PATH Study Interim Report

Attach20. PATH Study Interim Report.pdf

Population Assessment of Tobacco and Health (PATH) Study - Fourth Wave of Data Collection (NIDA)

PATH Study Interim Report

OMB: 0925-0664

Document [pdf]

Download: pdf | pdf

Population Assessment of Tobacco and Health (PATH) Study (NIDA)

Attachment 20

PATH Study Interim
Report

Population Assessment of
Tobacco and Health Study
Interim Report to the Office of
Management and Budget on Data
and Biospecimen Collection from
Waves 1, 2, and 3
June 24, 2016

Submitted by:
Kevin P. Conway, Ph.D.
Deputy Director
Division of Epidemiology, Services, and Prevention Research
National Institute on Drug Abuse
6001 Executive Blvd., Room 5185
Rockville, MD 20852
Phone: 301-443-8755

Table of Contents
Chapter
1

Page
Introduction ........................................................................................................

1.1
1.2

1
2

Purpose of Interim Report (Terms of Clearance) ............................
Sample Design .......................................................................................
1.2.1

Overview of Sample Design for Wave 1
(Baseline Wave) .....................................................................
Overview of Sample Design for Waves 2 and 3 ..............

2
3

Summary of Wave 1 Findings .............................................................

1.3.1
1.3.2
1.3.3

5
6

1.2.2
1.3

Wave 1 Response Rates .......................................................
Wave 1 Nonresponse Bias Analysis ...................................
Wave 1 Statistical Approach for Addressing
Nonresponse .........................................................................
Wave 2 ..................................................................................................................

7
9

2.1

Response Rates ......................................................................................
2.1.1
2.1.2
2.1.3

2.2

2.3

2.4

Retention Rates for Continuing Adults and
Continuing Youth .................................................................
Recruitment Rates for Aged-up Adults and
Aged-up Youth .....................................................................
Biospecimen Collections .....................................................

11
15
19

Nonresponse Bias Analysis .................................................................

2.2.1
2.2.2

Method ...................................................................................
Results ....................................................................................

23
25

Statistical Approach for Addressing Nonresponse ..........................

2.3.1
2.3.2

Interviews...............................................................................
Biospecimens.........................................................................

40
56

Summary of Findings ...........................................................................

Contents (continued)

Chapter
3

Page
Wave 3 ..................................................................................................................

3.1

Predicted Response Rates ....................................................................
3.1.1

Predicted Retention Rates among Wave 2
Respondents for Continuing Adults and
Continuing Youth .................................................................
Predicted Recruitment Rates among Wave 2
Respondents for Aged-up Adults and Aged-up
Youth ......................................................................................
Predicted Retention and Recruitment Rates
among Wave 1 Respondents ...............................................
Biospecimen Collections .....................................................

69
70

Nonresponse Bias Analysis .................................................................
Statistical Approach for Addressing Nonresponse ..........................
Summary of Findings ...........................................................................

73
82
83

Conclusions and Implications for Study Going Forward .............................

References ...............................................................................................................................

3.1.2
3.1.3
3.1.4
3.2
3.3
3.4
4

61
65

Appendix
A

Cigarette Smoking Questions in the PATH Study and Other
Surveys .................................................................................................................

A-1

1-1

Summary of PATH Study Wave 1 response rates .........................................

2-1

Case counts by Wave 1 adult/youth/shadow youth classification
and Wave 2 adult/youth/shadow youth classification..................................

2-2

PATH Study Wave 2 retention rates by Wave 1 characteristics:
Adult Interview (continuing adults) .................................................................

Table

iii

Contents (continued)

Table

Page

2-3

PATH Study Wave 2 retention rates by Wave 1 characteristics:
Youth Interview (continuing youth) ................................................................

2-4

PATH Study Wave 2 recruitment rates by Wave 1 characteristics:
Adult Interview (aged-up adults) ......................................................................

2-5

PATH Study Wave 2 recruitment rates by Wave 1 characteristics:
Youth Interview (aged-up youth) .....................................................................

2-6

PATH Study Wave 2 response rates by respondent
characteristics: Urine collection (continuing adults) ......................................

2-7

PATH Study Wave 2 response rates by respondent
characteristics: Biospecimen collections (aged-up adults) ............................

2-8

Comparison of Wave 2 continuing Adult Interview respondents
with nonrespondents ..........................................................................................

2-9

Comparison of Wave 1 tobacco use rates for Wave 2 continuing
Adult Interview respondents with nonrespondents ......................................

2-10

Comparison of Wave 2 continuing Youth Interview respondents
with nonrespondents ..........................................................................................

2-11

Comparison of Wave 1 tobacco use rates for Wave 2 continuing
Youth Interview respondents with nonrespondents .....................................

2-12

Comparison of Wave 2 aged-up Adult Interview respondents
with nonrespondents ..........................................................................................

2-13

Comparison of Wave 1 tobacco use rates for Wave 2 aged-up
Adult Interview respondents with nonrespondents ......................................

2-14

Comparison of Wave 2 aged-up Youth Interview respondents
with nonrespondents ..........................................................................................

2-15

Comparison of Wave 2 aged-up adult urine specimen providers
with non-providers of urine ..............................................................................

2-16

Comparison of Wave 2 aged-up adult blood specimen providers
with non-providers of blood .............................................................................

Contents (continued)

Table

Page

2-17

Current cigarette smoking based on Wave 2 Adult Interview
respondents: IPS weights...................................................................................

2-18

Cigarette smoking based on Wave 2 Youth Interview
respondents: IPS weights...................................................................................

2-19

Comparison of Wave 2 continuing Adult Interview respondents
with Wave 1 Adult Interview respondents .....................................................

2-20

Comparison of Wave 1 tobacco use rates for Wave 2 continuing
Adult Interview respondents with Wave 1 Adult Interview
respondents .........................................................................................................

Comparison of Wave 2 continuing Youth Interview respondents
with Wave 1 youth respondents who were continuing youth at
Wave 2 ..................................................................................................................

Comparison of Wave 1 tobacco use rates for Wave 2 continuing
Youth Interview respondents with Wave 1 youth respondents
who were continuing youth at Wave 2 ............................................................

Comparison of Wave 2 aged-up Adult Interview respondents
with Wave 1 youth respondents who were aged-up adults at
Wave 2 ..................................................................................................................

Comparison of Wave 1 tobacco use rates for Wave 2 aged-up
Adult Interview respondents with Wave 1 youth respondents
who were aged-up adults at Wave 2.................................................................

2-25

Comparison of Wave 2 aged-up Youth Interview respondents
with Wave 1 shadow youth who were aged-up youth at Wave 2 ................

2-26

Current cigarette smoking based on Wave 2 Adult Interview
respondents .........................................................................................................

2-27

Cigarette smoking based on Wave 2 Youth Interview
respondents .........................................................................................................

2-28

Summary of PATH Study Wave 2 response rates .........................................

2-21

2-22

2-23

2-24

Contents (continued)

Table

Page

3-1

Status of Wave 3 released cases from replicate group 1, as of
April 29, 2016 ......................................................................................................

3-2

PATH Study Wave 3 predicted retention rates by respondent
characteristics: Adult Interview (continuing adults) ......................................

3-3

PATH Study Wave 3 predicted retention rates by respondent
characteristics: Youth Interview (continuing youth) .....................................

3-4

PATH Study Wave 3 predicted recruitment rates by respondent
characteristics: Adult Interview (aged-up adults) ...........................................

3-5

PATH Study Wave 3 predicted recruitment rates by respondent
characteristics: Youth Interview (aged-up youth) ..........................................

3-6

PATH Study Wave 3 predicted retention and recruitment rates
among Wave 1 respondents ..............................................................................

3-7

PATH Study Wave 3 response rates by respondent
characteristics: Urine collection (continuing adults) ......................................

3-8

PATH Study Wave 3 response rates by respondent
characteristics: Biospecimen collections (aged-up adults) ............................

3-9

Comparison of Wave 3 Adult Interview respondents with
finalized and provisional nonrespondents (Wave 1 adults) ..........................

3-10

Comparison of Wave 1 tobacco use rates for Wave 3 Adult
Interview respondents with finalized and provisional
nonrespondents (Wave 1 adults) ......................................................................

3-11

Comparison of Wave 3 interview respondents with finalized and
provisional nonrespondents (Wave 1 youth)..................................................

3-12

Comparison of Wave 1 tobacco use rates for Wave 3 respondents
with finalized and provisional nonrespondents (Wave 1 youth) .................

3-13

Comparison of Wave 3 Youth Interview respondents with
provisional nonrespondents (Wave 1 shadow youth) ...................................

Contents (continued)

Table

Page

3-14

Summary of PATH Study predicted response rates for Wave 3 .................

A-1

Questions used to define adult current cigarette smoking in the
PATH Study, TUS-CPS, NHIS, NHANES, and NSDUH .........................

A-3

A-2

Questions used to define youth cigarette smoking in the PATH
Study, NHANES, NSDUH, and NYTS .........................................................

A-5

vii

Population Assessment of Tobacco and Health Study

Introduction

The National Institute on Drug Abuse (NIDA), in partnership with the Food and Drug
Administration (FDA), has prepared this report in response to the terms of clearance by the Office
of Management and Budget (OMB) in its approval of Wave 3 of the Population Assessment of
Tobacco and Health (PATH) Study.

1.1

Purpose of Interim Report (Terms of Clearance)

OMB approved the PATH Study’s Revision Request for Wave 3 on August 17, 2015 (0925-0664).
The terms of clearance of OMB’s approval state: “Before submitting the information collection
request for Wave 4 to OMB, NIDA/FDA should report to OMB regarding: a) the response rates
associated with the full baseline wave [and full Wave 2], including screening, interview completion,
and bio-specimen response; b) Wave 3 retention, recruitment rates for the “age in to adult” and “age
in of shadow” subsamples; c) the results of nonresponse analysis and statistical approach for
addressing non-response, as well as implications for the study going forward; and d) the statistical
approach to be applied to the bio-specimen data to address potential non-response bias from lower
consent and cooperation rates with this aspect of the study.”
This report is organized in sections that correspond to OMB’s terms of clearance: Section 1 includes
a summary of findings from the 2015 Interim Report to OMB on the full Wave 1 response rates,
nonresponse analysis, and statistical approach for addressing nonresponse. Section 2 presents the
full Wave 2 response rates (retention and recruitment rates), results of a nonresponse analysis, and
statistical approach for addressing nonresponse. Section 3 presents the predicted Wave 3 response
rates (retention and recruitment rates), results of an interim nonresponse analysis, and statistical
approach for addressing nonresponse. Section 4 summarizes the findings and considers their
implications.
The report covers the full Waves 1 and 2 of the PATH Study, and data collected from a probability
subsample of Wave 3 between October 19, 2015 and April 29, 2016. Response rates for Wave 2 and
Wave 3 are compared throughout this report with corresponding rates projected in the PATH
Study’s Revision Requests to OMB for Wave 2 and Wave 3, respectively.

Population Assessment of Tobacco and Health Study

1.2

Sample Design

This section provides an overview of the sample design for the PATH Study and a description of
replicate group 1; results in this report for Wave 3 are based on replicate group 1. Information on
the study background and overall design is provided in Supporting Statement A of the PATH
Study’s Revision Request to OMB for Wave 3.

