
other way
See notes at end of table.

28.6a*

(9.04*)

63.9*
*

(8.27 )

0.0*

(0.00*)

0.0*

11.6

*
*

23.4

19.3

(7.67 )
*

10.1

(11.68*)

(6.65 )

(0.00*)

(0.00 )

(8.44 )

(8.17 )

0.0

3.8*

(2.71*)

5.9*

(4.24*)

0.0*

(0.00*)

0.0*

(0.00*)

5.5*

(4.46*)

8.5*

(8.08*)

(0.00 )

0.0

(4.94)

121

32.2*

(7.40*)

1.9*

(1.93*)

121

2.3*

(2.24*)

0.3

(0.28)

121

0.0*

(0.00*)

46.2

(5.27)

121

40.0*

(7.25*)

12.0

(3.06)

121

9.1*

(2.28*)

4.1

(1.87)

121

2.8*

(2.57*)

5.9

(1.69)

121

8.6*

(4.65*)

4.7*

(2.93*)

121

4.9*

(2.61*)

79
(continued)

0.0

24.9

H-16

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How did your friend or relative get
the [PRLASTFILL]? (PRY42C)5
He or she got a prescription for
the [PRLASTFILL] from just
one doctor
He or she got prescriptions for
the [PRLASTFILL] from more
than one doctor
He or she stole the
[PRLASTFILL] from a
doctor's office, clinic, hospital,
or pharmacy
He or she got the
[PRLASTFILL] from a friend
or relative for free
He or she bought the
[PRLASTFILL] from a friend
or relative
He or she took the
[PRLASTFILL] from a friend
or relative without asking
He or she bought the
[PRLASTFILL] from a drug
dealer or other stranger
He or she got the
[PRLASTFILL] in some other
way
Have you ever, even once, used any
prescription tranquilizer in any
way a doctor did not direct you to
use it? (TRL01 and TRL02)
In the past 12 months, did you use
Xanax in any way a doctor did not
direct you to use it? (TRY01)
How old were you when you first
used Xanax in a way a doctor did
not direct you to use it?
(TRY01a)7
In the past 12 months, did you use
Xanax XR in a way a doctor did
not direct you to use it? (TRY02)
How old were you when you first
used Xanax XR in a way a doctor
did not direct you to use it?
(TRY02a)7
In the past 12 months, did you use
alprazolam in any way a doctor
did not direct you to use it?
(TRY03)
See notes at end of table.

90.2*

(4.97*)

94.8*

(5.08*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

2.7*

(1.96*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

1.2*

(1.20*)

5.2*

(5.08*)

1.5*

(1.51*)

0.0*

(0.00*)

4.4*

(4.27*)

0.0*

(0.00*)

5.9

(0.84)

1,683

5.4

(0.88)

1,309

1.3

(0.29)

1,687

1.1

(0.29)

1,315

21.0

(1.66)

N/A

23.5

(2.59)

N/A

0.2

(0.13)

1,687

0.0

(0.01)

1,315

25.7*

(6.48*)

N/A

15.0*

(0.00*)

N/A

0.3

(0.11)

1,687

0.4

(0.28)

1,315
(continued)

H-17

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used alprazolam in a way a doctor
did not direct you to use it?
(TRY03a)7
In the past 12 months, did you use
extended-release alprazolam in
any way a doctor did not direct
you to use it? (TRY04)
How old were you when you first
used extended-release alprazolam
in a way a doctor did not direct
you to use it? (TRY04a)7
In the past 12 months, did you use
Ativan in any way a doctor did not
direct you to use it? (TRY05)
How old were you when you first
used Ativan in a way a doctor did
not direct you to use it?
(TRY05a)7
In the past 12 months, did you use
Klonopin in any way a doctor did
not direct you to use it? (TRY06)
How old were you when you first
used Klonopin in a way a doctor
did not direct you to use it?
(TRY06a)7
In the past 12 months, did you use
lorazepam in any way a doctor did
not direct you to use it? (TRY07)
How old were you when you first
used lorazepam in a way a doctor
did not direct you to use it?
(TRY07a)7
In the past 12 months, did you use
clonazepam in any way a doctor
did not direct you to use it?
(TRY08)
How old were you when you first
used clonazepam in a way a doctor
did not direct you to use it?
(TRY08a)7
In the past 12 months, did you use
Valium in any way a doctor did
not direct you to use it? (TRY09)
How old were you when you first
used Valium in a way a doctor did
not direct you to use it?
(TRY09a)7
See notes at end of table.

20.8*

(4.15*)

N/A

29.2*

(7.57*)

N/A

0.0

(0.03)

1,687

0.1

(0.11)

1,315

13.0a*

(0.00*)

N/A

17.9*

(0.14*)

N/A

0.2a

(0.07)

1,687

0.0*

(0.00*)

1,315

24.8*

(4.67*)

N/A

0.0*

(0.00*)

N/A

0.5

(0.19)

1,687

0.2

(0.13)

1,315

18.7

(0.84)

N/A

23.4*

(3.93*)

N/A

0.5

(0.15)

1,687

0.4

(0.24)

1,315

26.3

(4.40)

N/A

40.2*

(9.67*)

N/A

0.2

(0.08)

1,687

0.2

(0.11)

1,315

16.0a*

(1.01*)

N/A

28.0*

(4.73*)

N/A

0.6

(0.18)

1,687

0.3

(0.13)

1,315

20.6

(2.50)

N/A

22.6*

(2.87*)

N/A

5
(continued)

H-18

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
Librium in any way a doctor did
not direct you to use it? (TRY10)
How old were you when you first
used Librium in a way a doctor
did not direct you to use it?
(TRY10a)7
In the past 12 months, did you use
Tranxene in any way a doctor did
not direct you to use it? (TRY11)
How old were you when you first
used Tranxene in a way a doctor
did not direct you to use it?
(TRY11a)7
In the past 12 months, did you use
diazepam in any way a doctor did
not direct you to use it? (TRY12)
How old were you when you first
used diazepam in a way a doctor
did not direct you to use it?
(TRY12a)7
In the past 12 months, did you use
oxazepam, also known as Serax,
in any way a doctor did not direct
you to use it? (TRY13)
How old were you when you first
used oxazepam in a way a doctor did
not direct you to use it?
(TRY13a)7
In the past 12 months, did you use
Flexeril in any way a doctor did
not direct you to use it? (TRY14)
How old were you when you first
used Flexeril in a way a doctor did
not direct you to use it?
(TRY14a)7
In the past 12 months, did you use
Soma in any way a doctor did not
direct you to use it? (TRY15)
How old were you when you first
used Soma in a way a doctor did
not direct you to use it?
(TRY15a)7
In the past 12 months, did you use
buspirone, also known as BuSpar,
in any way a doctor did not direct
you to use it? (TRY16)
How old were you when you first
used buspirone, also known as
BuSpar, in a way a doctor did not
direct you to use it? (TRY16a)7
In the past 12 months, did you use
hydroxyzine, also known as
Atarax or Vistaril, in any way a
doctor did not direct you to use it?
(TRY17)
See notes at end of table.

(0.02)

1,687

0.0*

(0.00*)

1,315

17.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,315

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.08)

1,687

0.3

(0.25)

1,315

20.9*

(2.58*)

N/A

21.0*

(1.27*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,315

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.4a

(0.15)

1,687

0.0

(0.02)

1,315

29.9*

(4.47*)

N/A

26.2*

(0.40*)

N/A

0.4

(0.12)

1,687

0.7

(0.33)

1,315

19.6a

(1.11)

N/A

30.3*

(4.73*)

N/A

0.0

(0.03)

1,687

0.2

(0.19)

1,315

13.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.03)

1,687

0.0*

(0.00*)

1,315
(continued)

H-19

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used hydroxyzine, also known as
Atarax or Vistaril, in a way a
doctor did not direct you to use it?
(TRY17a)7
In the past 12 months, did you use
meprobamate, also known as
Equanil or Miltown, in any way a
doctor did not direct you to use it?
(TRY18)
How old were you when you first
used meprobamate, also known as
Equanil or Miltown, in a way a
doctor did not direct you to use it?
(TRY18a)7
In the past 12 months, did you use
any other prescription tranquilizer
in a way a doctor did not direct
you to use it? (TRY19)
How old were you when you first
used any other prescription
tranquilizer in a way a doctor did not
direct you to use it?
(TRY18a)7
In the past 30 days, did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it?
(TRM01)
During the past 30 days, on how
many days did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it?
(TRM02)7
During the past 30 days, did you use
[TRNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (TRM03)
Which of these statements describe
your use of [TRNAMEFILL] at
any time in the past 12 months?
(TRY20)5
I used [TRNAMEFILL] without
a prescription of my own.
I used [TRNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [TRNAMEFILL] more
often than it was/they were
prescribed.
I used [TRNAMEFILL] for
longer than it was/they were
prescribed.
I used [TRNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
See notes at end of table.

16.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.03)

1,687

0.0

(0.01)

1,315

13.0*

(0.00*)

N/A

17.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.1

(0.12)

1,316

0.0*

(0.00*)

N/A

40.0*

(0.00*)

N/A

0.9

(0.26)

1,687

0.7

(0.30)

1,314

5.6a

(1.53)

N/A

11.0

(2.05)

N/A

0.4

(0.16)

1,605

0.1

(0.04)

1,238

78.9*

(5.76*)

59.3*

(10.73*)

20.0*

(5.57*)

20.8*

(10.10*)

7.4*

(3.17*)

3.5*

(3.26*)

2.9*

(2.14*)

2.7*

(2.21*)

9.0*

(3.29*)

25.8*

(9.33*)

41
(continued)

H-20

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
What were the reasons you used
[TRLASTFILL2] that time?
(TRYMOTIV)5
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
Which was the main reason you
used [TRLASTFILL2] that time?
(TRYMOT1)5
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
Now think about the last time you
used [TRLASTFILL2] in any way
a doctor did not direct you to use
it/them. How did you get the
[TRLASTFILL]? (TRY21B)5
I got a prescription for the
[TRLASTFILL] from just one
doctor
I got prescriptions for the
[TRLASTFILL] from more
than one doctor
I stole the [TRLASTFILL] from
a doctor's office, clinic,
hospital, or pharmacy
I got the [TRLASTFILL] from a
friend or relative for free
I bought the [TRLASTFILL]
from a friend or relative
See notes at end of table.