1.2.1

Overview of Sample Design for Wave 1 (Baseline Wave)

The target population of the PATH Study at Wave 1 was the civilian, non-institutionalized U.S.
population (i.e., including the 50 states and the District of Columbia) 12 years of age and older at
that point in time. Thus, active duty military personnel and those residing in an institutional setting
were excluded. College students living away from home during the school year were identified as
members of their permanent residence (e.g., parents’ home). For Wave 1, a four-stage stratified area
probability sample design was used with a two-phase design for sampling the adult cohort at the
final stage. The sampling rates for adults varied by age, race, and tobacco use status. At the first
stage, a stratified sample of geographical primary sampling units (PSUs) was selected, in which a
PSU was a county or group of counties. For the second stage, within each selected PSU, smaller
geographical segments (consisting of one or more census blocks) were formed and then a sample of
these segments was drawn. At the third stage, a sample of addresses within sampled segments was
drawn from listings of addresses; the main source of these addresses was obtained from the Postal
Service (USPS) Computerized Delivery Sequence Files (CDSFs). The CDSFs provide very high
coverage of the residential addresses in the U.S.
The fourth stage was the random selection of persons within the sampled households. A roster of all
the members in the sampled household was constructed using the Household Screener. An adult
household member, the household screener respondent, was asked to list members of the household
and provide demographic as well as, for each adult, tobacco use information. This information was
used in sampling three main groups of interest:


Adults (up to two adults per household);



Children ages 12 to 17 (referred to as “youth,” generally up to two per household); and

Population Assessment of Tobacco and Health Study



Children ages 9 to 11 (referred to as “shadow youth,” generally up to two per
household) to be enrolled in the youth cohort in later waves of the study on reaching 12
years of age.

Two-phase sampling was used for adult selection due to potential misreporting by the household
screener respondent of the tobacco use status of other adult household members. The Phase 1
sampling depended on the age, race, and tobacco use information obtained from the Household
Screener. The Phase 2 sampling was based on self-reported age, race, and tobacco use status,
obtained by interviewing the individuals sampled at Phase 1. The sampling rates for the two phases
were designed to achieve large enough sample sizes for young adults (ages 18 to 24) and adult
tobacco users of all ages.
Because the full sample was selected using probability sampling methods, it is representative of the
U.S. civilian non-institutionalized population 12 years of age and older. The PATH Study Wave 1
sample was divided into four replicate groups, consisting of probability samples of approximately 20
percent, 30 percent, 30 percent, and 20 percent of the sampled segments, respectively, within each
sampled PSU. Each separate replicate group was a probability sample from the set of segments in
the frame and, therefore, also representative of the civilian non-institutionalized U.S. population.
The replicate groups were released to the field in a sequential manner (replicate group 1 in
September 2013, replicate group 2 in November 2013, replicate group 3 in February and March
2014, and replicate group 4 in May 2014).
The PATH Study completed 32,320 Adult Interviews and 13,651 Youth Interviews in Wave 1. All
adult interview respondents were asked to provide urine and blood specimens; 21,801 provided a
urine specimen and 14,520 provided a blood specimen.

1.2.2

Overview of Sample Design for Waves 2 and 3

Wave 2 of the PATH Study was the first follow-up wave for participants in Wave 1. The target
population for Wave 2 was the Wave 1 target population residing in the U.S. at Wave 2 with the
exception of those who were incarcerated (with a corresponding definition for the target population
at Wave 3). Thus, Wave 1 respondents who later joined the military or entered a health care
institution (e.g., nursing home) were members of the target population and eligible for data
collection for the PATH Study. At Wave 2 only Wave 1 respondents who died, resided outside the
U.S., or were in a correctional facility were ineligible for a Wave 2 interview (and similarly for
Wave 3).
3

Population Assessment of Tobacco and Health Study

Attempts were made to contact the Wave 1 youth and adult respondents as well as members of the
shadow youth sample established at Wave 1. Youth from the shadow youth sample who turned age
12 by Wave 2 (aged-up youth) and were permitted by a parent or guardian to participate in the study
were asked for assent to be interviewed for the first time at Wave 2. Similarly, persons in the youth
sample at Wave 1 who reached age 18 by Wave 2 (aged-up adults) were asked to complete the adult
instrument as well as to provide urine and blood specimens.
The PATH Study completed 28,375 Adult Interviews and 12,172 Youth Interviews in Wave 2. The
study subsampled 14,465 adults for urine collection from adults who provided urine at Wave 1;
among these subsampled adults, 12,569 completed the Wave 2 interview and 12,113 provided a
urine specimen again at Wave 2. The study also collected urine and blood specimens from
consenting aged-up adults (1,587 urine specimens and 908 blood specimens).
The PATH Study is currently conducting Wave 3, the second follow-up wave for participants in
Wave 1. In addition, youth from the shadow youth sample established at Wave 1 who turn age 12 by
Wave 3 and who are permitted by a parent or guardian to participate in the study are asked for
assent to be interviewed for the first time at Wave 3. Similarly, persons in the youth sample at
Wave 1 who reach age 18 by Wave 3 are asked to complete the adult instrument as well as to
provide urine and blood specimens for the first time. The same follow-up rules apply for Wave 3 as
for Wave 2:


Wave 1 respondents who reside in the U.S. and are not incarcerated at Wave 3 are
eligible for Wave 3 data collection, including those in the military or living in an
institution at Wave 3; and



In addition to Wave 2 respondents, Wave 2 nonrespondents are fielded for Wave 3 data
collection, unless the nonresponse at Wave 2 was due to a firm or hostile refusal,
inability to complete the Wave 2 interview in English or Spanish, death, or a physical or
mental disability that prevents participation in the study.

For Wave 3, the PATH Study is subsampling 16,138 adults for urine collection from adults who
provided urine at a previous wave. Among these adults, an estimated 12,000 are expected to provide
a urine specimen again at Wave 3. The study is also collecting urine and blood specimens from
consenting aged-up adults.
This Interim Report reviews the results for Wave 3 based on data collected from replicate group 1
through April 29, 2016. Wave 3 cases are released for follow up in monthly groups with the goal of

Population Assessment of Tobacco and Health Study

completing the Wave 3 interview as close as possible to the one-year anniversary date of the Wave 2
interview (or to the two-year anniversary date of the Wave 1 interview if no interview was obtained
in Wave 2). Cases are fielded at the beginning of the calendar month prior to the month in which
the anniversary date falls. 1 If there is more than one sampled person in the same household, the
cases from that household are typically clustered for simultaneous release based on the earliest date
of a completed interview for any sampled adult or youth in the household in Wave 2 (or in Wave 1,
if nobody in the household completed a Wave 2 interview). 2 Thus, while the Wave 3 sample release
is not explicitly tied to the release of the Wave 1 replicate groups, it does correlate with the
distribution of completion dates for Wave 1 interviews. That is, a high percentage of cases in the
first Wave 3 release groups came from replicate group 1, and replicate group 1 has higher
percentages of released and finalized cases than the other replicate groups.

1.3

Summary of Wave 1 Findings

The 2015 Interim Report to OMB provided detailed findings on the full Wave 1 response rates,
nonresponse analysis, and statistical approach for addressing nonresponse. This section summarizes
those findings.

1.3.1

Wave 1 Response Rates

Sections 2.2 and 2.3 of the 2015 Interim Report describe the weight construction for Wave 1 of the
PATH Study, and a brief summary also appears in Section 2.3.1 of this report. Inverse probability of
selection (IPS) weights were used to compute the Wave 1 weighted response rates and to examine
nonresponse bias. The final raked weights from Wave 1 were then used to demonstrate the
effectiveness of the nonresponse weighting adjustments in reducing potential nonresponse bias.

However, cases with anniversary dates in September through December 2016 (as determined by Wave 2 or Wave 1
interview dates) will all be released in the beginning of August, because Wave 3 data collection will end on October
31, 2016. Similarly, cases with anniversary dates in September through November 2015 were all released for data
collection in October 2015 and were available for data collection beginning on the first day of Wave 3 data collection,
October 19, 2015.

If the gap between interviews within the household is greater than one month, interviewers are instructed to hold data
collection for a sampled person with a later interview to a date closer to the individual’s anniversary date whenever
feasible and the interview would not be lost to nonresponse.

Population Assessment of Tobacco and Health Study

The weighted response rates for the PATH Study Household Screener, Adult and Youth Interviews,
and biospecimen collections in Wave 1 are provided in Table 1-1.
Table 1-1.

Summary of PATH Study Wave 1 response rates

Collection
Household Screener
Adult Interview
Youth Interview
Urine
Blood

Unweighted response rate, based on full
Wave 1 sample
54.1%
74.8%
78.2%
67.5%
44.9%

Weighted response rate, based on full
Wave 1 sample
54.0%
74.0%
78.4%
63.6%
43.0%

Differences in weighted response rates were modest for tobacco use status and demographic
subgroups. The largest differential weighted response rate, 11.5 percentage points, was for the age of
adults who provided urine specimens, which suggests a heightened potential for nonresponse bias.
Notably, the differential weighted response rates for blood collection, ranging from 3.2 percentage
points for ethnicity to 5.8 percentage points for race, were more consistent with those of other
PATH Study collections at Wave 1.

1.3.2

Wave 1 Nonresponse Bias Analysis

A nonresponse bias analysis indicated that estimates of key demographic and tobacco use variables
calculated from the PATH Study Wave 1 sample with the inverse probability of selection weights
were comparable to those produced by national cross-sectional surveys. However, the completed
household screening interviews from the Wave 1 sample appeared to underrepresent single- and
two-person households relative to the 1-year 2013 American Community Survey (ACS) counts. The
estimated percentage of persons who were non-Black and 25 years of age or older, from the
household rosters, was also smaller than the corresponding estimate from the ACS.
Estimated distributions of demographic characteristics for adults completing the Adult Interview
were similar to those from the 1-year 2013 ACS for race (except for persons in the “other race”
category). Persons in the “other race” category were also underrepresented among the persons
providing blood or urine specimens. The percentage of adults who were Hispanic estimated from
those who provided blood specimens was similar to ACS population estimates, but, compared to
ACS, Hispanics were overrepresented among adults who responded to the Adult Interview and
those who provided urine specimens. In addition, the percentages of adults who were between 18

Population Assessment of Tobacco and Health Study

and 24 years old or between 25 and 44 years old as estimated from PATH Study interview
respondents and from those who provided urine specimens were higher than the ACS estimates for
these age groups. Males were underrepresented among respondents to the Adult Interview, and also
among the persons who provided blood or urine specimens.
When compared to national cross-sectional surveys that measure adult tobacco use [the Tobacco
Use Supplement to the Current Population Survey (TUS-CPS), the National Health Interview
Survey (NHIS), the National Health and Nutrition Examination Survey (NHANES), and the
National Survey on Drug Use and Health (NSDUH)], estimates of adult cigarette smoking from the
PATH Study Wave 1 sample were roughly in the middle of the range of estimates for cigarette
smoking. There is no indication of nonresponse bias with respect to this measure.
Estimates of demographic characteristics of youth in Wave 1 aligned with the 1-year 2013 ACS for
most demographic characteristics. However, the estimated percentage of youth who were Hispanic
from the PATH Study was significantly higher than the corresponding percentage estimated from
the ACS.
PATH Study estimates of the selected youth cigarette smoking measure from the full Wave 1 sample
were at the low end of estimates in comparison with national cross-sectional surveys that measure
youth tobacco use (NHANES, NSDUH, and the National Youth Tobacco Survey (NYTS)).
However, estimates based on these comparison surveys were from 2011 through 2013 while those
based on the PATH Study were from September 2013 through December 2014, and evidence
suggests that the use of traditional cigarettes is declining among youth. Although the difference in
the time periods of the surveys is not by itself large enough to account for the different estimates, it
is one of a number of factors that may explain the different estimates.

1.3.3

Wave 1 Statistical Approach for Addressing Nonresponse

The approach used to reduce potential nonresponse bias in the PATH Study is to adjust the weights
of respondents at the household, adult, and youth levels to account for nonrespondents. Results of
applying this approach to the full Wave 1 sample indicated the nonresponse adjustments were
successful for reducing the discrepancy between the PATH Study estimates and 1-year estimates
from the 2013 ACS with respect to demographic characteristics. Raked weights used for adults
responding to the Adult Interview reduced differences between the PATH Study and ACS for adults
providing biospecimens as well, for sex and ethnicity. The raking did not reduce differences in the
7

Population Assessment of Tobacco and Health Study

age distributions for the persons providing blood specimens, however. Additional sets of
nonresponse-adjusted weights were therefore created for the persons who provided biospecimens,
separately for urine and blood. Further biospecimen weights were created for the sets of person for
whom urine and/or blood specimens were sent for laboratory analysis, using the procedure
described in the memo on Wave 1 biospecimen weighting procedures submitted to OMB and
approved on October 9, 2015.
Estimates of adult cigarette smoking using the IPS weights (before nonresponse adjustment) were in
line with estimates from other surveys; agreement in these estimates was preserved using the
nonresponse-adjusted weights. Weighting adjustments for youth corrected for the slight
overestimate of the percentage of Hispanics among youth in Wave 1 but had little effect on the
other demographic characteristics (i.e., IPS-weighted estimates already agreed with the ACS values)
and estimates of youth cigarette smoking.

Population Assessment of Tobacco and Health Study

Wave 2

The PATH Study completed the Wave 2 data and biospecimen collections in October 2015. 3 This
section presents findings on the response rates for Wave 2, on the nonresponse analysis, and on the
study’s statistical approach for addressing nonresponse.

2.1

Response Rates

This section summarizes the retention and recruitment rates of the PATH Study for Wave 2, as well
as the response rates for the biospecimen collections. The PATH Study Wave 2 Adult Interview and
Youth Interview collected extensive self-report information through in-person data collection using
audio computer-assisted self-interviewing (ACASI). The Adult Interview gathered information from
adults (18 years old and older) about tobacco use behaviors, attitudes, knowledge, and health
conditions, as well as information on demographics, environmental factors, family and peer
influences, substance use, and general physical and mental health status. The Youth Interview
gathered information from youth (12 to 17 years old) on topics similar to those in the Adult
Interview. Youth were asked about their tobacco use and attitudes about tobacco, and for
information on demographics, environmental factors, family and peer influences, substance use, and
mental health. The PATH Study completed 28,375 Adult Interviews and 12,172 Youth Interviews in
Wave 2.
Retention rates for Wave 2 apply to persons who completed the Adult Extended Interview in
Wave 1 and persons who completed the Youth Extended Interview in Wave 1 and who were age 17
or younger at Wave 2. Recruitment rates for Wave 2 apply to those who had aged up, either as
shadow youth who had turned age 12 and were eligible for the Wave 2 Youth Interview, or as youth
in Wave 1 who had turned age 18, thus becoming eligible to participate as adults in Wave 2.
Specifically,


All the adults from Wave 1 are continuing adults and, if eligible, were asked to complete
an Adult Interview in Wave 2.

A small number of blood collections for Wave 2 were completed in November 2015.

Population Assessment of Tobacco and Health Study



The youth from Wave 1 who completed a Youth Interview at Wave 2 are continuing
youth, and those who completed an Adult Interview at Wave 2 are aged-up adults.



The Wave 1 shadow youth who completed the Wave 2 Youth Interview are aged-up
youth.

A Wave 1 participant who was a nonrespondent at Wave 2, however, does not have a Wave 2
interview date, so his/her Wave 2 age was determined using the date of birth or age information
collected in Wave 1 as well as any updated information obtained between Waves 1 and 2. Each
Wave 1 respondent had an “anniversary month” for Wave 2, which was approximately one year
after the Wave 1 interview was completed. The age classification date for a Wave 2 nonrespondent is
two months after the last day of his/her anniversary month or the final date of the Wave 2 data
collection (October 31, 2015), whichever is earlier. 4 Age as of the age classification date was used to
categorize Wave 1 youth who were nonrespondents in Wave 2. That is, a Wave 1 youth who did not
respond in Wave 2 was classified as a continuing youth if his/her age was determined to be 17 or
younger on the age classification date; otherwise he/she was classified as an aged-up adult. A similar
classification rule was used for persons who were shadow youth at Wave 1. A Wave 1 shadow youth
who did not respond in Wave 2 was classified as an aged-up youth if he/she was determined to have
attained age 12 on or before the age classification date.
Table 2-1 displays the case counts according to their adult/youth/shadow youth classification in
Wave 1 and their classification in Wave 2.
Table 2-1.

Case counts by Wave 1 adult/youth/shadow youth classification and Wave 2
adult/youth/shadow youth classification

Wave 1 classification
Adult
Youth
Youth
Shadow youth
Shadow youth

Wave 2 classification
Continuing adult
Aged-up adult
Continuing youth
Aged-up youth
Shadow youth

Unweighted count
32,320
2,239
11,412
2,555
4,652

The purpose of the “anniversary month” concept is to provide a reasonable target period for completing the Wave 2
interview that would improve the likelihood of the interviews taking place one year after each individual’s Wave 1
interview, by giving the field interviewers a target for their data collection efforts. The Wave 2 anniversary month is
defined as the calendar month containing the date of the earliest Wave 1 interview completed by a member of the
study participant’s household, one year after the Wave 1 interview. Given the challenges of contacting and scheduling
the interviews, the target period encompassed four months, starting with the month before the anniversary month
and ending two months after the anniversary month; however, if necessary, efforts to complete the interview could
continue past this period. In theory, data collection efforts for nonrespondents could continue up to the last day of
the data collection period; because a reasonable cutoff date was needed for defining the ages of nonrespondents, the
end of the target period was chosen as a consistent and justifiable standard.

Population Assessment of Tobacco and Health Study

The interview retention and recruitment rates are presented for demographic and tobacco use
subgroups defined using Wave 1 characteristics. This is necessary because Wave 2 characteristics are
not available for Wave 2 nonrespondents. All persons asked to provide a biospecimen at Wave 2
completed the Wave 2 Adult Interview so the subgroup definitions for the biospecimen response
rates use Wave 2 data for characteristics where information was updated (age and tobacco use
status), and Wave 1 data otherwise.

2.1.1

Retention Rates for Continuing Adults and Continuing Youth

This section reports retention rates for continuing adults, who completed the Adult Interview at
Wave 1, and continuing youth, who completed the Youth Interview at Wave 1 and remained eligible
for the Youth Interview at Wave 2.

Method
Consistent with the response rate calculation guidelines specified by the Office of Management and
Budget in its “Standards and Guidelines for Statistical Surveys” (2006), the retention rate for Wave 2
was calculated as the ratio of the number of Wave 2 complete cases (or sufficient partials) to the
number of cases eligible for the Wave 2 interview, which is essentially the American Association for
Public Opinion Research (AAPOR) RR3 (AAPOR, 2015) response rate.
The retention rate for continuing adults, denoted as RRCA, was calculated using the equations below:
RRCA = CCA /(CCA+ NCA+ eCA * UCA)
eCA = CCA /(CCA+ ICA)
where
CCA = number of Wave 2 completed cases or sufficient partials among continuing adults;
NCA = number of Wave 2 nonrespondents known to be eligible among continuing
adults;
UCA = number of Wave 2 nonrespondents with eligibility unknown among continuing
adults;
ICA = number of Wave 2 ineligible cases among continuing adults; and

Population Assessment of Tobacco and Health Study

eCA = estimated proportion of nonrespondents with unknown eligibility who were
eligible among continuing adults.
The retention rate for continuing youth, denoted as RRCY, was calculated using the equations below:
RRCY = CCY /(CCY+ NCY+ eCY * UCY)
eCY = CCY /(CCY+ ICY)
where
CCY = number of Wave 2 completed cases or sufficient partials among continuing youth;
NCY = number of Wave 2 nonrespondents known to be eligible among Wave 1 youth
respondents who were age 17 or younger on the age classification date;
UCY = number of Wave 2 nonrespondents with eligibility unknown among Wave 1
youth respondents who were age 17 or younger on the age classification date;
ICY = number of Wave 2 ineligible cases among Wave 1 youth respondents who were
age 17 or younger on the age classification date; and
eCY = estimated proportion of nonrespondents with unknown eligibility who were
eligible among Wave 1 youth respondents who were age 17 or younger on the age
classification date.
Both unweighted and weighted retention rates were calculated. For the unweighted retention rates,
the numbers of cases used in the calculations are the actual case counts. For the weighted retention
rates, the numbers of cases used in the calculations are the sums of Wave 1 IPS weights, in
accordance with AAPOR guidance (see AAPOR, 2015, p.51). The unweighted retention rate
measures the success of field operations in obtaining responses from the sample group. The
weighted retention rate estimates the proportion of the population represented by the sample group
that would have responded if they all had been asked to participate in the study, and provides a
measure of the potential impact of nonresponse on the quality of weighted estimates.

Results
Tables 2-2 and 2-3 provide unweighted and weighted retention rates by Wave 1 characteristics for
continuing adults and continuing youth. In addition to the overall row, each table includes rows on
age, sex, race, and ethnicity subgroups based on self-reported data from the Wave 1 Extended
Interviews. Both tables also include rows on tobacco use status. Persons with missing values for
these characteristics from Wave 1 were excluded from the response rate calculation for the particular
characteristic.
12

Population Assessment of Tobacco and Health Study

Table 2-2.

PATH Study Wave 2 retention rates by Wave 1 characteristics: Adult Interview
(continuing adults)

Characteristic at Wave 1a
Overall
Tobacco use statusb
Current established
user
Not current
established user
Age
18-24
25-44
45-64
65+
Sex
Male
Female
Race
White alone
Black alone or in
combination with
some other race
Other
Ethnicity
Hispanic
Non-Hispanic

ICA:
Ineligible
(n)
263

NCA:
Nonresponse
known to be
eligible
(n)
4,047

UCA:
Nonresponse
with unknown
eligibility
(n)
1,551

Unweighted
RRCA:
Unweighted
retention
rate for
Wave 2
(%)
82.6

11,693

135

1,753

788

82.2

82.1

14,039

112

2,136

728

83.1

83.8

7,330
9,233
7,375
2,516

40
53
87
83

1,097
1,351
1,110
483

645
632
246
28

80.8
82.3
84.5
83.1

81.2
82.6
84.6
82.9

13,080
13,356

162
101

2,167
1,875

900
650

81.1
84.1

81.8
84.2

18,997

191

3,086

970

82.4

83.3

4,647

516

338

84.5

84.4

2,170

341

193

80.3

80.2

4,528
21,539

43
213

580
3,391

385
1,145

82.5
82.6

82.7
83.3

CCA:
Adult
Interview,
completed
(n)
26,459

Weighted
RRCA:
Weighted
retention
rate for
Wave 2
(%)
83.1

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the
categories may not be equal to the count in the overall row due to missing values.

A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview.
A ‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days and: for
cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used that
product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little
filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 2-3.