67.0*

(6.92*)

52.2*

(9.27*)

10.0*

(4.10*)

2.0*

(1.84*)

25.3

*
*

23.3

(6.03 )

(9.36 )

(7.90 )

28.9

(7.91 )

26.6

20.9*

(5.65*)

14.8*

(6.06*)

10.2a*

(4.87*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

2.3*

(2.28*)

7.1*

(5.01*)

49.4*

(11.24*)

35.9*

(17.83*)

5.2*

(5.06*)

0.0*

(0.00*)

6.6

(4.57 )

21.9

(14.18 )

18.1*

(11.72*)

42.2*

(19.85*)

13.9a*

(6.99*)

0.0*

(0.00*)

6.8*

(5.53*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

(0.00 )

0.0

17.5*

(7.08*)

26.2*

(10.44*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

52.7*

(7.13*)

39.4*

(10.70*)

10.5*

(3.90*)

12.1*

(5.54*)

43
(continued)

0.0

H-21

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
I took the [TRLASTFILL] from
a friend or relative without
asking
I bought the [TRLASTFILL]
from a drug dealer or other
stranger
I got the [TRLASTFILL] in
some other way
How did your friend or relative get
the [TRLASTFILL]? (TRY21C)5
He or she got a prescription for
the [TRLASTFILL] from just
one doctor
He or she got prescriptions for
the[TRLASTFILL] from more
than one doctor
He or she stole the
[TRLASTFILL] from a
doctor's office, clinic, hospital,
or pharmacy
He or she got the
[TRLASTFILL] from a friend
or relative for free
He or she bought the
[TRLASTFILL] from a friend
or relative
He or she took the
[TRLASTFILL] from a friend
or relative without asking
He or she bought the
[TRLASTFILL] from a drug
dealer or other stranger
He or she got the
[TRLASTFILL] in some other
way
Have you ever, even once, used any
prescription stimulant in any way
a doctor did not direct you to use
it? (STL01 and STL02)
In the past 12 months, did you use
Adderall in any way a doctor did
not direct you to use it? (STY01)
How old were you when you first
used Adderall in a way a doctor
did not direct you to use it?
(STY01a)7
In the past 12 months, did you use
Adderall XR in any way a doctor
did not direct you to use it?
(STY02)
See notes at end of table.

(5.70*)

6.0*

(3.53*)

5.0*

(3.25*)

13.6*

(7.09*)

2.0*

(2.06*)

2.6*

(2.59*)

89.1*

(5.48*)

97.6*

(1.65*)

3.0*

(2.97*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

2.2*

(2.25*)

0.0*

(0.00*)

5.7*

(4.08*)

2.4*

(1.65*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

4.3

(0.64)

1,684

4.4

(0.91)

1,311

1.3

(0.31)

1,687

1.8

(0.48)

1,314

19.2

(0.60)

N/A

20.9

(1.12)

N/A

0.6

(0.16)

1,687

0.3

(0.17)

1,314
(continued)

H-22

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used Adderall XR in a way a
doctor did not direct you to use it?
(STY02a)7
In the past 12 months, did you use
Dexedrine in any way a doctor did
not direct you to use it? (STY03)
How old were you when you first
used Dexedrine in a way a doctor
did not direct you to use it?
(STY03a)7
In the past 12 months, did you use
dextroamphetamine in any way a
doctor did not direct you to use it?
(STY04)
How old were you when you first
used dextroamphetamine in a way
a doctor did not direct you to use
it? (STY04a)7
In the past 12 months, did you use
mixed amphetamine
dextroamphetamine pills other
than Adderall in any way a doctor
did not direct you to use them?
(STY05)
How old were you when you first
used mixed amphetamine
dextroamphetamine pills other
than Adderall in a way a doctor
did not direct you to use them?
(STY05a)7
In the past 12 months, did you use
Ritalin in any way a doctor did not
direct you to use it? (STY06)
How old were you when you first
used Ritalin in a way a doctor did
not direct you to use it? (STY06a)7
In the past 12 months, did you use
Ritalin SR or Ritalin LA in any
way a doctor did not direct you to
use it? (STY07)
How old were you when you first
used Ritalin SR or Ritalin LA in a
way a doctor did not direct you to
use it? (STY07a)7
In the past 12 months, did you use
Concerta in any way a doctor did
not direct you to use it? (STY08)
How old were you when you first
used Concerta in a way a doctor
did not direct you to use it?
(STY08a)7
See notes at end of table.

18.6

(0.85)

N/A

26.2*

(3.89*)

N/A

0.1

(0.09)

1,687

0.0*

(0.00*)

1,314

17.6*

(0.44*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.10)

1,687

0.1

(0.07)

1,314

18.3a*

(0.26*)

N/A

19.0*

(0.00*)

N/A

0.3

(0.14)

1,687

0.1

(0.05)

1,314

20.2*

(1.26*)

N/A

24.1*

(3.98*)

N/A

0.3

(0.11)

1,687

0.1

(0.05)

1,314

26.3*

(6.68*)

N/A

18.6*

(0.58*)

N/A

0.2a

(0.09)

1,687

0.0*

(0.00*)

1,315

18.2*

(0.63*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.09)

1,687

0.2

(0.14)

1,314

17.5a*

(0.79*)

N/A

22.0*

(0.00*)

N/A

1
(continued)

H-23

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
In the past 12 months, did you use
Daytrana in any way a doctor did
not direct you to use it? (STY09)
How old were you when you first
used Daytrana in a way a doctor
did not direct you to use it?
(STY09a)7
In the past 12 months, did you use
methylphenidate in any way a
doctor did not direct you to use it?
(STY10)
How old were you when you first
used methylphenidate in a way a
doctor did not direct you to use it?
(STY10a)7
In the past 12 months, did you use
Metadate CD in any way a doctor
did not direct you to use it?
(STY11)
How old were you when you first
used Metadate CD in a way a
doctor did not direct you to use it?
(STY11a)7
In the past 12 months, did you use
Metadate ER in any way a doctor
did not direct you to use it?
(STY12)
How old were you when you first
used Metadate ER in a way a
doctor did not direct you to use it?
(STY12a)7
In the past 12 months, did you use
Focalin in any way a doctor did
not direct you to use it? (STY13)
How old were you when you first
used Focalin in a way a doctor did
not direct you to use it? (STY13a)7
In the past 12 months, did you use
Focalin XR in any way a doctor
did not direct you to use it?
(STY14)
How old were you when you first
used Focalin XR in a way a doctor
did not direct you to use it?
(STY14a)7
In the past 12 months, did you use
dexmethylphenidate in any way a
doctor did not direct you to use it?
(STY15)
How old were you when you first
used dexmethylphenidate in a way
a doctor did not direct you to use
it? (STY15a)7
In the past 12 months, did you use
benzphetamine in any way a
doctor did not direct you to use it?
(STY16)
See notes at end of table.

(0.02)

1,687

0.0*

(0.00*)

1,315

19.6*

(2.47*)

N/A

0.0*

(0.00*)

N/A

0.2

(0.10)

1,687

0.0

(0.01)

1,315

30.1*

(11.21*)

N/A

17.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.1

(0.06)

1,687

0.1

(0.05)

1,314

17.7a*

(1.05*)

N/A

21.5*

(0.71*)

N/A

0.1

(0.06)

1,687

0.0

(0.03)

1,314

17.3a*

(0.45*)

N/A

13.2*

(0.29*)

N/A

0.1

(0.06)

1,687

0.0*

(0.00*)

1,314

17.4*

(0.92*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314
(continued)

H-24

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used benzphetamine in a way a
doctor did not direct you to use it?
(STY16a)7
In the past 12 months, did you use
Didrex in any way a doctor did not
direct you to use it? (STY17)
How old were you when you first
used Didrex in a way a doctor did not
direct you to use it? (STY17a)7
In the past 12 months, did you use
diethylpropion in any way a
doctor did not direct you to use it?
(STY18)
How old were you when you first
used diethylpropion in a way a
doctor did not direct you to use it?
(STY18a)7
In the past 12 months, did you use
phendimetrazine in any way a
doctor did not direct you to use it?
(STY19)
How old were you when you first
used phendimetrazine in a way a
doctor did not direct you to use it?
(STY19a)7
In the past 12 months, did you use
phentermine in any way a doctor
did not direct you to use it?
(STY20)
How old were you when you first
used phentermine in a way a
doctor did not direct you to use it?
(STY20a)7
In the past 12 months, did you use
Provigil in any way a doctor did
not direct you to use it? (STY21)
How old were you when you first
used Provigil in a way a doctor did not
direct you to use it? (STY21a)7
In the past 12 months, did you use
Tenuate in any way a doctor did
not direct you to use it? (STY22)
How old were you when you first
used Tenuate in a way a doctor did not
direct you to use it? (STY22a)7
In the past 12 months, did you use
Vyvanse in any way a doctor did
not direct you to use it? (STY23)
How old were you when you first
used Vyvanse in a way a doctor
did not direct you to use it?
(STY23a)7
In the past 12 months, did you use
any other prescription stimulant in
a way a doctor did not direct you
to use it? (STY24)
See notes at end of table.

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.03)

1,687

0.0*

(0.00*)

1,314

12.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.03)

1,687

0.0*

(0.00*)

1,314

21.4*

(1.06*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,687

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.3

(0.10)

1,687

0.2

(0.12)

1,314

17.9a*

(0.64*)

N/A

22.3*

(1.39*)

N/A

0.1

(0.09)

1,687

0.0

(0.02)

1,316
(continued)

H-25

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used any other prescription
stimulant in a way a doctor did not
direct you to use it? (STY24a)7
In the past 30 days, did you use
[STNAMEFILL] in any way a
doctor did not direct you to use it?
(STM01)
During the past 30 days, on how
many days did you use
[STNAMEFILL] in any way a
doctor did not direct you to use it?
(STM02)7
During the past 30 days, did you use
[STNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (STM03)
Which of these statements describe
your use of [STNAMEFILL] at
any time in the past 12 months?
(STY25)5
I used [STNAMEFILL] without
a prescription of my own.
I used [STNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [STNAMEFILL] more
often than it was/they were
prescribed.
I used [STNAMEFILL] for
longer than it was/they were
prescribed.
I used [STNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
At any time in the past 12 months,
did you ever use a needle to inject
[STNAMEFILL]? (STY25a)
How long has it been since you last
used a needle to inject
[STNAMEFILL]? (STY25b)5
Within the past 30 days
More than 30 days ago but
within the past 12 months
More than 12 months ago
What were the reasons you used
[STLASTFILL2] that time?
(STYMOTIV)5
To help me lose weight
To help me concentrate
To help me be alert or stay
awake
To help me study
To experiment or to see what it's
like
See notes at end of table.

20.8a*

(1.17*)

N/A

0.5

(0.14)

1,687

10.4

(3.66)

N/A

0.2

(0.11)

1,609

29.0*

(0.00*)

N/A

0.3

(0.12)

1,313

5.1*

(0.94*)

N/A

0.2

(0.10)

1,249

81.4*

(5.84*)

84.6*

(7.71*)

23.1a*

(6.89*)

2.7*

(2.02*)

12.5*

(5.42*)

3.5*

(2.20*)

10.0*

(5.60*)

1.1*

(1.11*)

13.6*

(4.58*)

17.0*

(7.79*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0

(0.00 )

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

(0.00 )

0.0

(0.00 )

0.0

8.4*

(3.84*)

14.3*

(7.37*)

(10.45 )

46.6

(9.03 )

49.5

53.3*

(6.22*)

47.5*

(7.08*)

(11.84 )

(7.18*)

36
(continued)

38.4

(9.54 )

30.8

12.5*

(4.31*)

13.5*

H-26

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
To feel good or get high
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
Which was the main reason you
used [STLASTFILL2] that time?
(STYMOT1)5
To help me lose weight
To help me concentrate
To help me be alert or stay
awake
To help me study
To experiment or to see what it's
like
To feel good or get high
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
I used it/them for some other
reason
How did you get the
[STLASTFILL]? (STY26b)5
I got a prescription for the
[STLASTFILL] from just one
doctor
I got prescriptions for the
[STLASTFILL] from more
than one doctor
I stole the [STLASTFILL] from
a doctor's office, clinic,
hospital, or pharmacy
I got the [STLASTFILL] from a
friend or relative for free
I bought the [STLASTFILL]
from a friend or relative
I took the [STLASTFILL] from a
friend or relative without
asking
I bought the [STLASTFILL]
from a drug dealer or other
stranger
I got the [STLASTFILL] in some
other way
How did your friend or relative get
the [STLASTFILL]? (STY26c)5
He or she got a prescription for
the [STLASTFILL] from just
one doctor
He or she got prescriptions for
the [STLASTFILL] from more
than one doctor
See notes at end of table.

(6.36*)

9.6*

(5.13*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

5.3*

(3.16*)

3.9*

(2.93*)

6.9*

(5.00*)

0.0*

(0.00*)

23.1

(11.72 )

43.1

(12.09 )

15.3*

(8.97*)

24.4*

(16.17*)

(15.03 )

(0.00*)

8.2*
*

46.8

0.0*

20.8

(11.24 )

(8.15*)

(0.00 )

7.8

(6.03 )

0.0

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

3.5*

(3.64*)

8.8*

(3.99*)

2.4*

(1.84*)

3.4*

(3.31*)

7.3*

(6.43*)

0.0*

(0.00*)

0.0*

(0.00*)

59.3*

(7.45*)

54.1*

(11.42*)

14.8*

(4.99*)

28.8*

(8.93*)

3.1*

(2.12*)

3.6*

(3.63*)

6.2*

(4.09*)

3.8*

(3.00*)

4.4*

(4.22*)

0.0*

(0.00*)

81.1*

(7.49*)

85.9*

(8.70*)

0.0*

(0.00*)

0.0*

(0.00*)

21
(continued)

H-27

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
He or she stole the
[STLASTFILL] from a
doctor's office, clinic, hospital,
or pharmacy
He or she got the
[STLASTFILL] from another
friend or relative for free
He or she bought the
[STLASTFILL] from another
friend or relative
He or she took the
[STLASTFILL] from another
friend or relative without
asking
He or she bought the
[STLASTFILL] from a drug
dealer or other stranger
He or she got the
[STLASTFILL] in some other
Way
Have you ever, even once, used any
prescription sedative in any way a
doctor did not direct you to use it?
(SVL01 and SVL02)
In the past 12 months, did you use
Ambien in any way a doctor did
not direct you to use it? (SVY01)
How old were you when you first
used Ambien in a way a doctor
did not direct you to use it?
(SVY01a)7
In the past 12 months, did you use
Ambien CR in a way a doctor did
not direct you to use it? (SVY02)
How old were you when you first
used Ambien CR in a way a
doctor did not direct you to use it?
(SVY02a)7
In the past 12 months, did you use
zolpidem in any way a doctor did
not direct you to use it? (SVY03)
How old were you when you first
used zolpidem in a way a doctor
did not direct you to use it?
(SVY03a)7
In the past 12 months, did you use
extended-release zolpidem in any
way a doctor did not direct you to
use it? (SVY04)
How old were you when you first
used extended-release zolpidem in a
way a doctor did not direct you to use
it? (SVY04a)7
In the past 12 months, did you use
Lunesta in any way a doctor did
not direct you to use it? (SVY05)
See notes at end of table.

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

7.1*

(7.18*)

6.3*

(3.60*)

1.9*

(1.90*)

2.9*

(2.87*)

0.0*

(0.00*)

4.7*

(4.67*)

1.0*

(1.00*)

5.0*

(4.79*)

4.2*

(4.14*)

3.7

(0.62)

1,683

3.2

(0.75)

1,309

0.4

(0.17)

1,688

0.7

(0.30)

1,314

25.2a*

(2.58*)

N/A

49.5*

(4.45*)

N/A

0.0

(0.02)

1,688

0.0

(0.03)

1,314

18.9a*

(2.12*)

N/A

35.0*

(0.00*)

N/A

0.4a

(0.20)

1,688

0.0*

(0.00*)

1,314

45.4*

(7.55*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.0

(0.03)

1,314

0.0*

(0.00*)

N/A

11.0*

(0.00*)

N/A

0.1

(0.11)

1,688

0.0*

(0.00*)

1,314
(continued)

H-28

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used Lunesta in a way a doctor did
not direct you to use it?
(SVY05a)7
In the past 12 months, did you use
Sonata in any way a doctor did not
direct you to use it? (SVY06)
How old were you when you first
used Sonata in a way a doctor did
not direct you to use it?
(SVY06a)7
In the past 12 months, did you use
zaleplon in any way a doctor did
not direct you to use it? (SVY07)
How old were you when you first
used zaleplon in a way a doctor did
not direct you to use it?
(SVY07a)7
In the past 12 months, did you use
Dalmane in any way a doctor did
Not direct you to use it? (SVY08)
How old were you when you first
used Dalmane in a way a doctor did
not direct you to use it?
(SVY08a)7
In the past 12 months, did you use
Halcion in any way a doctor did
not direct you to use it? (SVY09)
How old were you when you first
used Halcion in a way a doctor did
not direct you to use it?
(SVY09a)7
In the past 12 months, did you use
triazolam in any way a doctor did
not direct you to use it? (SVY11)
How old were you when you first
used triazolam in a way a doctor did
not direct you to use it?
(SVY10a)7
In the past 12 months, did you use
Restoril in any way a doctor did
not direct you to use it? (SVY12)
How old were you when you first
used Restoril in a way a doctor
did not direct you to use it?
(SVY12a)7
In the past 12 months, did you use
temazepam in any way a doctor
did not direct you to use it?
(SVY13)
How old were you when you first
used temazepam in a way a doctor did
not direct you to use it?
(SVY13a)7
In the past 12 months, did you use
Butisol in any way a doctor did
not direct you to use it? (SVY14)
See notes at end of table.

57.0*

(12.65*)

N/A

0.0*

(0.00*)

N/A

0.1

(0.07)

1,688

0.0*

(0.00*)

1,314

16.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.1

(0.12)

1,314

0.0*

(0.00*)

N/A

45.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.1

(0.06)

1,314

0.0*

(0.00*)

N/A

24.0*

(0.00*)

N/A

0.1

(0.07)

1,688

0.0*

(0.00*)

1,314

16.2*

(0.22*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.0

(0.02)

1,314

0.0*

(0.00*)

N/A

66.0*

(0.00*)

N/A

0.0

(0.03)

1,688

0.0*

(0.00*)

1,314
(continued)

H-29

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when you first
used Butisol in a way a doctor did
not direct you to use it? (SVY14a)7
In the past 12 months, did you use
Seconal in any way a doctor did
Not direct you to use it? (SVY15)
How old were you when you first
used Seconal in a way a doctor did
not direct you to use it?
(SVY15a)7
In the past 12 months, did you use
phenobarbital in any way a doctor
did not direct you to use it?
(SVY16)
How old were you when you first
used phenobarbital in a way a
doctor did not direct you to use it?
(SVY16a)7
In the past 12 months, did you use
other prescription sedative in any way
a doctor did not direct you to use it?
(SVY17)
How old were you when you first
used any other prescription
sedative in a way a doctor did not
direct you to use it? (SVY17a)7
In the past 30 days, did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it?
(SVM01)
During the past 30 days, on how
Many days did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it?
(SVM02)7
During the past 30 days, did you use
[SVNAMEFILL] in any way a
doctor did not direct you to use it
while you were drinking alcohol
or within a couple of hours of
drinking? (SVM03)
Which of these statements describe
your use of [SVNAMEFILL] at
any time in the past 12 months?
(SVY18)5
I used [SVNAMEFILL] without
a prescription of my own.
I used [SVNAMEFILL] in
greater amounts than it
was/they were prescribed.
I used [SVNAMEFILL] more
often than it was/they were
prescribed
I used [SVNAMEFILL] for
longer than it was/they were
prescribed.
I used [SVNAMEFILL] in some
other way a doctor did not
direct me to use it/them.
See notes at end of table.

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

1,688

0.0*

(0.00*)

1,314

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0

(0.02)

1,688

0.0

(0.04)

1,314

20.0*

(0.00*)

N/A

16.0*

(0.00*)

N/A

0.0

(0.02)

1,688

0.0*

(0.00*)

1,313

16.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.3

(0.17)

1,688

0.2

(0.11)

1,313

11.2*

(5.80*)

N/A

17.4*

(9.81*)

N/A

0.2

(0.11)

1,639

0.1

(0.07)

1,266

54.8*

(14.38*)

59.2*

(20.40*)

23.2*

(12.30*)

28.1*

(21.55*)

16.8*

(11.93*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.8

(0.24)

1,688

0.8

(0.31)

1,313
(continued)

H-30

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
What were the reasons you used
[SVLASTFILL2] that time?
(SVYMOTIV)5
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
Which was the main reason you
used [SVLASTFILL] that time?
(SVYMOT1)5
To relax or relieve tension
To experiment or to see what it's/
they're like
To feel good or get high
To help with my sleep
To help me with my feelings or
emotions
To increase or decrease the
effect(s) of some other drug
Because I am "hooked" or I have
to have it/them
The other reason I reported
How did you get the
[SVLASTFILL]? (SVY19B)5
I got a prescription for the
[SVLASTFILL] from just one
doctor
I got prescriptions for the
[SVLASTFILL] from more
than one doctor
I stole the [SVLASTFILL] from
a doctor's office, clinic,
hospital, or pharmacy
I got the [SVLASTFILL] from a
friend or relative for free
I bought the [SVLASTFILL]
from a friend or relative
I took the [SVLASTFILL] from
a friend or relative without
asking
I bought the [SVLASTFILL]
from a drug dealer or other
stranger
I got the [SVLASTFILL] in
some other way
See notes at end of table.

29.7

5.7*

(13.41 )

18.6*

(13.60*)

(4.18*)

3.7*

(3.88*)

0.0

(0.00 )

75.2

(14.48 )

(4.23 )

(10.34 )

2.0*

(1.92*)

0.0*

(0.00*)

3.9*

(2.70*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

(2.65 )

7.1

76.8

(0.00 )

2.5

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0

*
*

0.0

23.8

(22.23 )

(0.00 )

(22.23 )

0.0

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

(0.00 )

76.2

0.0

(0.00 )

0.0

45.2*

(14.38*)

40.8*

(20.40*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

38.8*

(13.62*)

45.9*

(18.97*)

5.5*

(4.03*)

4.9*

(5.01*)

0.0*

(0.00*)

0.0*

(0.00*)

8.5*

(8.13*)

0.0*

(0.00*)

1.9*

(1.88*)

8.4*

(8.33*)

11
(continued)

H-31

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How did your friend or relative get
the [SVLASTFILL]? (SVY19C)5
He or she got a prescription for
the [SVLASTFILL] from just
one doctor
He or she got prescriptions for
the [SVLASTFILL] from more
than one doctor
He or she stole the
[SVLASTFILL] from a
doctor's office, clinic, hospital,
or pharmacy
He or she got the
[SVLASTFILL] from another
friend or relative for free
He or she bought the
[SVLASTFILL] from another
friend or relative
He or she took the
[SVLASTFILL] from another
friend or relative without
asking
He or she bought the
[SVLASTFILL] from a drug
dealer or other stranger
He or she got the
[SVLASTFILL] in some other
way
Have you ever, even once, used a
needle to inject any drug that was
not prescribed for you? (SD15)
Was any of your marijuana use in
the past 12 months recommended
by a doctor? (MJMM)
Was all of your marijuana use in the
past 12 months recommended by a
doctor? (MJMM01)5
During the past 12 months, was
there a month or more when you
spent a lot of your time getting or
using methamphetamine?
(DRME01)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting
over the effects of the
methamphetamine you used?
(DRME02)
During the past 12 months, did you
try to set limits on how often or
how much methamphetamine you
would use? (DRME04)
Were you able to keep to the limits
you set, or did you often use
methamphetamine more than you
intended to? (DRME05)
See notes at end of table.