PATH Study Wave 2 retention rates by Wave 1 characteristics: Youth Interview
(continuing youth)

Characteristic at Wave 1a
Overall
Tobacco use statusb
Ever user
Never user
Age
12-13
14-17
Sex
Male
Female
Race
White alone
Black alone or in
combination with
some other race
Other
Ethnicity
Hispanic
Non-Hispanic

CCY:
Youth
Interview,
completed
(n)
10,081

Unweighted
RRCY:
Unweighted
retention
rate for
Wave 2
(%)
88.5

Weighted
RRCY:
Weighted
retention
rate for
Wave 2
(%)
88.4

ICY:
Ineligible
(n)
15

NCY:
Nonresponse
known to be
eligible
(n)
964

UCY:
Nonresponse
with unknown
eligibility
(n)
352

1,665
8,006

4
11

170
737

91
245

86.5
89.1

86.6
89.0

4,164
5,917

6
9

381
582

133
219

89.0
88.1

88.9
88.0

5,164
4,892

11
4

510
445

155
196

88.6
88.4

88.5
88.4

6,497

662

190

88.4

1,882

149

107

88.0

87.7

1,143

89.7

89.4

2,907
6,982

7
7

244
698

131
213

88.6
88.5

88.5
88.4

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the
categories may not be equal to the count in the overall row due to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A
‘never user’ is someone who has never used any of those tobacco products. Ever use of a tobacco product is defined as having ever
used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional
cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

The weighted retention rates are approximately 83 percent for continuing adults and 88 percent for
continuing youth. The unweighted retention rates are approximately the same as the weighted
retention rates. These rates are moderately lower than the projected retention rates reported in the
Revision Request to OMB for Wave 2 (i.e., 86 percent for continuing adults and 90 percent for
continuing youth). For continuing adults, it appears that females had higher retention rates than
males, 18-24 years olds had lower retention rates than older age groups, adults who identified as
Black or White alone had higher retention rates than those of “other” race, and current established
tobacco users at Wave 1 had slightly lower retention rates than those who were not. The retention
rates are similar for most the subgroups of continuing youth, but are slightly higher among Wave 1
never users of tobacco. Slight variation in response rates by subgroups is to be expected in largescale data collection efforts. Although there is some variability among the retention rates for the
various subgroups of interest, none of the subgroup differences in Tables 2-2 and 2-3 suggests a
reason for concern.

2.1.2

Recruitment Rates for Aged-up Adults and Aged-up Youth

This section reports recruitment rates for aged-up adults, who completed the Youth Interview at
Wave 1 and were eligible for the Adult Interview at Wave 2, and aged-up youth, who were shadow
youth at Wave 1 and were eligible for the Youth Interview at Wave 2. The Wave 2 Youth Interview
was the first interview for responding aged-up youth, and aged-up adult respondents completed the
Adult Interview for the first time.

Method
The methods described in Section 2.1.1 for calculating the retention rates were also used to calculate
the recruitment rates for aged-up adults and aged-up youth. For aged-up adults, the recruitment rate,
denoted as RRAUA, was calculated using the equations below:
RRAUA = CAUA /(CAUA+ NAUA+ eAUA * UAUA)
eAUA = CAUA /(CAUA+ IAUA)

Population Assessment of Tobacco and Health Study

where
CAUA = number of Wave 2 completed cases or sufficient partials among persons who
completed the Youth Interview at Wave 1 and were administered the Adult
Interview at Wave 2;
NAUA = number of Wave 2 nonrespondents known to be eligible among Wave 1 youth
respondents who attained age 18 by the age classification date;
UAUA = number of Wave 2 nonrespondents with eligibility unknown among Wave 1
youth respondents who attained age 18 by the age classification date;
IAUA = number of Wave 2 ineligible cases among Wave 1 youth respondents who
attained age 18 by the age classification date; and
eAUA = estimated proportion of nonrespondents with unknown eligibility who were
eligible among Wave 1 youth respondents who attained age 18 by the age
classification date.
The recruitment rate for aged-up youth, denoted as RRAUY, was calculated using the equations
below:
RRAUY = CAUY /(CAUY+ NAUY+ eAUY * UAUY)
eAUY = CAUY /(CAUY+ IAUY)
where
CAUY = number of Wave 2 completed cases or sufficient partials among persons who
were shadow youth at Wave 1 and were administered the Youth Interview at
Wave 2;
NAUY = number of Wave 2 nonrespondents known to be eligible among Wave 1 shadow
youth who were age 12 by the age classification date;
UAUY = number of Wave 2 nonrespondents with eligibility unknown among Wave 1
shadow youth who were age 12 by the age classification date;
IAUY = number of Wave 2 ineligible cases among Wave 1 shadow youth who were age 12
by the age classification date; and
eAUY = estimated proportion of nonrespondents with unknown eligibility who were
eligible among Wave 1 shadow youth who were age 12 by the age classification
date.
As for the retention rates, both unweighted and weighted recruitment rates were calculated. For the
unweighted recruitment rates, the numbers of cases used in the calculations are the actual case
counts. For the weighted recruitment rates, the numbers of cases used in the calculations are the
sums of Wave 1 IPS weights (AAPOR, 2015).

Population Assessment of Tobacco and Health Study

Results
Table 2-4 provides recruitment rates by Wave 1 characteristics for the Adult Interview for aged-up
adults, and Table 2-5 provides recruitment rates by Wave 1 characteristics for the Youth Interview
for aged-up youth. In addition to the overall row, each table includes rows on sex, race, and ethnicity
subgroups; Table 2-4 also includes rows on tobacco use status (which is not available for aged-up
youth). There are no rows corresponding to age subgroups in Table 2-4 or Table 2-5, because most
of the aged-up adults are 18 years old and most of the aged-up youth are 12 years old. Information
from the Wave 1 Extended Youth Interview was used to define the demographic and tobacco-use
characteristics for the aged-up adults, and information from the Wave 1 Household Screener was
used to define the demographic characteristics for the aged-up youth. Persons with missing values
for these characteristics were excluded from the response rate calculation for that characteristic.
The weighted recruitment rate is approximately 86 percent for aged-up adults, which is slightly
higher than the projected recruitment rate for this group (85 percent) provided in the Revision
Request to OMB for Wave 2. The weighted recruitment rate is approximately 82 percent for agedup youth, which is lower than the projected recruitment for this group (88 percent) in the Wave 2
Revision Request. For both aged-up adults and aged-up youth, the variability of recruitment rates by
subgroups is small. For aged-up adults, the differences in recruitment rates by subgroup range from
0.8 percentage points for tobacco use status to 2.3 percentage points for race/ethnicity. For aged-up
adults, the differences in recruitment rates by subgroup are 0.8 percentage points for sex and 1.1
percentage points for race/ethnicity.

Population Assessment of Tobacco and Health Study

Table 2-4.

PATH Study Wave 2 recruitment rates by Wave 1 characteristics: Adult Interview
(aged-up adults)

Characteristic at Wave 1a
Overall
Tobacco use statusb
Ever user
Never user
Sex
Male
Female
Race/ethnicity
Non-Hispanic White
alone
Other

UAUA:
Nonresponse
with unknown
eligibility
(n)
96

Unweighted
RRAUA:
Unweighted
recruitment
rate for
Wave 2
(%)
85.9

Weighted
RRAUA:
Weighted
recruitment
rate for
Wave 2
(%)
85.7

CAUA:
Adult
Interview,
completed
(n)
1,916

IAUA:
Ineligible
(n)
8

NAUA:
Nonresponse
known to be
eligible
(n)
219

811
1,061

2
6

80
133

46
47

86.6
85.5

86.2
85.4

957
956

5
2

123
96

46
50

85.0
86.8

84.7
86.6

954

136

84.7

84.5

945

87.0

86.8

Table 2-5.

PATH Study Wave 2 recruitment rates by Wave 1 characteristics: Youth Interview
(aged-up youth)

Characteristic at
Wave 1a
Overall
Sex
Male
Female
Race/ethnicity
Non-Hispanic White
alone
Other
a

UAUY:
Nonresponse
with
unknown
eligibility
(n)
92

Unweighted
RRAUY:
Unweighted
recruitment
rate for
Wave 2
(%)
82.0

Weighted
RRAUY:
Weighted
recruitment
rate for
Wave 2
(%)
82.1

CAUY:
Youth
Interview,
completed
(n)
2,091

IAUY:
Ineligible
(n)
5

NAUY:
Nonresponse
known to be
eligible
(n)
367

1,055
1,036

3
2

179
188

58
34

81.7
82.4

81.7
82.5

1,008

190

82.4

82.6

1,078

176

81.6

81.5

The characteristics are as reported in the Household Screener at Wave 1. For each characteristic, the sum of the counts in all the
categories may not be equal to the count in the overall row due to missing values.

Population Assessment of Tobacco and Health Study

2.1.3

Biospecimen Collections

This section addresses the response rates for the collection of urine and blood specimens from
continuing adults and aged-up adults who completed a Wave 2 Adult Interview and were asked to
provide a specimen. Biospecimens are intended to provide a basis for the assessment of betweenperson differences and within-person changes in markers of tobacco exposure, and to detect and
compare indicators of conditions and related disease processes associated with the use of tobacco
products. Field interviewers collected the urine specimens; on separate visits, phlebotomists
collected the blood specimens. Among Wave 2 Adult Interview respondents, a subsample of
continuing adults were asked to provide a urine specimen and all aged-up adults were asked to
provide urine and blood specimens.

Method
Tables 2-6 and 2-7 provide overall unweighted response rates 5 for the biospecimen collections, and
response rates by tobacco use status and demographic subgroups, for continuing adults and aged-up
adults, respectively. The response rates are conditional on a completed Wave 2 Adult Interview and
the adult having been asked to provide a specimen. The response rates were calculated using the
following formula:
RRU = (Number of adults who provided a specimen)/(Number of adults from whom a
specimen was requested)
This is the same formula used to compute the biospecimen response rates for Wave 1. However, all
Adult Interview respondents were asked to provide urine and blood biospecimens in Wave 1. The
denominator for the Wave 2 urine response rate in Table 2-6 is the 12,569continuing adults who
completed the Wave 2 Adult Interview and were asked for a urine specimen. The denominator for
the Wave 2 urine and blood response rates in Table 2-7 is the 1,916 aged-up adults who completed
the Wave 2 Adult Interview.

Weighted response rates are not provided because the subset of continuing adults asked to provide a urine specimen
at Wave 2 does not represent a readily interpretable portion of the population. Similarly, while most aged-up adults
were age 18 at Wave 2, some were older due to the actual time elapsed between the Wave 1 and Wave 2 interviews.

Population Assessment of Tobacco and Health Study

Table 2-6.

PATH Study Wave 2 response rates by respondent characteristics: Urine collection
(continuing adults)

Characteristica
Overall
Tobacco use statusb
Current established user
Not current established user
Age
18-24
25-44
45-64
65+
Sex
Male
Female
Race
White alone
Black alone or in combination
Other
Ethnicity
Hispanic
Non-Hispanic

A:
Adults requested to
provide urine
(n)
12,569

B:
Collected
(n)
12,113

Urine
Unweighted response
rate for Wave 2c
(%)
96.4

7,977
4,533

7,725
4,332

96.8
95.6

3,326
4,948
3,450
845

3,195
4,789
3,325
804

96.1
96.8
96.4
95.1

6,491
6,070

6,229
5,878

96.0
96.8

8,855
2,400
1,032

8,540
2,316
989

96.4
96.5
95.8

2,125
10,272

2,025
9,927

95.3
96.6

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Adult Extended Interview. The age information and tobacco
use status are as reported in the Wave 2 Adult Interview. For each characteristic, the sum of the counts in all the categories may not
be equal to the count in the overall row due to missing values.

A tobacco user is defined as someone who uses one or more of the following tobacco products covered by the Wave 2 Adult Interview:
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable
tobacco. A ‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days
and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever
used that product regularly.

Response rate = B/A.

Population Assessment of Tobacco and Health Study

Table 2-7.