79.6*

(13.03*)

91.9*

(8.46*)

5.0*

(5.18*)

8.1*

(8.46*)

0.0*

(0.00*)

0.0*

(0.00*)

15.4*

(11.58*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.0*

(0.00*)

0.9

(0.29)

1,692

0.9

(0.50)

1,318

0.5

(0.17)

1,692

0.0*

(0.00*)

40.3*

(16.91*)

0.0*

(0.00*)

0.2

(0.08)

1,691

0.4

(0.24)

1,319

0.0*

(0.00*)

1,691

0.0*

(0.00*)

1,319

0.1

(0.04)

1,691

0.5

(0.21)

1,319

0.0

(0.03)

1,691

0.4

(0.20)

1,319
(continued)

H-32

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 12 months, did you
need to use more
methamphetamine than you used
in order to get the effect you
wanted? (DRME06)
During the past 12 months, did you
notice that using the same amount
of methamphetamine had less
effect on you than it used to?
(DRME07)
During the past 12 months, did you
want to or try to cut down or stop
using methamphetamine?
(DRME08)
During the past 12 months, were
you able to cut down or stop
using methamphetamine every
time you wanted to or tried to?
(DRME09)
During the past 12 months, have you
felt kind of blue or down when
you cut down or stopped using
methamphetamine? (DRME10)
During the past 12 months, have you
felt kind of blue or down when
you cut down or stopped using
methamphetamine? (DRME10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped
using methamphetamine?
(DRME11)
During the past 12 months, did you
have 2 or more of these symptoms
at the same time that lasted for
longer than a day after you cut
back or stopped using
methamphetamine? (DRME12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of
methamphetamine? (DRME13)
Did you continue to use
methamphetamine even though
you thought it was causing you to
have problems with your
emotions, nerves, or mental
health? (DRME14)
See notes at end of table.

0.2

(0.13)

1,691

0.5

(0.22)

1,319

0.1

(0.06)

1,691

0.0*

(0.00*)

1,319

0.2

(0.13)

1,691

0.4

(0.19)

1,319

0.2

(0.13)

1,691

0.4

(0.19)

1,319

0.1

(0.06)

1,691

0.1

(0.13)

1,319

0.2

(0.13)

1,691

0.2

(0.15)

1,319

0.2

(0.13)

1,691

0.1

(0.10)

1,319

0.2

(0.13)

1,691

0.1

(0.10)

1,319

0.2

(0.13)

1,691

0.1

(0.10)

1,319

0.1

(0.03)

1,691

0.1

(0.10)

1,319
(continued)

H-33

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of
methamphetamine? (DRME15)
Did you continue to use
methamphetamine even though
this was causing you to have
physical problems? (DRME16)
During the past 12 months, did
using methamphetamine cause
you to give up or spend less time
doing these types of important
activities? (DRME17)
During the past 12 months, did
using methamphetamine cause
you to have serious problems
either at home, work, or school?
(DRME18)
During the past 12 months, did you
regularly use methamphetamine
and then do something where
using methamphetamine might
have put you in physical danger?
(DRME19)
During the past 12 months, did
using methamphetamine cause
you to do things that repeatedly
got you in trouble with the law?
(DRME20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused
by your use of methamphetamine?
(DRME21)
Did you continue to use
methamphetamine even though
you thought it caused problems
with family or friends? (DRME22)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting or
using prescription stimulants?
(DRST01)
During the past 12 months, was
there a month or more when you
spent a lot of your time getting
over the effects of the prescription
stimulants you used? (DRST02)
During the past 12 months, did you
try to set limits on how often or
how much prescription stimulants
you would use? (DRST04)
See notes at end of table.

0.0*

(0.00*)

1,691

0.0*

(0.00*)

1,319

0.0*

(0.00*)

1,691

0.0*

(0.00*)

1,319

0.0

(0.02)

1,691

0.1

(0.10)

1,319

0.0

(0.02)

1,691

0.1

(0.10)

1,319

0.1

(0.03)

1,691

0.1

(0.10)

1,319

0.0

(0.02)

1,691

0.1

(0.10)

1,319

0.1

(0.06)

1,691

0.1

(0.10)

1,319

0.0

(0.02)

1,691

0.1

(0.10)

1,319

0.2

(0.07)

1,684

0.3

(0.15)

1,311

0.0*

(0.00*)

1,684

0.0*

(0.00*)

1,311

0.5

(0.17)

1,684

0.3

(0.15)

1,311
(continued)

H-34

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Were you able to keep to the limits
you set, or did you often use
prescription stimulants more than
you intended to? (DRST05)
During the past 12 months, did you
need to use more prescription
stimulants than you used to in
order to get the effect you wanted?
(DRST06)
During the past 12 months, did you
notice that using the same amount
of prescription stimulants had less
effect on you than it used to?
(DRST07)
During the past 12 months, did you
want to or try to cut down or stop
using prescription stimulants?
(DRST08)
During the past 12 months, were
you able to cut down or stop
Using prescription stimulants
every time you wanted to or tried
to? (DRST09)
During the past 12 months, did you
cut down or stop using
prescription stimulants at least one
time? (DRST10)
During the past 12 months, have
you felt kind of blue or down
when you cut down or stopped
using prescription stimulants?
(DRST10a)
During the past 12 months, did you
have 2 or more of these symptoms
after you cut back or stopped
using prescription stimulants?
(DRST11)
During the past 12 months, did you
have 2 or more of these symptoms
at the same time that lasted for
longer than a day after you cut
back or stopped using prescription
stimulants? (DRST12)
During the past 12 months, did you
have any problems with your
emotions, nerves, or mental health
that were probably caused or made
worse by your use of prescription
stimulants? (DRST13)
See notes at end of table.

0.4

(0.16)

1,684

0.2

(0.14)

1,311

0.4

(0.13)

1,684

0.2

(0.10)

1,311

0.1

(0.08)

1,684

0.1

(0.11)

1,311

0.6

(0.18)

1,684

0.5

(0.19)

1,311

0.5

(0.17)

1,684

0.5

(0.19)

1,311

0.3

(0.11)

1,684

0.3

(0.18)

1,311

0.3

(0.12)

1,684

0.1

(0.09)

1,311

0.3

(0.12)

1,684

0.1

(0.06)

1,311

0.2

(0.08)

1,684

0.1

(0.06)

1,311

0.2

(0.10)

1,684

0.1

(0.06)

1,311
(continued)

H-35

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Did you continue to use prescription
stimulants even though you
thought this was causing you to
have problems with your
emotions, nerves, or mental
health? (DRST14)
During the past 12 months, did you
have any physical health problems
that were probably caused or made
worse by your use of prescription
stimulants? (DRST15)
Did you continue to use prescription
stimulants even though this was
causing you to have physical
problems? (DRST16)
During the past 12 months, did
using prescription stimulants
cause you to give up or spend less
time doing these types of
important activities? (DRST17)
During the past 12 months, did
using prescription stimulants
cause you to have serious
problems either at home, work, or
school? (DRST18)
During the past 12 months, did you
regularly use prescription
stimulants and then do something
where using prescription
stimulants might have put you in
physical danger? (DRST19)
During the past 12 months, did
using prescription stimulants
cause you to do things that
repeatedly got you in trouble with
the law? (DRST20)
During the past 12 months, did you
have any problems with family or
friends that were probably caused
by your use of prescription
stimulants? (DRST21)
Did you continue to use prescription
stimulants even though you
thought this caused problems with
family or friends? (DRST22)
See notes at end of table.

0.1

(0.09)

1,684

0.0

(0.02)

1,311

0.0

(0.04)

1,684

0.1

(0.08)

1,311

0.0

(0.04)

1,684

0.0*

(0.00*)

1,311

0.0*

(0.00*)

1,684

0.0*

(0.00*)

1,311

0.0

(0.02)

1,684

0.0*

(0.00*)

1,311

0.0*

(0.00*)

1,684

0.1

(0.11)

1,311

0.0*

(0.00*)

1,684

0.0*

(0.00*)

1,311

0.0*

(0.00*)

1,684

0.1

(0.11)

1,311

0.0*

(0.00*)

1,684

0.1

(0.11)

1,311
(continued)

H-36

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you the last time you
used any methamphetamine for
kicks or to get high? (LU17)7
Height in inches (HLTH05HLTH08)7,9
7,10

Weight in pounds (HLTH10-14)
During the past 12 months, how
many times have you visited a
doctor, nurse, physician assistant
or nurse practitioner about your
own health at a doctor's office, a
clinic, or some other place?
(HLTH19)7,8
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you smoke
cigarettes or use any other tobacco
products? (HLTH20a)5
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you drink alcohol?
(HLTH20b)5
During the past 12 months, did any
doctor or other health care
professional ask, either in person
or on a form, if you use illegal
drugs? (HLTH20c)5
During the past 12 months, did any
doctor or other health care
professional advise you to quit
smoking cigarettes or quit using
any other tobacco products?
(HLTH21)5
Choose the statement or statements
below that describe any
discussions you may have had in
person with a doctor or other
health professional about your
alcohol use. (HLTH22)5
The doctor asked how much I
drink.
The doctor asked how often I
drink.
The doctor asked if I have any
problems because of my
drinking.
The doctor advised me to cut
down on my drinking.
The doctor offered to give me
more information about alcohol
use and treatment for problems
with alcohol use.
The doctor didn't discuss my
alcohol use with me in the past
12 months.
See notes at end of table.

(0.87)

N/A

27.4

(1.55)

N/A

66.7

(0.28)

N/A

1,668

66.6

(0.19)

N/A

1,296

176.6

(1.51)

N/A

1,661

174.0

(1.83)

N/A

1,290

5.5

(0.69)

N/A

6.1

(0.33)

N/A

72.1

(1.46)

992

1,415

73.2

(1.93)

744

1,063

68.4

(1.61)

928

1,414

67.2

(1.83)

685

1,060

51.5

(1.75)

760

1,413

51.0

(2.52)

551

1,059

52.8

(2.51)

252

520

51.9

(3.82)

153

344

33.9

(2.07)

292

888

31.1

(2.64)

213

658

32.1a

(2.12)

281

888

25.7

(1.91)

199

658

5.8

(0.92)

888

4.5

(1.06)

658

2.4

(0.58)

888

1.1

(0.44)

658

0.9

(0.29)

888

0.6

(0.32)

658

47.8

(1.89)

477

1,032

49.5

(2.55)

364

806
(continued)

H-37

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 12 months, did any
doctor or other health care
professional talk to you about your
use of marijuana, cocaine, crack,
heroin, inhalants, hallucinogens,
or methamphetamine? (HLTH23)5
During the past 12 months, did you
have a sexually transmitted
disease such as chlamydia,
gonorrhea, herpes or syphilis?
(HLTH24)
Conditions that a doctor or other
health care professional has ever
told you that you had (HLTH25)
Any kind of heart condition or
heart disease

18.4

(3.30)

237

23.4

1.7

(0.33)

1,688

1.6

(4.73)

(0.46)

176

1,314

11.5

(1.16)

118

1,680

8.7

(1.30)

1,307

Diabetes or sugar diabetes
Chronic bronchitis, emphysema,
chronic obstructive pulmonary
disease, also called COPD

9.2

(1.06)

1,680

9.2

(1.19)

1,307

3.6

(0.64)

1,680

4.7

(1.05)

1,307

Cirrhosis of the liver

0.1

(0.13)

1,680

0.2

(0.19)

1,307

Hepatitis
Kidney disease, not including
bladder infection or
Incontinence

2.0

(0.55)

1,680

1.8

(0.59)

1,307

1.4

(0.39)

1,680

2.6

(0.83)

1,307

Asthma
HIV or AIDS
Cancer or a malignancy of any
kind
Hypertension, also called high
blood pressure
None of the above – I have never
had any of these conditions
What kind of cancer was it?
(HLTH26)5
Bladder
Blood
Bone

11.4

(0.87)

214

1,680

12.1

(1.37)

161

1,307

0.0*

(0.00*)

1,680

0.1

(0.07)

1,307

6.4

(0.94)

1,680

5.5

(1.04)

1,307

18.5

(1.26)

181

1,680

16.8

(1.94)

132

1,307

56.0

(1.69)

1,126

1,680

60.4

(1.69)

912

1,307

(0.00*)

0.0*

(0.00*)

3.7

0.0

*
*

0.0*
1.8

*
*

0.0

(1.73 )
*

(0.00 )

(5.17*)

(3.62 )
(0.00 )

Brain

2.0

(1.99 )

0.0

Breast

23.5*

(6.55*)

9.9*

19.6

3.0

3.7

0.0

*
*

(0.00 )

(3.62*)

38
(continued)

Cervix (Females Only)
Colon
Esophagus
Gallbladder
Kidney

21.7

5.5
3.8

*
*

0.0

3.2

(7.40 )
*

(2.57 )
*

(2.39 )
*

(0.00 )
*

(2.22 )

5
3
0
2

56
56
56
56

Larynx/Windpipe

0.0

(0.00 )

0.0

Leukemia

2.2*

(1.79*)

3.9*

0.0

12.6

6.6

Liver
Lung
Lymphoma
See notes at end of table.

0.0
3.5

*
*

9.8

(0.00 )
*

(2.52 )
*

(5.04 )

0
2
4

H-38

56
56

(9.76 )
(2.20 )
(3.62 )
(0.00 )
(0.00 )

(0.00 )
(7.25 )
(5.27 )

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Melanoma
Mouth/Tongue/Lip
Ovary (Females Only)
Pancreas
Prostate (Males Only)
Rectum
Skin (Not Melanoma)

*
*

3.1

3.8

12.7

0.0

18.1

(5.05*)
*

(0.00 )
*

(3.03 )
*

(3.70 )
*

(8.14 )
*

(0.00 )
*

(5.49 )

0
1
1
2
0
8

20.4*

(8.26*)

3.7

(3.62*)

8.2

0.0

15.4

3.7

33.0

11.5

(7.69 )

(1.02*)

35
56
21
56
56

(6.68 )
(0.00 )
(9.19 )
(3.62 )
(9.12 )

Skin (Don't Know Which Kind)

4.8

(4.53 )

Soft Tissue (Muscle or Fat)

0.0*

(0.00*)

1.2*

3.7

0.0

4.2

(3.02 )

21.8

(5.31 )

Stomach
Testis (Males Only)
Throat/Pharynx
Thyroid
Uterus (Females Only)
Other
How old were you when your bladder
cancer was first diagnosed?
(HLTH27)7
How old were you when your blood
cancer was first diagnosed?
(HLTH28a)7
How old were you when your bone
cancer was first diagnosed?
(HLTH28b)7
How old were you when your brain
cancer was first diagnosed?
(HLTH28c)7
How old were you when your breast
cancer was first diagnosed?
(HLTH28d)7
How old were you when your
cervical cancer was first
diagnosed? (HLTH28e)7
How old were you when your colon
cancer was first diagnosed?
(HLTH28f)7
How old were you when your
esophageal cancer was first
diagnosed? (HLTH28g)7
How old were you when your
gallbladder cancer was first
diagnosed? (HLTH28h)7
How old were you when your
kidney cancer was first
diagnosed? (HLTH28i)7
How old were you when your
larynx/windpipe cancer was first
diagnosed? (HLTH28j)7
How old were you when your
leukemia was first diagnosed?
(HLTH28k)7
See notes at end of table.

0.0

0.0
2.8

*
*

5.7

(0.00 )
*

(2.14 )
*

(5.53 )

0
0
3
1

21
56
56
35

(3.62 )
(0.00 )
(0.00 )
(11.07 )

(2.51 )

5.5

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

13.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

50.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

51.2

(3.53)

N/A

46.2*

(3.63*)

N/A

35.1*

(4.02*)

N/A

29.9*

(2.20*)

N/A

51.1*

(5.49*)

N/A

45.4*

(8.35*)

N/A

63.4*

(9.11*)

N/A

62.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

44.8*

(6.58*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

28.6a*

(7.09*)

N/A

13.2*

(0.28*)

N/A

2
(continued)

3.6

H-39

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
How old were you when your liver
cancer was first diagnosed?
(HLTH28l)7
How old were you when your lung
cancer was first diagnosed?
(HLTH28m)7
How old were you when your
lymphoma was first diagnosed?
(HLTH28n)7
How old were you when your
melanoma was first diagnosed?
(HLTH28o)7
How old were you when your
mouth/tongue/lip cancer was first
diagnosed? (HLTH28p)7
How old were you when your
ovarian cancer was first
diagnosed? (HLTH28q)7
How old were you when your
pancreatic cancer was first
diagnosed? (HLTH28r)7
How old were you when your
prostate cancer was first
diagnosed? (HLTH28s)7
How old were you when your
rectum cancer was first
diagnosed? (HLTH28t)7
How old were you when your skin
[not melanoma] cancer was first
diagnosed? (HLTH28u)7
How old were you when your skin
cancer was first diagnosed?
(HLTH28v)7
How old were you when your soft tissue
cancer was first diagnosed?
(HLTH28w)7
How old were you when your stomach
cancer was first diagnosed?
(HLTH28x)7
How old were you when your testis
cancer was first diagnosed?
(HLTH28y)7
How old were you when your
throat/pharynx cancer was first
diagnosed? (HLTH28z)7
How old were you when your
thyroid cancer was first
diagnosed? (HLTH28aa)7
How old were you when your
uterine cancer was first
diagnosed? (HLTH28bb)7
How old were you when the type of
cancer listed below was first
diagnosed? (HLTH28cc)7
Did you have cancer during the past
12 months? (HLTH29)5
How old were you when your heart
condition or heart disease was first
diagnosed? (HLTH30)7,8
See notes at end of table.

(0.00*)

N/A

0.0*

(0.00*)

N/A

58.7*

(10.48*)

N/A

63.2*

(2.61*)

N/A

56.0a*

(5.42*)

N/A

78.6*

(3.82*)

N/A

38.0*

(4.13*)

N/A

39.7*

(7.51*)

N/A

0.0*

(0.00*)

N/A

67.0*

(0.00*)

N/A

59.0a*

(0.00*)

N/A

36.5*

(9.88*)

N/A

64.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

64.5*

(0.35*)

N/A

70.3*

(3.15*)

N/A

0.0*

(0.00*)

N/A

62.0*

(0.00*)

N/A

46.0a*

(0.00*)

N/A

73.7*

(7.38*)

N/A

54.5*

(2.99*)

N/A

47.0*

(4.83*)

N/A

0.0*

(0.00*)

N/A

13.6*

(0.76*)

N/A

0.0*

(0.00*)

N/A

32.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

0.0*

(0.00*)

N/A

35.6*

(2.48*)

N/A

43.4*

(7.19*)

N/A

40.0*

(0.00*)

N/A

57.7*

(13.64*)

N/A

47.7a*

(10.47*)

N/A

79.0*

(0.00*)

N/A

30.6*

(8.17*)

32.0*

(10.04*)

44.1

(1.92)

N/A

116

48.9

(3.71)

N/A

73
(continued)

H-40

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Did you have any kind of heart
condition or heart disease in the
past 12 months? (HLTH31)5
How old were you when your
diabetes or sugar diabetes was first
diagnosed? (HLTH32)7,8
How old were you when your
chronic bronchitis, emphysema, or
chronic obstructive pulmonary
disease, also called COPD were
first diagnosed? (HLTH33)7
How old were you when your
cirrhosis of the liver was first
diagnosed? (HLTH34)7
How old were you when your
hepatitis was first diagnosed?
(HLTH35)7
How old were you when your
kidney disease was first
diagnosed? (HLTH36)7
How old were you when your
asthma was first diagnosed?
(HLTH37)7
Do you still have asthma?
(HLTH38)5
How old were you when you found out
you had HIV/AIDS? (HLTH39)7
Are you currently taking
prescription medicine for your
high blood pressure? (HLTH40)5
How old were you when your high
blood pressure was first
diagnosed? (HLTH41)7
How many times in the past 12
months have you moved?
(QD13)7,8
Were you born in the United States?
(QD14)
Have you lived in the United States
for at least one year? (QD16a)5
For how many years have you lived
in the United States? (QD16b)7
For how many months have you
lived in the United States?
(QD16c)7
Are you now attending or are you
currently enrolled in school?
(QD17)
What grade or year of school are
you now attending? (QD18)5

(5.73)

116

58.4*

(7.90*)

44.1

(1.74)

N/A

39.6

(2.50)

N/A

36.1a

(3.32)

N/A

45.7

(3.24)

N/A

52.0*

(0.00*)

N/A

42.0*

(0.00*)

N/A

29.1

(4.32)

N/A

38.7

(3.43)

N/A

41.9

(4.89)

N/A

44.4

(6.05)

N/A

19.4

(1.90)

N/A

199

17.1

(2.43)

N/A

148

65.4

(4.29)

149

210

56.4

(5.45)

159

(0.00*)

N/A

(0.00*)

N/A

87.4

(2.37)

142

181

90.2

(2.87)

109

132

45.2a

(1.07)

N/A

137

49.1

(1.46)

N/A

105

0.4

(0.03)

N/A

1,674

(0.65*)

N/A

1,301

90.4

(1.37)

1,553

1,691

88.9

(2.00)

1,178

1,319

95.4

(2.06)

131

138

98.3

(0.99)

134

141

24.8

(1.74)

N/A

131

22.9

(2.66)

N/A

132

(2.58*)

N/A

(2.32*)

N/A

17.5

(1.10)

634

1,691

(1.31)

446

1,316

1st Grade

0.4

(0.28)

633

0.0*

(0.00*)

445

2nd Grade

0.0*

(0.00*)

633

0.2

(0.16)

445

3rd Grade

0.0*

(0.00*)

633

0.0*

(0.00*)

4th Grade

633

0.0

*
*

5th Grade
6th Grade
See notes at end of table.