PATH Study Wave 2 response rates by respondent characteristics: Biospecimen
collections (aged-up adults)

Characteristica
Overall
Tobacco use statusb
Current established user
Not current established user
Sex
Male
Female
Race
White alone
Black alone or in combination
Other
Ethnicity
Hispanic
Non-Hispanic

A:
Adult
Interviews
completed
(n)
1,916

Urine
Unweighted
B:
response rate
Collected
for Wave 2c
(n)
(%)
1,587
82.8

Blood
Unweighted
B:
response rate
Collected
for Wave 2c
(n)
(%)
908
47.4

328
1,578

295
1,285

89.9
81.4

177
727

54.0
46.1

957
956

792
792

82.8
82.8

457
450

47.8
47.1

1,257
348
212

1,034
299
173

82.3
85.9
81.6

596
161
109

47.4
46.3
51.4

515
1,384

426
1,145

82.7
82.7

251
648

48.7
46.8

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Youth Extended Interview. The tobacco use status is as
reported in the Wave 2 Adult Interview. For each characteristic, the sum of the counts in all the categories may not be equal to the
count in the overall row due to missing values.

Response rate = B/A.

In addition to the overall row, each table includes rows on tobacco use status, sex, race, and
ethnicity subgroups. Table 2-6 includes rows on age subgroups; this is not necessary for Table 2-7
because the age range among Wave 2 aged-up adults is narrow. Information from the Wave 2 Adult
Interview was used to define the age categories and tobacco use status; information from the
Wave 1 Adult Interview was used to define the sex, race, and ethnicity categories (because these
characteristics were not re-asked of Wave 2 adults). Adults with missing values for such
characteristics were excluded from the response rate calculation for that characteristic.

Results
As shown in Table 2-6, the unweighted response rate for urine was approximately 96 percent
among continuing adults from whom a specimen was requested. This is considerably higher than the

Population Assessment of Tobacco and Health Study

80 percent response rate projected in the Revision Request to OMB for Wave 2. The response rates
vary little across the various tobacco use status and demographic subgroups.
As shown in Table 2-7, the unweighted response rate for urine among aged-up adults was
approximately 83 percent. This is considerably higher than the projected response rate of 69 percent.
It appears that current established tobacco users at Wave 2 had a higher tendency to provide urine
specimens than those who were not, and Black aged-up adults had a higher response rate for the
urine specimen collection than aged-up adults of other races. There is little variation in urine
specimen response rates by sex or ethnicity among aged-up adults.
The unweighted response rate for blood among aged-up adults was approximately 47 percent. This
is higher than the projected response rate of 45 percent. The differential unweighted response rate
for subgroups of respondents ranges from 1 percentage point for sex to 8 percentage points for
tobacco use status.

2.2

Nonresponse Bias Analysis

The standard approach for an analysis of nonresponse bias in the first follow-up of a longitudinal
cohort study such as the PATH Study is to compare Wave 2 respondents with Wave 2
nonrespondents with respect to characteristics from Wave 1 (Bose and West, 2002; Javitz and
Wagner, 2005; Brownstein et al., 2009). By so doing, the study can assess the extent to which
differential nonresponse among population subgroups may affect estimates. Results are presented
on the characteristics of respondents to the Wave 2 Adult and Youth Interviews, and on aged-up
adults from whom biospecimens were collected at Wave 2. The response rates for urine collection
from continuing adults are above 95 percent and vary little for all the subgroups in Table 2-6; further
analysis of potential nonresponse bias for this aspect of Wave 2 biospecimen collection is therefore
not warranted.
Analyses of bias for the Wave 2 interviews are presented for demographic and tobacco use
subgroups based on their Wave 1 characteristics because Wave 2 characteristics are not available for
Wave 2 nonrespondents. All persons asked to provide a biospecimen at Wave 2 completed the
Wave 2 Adult Interview so the subgroup definitions for the biospecimen nonresponse analysis use
Wave 2 data for characteristics where information was updated (tobacco use status), and Wave 1
data otherwise.

Population Assessment of Tobacco and Health Study

The PATH Study measures a range of tobacco use behaviors; many of these variables are not
available in other studies. However, responses to the PATH Study questions on current cigarette
smoking can be compared with estimates from other surveys that ask about cigarette smoking
behavior. A separate component of the nonresponse bias analysis investigates possible differences
between cigarette smoking estimates calculated from Wave 2 of the PATH Study and independent
estimates of those quantities from other studies for the most similar time periods for which
estimates were available.
Both components of the analysis of nonresponse bias are based on the full set of Wave 2 data.

2.2.1

Method

The method used in the PATH Study to assess potential interview nonresponse bias begins by
comparing estimates of Wave 1 demographic characteristics and tobacco use for Wave 2
respondents with corresponding estimates for Wave 2 nonrespondents. 6 These comparisons are
made separately for continuing adults, continuing youth, aged-up adults, and aged-up youth. To
asses potential nonresponse bias associated with the collection of urine and blood specimens from
aged-up adults at Wave 2, estimates of Wave 1 demographic characteristics and Wave 2 tobacco use
are compared for specimen providers and non-providers, separately for each type of biospecimen.
Sections 2.2 and 2.3 of the 2015 Interim Report describe the weight construction for Wave 1 of the
PATH Study. The final raked weights from Wave 1 were designed to reduce the potential
nonresponse bias from Wave 1. For Wave 2, the interview nonresponse bias analysis uses the raked
weights from Wave 1. Differences between the weighted estimates of Wave 1 characteristics for
Wave 2 respondents and nonrespondents therefore identify characteristics that might be associated
with nonresponse bias due to attrition between the first two waves of the study, after compensating
for Wave 1 nonresponse and possible undercoverage.
Section 2.3.1 of this Interim Report describes the weight construction for Wave 2 of the PATH
Study. The final raked weights from Wave 2 were designed to reduce the potential nonresponse bias
from Wave 2. For Wave 2, the biospecimen nonresponse bias analysis uses the raked weights from
Wave 2 because all aged-up adults asked to provide urine and blood specimens were Wave 2 Adult
6

Persons determined to be ineligible for Wave 2 (see Section 1.2.2) are not included in this analysis.

Population Assessment of Tobacco and Health Study

Interview respondents. Differences between the weighted estimates of characteristics for Wave 2
specimen providers and non-providers therefore identify characteristics that might be associated
with bias due to biospecimen nonresponse at Wave 2, after accounting for Wave 1 nonresponse and
possible undercoverage, and attrition between the first two waves of the study.
Wave 2 cigarette smoking estimates were compared to estimates based on data from the following
surveys: TUS-CPS, 2010-2011; NHIS, 2014; NHANES, 2013-2014; NSDUH, 2014; and NYTS,
2014. Appendix A describes the questions used to define current cigarette smoking on each of these
surveys as well as the PATH Study, and outlines differences in target populations among these
surveys and the PATH Study. This second component of the nonresponse bias analysis uses the
Wave 1 IPS weights. Differences between the weighted estimates of current cigarette smoking
behavior reported by Wave 2 respondents therefore identify characteristics that might be associated
with bias due to nonresponse at Wave 1 or attrition between the first two waves of the study.
In the tables presented in Section 2.2.2, the unweighted counts and estimates of percentages
calculated using weights exclude respondents with missing values for that item. The estimates
calculated from other surveys that are used for comparison purposes also exclude missing values.
The proportions of item missingness are generally very low in both the PATH Study and the surveys
that are used for comparison purposes.
Point estimates for the PATH Study were calculated using the Wave 1 final weights or Wave 1 IPS
weights or Wave 2 final weights, as described above. The corresponding replicate weights were used
to calculate variances, and account for the complex sampling features of stratification and clustering.
Precisions for the PATH Study estimates are reported using 95 percent confidence intervals based
on the modified Wilson confidence interval approach (Wilson, 1927; SAS Institute, 2013). Estimates
from TUS-CPS, NHANES, NHIS, NSDUH, and NYTS 7 also have sampling error, so 95 percent
confidence intervals are reported for the estimates from those surveys as well. 8 SAS software version
9.4 was used to calculate all point estimates and confidence intervals.
7

Estimates from TUS-CPS were obtained from United States Department of Commerce (2012) published tables.
Estimates from NHANES, NHIS, NYTS, and NSDUH were calculated from their respective public use files
(Centers for Disease Control and Prevention (2015a, 2015b, 2015c), Center for Behavioral Health Statistics and
Quality (2015a)).

The modified Wilson confidence interval approach was used to compute the 95 percent confidence intervals for
NHANES, NHIS, NSDUH, and NYTS; the TUS-CPS documentation does not state what method was used to
produce the confidence intervals that appear in the published tables. If the confidence interval from the PATH Study
estimate does not overlap with the confidence interval from the comparison study, then the results are significantly
different at the 0.05 significance level. Schenker and Gentleman (2001) show that this results in a conservative test. In

Population Assessment of Tobacco and Health Study

2.2.2

Results

Tables 2-8 and 2-9 compare Wave 1 demographic characteristics 9 and tobacco use rates,
respectively, for Wave 2 continuing adult respondents with nonrespondents. Tables 2-10 and 2-11
present similar comparisons for continuing youth at Wave 2. Tables 2-12 and 2-13 present the
comparisons for Wave 2 aged-up adult respondents and nonrespondents. Table 2-14 compares
Wave 1 demographic characteristics for Wave 2 aged-up youth respondents with nonrespondents
(tobacco use was not measured for shadow youth at Wave 1). If nonresponse at Wave 2 is not
associated with demographic characteristics or Wave 1 tobacco use, then the estimates calculated
using the Wave 1 final weights would be expected to be similar for Wave 2 respondents and
nonrespondents.
The results in Table 2-8 show that males, 18-24 year-olds, and those with high school education are
underrepresented among continuing adult respondents; and 45-64 year-olds, persons with health
insurance, and those with at least a bachelor’s degree are overrepresented among respondents. Table
2-9 shows that estimates of Wave 1 current established tobacco use are lower overall, and for males,
18-44 year-olds, and non-Hispanic Whites among continuing adult respondents. Tables 2-10 and
2-11 show that estimates of Wave 1 ever use of tobacco are lower for continuing youth respondents
than nonrespondents. Table 2-11 shows that estimates of Wave 1 ever use of tobacco are also lower
for females, 14-17 year-olds, and non-Hispanic Whites among continuing youth respondents. If no
further weighting adjustments were performed then to the extent that these characteristics are
associated with the PATH Study’s outcomes, those outcomes may be affected by nonresponse bias.
However, this concern is addressed by the weighting adjustments and results described in Section
2.3.
No evidence of potential nonresponse bias was found for aged-up adults or aged-up youth based on
the estimates in Tables 2-12, 2-13, and 2-14 because all the confidence intervals for estimates of
difference between Wave 2 respondents and nonrespondents include zero.

general, if a 95 percent confidence interval for the percentage of adults who are current cigarette smokers from the
PATH Study includes a fixed value X, then a hypothesis test of the null hypothesis that the percentage of adults who
are current smokers equals X would have p-value > 0.05 and therefore the difference between the PATH Study
estimate and the estimate from the external survey is not statistically significant. No adjustments were made for
multiple testing; however, results are reported here for all of the comparisons performed, and all of these were preplanned comparisons.
9

Health insurance coverage is also included in Table 2-8.

Population Assessment of Tobacco and Health Study

Table 2-8.

Comparison of Wave 2 continuing Adult Interview respondents with nonrespondents
Wave 2 continuing adult respondents

Characteristic at Wave 1a
Sex

Unweighted
count

Male

13,080

Female

13,356

Age group
18-24

7,330

25-44

9,233

45-64

7,375

65+

2,516

Race/ethnicity
Non-Hispanic White alone

15,767

Other

10,252

Health insurance
Yes

21,269

4,924

Education
< HS or GED

5,279

6,056

Some college, no degree

9,308

Wave 2 continuing adult nonrespondents

Weighted percentage, using
adult Wave 1 final weights
[95% confidence interval]

Unweighted
count

47.2%
[46.6%, 47.8%]
52.8%
[52.2%, 53.4%]

3,067
2,525

12.8%
[12.4%, 13.2%]
34.2%
[33.6%, 34.7%]
35.1%
[34.6%, 35.7%]
17.9%
[17.4%, 18.4%]

1,742
1,983
1,356
511

66.2%
[65.6%, 66.7%]
33.8%
[33.3%, 34.4%]

3,371
2,116

85.9%
[85.3%, 86.5%]
14.1%
[13.5%, 14.7%]

4,281
1,186

16.4%
[15.9%, 16.8%]
23.8%
[23.3%, 24.3%]
31.2%
[30.6%, 31.7%]

1,097
1,421
1,923

Weighted percentage, using
adult Wave 1 final weights
[95% confidence interval]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]

51.4%
[49.6%, 53.2%]
48.6%
[46.8%, 50.4%]

-4.2%
[-6.4%, -2.0%]
4.2%
[2.0%, 6.4%]

14.2%
[13.3%, 15.1%]
35.5%
[34.2%, 36.9%]
31.9%
[30.4%, 33.5%]
18.4%
[16.9%, 20.0%]

-1.4%
[-2.4%, -0.3%]
-1.3%
[-3.0%, 0.3%]
3.2%
[1.3%, 5.1%]
-0.5%
[-2.4%, 1.4%]

65.1%
[63.5%, 66.6%]
34.9%
[33.4%, 36.5%]

1.1%
[-0.8%, 3.0%]
-1.1%
[-3.0%, 0.8%]

83.3%
[81.8%, 84.6%]
16.7%
[15.4%, 18.2%]

2.6%
[1.1%, 4.2%]
-2.6%
[-4.2%, -1.1%]

17.8%
[16.5%, 19.1%]
26.5%
[24.9%, 28.1%]
30.7%
[29.2%, 32.2%]

-1.4%
[-2.9%, 0.1%]
-2.7%
[-4.5%, -0.8%]
0.5%
[-1.3%, 2.2%]

Population Assessment of Tobacco and Health Study

Table 2-8.

Comparison of Wave 2 continuing Adult Interview respondents with nonrespondents (continued)

Characteristic at Wave 1a
Education (continued)

Wave 2 continuing adult respondents
Weighted percentage,
using adult Wave 1 final
Unweighted
weights
count
[95% confidence interval]

Wave 2 continuing adult nonrespondents
Weighted percentage,
using adult Wave 1 final
Unweighted
weights
count
[95% confidence interval]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]

5,684

28.6%
[28.1%, 29.2%]

1,087

25.0%
[23.4%, 26.8%]

3.6%
[1.6%, 5.6%]

Current established user

11,693

23.5%
[22.9%, 24.2%]

2,541

25.6%
[24.5%, 26.8%]

-2.1%
[-3.4%, -0.8%]

Not current established
user

14,039

76.5%
[75.8%, 77.1%]

2,864

74.4%
[73.2%, 75.5%]

2.1%
[0.8%, 3.4%]

Bachelor degree +
Tobacco use statusb

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A ‘current established user’ of a given tobacco
product is someone who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has
reported they ever used that product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes,
smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 2-9.

Comparison of Wave 1 tobacco use* rates for Wave 2 continuing Adult Interview respondents with nonrespondents
Wave 2 continuing adult respondents

Characteristic at Wave 1a
Overall

Unweighted
count
25,732

Sex
Male

12,738

Female

12,973

Age group
18-24

7,214

25-44

9,052

45-64

7,113

65+

2,348

Race/ethnicity
Non-Hispanic White alone
Other
a

15,428
9,923

Wave 2 continuing adult nonrespondents

Weighted percentage, using
adult Wave 1 final weights
[95% confidence interval]
23.5%
[22.9%, 24.2%]

Unweighted
count
5,405

29.0%
[28.0%, 30.0%]
18.6%
[17.8%, 19.3%]

2,961
2,439

28.2%
[26.7%, 29.7%]
28.6%
[27.5%, 29.6%]
23.0%
[22.0%, 24.0%]
10.8%
[9.6%, 12.1%]

1,708
1,938
1,293
460

24.6%
[23.7%, 25.6%]
21.3%
[20.4%, 22.2%]

3,272
2,039

25.6%
[24.5%, 26.8%]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]
-2.1%
[-3.4%, -0.8%]

33.2%
[31.5%, 34.9%]
17.7%
[16.2%, 19.2%]

-4.1%
[-6.1%, -2.1%]
0.9%
[-0.6%, 2.4%]

32.6%
[30.2%, 35.1%]
31.1%
[29.1%, 33.2%]
24.5%
[22.2%, 26.9%]
10.4%
[7.9%, 13.5%]

-4.4%
[-6.9%, -2.0%]
-2.5%
[-4.9%, -0.1%]
-1.5%
[-3.7%, 0.7%]
0.4%
[-1.9%, 2.7%]

28.7%
[27.1%, 30.2%]
20.3%
[18.4%, 22.4%]

-4.1%
[-5.6%, -2.5%]
1.0%
[-1.3%, 3.3%]

Weighted percentage, using
adult Wave 1 final weights
[95% confidence interval]

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

* Defined as current established use of tobacco. A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A
‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their
lifetime and, for any other tobacco product, has reported they ever used that product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional
cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 2-10.

Comparison of Wave 2 continuing Youth Interview respondents with nonrespondents

Characteristic at Wave 1a
Sex

Wave 2 continuing youth respondents
Weighted percentage,
using youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]

Male

5,164

Female

4,892

Age group
12-13

4,164

14-17

5,917

Race/ethnicity
Non-Hispanic White
alone
Other

4,731
5,141

Tobacco use statusb
Ever user

1,665

Never user

8,006

Wave 2 continuing youth nonrespondents
Unweighted
count

51.4%
[50.5%, 52.4%]
48.6%
[47.6%, 49.5%]

665
641

41.1%
[40.1%, 42.1%]
58.9%
[57.9%, 59.9%]

514
801

54.6%
[53.6%, 55.5%]
45.4%
[44.5%, 46.4%]

616
667

16.9%
[16.0%, 17.9%]
83.1%
[82.1%, 84.0%]

261
982

Weighted percentage, using
youth Wave 1 final weights
[95% confidence interval]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]

50.6%
[47.9%, 53.3%]
49.4%
[46.7%, 52.1%]

0.8%
[-2.3%, 3.9%]
-0.8%
[-3.9%, 2.3%]

39.4%
[36.8%, 42.1%]
60.6%
[57.9%, 63.2%]

1.7%
[-1.0%, 4.3%]
-1.7%
[-4.3%, 1.0%]

53.8%
[50.7%, 57.0%]
46.2%
[43.0%, 49.3%]

0.7%
[-2.9%, 4.3%]
-0.7%
[-4.3%, 2.9%]

20.4%
[18.2%, 22.7%]
79.6%
[77.3%, 81.8%]

-3.4%
[-5.7%, -1.2%]
3.4%
[1.2%, 5.7%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A ‘never user’ is someone who has never used any of
those tobacco products. Ever use of a tobacco product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 2-11.

Comparison of Wave 1 tobacco use* rates for Wave 2 continuing Youth Interview respondents with nonrespondents

Characteristic at Wave 1a
Overall
Sex

Wave 2 continuing youth respondents
Weighted percentage,
using youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]
16.9%
9,671
[16.0%, 17.9%]

Male

4,922

Female

4,727

Age group
12-13

3,923

14-17

5,748

Race/ethnicity

Non-Hispanic White alone

4,555

Other

4,930

Wave 2 continuing youth nonrespondents
Unweighted
count
1,243

18.0%
[16.9%, 19.2%]
15.8%
[14.7%, 17.0%]

619
616

7.8%
[6.9%, 9.0%]
23.1%
[21.9%, 24.3%]

473
769

17.6%
[16.4%, 19.0%]
16.3%
[15.1%, 17.4%]

583
633

Weighted percentage, using
youth Wave 1 final weights
[95% confidence interval]
20.4%
[18.2%, 22.7%]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]
-3.4%
[-5.7%, -1.2%]

21.4%
[18.2%, 25.0%]
19.4%
[16.5%, 22.8%]

-3.4%
[-7.0%, 0.2%]
-3.7%
[-6.9%, -0.4%]

7.9%
[5.7%, 10.7%]
28.1%
[24.9%, 31.6%]

-0.0%
[-2.7%, 2.7%]
-5.1%
[-8.4%, -1.8%]

24.0%
[20.6%, 27.8%]
16.6%
[13.7%, 20.1%]

-6.4%
[-9.8%, -2.9%]
-0.4%
[-3.7%, 3.0%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

* Defined as ever use of tobacco. An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. Ever use of a tobacco
product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional cigars, cigarillos, little
filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 2-12.

Comparison of Wave 2 aged-up Adult Interview respondents with nonrespondents

Characteristic at Wave 1a
Sex

Wave 2 aged-up adult respondents
Weighted percentage,
using youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]

Male

957

Female

956

Race/ethnicity
Non-Hispanic White alone

954

Other

945

Tobacco use statusb
Ever user
Never user

811
1,061

Wave 2 aged-up adult nonrespondents
Unweighted
count

50.4%
[48.1%, 52.6%]
49.6%
[47.4%, 51.9%]

169
146

54.7%
[52.4%, 56.9%]
45.3%
[43.1%, 47.6%]

173
141

42.9%
[40.4%, 45.5%]
57.1%
[54.5%, 59.6%]

126
180

Weighted percentage, using
youth Wave 1 final weights
[95% confidence interval]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]

54.3%
[48.8%, 59.8%]
45.7%
[40.2%, 51.2%]

-4.0%
[-10.3%, 2.3%]
4.0%
[-2.3%, 10.3%]

58.9%
[52.5%, 65.0%]
41.1%
[35.0%, 47.5%]

-4.2%
[-11.4%, 3.0%]
4.2%
[-3.0%, 11.4%]

40.5%
[34.7%, 46.5%]
59.5%
[53.5%, 65.3%]

2.4%
[-4.3%, 9.2%]
-2.4%
[-9.2%, 4.3%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A ‘never user’ is someone who has never used any of
those tobacco products. Ever use of a tobacco product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 2-13.

Comparison of Wave 1 tobacco use* rates for Wave 2 aged-up Adult Interview respondents with nonrespondents

Characteristic at Wave 1a
Overall
Sex

Wave 2 aged-up adult respondents
Weighted percentage,
using youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]
42.9%
1,872
[40.4%, 45.5%]

Male

933

Female

936

Race/ethnicity

Non-Hispanic White alone

936

Other

919

Wave 2 aged-up adult nonrespondents
Unweighted
count
306

46.6%
[42.8%, 50.3%]
39.4%
[36.3%, 42.5%]

164
142

46.7%
[42.9%, 50.5%]
38.1%
[34.2%, 42.3%]

169
136

Weighted percentage, using
youth Wave 1 final weights
[95% confidence interval]
40.5%
[34.7%, 46.5%]

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]
2.4%
[-4.3%, 9.2%]

44.0%
[35.7%, 52.5%]
36.4%
[27.9%, 45.9%]

2.6%
[-7.4%, 12.6%]
3.0%
[-6.5%, 12.5%]

43.8%
[36.5%, 51.3%]
36.0%
[26.8%, 46.4%]

2.9%
[-4.8%, 10.6%]
2.2%
[-9.4%, 13.7%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

Population Assessment of Tobacco and Health Study

Table 2-14.