0.0*

7.6*

0.0

(0.00 )

1.0*

1.3*

3.4*
17.1

0.0

(0.00 )

633

0.0

1.1

(0.47)

633

0.8

H-41

445

(0.00 )

445

(0.40)

445
(continued)

(0.00 )

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item

7th Grade

7.9

(1.06)

633

5.8

(1.12)

445

8th Grade

9.0

(1.19)

633

8.2

(2.36)

445

9th Grade

9.1

(1.31)

633

7.8

(1.44)

445

10th Grade

8.4

(1.02)

633

15.2

(2.30)

445

11th Grade

7.9

(1.02)

633

5.8

(1.21)

445

12th Grade

9.2

(1.14)

633

7.7

(1.30)

445

(1.74)

633

5.8

(1.26)

445

College or University/1st Year

12.6

College or University/2nd Year

9.3

(1.57)

633

13.7

(2.69)

445

College or University/3rd Year

8.3

(1.52)

633

9.6

(2.35)

445

College or University/4th Year
College or University/5th Year or
Higher
Are you a full-time student or a part
time student? (QD19)5

5.6

(1.45)

633

8.1

(2.11)

445

11.2

(2.16)

633

11.4

(2.81)

445

80.7

(2.39)

550

628

80.9

(2.78)

374

441

19.3

(2.39)

628

19.1

(2.78)

441

0.7

(0.18)

N/A

465

0.7

(0.13)

N/A

332

0.4

(0.08)

N/A

468

0.4

(0.10)

N/A

333

52.0

(2.12)

557

1,482

47.0

(2.35)

426

1,191

5.2

(0.89)

1,480

7.4

(1.36)

1,189

14.3

(1.28)

157

1,480

14.8

(1.51)

129

1,190

28.8

(1.55)

734

1,485

31.0

(2.17)

590

1,191

1.4

(0.03)

N/A

749

1.3

(0.05)

N/A

600

6.0

(0.73)

116

1,674

8.1

(1.15)

124

1,307

6.0*

(2.84*)

10.0*

(4.12*)

120

1.1

(0.73)

120

2.1

(1.07)

120

5.2

(2.05)

120

Full-Time
Part-Time
During the past 30 days, how many
whole days of school did you miss
because you were sick or injured?
(QD20)7,8
During the past 30 days, how many
whole days of school did you miss
because you skipped or "cut" or
just didn't want to be there?
(QD21)7,8
Are you now married, widowed,
divorced or separated, or have you
never married? (QD07)5
Married
Widowed
Divorced or Separated
Have Never Married
How many times have you been
married? (QD08)7
Is anyone in your immediate family
currently serving in the United
States military? (QD10d)11
Which member or members of your
immediate family are currently in
the United States military? (QD10e)5,11
My spouse
Unmarried partner
My mother
My father
My son or sons
My daughter or daughters
My brother or brothers
My sister or sisters
See notes at end of table.

3.3

1.3

3.2

35.8

(1.93 )
(0.76 )
(1.39 )
*

(7.02 )
*

16.6

*
*

(4.58 )
*

4.1

(3.05 )

0.0

(0.00 )

120

48.2a*

(6.48*)

23.9*

(5.44*)

120

0.8*

(0.57*)

7.1*

(3.52*)

120
(continued)

H-42

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Another member of my immediate
family
Did you work at a job or business at
any time last week? (QD26)5
Even though you did not work at
any time last week, did you have
a job or business? (QD27)5
How many hours did you work last
week at all jobs or businesses?
(QD28)7
Do you usually work 35 hours or
more per week at all jobs or
businesses? (QD29)5
Which one of these reasons best
describes why you did not work
last week? (QD30)5
Vacation/Sick/Furlough/Strike/
Other Temporary
Absence/Maternity Leave
Layoff, Not Looking for Work
Layoff, Looking for Work
Waiting to Report to New Job
Self-Employed, No Business
Last Week
Going to School/Training
Some Other Reason
Which one of these reasons best
describes why you did not have a
job or business last week?
(QD31)5
Looking for Work
On Layoff, Not Looking for
Work
Keeping House/Caring for
Children Full Time
Going to School/Training

(N/A)

N/A

44.9*

(6.07*)

120

58.9

(1.89)

858

1,484

55.4

(2.28)

646

1,186

11.9

(1.76)

623

11.7

(1.91)

540

38.6

(0.55)

N/A

854

38.5

(0.78)

N/A

645

76.6

(1.69)

669

940

75.9

(2.36)

518

724

35.1*

(6.41*)

31.6*

(7.35*)

(0.00 )

(2.96*)

4.0

(2.41 )

0.0

10.8*

(4.79*)

6.2*
*

(2.52 )

(1.64*)

4.1

(2.02 )

4.5

15.7*

(5.98*)

4.8*
*

9.5

(3.52 )

9.6

(3.22 )

20.8a*

(6.18*)

43.3*

(8.10*)

15.9

(1.94)

130

537

13.6

(2.02)

101

460

1.4

(0.44)

537

2.8

(1.28)

460

11.5

(1.96)

537

10.5

(1.74)

460

9.1

(1.09)

122

537

10.0

(1.62)

460

Retired

39.7

(3.07)

100

537

35.6

(4.00)

460

Disabled

15.3

(2.11)

537

14.9

(2.88)

460

1.9

(0.50)

537

2.9

(0.87)

460

5.1

(1.03)

537

9.7

(2.33)

460

81.9

(4.05)

101

130

86.4*

(4.35*)

100

18.1

(2.17)

135

540

16.2

(2.42)

104

460

1.4

(0.05)

N/A

1,066

1.4

(0.07)

N/A

820
(continued)

Didn't Want A Job
Some Other Reason
During the past 30 days, did you
make specific efforts to find
work? (QD32)5
Did you work at a job or business at
any time during the past 12
months? (QD33)5
How many different employers have
you had in the past 12 months?
(QD35 and QD36)7
See notes at end of table.

H-43

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 12 months, was
there ever a time when you did not
have at least one job or business?
(QD37)5
In how many weeks during the past
12 months did you not have at
least one job or business?
(QD38)7
During the past 30 days, how many
whole days of work did you miss
because you were sick or injured?
(QD40)7,8
During the past 30 days, how many
whole days of work did you miss
because you just didn't want to be
there? (QD41)7,8
Thinking about the location where
you work, how many people work
for your employer out of this
office, store, etc.? (QD42)5

14.7

(1.48)

197

939

N/A

(N/A)

N/A

14.8

(1.21)

N/A

185

16.7

(2.48)

N/A

137

0.7

(0.13)

N/A

931

0.7

(0.21)

N/A

717

0.2

(0.03)

N/A

932

0.2

(0.03)

N/A

719

Less Than 10 People

30.6

(2.01)

278

929

N/A

(N/A)

N/A

10 to 24 People

18.0

(1.46)

193

929

N/A

(N/A)

N/A

25 to 99 People

18.5

(1.44)

186

929

N/A

(N/A)

N/A

100 to 499 People

18.8

(1.78)

163

929

N/A

(N/A)

N/A

(0.00 )

500 People or More
At your workplace, is there a written
policy about employee use of
alcohol or drugs? (QD43)5
Does this policy cover only alcohol,
only drugs, or both alcohol and
drugs? (QD44)5

14.1

(1.78)

109

929

0.0

79.7

(1.75)

707

907

75.6

(2.16)

520

703

(0.53)

702

0.0

(0.03)

517

2.2

(0.53)

702

4.8

(1.14)

517

96.8

(0.75)

675

702

95.2

(1.14)

490

517

Yes

32.7

(2.07)

273

934

29.9

(2.41)

192

721

49.9

(2.17)

490

934

51.7

(2.41)

392

721

17.4

(1.45)

171

934

18.4

(1.79)

137

721

54.7

(2.07)

426

870

52.0

(2.77)

318

679
(continued)

Only Alcohol
Only Drugs
Both Alcohol and Drugs
At your workplace, have you ever
been given any educational
information regarding the use of
alcohol or drugs? (QD45)5

Don't Remember
Through your workplace, is there
access to any type of employee
assistance program or other type
of counseling program for
employees who have alcohol or
drug-related problems? (QD46)5
See notes at end of table.

1.0
a

H-44

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Does your workplace ever test its
employees for alcohol use?
(QD47)5
Does your workplace ever test its
employees for drug use? (QD48)5
Does your workplace test its
employees for drug or alcohol use
as part of the hiring process?
(QD49)5
Does your workplace test its
employees for drug or alcohol use
on a random basis? (QD50)5
According to the policy at your
workplace, what happens to an
employee the first time he or she
tests positive for illicit drugs?
(QD51)5
Handled On Individual
Basis/Policy Does Not Specify
What Happens
Employee is Fired
Employee Referred for
Treatment/Counseling
Nothing Happens

(1.68)

268

905

26.9

(2.24)

187

700

47.6

(2.19)

432

914

47.5

(3.01)

310

699

86.9

(1.95)

365

429

85.4

(2.64)

274

314

58.1

(3.44)

256

420

61.3

(3.76)

181

308

24.6

(2.77)

388

32.7

(4.03)

283

45.6

(2.93)

191

388

41.2

(4.45)

130

283

24.6

(2.46)

388

22.8

(3.81)

283

(0.56 )

283

1.8

(0.97)

388

0.6

3.3

(1.08)

388

2.7

(1.30)

283

More Likely

48.1

(1.98)

430

937

43.6

(3.26)

307

721

Less Likely

7.5

(0.91)

937

7.1

(1.39)

721

44.3

(1.70)

429

937

49.3

(3.10)

352

721

42.9

(1.94)

381

938

37.7

(3.24)

257

720

Something Else Happens
Would you be more or less likely to
want to work for an employer that
tests its employees for drug use as
part of the hiring process?
(QD52)5

Would Make No Difference
Would you be more or less likely to
want to work for an employer that
tests its employees for drug or
alcohol use on a random basis?
(QD53)5
More Likely
Less Likely
Would Make No Difference
How well do you speak English?
(QD55)
Very well

11.5

(1.30)

117

938

10.7

(1.61)

720

45.5

(1.92)

440

938

51.6

(3.23)

369

720

92.0a

(0.94)

1,583

1,691

88.2

(1.36)

1,194

1,315

Well

7.8

(0.96)

101

1,691

9.4

(1.20)

101

1,315

Not well

0.2a

(0.07)

1,691

2.4

(0.77)

1,315

0.0

(0.04)

1,691

0.0

(0.02)

1,315
(continued)

Not at all
See notes at end of table.

H-45

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Are you deaf or do you have serious
difficulty hearing? (QD56)
Are you blind or do you have
serious difficulty seeing, even
when wearing glasses? (QD57)
Because of a physical, mental or
emotional condition, do you have
serious difficulty concentrating,
remembering, or making
decisions? (QD58)
Do you have serious difficulty
walking or climbing stairs?
(QD59)
Do you have difficulty dressing or
bathing? (QD60)
Because of a physical, mental or
emotional condition, do you have
difficulty doing errands alone such
as visiting a doctors' office or
shopping? (QD61)5
Covered by Medicare? (QHI01)
Covered by Medicaid/[CHIPFILL]
(QHI02 and QHI02a)
Covered by TRICARE, CHAMPUS
CHAMPVA, VA, Military Health
Care (QHI03)
Covered by Private Health
Insurance (QHI06)12
Was [MEMBER] private health
insurance obtained through work,
such as through an employer,
union, or professional association?
(QHI07)5
Does [MEMBER] private health
insurance include coverage for
treatment for alcohol abuse or
alcoholism? (QHI08)5
Does [MEMBER] private health
insurance include coverage for
treatment for drug abuse?
(QHI09)5
Does [MEMBER] private health
insurance include coverage for
treatment for mental or emotional
problems? (QHI10)5
[MEMBER] currently covered by
any kind of health insurance,
including Indian Health
Insurance? (QHI11)5
Any Health Insurance Coverage
(Recode)
See notes at end of table.