Comparison of Wave 2 aged-up Youth Interview respondents with nonrespondents

Characteristic at Wave 1a
Sex

Wave 2 aged-up youth respondents
Weighted percentage,
using shadow youth
Unweighted
Wave 1 final weights
count
[95% confidence interval]

Male

1,055

Female

1,036

Race/ethnicity

Non-Hispanic White alone

1,008

Other

1,078

Wave 2 aged-up youth nonrespondents
Weighted percentage, using
shadow youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]

49.9%
[47.8%, 52.1%]
50.1%
[47.9%, 52.2%]

237
222

52.4%
[50.2%, 54.5%]
47.6%
[45.5%, 49.8%]

215
243

Difference in weighted
percentages
[respondents –
nonrespondents]
[95% confidence interval]

51.4%
[46.8%, 55.9%]
48.6%
[44.1%, 53.2%]

-1.4%
[-6.9%, 4.0%]
1.4%
[-4.0%, 6.9%]

51.3%
[46.7%, 55.9%]
48.7%
[44.1%, 53.3%]

1.1%
[-4.4%, 6.5%]
-1.1%
[-6.5%, 4.4%]

The characteristics are as reported in the Household Screener at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall
row due to missing values.

Population Assessment of Tobacco and Health Study

Table 2-15 compares the characteristics of aged-up adults who provided a urine specimen with those
of aged-up adults who did not provide a urine specimen. Table 2-16 presents a similar analysis for
the collection of blood specimens from aged-up adults. If nonresponse to the Wave 2 biospecimen
collection is not associated with demographic characteristics or Wave 2 tobacco use, then the
estimates calculated using the Wave 2 final weights would be expected to be similar for Wave 2
specimen providers and non-providers. Based on the results in Tables 2-15 and 2-16, no evidence of
nonresponse bias was found for sex and race/ethnicity. However, for both urine and blood, current
established tobacco users are overrepresented among aged-up adults who provided specimens at
Wave 2. This finding is consistent with the biospecimen response rates shown in Table 2-7.
Tables 2-17 and 2-18 address the second component of the nonresponse bias analysis and compare
estimates of cigarette smoking calculated from Wave 2 of the PATH Study to independent estimates
of those quantities from other studies.
Table 2-17 presents the estimates of prevalence of current cigarette smoking 10 for adults based on
the Wave 2 Adult Interview, for the adult population as a whole and for subgroups. These estimates
are accompanied by 95 percent confidence intervals for the percentage of current cigarette smokers
for the PATH Study estimates. The unweighted estimates are much higher than the weighted
estimates due to the oversampling of adult tobacco users at Wave 1. The last five columns are the
estimates of smoking prevalence from TUS-CPS, NHIS, NHANES, and NSDUH, respectively,
along with 95 percent confidence intervals from those surveys. Note that these estimates exclude
missing values.
The estimates of current smoking prevalence differ from survey to survey. Many potential reasons
can explain these disparities, including that each survey has sampling error. Beyond that, however,
the surveys differ in question order, context, design, mode of administration, and year of most
recent data collection.
In general, the TUS-CPS estimates of smoking prevalence are lower than estimates from the other
surveys, including the PATH Study. This may be related to the proxy responses used in the TUSCPS. The rotation group structure of the TUS-CPS may result in underestimates of smoking
prevalence, as smokers are more likely to drop out over the course of the panel survey (Song, 2013).
10

For the PATH Study, following common practice for tobacco surveys, a current cigarette smoker is someone who (1)
has smoked at least 100 cigarettes in his or her lifetime and (2) currently smokes cigarettes every day or some days.
The questions used to define current cigarette smoking for each survey are provided in Appendix A.

Table 2-15.

Population Assessment of Tobacco and Health Study

Comparison of Wave 2 aged-up adult urine specimen providers with non-providers of urine

Wave 2 aged-up adult non-providers of
urine

Wave 2 aged-up adult urine providers

Characteristica

Unweighted
count

Sex
Male

792

Female

792

Race/ethnicity
Non-Hispanic White alone

781

Other

790

Tobacco use statusb
Current established user
Not current established user

295
1,285

Weighted percentage, using
adult Wave 2 final weights
[95% confidence interval]
50.8%
[48.3%, 53.2%]
49.2%
[46.8%, 51.7%]

Unweighted
count
165
164

54.8%
[52.4%, 57.3%]
45.2%
[42.7%, 47.6%]

173
155

18.7%
[16.5%, 21.0%]
81.3%
[79.0%, 83.5%]

33
293

Weighted percentage, using
adult Wave 2 final weights
[95% confidence interval]

Difference in weighted
percentages
[providers – non-providers]
[95% confidence interval]

50.5%
[45.0%, 56.0%]
49.5%
[44.0%, 55.0%]

0.3%
[-6.3%, 6.9%]
-0.3%
[-6.9%, 6.3%]

56.2%
[50.8%, 61.5%]
43.8%
[38.5%, 49.2%]

-1.4%
[-7.3%, 4.5%]
1.4%
[-4.5%, 7.3%]

10.7%
[6.9%, 16.1%]
89.3%
[83.9%, 93.1%]

8.0%
[2.7%, 13.2%]
-8.0%
[-13.2%, -2.7%]

The sex and race/ethnicity characteristics are as reported in the Wave 1 Youth Extended Interview. The tobacco use status is as reported in the Wave 2 Adult Interview. For each characteristic, the sum
of the counts in all the categories may not be equal to the count in the overall row due to missing values.

A tobacco user is defined as someone who uses one or more of the following tobacco products covered by the Wave 2 Adult Interview: cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes,
smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco. A ‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days and:
for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used that product regularly.

Table 2-16.

Population Assessment of Tobacco and Health Study

Comparison of Wave 2 aged-up adult blood specimen providers with non-providers of blood

Wave 2 aged-up adult non-providers of
blood

Wave 2 aged-up adult blood providers

Characteristica

Unweighted
count

Sex
Male

457

Female

450

Race/ethnicity
Non-Hispanic White alone

445

Other

454

Tobacco use statusb
Current established user

177

Not current established user

727

Weighted percentage, using
adult Wave 2 final weights
[95% confidence interval]
52.0%
[48.7%, 55.2%]
48.0%
[44.8%, 51.3%]

Unweighted
count
500
506

55.1%
[51.8%, 58.3%]
44.9%
[41.7%, 48.2%]

509
491

20.2%
[17.2%, 23.5%]
79.8%
[76.5%, 82.8%]

151
851

Weighted percentage, using
adult Wave 2 final weights
[95% confidence interval]

Difference in weighted
percentages
[providers – non-providers]
[95% confidence interval]

49.6%
[46.6%, 52.7%]
50.4%
[47.3%, 53.4%]

2.4%
[-2.7%, 7.4%]
-2.4%
[-7.4%, 2.7%]

55.1%
[52.0%, 58.2%]
44.9%
[41.8%, 48.0%]

-0.0%
[-5.4%, 5.4%]
0.0%
[-5.4%, 5.4%]

14.7%
[12.5%, 17.3%]
85.3%
[82.7%, 87.5%]

5.4%
[1.4%, 9.5%]
-5.4%
[-9.5%, -1.4%]

A tobacco user is defined as someone who uses one or more of the following tobacco products covered by the Wave 2 Adult Interview: cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes,
smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco. A ‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days and:
for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used that product regularly.

Table 2-17.

Population Assessment of Tobacco and Health Study

Current cigarette smoking based on Wave 2 Adult Interview respondents: IPS weights

Unweighted
count

PATH
Study:
Unweighted
percentage

28,337

34.2%

14,014

35.3%

14,297

33.3%

8,173

22.9%

9,872

40.3%

7,525

42.4%

2,763

23.9%

5,033

25.0%

16,707

37.1%

6,143

33.7%

Current every-day smoker

28,337

26.3%

Current some-days smoker

28,337

7.9%

Characteristic at Wave 2
Current smoker
Current smoker,
male
Current smoker,
female
Current smoker,
age 18-24
Current smoker,
age 25-44
Current smoker,
age 45-64
Current smoker,
age 65+
Current smoker,
Hispanic
Current smoker,
White non-Hispanic
Current smoker,
other non-Hispanic

PATH Study:
Weighted
percentage,
using adult IPS
weights
[95%
confidence
interval]
19.0%
[18.1%, 19.9%]
21.2%
[20.2%, 22.3%]
17.2%
[16.2%, 18.1%]
18.3%
[17.2%, 19.4%]
23.5%
[22.3%, 24.8%]
20.1%
[19.0%, 21.2%]
8.2%
[7.2%, 9.3%]
14.4%
[13.3%, 15.5%]
19.4%
[18.2%, 20.6%]
21.8%
[20.6%, 23.1%]
14.5%
[13.7%, 15.4%]
4.4%
[4.2%, 4.7%]

Percentage
from 20102011 TUS-CPS
[95%
confidence
interval]
16.1%
[15.8%, 16.3%]
18.0%
[17.7%, 18.4%]
14.2%
[13.9%, 14.5%]
17.1%
[16.4%, 17.8%]
17.9%
[17.5%, 18.4%]
17.8%
[17.4%, 18.2%]
7.8%
[7.5%, 8.2%]
10.9%
[10.4%, 11.5%]
17.5%
[17.2%, 17.8%]
NA
12.7%
[12.4%, 12.9%]
3.4%
[3.3%, 3.5%]

Percentage from
2014 NHIS
[95% confidence
interval]
16.7%
[16.1%, 17.4%]
18.8%
[18.0%, 19.7%]
14.8%
[13.9%, 15.7%]
16.7%
[14.2%, 19.5%]
20.0%
[19.0%, 21.0%]
18.0%
[17.0%, 19.0%]
8.5%
[7.7%, 9.4%]
11.2%
[10.2%, 12.2%]
18.2%
[17.3%, 19.1%]
16.2%
[15.2%, 17.2%]
12.8%
[12.3%, 13.4%]
3.9%
[3.6%, 4.2%]

Percentage from
2013-2014
NHANES
[95% confidence
interval]
20.0%
[17.8%, 22.3%]
21.6%
[19.4%, 24.1%]
18.4%
[15.6%, 21.6%]
23.4%
[19.2%, 28.3%]
23.1%
[20.7%, 25.7%]
21.6%
[17.6%, 26.2%]
8.4%
[6.9%, 10.1%]
13.7%
[11.4%, 16.3%]
20.9%
[17.7%, 24.5%]
21.7%
[18.9%, 24.8%]
16.1%
[14.2%, 18.2%]
3.9%
[3.2%, 4.7%]

Percentage from
2014 NSDUH,
original definitiona
[95% confidence
interval]
22.7%
[22.1%, 23.2%]
25.5%
[24.7%, 26.3%]
20.1%
[19.3%, 20.8%]

Percentage from
2014 NSDUH,
modified
definitiona
[95% confidence
interval]
20.9%
[20.3%, 21.4%]
23.6%
[22.8%, 24.4%]
18.3%
[17.6%, 19.2%]

NAb

18.5%
[17.0%, 20.1%]
23.9%
[23.2%, 24.7%]
21.8%
[20.4%, 23.3%]

15.2%
[13.8%, 16.6%]
22.7%
[21.9%, 23.4%]
19.4%
[18.0%, 20.9%]

NSDUH’s definition of a current cigarette smoker is someone who has smoked part or all of a cigarette in the past 30 days, which is more expansive than the definition used in the other surveys.
However, NSDUH contains questions on lifetime smoking and current smoking. The modified definition uses these questions to construct a measure of “current smoking” that is comparable to that of
the other surveys (Ryan et al., 2012). The construction of this variable is described in Appendix A.

Detailed age information was not available in the public use file for NSDUH 2014.

Table 2-18.