(0.64)

1,690

4.7

(0.79)

1,317

3.5

(0.65)

1,689

3.6

(0.73)

1,317

6.8

(0.72)

143

1,687

9.0

(1.14)

126

1,312

6.8a

(0.99)

1,690

10.5

(1.52)

1,316

1.7

(0.40)

1,691

2.8

(0.66)

1,315

4.3

(0.73)

1,483

4.8

(0.88)

1,188

19.6

(1.69)

170

1,692

22.0

(2.26)

150

1,320

12.9

(1.25)

310

1,692

13.8

(1.43)

242

1,320

5.2

(0.81)

1,692

3.8

(0.91)

1,320

63.4

(2.01)

1,009

1,692

60.0

(2.29)

768

1,320

88.2

(1.60)

912

994

86.0

(2.21)

670

759

73.9

(2.21)

514

714

71.0

(3.32)

380

556

72.6

(2.29)

503

709

68.4

(3.48)

367

547

84.8

(1.78)

693

816

83.9

(2.18)

509

629

24.6

(3.18)

306

22.2

(4.08)

258

87.1

(1.06)

1,451

1,692

86.4

(1.62)

1,119

1,320
(continued)

H-46

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
During the past 12 months, was
there any time when [MEMBER]
did not have any kind of health
insurance or coverage? (QHI13)5
During the past 12 months, about
How many months without any
kind of health insurance or
coverage? (QHI14)7
About how long has it been since
[MEMBER] last had any kind of
health care coverage? (QHI15)5
Within the Past 6 Months
More Than 6 Months Ago but
Within the Past Year
More Than 1 Year Ago but
Within the Past 3 Years
More Than 3 Years Ago
Never Had Coverage
Which of these reasons is the main
reason why [MEMBER] stopped
being covered by health
insurance? (QHI17)5
Person in Family with Health
Insurance Lost Job/Changed
Employer
Lost Medicaid Coverage Because
of New Job/Increase in Income
Lost Medicaid Coverage For
Some Other Reason
Cost Is Too High/Can't Afford
Premiums
Became Ineligible Because of
Age/Leaving School
Employer Does Not Offer
Coverage or Not Eligible For
Coverage
Divorced/Separated From Person
With Insurance
Death of Spouse/Parent
Insurance Company Refused
Coverage
Don't Need It
Received Medicaid/Insurance
Only While Pregnant
Some Other Reason
See notes at end of table.

7.0

(0.81)

123

1,430

6.1

4.3

(0.44)

N/A

122

3.8

(0.99)

(0.45)

1,105

N/A

17.2

(2.88)

231

18.3

(3.50)

197

8.4

(1.90)

231

7.5

(2.10)

197

22.6

(3.42)

231

27.4

(4.89)

197

37.5

(3.75)

231

26.4

(4.58)

197

14.4

(2.71)

231

20.4

(3.60)

197

29.3

(4.62)

193

23.0

(5.04)

155

6.1

(1.57)

193

8.6*

(3.46*)

155

4.7

(1.56)

193

6.7

(2.26)

155

27.6

(4.05)

193

19.6

(4.48)

155

10.0

(2.32)

193

10.1

(2.46)

155

4.0

(1.26)

193

7.2

(3.02)

155

(0.81)

193

1.8

(1.02)

155

1.4
*

193

0.0

(0.00 )

155

1.3*

(1.08*)

193

3.4*

(2.94*)

155

2.3

(1.25)

193

0.7

(0.50)

155

2.7

(1.11)

193

3.8

(1.72)

155

10.8

(2.72)

193

15.1

(4.12)

155
(continued)

H-47

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
Which of these reasons describe
why [SAMPLE MEMBER] never
had health insurance coverage?
(QHI18)5
Cost Too High/ Can't Afford
Premiums
Employer Does Not Offer
Coverage or Not Eligible For
Coverage
Insurance Company Refused
Coverage
Don't Need It
Some Other Reason
In [YEAR], did you receive Social
Security or Railroad Retirement
payments? (QI01N)
In [YEAR], did you receive
Supplemental Security Income or
SSI? (QI03N)13
In [YEAR], did you receive income
from wages or pay earned while
working at a job or business?
(QI05N)
In [YEAR], did you receive food
stamps? (QI07N)13
At any time during [YEAR], did
you receive any cash assistance
from a state or county welfare
program such as [TANFFILL]?
(QI08N)
In [YEAR], because of low income,
did you receive any other kind of
non-monetary welfare or public
assistance? (QI10N)
For how many months in [YEAR]
did you or your
[RELATIONSHIP] receive any
type of welfare or public
assistance, not including food
stamps? (QI12AN and QI12BN)7
Before taxes and other deductions,
was your total personal income
from all sources during [YEAR]
more or less than 20,000 dollars?
(QI20N)
$20,000 or More

47.9*

(8.67*)

62.0*

(8.68*)

6.8*

(4.06*)

22.7*

(8.43*)

0.0*

(0.00*)

0.0*

(0.00*)

16.8

*
*

(7.92 )

25.6

(2.40)

233

1,320

1,692

8.5

(1.16)

112

1,320

1,184

1,692

N/A

(N/A)

N/A

(1.63)

369

1,692

18.1

(1.67)

265

1,320

3.7

(0.62)

1,692

2.7

(0.57)

1,320

3.5

(0.58)

1,692

2.6

(0.60)

1,320

6.0a

(0.56)

N/A

136

9.3

(0.47)

N/A

1,276

(6.63 )

30.9

(11.88 )

17.9

27.6

(1.87)

307

1,692

8.9

(1.02)

149

68.7

(1.92)

17.1

14.4

(6.76 )

56.7

(1.71)

663

1,640

51.4

(2.51)

518

Less Than $20,000
Of these income groups, which
category best represents
[MEMBER] total personal income
during [YEAR]? (QI21A and QI21B)

43.2

(1.70)

973

1,640

48.3

(2.49)

754

Less Than $1,000
See notes at end of table.

13.9

1,276

(0.86)

437

H-48

1,580

12.7

(1.25)

314

1,223
(continued)

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
$6,000-$6,999
$7,000-$7,999
$8,000-$8,999
$9,000-$9,999
$10,000-$10,999
$11,000-$11,999
$12,000-$12,999
$13,000-$13,999
$14,000-$14,999
$15,000-$15,999
$16,000-$16,999
$17,000-$17,999
$18,000-$18,999
$19,000-$19,999
$20,000-$24,999
$25,000-$29,999
$30,000-$34,999
$35,000-$39,999
$40,000-$44,999
$45,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000-$149,999
$150,000 or More
Before taxes and other deductions,
was the total combined family
income during [YEAR] more or
less than 20,000 dollars? (QI22)
$20,000 or More
Less Than $20,000
Of these income groups, which
category best represents your total
combined family income during
[YEAR]? (QI23A and QI23B)
Less Than $1,000
$1,000-$1,999
$2,000-$2,999
$3,000-$3,999
$4,000-$4,999
$5,000-$5,999
See notes at end of table.

2012 QFT 2012 QFT 2012 QFT
2013 DR
2013 DR
2013 DR
2012 QFT Standard Unweighted Unweighted 2013 DR
Standard Unweighted Unweighted
Estimate1,2,3 Error
Total
Sample Size Estimate1,3,4
Error
Total
Sample Size
2.8
(0.42)
70
1,580
3.0
(0.66)
52
1,223
1.2
(0.25)
37
1,580
1.8
(0.45)
26
1,223
1.5
(0.31)
29
1,580
0.9
(0.26)
15
1,223
1.2
(0.31)
26
1,580
1.2
(0.29)
21
1,223
1.0
(0.26)
23
1,580
0.8
(0.30)
13
1,223
0.9
(0.29)
14
1,580
0.5
(0.18)
9
1,223
0.5
(0.21)
8
1,580
1.9
(0.75)
18
1,223
1.4
(0.35)
23
1,580
3.0
(0.82)
24
1,223
2.6
(0.56)
38
1,580
2.0
(0.64)
25
1,223
2.5
(0.48)
39
1,580
2.8
(0.61)
33
1,223
1.4
(0.38)
19
1,580
2.2
(0.69)
23
1,223
1.4
(0.39)
20
1,580
1.1
(0.37)
17
1,223
1.2
(0.38)
16
1,580
1.4
(0.60)
12
1,223
1.1
(0.30)
16
1,580
1.6
(0.61)
14
1,223
1.8
(0.39)
29
1,580
2.1
(0.64)
26
1,223
1.5
(0.34)
22
1,580
1.1
(0.37)
13
1,223
1.9
(0.45)
24
1,580
3.0
(0.61)
22
1,223
1.7
(0.40)
22
1,580
1.9
(0.55)
21
1,223
1.7
(0.39)
26
1,580
3.3
(0.93)
27
1,223
8.4
(0.92)
115
1,580
8.2
(1.41)
78
1,223
5.4
(0.74)
69
1,580
6.6
(0.90)
58
1,223
4.8
(0.76)
66
1,580
6.7
(1.18)
48
1,223
5.7
(0.84)
58
1,580
4.2
(0.75)
41
1,223
5.0
(0.87)
56
1,580
2.8
(0.75)
27
1,223
5.4
(0.84)
52
1,580
3.9
(0.61)
39
1,223
10.7
(1.09)
113
1,580
9.8
(1.16)
97
1,223
4.5
(0.79)
51
1,580
3.9
(0.72)
53
1,223
4.1
(0.94)
43
1,580
3.4
(0.91)
34
1,223
2.9
(0.98)
19
1,580
2.3
(0.78)
23
1,223

79.4a
20.6a

(1.66)
(1.66)

1,249
443

1,692
1,692

72.9
27.1

(2.14)
(2.14)

961
359

1,320
1,320

2.4
1.1
0.6
0.9
0.6
0.4

(0.44)
(0.30)
(0.17)
(0.26)
(0.20)
(0.17)

71
24
21
19
16
11

1,692
1,692
1,692
1,692
1,692
1,692

2.8
1.9
0.7
0.4
0.7
0.4

(0.63)
(0.48)
(0.24)
(0.20)
(0.22)
(0.17)

55
37
13
9
9
6

1,320
1,320
1,320
1,320
1,320
1,320
(continued)

H-49

Table H.1 Estimates and Standard Errors for New, Moved, or Revised Items in the 2012 Questionnaire Field Test
and 2013 Dress Rehearsal among Persons Aged 12 or Older (continued)

Instrument Item

$6,000-$6,999

0.6

(0.26)

1,692

0.2

(0.14)

1,320

$7,000-$7,999

0.2

(0.09)

1,692

0.9

(0.41)

1,320

$8,000-$8,999

0.6

(0.23)

1,692

1.5

(0.54)

1,320

$9,000-$9,999

0.7

(0.18)

1,692

1.1

(0.51)

1,320

$10,000-$10,999

1.1

(0.28)

1,692

1.9

(0.47)

1,320

$11,000-$11,999

0.6

(0.20)

1,692

2.0

(0.66)

1,320

$12,000-$12,999

0.8

(0.18)

1,692

0.9

(0.34)

1,320

$13,000-$13,999

0.7

(0.31)

1,692

1.9

(0.72)

1,320

$14,000-$14,999

1.6

(0.38)

1,692

0.8

(0.34)

1,320

$15,000-$15,999

1.0

(0.25)

1,692

1.3

(0.42)

1,320

$16,000-$16,999

0.7

(0.19)

1,692

0.6

(0.26)

1,320

$17,000-$17,999

1.8

(0.41)

1,692

1.5

(0.58)

1,320

$18,000-$18,999

1.2

(0.26)

1,692

1.7

(0.52)

1,320

$19,000-$19,999

2.1

(0.46)

1,692

3.5

(0.86)

1,320

$20,000-$24,999

7.6

(0.91)

127

1,692

7.1

(1.06)

100

1,320

$25,000-$29,999

3.8

(0.51)

1,692

5.0

(1.03)

1,320

$30,000-$34,999

4.8

(0.70)

1,692

6.1

(1.03)

1,320

$35,000-$39,999

5.1

(0.80)

1,692

5.3

(0.98)

1,320

$40,000-$44,999

6.2

(1.08)

1,692

4.3

(0.72)

1,320

$45,000-$49,999

4.7

(0.67)

1,692

3.8

(0.70)

1,320

$50,000-$74,999

16.8

(1.34)

242

1,692

15.2

(1.44)

192

1,320

$75,000-$99,999

11.1

(0.86)

178

1,692

11.1

(1.67)

149

1,320

$100,000-$149,999

12.6

(1.36)

202

1,692

9.7

(1.33)

155

1,320

7.9

(1.21)

110

1,692

5.8

(1.27)

1,320

65.8a

(1.79)

982

1,683

54.5

(2.58)

658

1,312

92.9

(0.87)

1,597

1,688

91.9

(1.21)

1,226

1,310

$150,000 or More
Is there at least one telephone at this
address that is not a cell phone?
(CELL1)
Do you or anyone at this address
have a working cell phone?
(CELL2)
*

Low precision; estimate would be suppressed due to not meeting the NSDUH suppression rule.
ACASI = audio computer-assisted self-interviewing; AMT = CHAMPUS = Civilian Health and Medical Program of the Uniformed Services; CHAMPVA =
Civilian Health and Medical Program of the Veterans Administration; CAPI = computer-assisted personal interviewing; CHAMPUS = Civilian Health and
Medical Program of the Uniformed Services; CHAMPVA = Civilian Health and Medical Program of the Veterans Administration; DMT = dimethyltryptamine; DR = 2013 Dress Rehearsal; GED = general equivalency diploma; N/A = not applicable; QFT = 2012 Questionnaire Field Test; R = respondent.
a
Difference between estimate and 2013 DR estimate is statistically significant at the 0.05 level.
1
Sample does not include Alaska or Hawaii and does not include Spanish-language interviews. Hispanic respondents who completed the interview in English
also have been excluded for these comparisons.
2
QFT data collected from September 1 through November 3, 2012.
3
Estimates are percentages of all persons aged 12 or older, except where noted.
4
DR data collected from September 1 through October 31, 2013.
5
Estimated percentage is based on respondents who were asked the question and exclude respondents with unknown or missing data.
6
Consistency check questions were revised to be consistent with the categories on educational attainment in the DR questionnaire.
7
Estimate is an average based on valid responses to the relevant question(s). Respondents with unknown or missing data were excluded.
8
The estimated mean includes zeroes.
9
The ranges for height in feet and inches were edited for accuracy in the DR questionnaire.
10
Includes pre-pregnancy weight of pregnant females as reported in HLTH13 and HLTH14.The upper weight limit was increased in the DR questionnaire.
11
The definition of "immediate family" was moved from the "Help" screen to the question text, minor wording changes were made to these questions for
clarity, and an "Other, Specify" item was to this series of questions in the DR questionnaire.
12
The "Help" instructions were removed and key terms were moving into the question itself in the DR questionnaire.
13
The wording was edited for accuracy in the DR questionnaire.
Source: SAMHSA, Center for Behavior Health Statistics and Quality, National Survey on Drug Use and Health, 2012 and 2013.

H-50

Appendix I: Notes on Analysis Variables for the Dress
Rehearsal

1. Key Illicit Drug Measures in Chapter 6 Tables
Measure
Use of Any Illicit Drug, Standard
Definition

Use of Any Illicit Drug, Alternate
Definition 1

Use of Any Illicit Drug, Alternate
Definition 2
Use of Any Illicit Drug, Alternate
Definition 3

Use of Illicit Drugs Other Than
Marijuana, Standard Definition

•
•
•
•
•
•
•

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Substances Included
Marijuana
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants2
Methamphetamine3
Prescription Drugs3
– Pain Relievers
– Tranquilizers
– Stimulants3
– Sedatives
Marijuana
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants
Marijuana
Cocaine (including crack)
Heroin
Marijuana
Cocaine (including crack)
Heroin
Methamphetamine3
Cocaine (including crack)
Heroin
Hallucinogens1
Inhalants2
Methamphetamine3
Prescription Drugs3
– Pain Relievers
– Tranquilizers
– Stimulants3
– Sedatives

See notes at end of table.

(continued)

I-1

1. Key Illicit Drug Measures in Chapter 6 Tables
Measure
Use of Illicit Drugs Other Than
Marijuana, Alternate Definition 1

Use of Illicit Drugs Other Than
Marijuana, Alternate Definition 2

•
•
•
•
•
•
•
•
•

Substances Included
Cocaine
Heroin
Hallucinogens1
Inhalants2
Cocaine
Heroin
Hallucinogens1
Inhalants2
Methamphetamine3

For the 2012 and 2013 comparison data, estimates are based on the use of any of the following hallucinogens: LSD,
also called "acid"; PCP, also called "angel dust" or phencyclidine; peyote; mescaline; psilocybin; or "Ecstasy," also
called MDMA; or any other hallucinogen. DR estimates are based on the use of any of the hallucinogens from the
2012 and 2013 comparison data, plus the following: ketamine, also called "Special K" or "Super K"; DMT, AMT,
or 5-MeO-DIPT ("Foxy"); or Salvia divinorum.
2
Lifetime estimates of inhalant use for the 2012 and 2013 comparison data are based on the use of any of the
following: amyl nitrite, "poppers," locker room odorizers, or "rush"; correction fluid, degreaser, or cleaning fluid;
gasoline or lighter fluid; glue, shoe polish, or toluene; halothane, ether, or other anesthetics; lacquer thinner or
other paint solvents; lighter gases, such as butane or propane; nitrous oxide or "whippits"; spray paints; other
aerosol sprays; or any other inhalant. DR estimates of lifetime use of inhalants are based on the use of any of the
inhalants from the 2012 and 2013 comparison data, plus the following: felt-tip pens, felt-tip markers, or magic
markers; and computer cleaner, also known as air duster.
3
Estimates of any prescription drug misuse, stimulant misuse, and methamphetamine use for the 2011 and 2012
comparison data include data from the new methamphetamine items added in 2005 and 2006 (i.e., core plus
noncore data). Estimates of stimulant misuse for the DR vary according to whether they include data from the
separate core methamphetamine module.

2. Stimulant Misuse:
•

The standard definition for the 2012 and 2013 comparison data and the DR includes
use of methamphetamine and misuse of prescription stimulants. Estimates for the
2012 and 2013 comparison data also include data from the new methamphetamine
items added in 2005 and 2006 (i.e., core plus noncore data).

•

The DR definition includes data only for misuse of prescription stimulants. A
corresponding measure is not available for the 2012 and 2013 comparison data.

3. Binge Alcohol Use – For the 2012 and 2013 comparison data, binge alcohol use is defined
for both males and females as drinking at least five or more drinks on the same occasion
(i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past
30 days. For the DR, binge alcohol use is defined for males as drinking five or more drinks
on the same occasion and for females as drinking four or more drinks on the same occasion
on at least 1 day in the past 30 days. Estimates in the DR for persons aged 12 or older and
by age group (i.e., regardless of gender) also take into account the lower threshold for
females.
4. Methamphetamine Dependence – For the DR sample, respondent s were classified with
past year methamphetamine dependence if they reported three of the following problems in
the past year because of their use of methamphetamine:

I-2

•

spent a great deal of time over a period of a month getting, using, or getting over the
effects of methamphetamine (METHLOTTM=1 or METHGTOVR=1, corresponding
to questions DRME01 and DRME02);

•

used methamphetamine more often than intended or was unable to keep set limits on
methamphetamine use (METHKPLMT=2, corresponding to DRME05);

•

inability to cut down or stop using methamphetamine every time tried or wanted to
(METHCUTEV=2, corresponding to DRME09);

•

continued to use methamphetamine even though it was causing problems with
emotions, nerves, mental health, or physical problems (METHEMCTD=1 or
METHPHCTD=1, corresponding to DRME14 and DRME16);

•

methamphetamine use reduced or eliminated involvement or participation in
important activities (METHLSACT=1, corresponding to DRME17); or

•

5. Methamphetamine Abuse – For the DR sample, respondents were classified with past year
abuse of methamphetamine if they had not been classified with past year methamphetamine
dependence and if they reported one or more of the following problems in the past year
because of their use of methamphetamine:
•

serious problems at home, work, or school caused by using methamphetamine, such
as
– neglecting their children,
–

missing work or school,

– doing a poor job at work or school,
– losing a job or dropping out of school
(METHSERPB=1, corresponding to DRME18);
•

used methamphetamine regularly and then did something that might have put you in
physical danger (METHPDANG=1, corresponding to DRME19);

•

use of methamphetamine caused you to do things that repeatedly got you in trouble
with the law (STMLAWTR=1, corresponding to DRME20); and

I-3

•

6. In the DR sample, a respondent was classified as having illicit drug dependence
(DEPNDILL) if he or she was classified as having dependence on any of the following:
marijuana, hallucinogens, inhalants, tranquilizers, cocaine, heroin, pain relievers, stimulants,
sedatives, or methamphetamine.
7. In the DR sample, a respondent was classified as having illicit drug abuse (ABUSEILL) if
he or she was not classified as having illicit drug dependence (DEPNDILL = 0) and met
abuse criteria for any of the following: marijuana, hallucinogens, inhalants, tranquilizers,
cocaine, heroin, pain relievers, stimulants, sedatives, or methamphetamine.
8. The following measures involving new survey items for comparisons between the DR
sample and the 2012 National Health Interview Survey (NHIS) were based on the raw
survey measures, as follows:
Measure
Living in a household with only cellular or no telephone
service
Number of visits to doctor or other health care
professional, past 12 months (none; 1; 2 to 3; 4 to 9; 10
or more)
Has been in a hospital overnight, past 12 months?
Emergency room visit in past 12 months?
Disability or Physical Limitation
Deaf or serious hearing difficulty
Blind or serious difficulty seeing
Serious difficulty concentrating, remembering, or
making decisions
Serious difficulty walking or climbing stairs
Difficulty dressing or bathing
Difficulty doing errands alone, such as visiting a
doctors' office or shopping
Conditions told to respondent by doctor or other
health care professional
Any kind of heart condition or heart disease
Diabetes or sugar diabetes
Chronic bronchitis, emphysema, chronic obstructive
pulmonary disease, also called COPD
Cirrhosis of the liver
Hepatitis
Kidney disease, not including bladder infection or
incontinence
Asthma
Cancer or a malignancy of any kind
Hypertension, also called high blood pressure

DR Survey Questions
CELL1 = 2
HLTH19, HLTH19a
HLTH17
HLTH16
QD56
QD57
QD58
QD59
QD60
QD61
HLTH25=1
HLTH25=2
HLTH25=3
HLTH25=4
HLTH25=5
HLTH25=6
HLTH25=7
HLTH25=9
HLTH25=10

I-4

File Type	application/pdf
Author	gmchenry
File Modified	2014-06-09
File Created	2014-06-09