Cigarette smoking* based on Wave 2 Youth Interview respondents: IPS weights

Characteristic at Wave 2
Ever tried cigarette smoking,
even one or two puffs

Unweighted
count

PATH Study:
Unweighted
percentage

12,148

11.8%

Ever tried smoking, male

6,208

11.8%

Ever tried smoking, female

5,911

11.7%

4,150

3.9%

7,998

15.8%

12,068

4.0%

Ever tried smoking,
age 12-13
Ever tried smoking,
age 14-17
Have smoked in past
30 days

PATH Study: Weighted
percentage,
using youth IPS weights
[95% confidence interval]
11.9%
[11.1%, 12.8%]
12.1%
[11.0%, 13.2%]
11.7%
[10.8%, 12.8%]
4.0%
[3.4%, 4.7%]
16.0%
[14.9%, 17.1%]
4.0%
[3.6%, 4.5%]

Population Assessment of Tobacco and Health Study

Percentage from
2013-2014 NHANES
[95% confidence
interval]
18.7%
[15.3%, 22.6%]
19.2%
[14.3%, 25.4%]
18.0%
[13.0%, 24.4%]
5.1%
[3.0%, 8.6%]
25.2%
[20.8%, 30.2%]
3.9%
[2.7%, 5.5%]

Percentage from
2014 NSDUH
[95% confidence
interval]
14.4%
[13.6%, 15.3%]
14.6%
[13.5%, 15.7%]
14.2%
[13.1%, 15.4%]
3.8%
[3.1%, 4.6%]
19.4%
[18.2%, 20.6%]
5.0%
[4.6%, 5.5%]

Percentage from
2014 NYTS
[95% confidence
interval]
21.7%
[20.2%, 23.3%]
22.5%
[20.8%, 24.2%]
21.0%
[19.2%, 22.9%]
10.3%
[8.6%, 12.3%]
27.4%
[25.4%, 29.4%]
5.8%
[5.2%, 6.5%]

* Defined as ever tried a cigarette, even one or two puffs. For comparison, an additional measure of current smoking commonly applied to youth (having smoked at all in the past 30 days) is also
included in this table.

Population Assessment of Tobacco and Health Study

The PATH Study and NSDUH both use ACASI administration for the tobacco use questions so
that the interviewer does not see responses to the questions. By contrast, TUS-CPS, NHIS, and
NHANES have direct questioning by an interviewer: NHIS and NHANES are conducted in person,
and TUS-CPS is conducted in person and by telephone. The contexts and purposes of these surveys
also differ: CPS is a general survey on unemployment, NHIS and NHANES are general health
surveys, and NSDUH is a cross-sectional survey on substance use (including tobacco use) and
health, including mental health. Unlike the cross-sectional prevalence surveys, the PATH Study is
designed for research purposes and uses a longitudinal cohort design to assess within-person change
and between-person differences in tobacco use behaviors and health over time. Other differences
among the questions used in the instruments of these different studies are outlined in Appendix A.
Table 2-17 indicates the IPS-weighted estimates of current smoking from Wave 2 of the PATH
Study are most similar to estimates from NHIS and NHANES. All of the 95 percent confidence
intervals for percent of current cigarette smokers constructed from the PATH Study overlap with
the confidence intervals for NHIS, NHANES, or both. Estimates from TUS-CPS tend to be below
the estimates from the PATH Study, NHIS, and NHANES; estimates from NSDUH tend to be
above the estimates from the PATH Study, NHIS, and NHANES. No evidence was found to
indicate nonresponse bias in the PATH Study with respect to current cigarette smoking behavior
among adults, because the PATH Study’s estimates are all within the range of estimates from
comparable surveys.
Table 2-18 provides estimates from the PATH Study for two common measures of cigarette
smoking prevalence among Wave 2 youth respondents compared with estimates from NHANES,
NSDUH, and NYTS. 11 Different measures of smoking are used in this report for youth than for
adults. The primary measure of cigarette smoking among youth in this report is whether the youth
has ever tried smoking a cigarette, even one or two puffs (see Appendix A). Another measure is
current smoking, defined as having smoked at all in the past 30 days. Both are shown in Table 2-18.
Differences in target populations and administration among the youth surveys might lead to
differences in their estimates. In addition, the youth survey estimates have sampling error, as
demonstrated by the confidence intervals about the estimates from the comparison surveys.

TUS-CPS does not interview persons younger than 18 about tobacco use.

Population Assessment of Tobacco and Health Study

Questions and their orderings also differ among the surveys, as described in Appendix A, as do the
modes of administration. The PATH Study, NHANES, and NSDUH use ACASI for the questions
about tobacco use by youth, and these are administered individually in a household or mobile
examination center setting. The NYTS is a pencil-and-paper survey that is self-administered in the
classroom. Currivan et al. (2004) found that even when telephone ACASI was used, estimates of
youth smoking prevalence were lower for a telephone survey of youth smoking than for a schoolbased survey of the same population (see also Fowler and Stringfellow, 2001, for a discussion of
higher smoking rates in school-based surveys). In addition, school-based surveys often include
students who are older than 17, which is the upper age limit for youth in the PATH Study.
The PATH Study’s estimates of youth smoking are lower than comparable estimates from
NHANES and NSDUH. Part of this difference may be sampling error and part may be attributable
to differences among the survey wordings and administrations. Moreover, the comparison surveys
are from different time periods. According to the Center for Behavioral Health Statistics and Quality
(CBHSQ) (2015b), cigarette smoking among teens is dropping (from 2013 to 2014, the percentage
of youth who had ever tried smoking dropped by 0.3 percentage points among 12-13 year-olds, 2.4
percentage points among 14-15 year-olds, and 2.1 percentage points among 16-17 year-olds, with
similar decreases from 2012 to 2013). The lower percentages found by the PATH Study may reflect,
in part, a continuation of this trend. However, some of the differences among the estimates of youth
smoking prevalence may be attributable to nonresponse bias or measurement error on the part of
one or more of the surveys.

2.3

Statistical Approach for Addressing Nonresponse

2.3.1

Interviews

The primary approach for addressing nonresponse is to use weight adjustments that account for
differential response propensities across subgroups. This section describes the weighting
adjustments used to address the areas of potential nonresponse bias identified in Section 2.2 and
evaluates their effectiveness.
Among numerous sources, the handbook on household surveys by the United Nations (2005,
Chapter 6) and Särndal and Lundström (2005) discuss the methods and theory of using weight
adjustments for nonresponse. For Wave 1, these adjustments were done at the household level and

Population Assessment of Tobacco and Health Study

at the person level. The Wave 1 household-level weighting adjustments calibrated the estimates to
household-level population estimates for census region and household composition and size
developed from the 1-year 2013 ACS (which provides high precision and accuracy due to a large
sample size and high response rate). Such weighting adjustments also correct for disparities among
other characteristics that might be associated with the characteristics involved in the weighting
adjustments. After accounting for household-level nonresponse, households with at least one person
sampled for the PATH Study were identified, and each sampled person within a household was
assigned the corresponding household weight. These weights were then adjusted to account for
nonresponse to the Wave 1 interview (for adults and youth) or to non-participation in the study (for
shadow youth). After this adjustment for nonresponse, the weights were calibrated using a raking
process to person-level population estimates also developed from the 1-year 2013 ACS. Outlier
values of the sample weights were trimmed if called for, and the weights were re-raked after any
such trimming.
For Wave 2, the same general approach was used for adults, youth, and shadow youth (all as defined
at Wave 1), separately. The final person-level weights assigned to Wave 1 respondents served as the
initial (“base”) weights for use in developing the Wave 2 weights. These weights were adjusted to
account for nonresponse to the Wave 2 interview and the resulting weights raked to control totals.
However, some of these control totals were sampled-based rather than population-based and they
reflected characteristics of the Wave 1 population (the baseline period for this sample).
Raking to sample-based control totals, often employed with longitudinal studies (see, for example,
Brick, Le, and West (2003)) can help limit drifting from some important baseline characteristics that
might arise through the applications of nonresponse adjustments over time. Lundström and Särndal
(1999) provide a theoretical discussion of the use of calibrating weights to sample-based controls as
well as providing empirical evidence that such calibration can serve to reduce both variance and
nonresponse bias.
In terms of more details on the Wave 2 weighting, the Wave 1 final weights were adjusted for
nonresponse at the person level in two stages. The first adjustment accounted for nonrespondents
whose eligibility status was unknown, largely those people who were unlocatable. A second
adjustment for nonresponse was undertaken to account for nonrespondents known to be eligible for
the Wave 2 interview. This group of nonrespondents consisted mainly of those who were contacted
but chose not to complete the Wave 2 interview.

Population Assessment of Tobacco and Health Study

Nonresponse adjustment cells were created separately for each of the two stages. Variables from
Wave 1 data collection were candidates for contributing to the formation of nonresponse
adjustment cells as were some variables based on paradata. The set of candidate variables varied
somewhat by age group (adult, youth, shadow youth). Examples of the types of candidate variables
are screener characteristics (e.g., household size), segment level characteristics (e.g., race percentages
in segment), Wave 1 tobacco use status variables (for youth and adults), and health-based
characteristics (e.g., BMI). The paradata variables included factors such as the number of contacts
required to complete the Wave 1 interview. A tree-based classification method was employed to
identify subgroups exhibiting differential nonresponse to the Wave 2 interview. [This general
approach is described by Roth et al. (2006) and Schouten and deNooij (2005).] A description of the
computation of each of the two nonresponse adjustment factors follows.
Suppose Wave 1 respondent i was assigned weight W1𝑖𝑖 for Wave 1 and was assigned to cell 𝑐𝑐𝑈𝑈
when adjusting for nonresponse among those whose eligibility status was not ascertained. The
Wave 2 weight accounting for such nonresponse is represented as W2NRUNK 𝑖𝑖 and was computed
as the product of the Wave 1 weight and the adjustment factor associated with cell 𝑐𝑐𝑈𝑈 :
W1𝑖𝑖 × sum of W1

𝑖𝑖

W2NRUNK 𝑖𝑖 =
sum of W1𝑖𝑖 weights for all Wave 1 respondents assigned to cell 𝑐𝑐𝑈𝑈

weights for all Wave 1 respondents with known eligibility status at Wave 2 assigned to cell 𝑐𝑐𝑈𝑈

After this adjustment, all those Wave 1 respondents known to be ineligible for the Wave 2 interview
were removed from the weighting process, leaving for further adjustment purposes only those
known to be eligible for Wave 2, whether respondent or nonrespondent. For this set of people, the
weight W2NRUNK 𝑖𝑖 was further adjusted for nonresponse among those known to be eligible for the
Wave 2 interview.
Let 𝑐𝑐𝐾𝐾 represent the cell to which respondent i was assigned for the nonresponse adjustment among
those known to be eligible for Wave 2. Then the final Wave 2 nonresponse-adjusted weight for
respondent i, represented as W2NRWT𝑖𝑖 , was computed as the product of W2NRUNK 𝑖𝑖 and the
adjustment factor associated with cell 𝑐𝑐𝐾𝐾 as follows
W2NRWT𝑖𝑖 =
sum of W2NRUNK𝑖𝑖 weights for all those eligible for Wave 2 assigned to cell 𝑐𝑐𝐾𝐾

W2NRUNK 𝑖𝑖 × sum of W2NRUNK

𝑖𝑖

weights for all Wave 2 eligible respondents assigned to cell 𝑐𝑐𝐾𝐾

Population Assessment of Tobacco and Health Study

Following the adjustments for nonresponse, the weights were raked to Wave 1 control totals.
Extreme weights were identified and trimmed and the weights were raked again. As mentioned
earlier, some of the control totals for raking were the population-based control totals used in the
Wave 1 raking while others were sampled based. An important aspect to note here is that the
sample-based control totals are subject to non-negligible sampling error which should be reflected in
the establishment of the replicate weights, and this was done as part of the Wave 2 weighting
(specifically, as part of the computation of the replicate weights).
For continuing adult respondents, the population-based controls were based on combinations of
Wave 1 census region, age, race/ethnicity, sex, and educational attainment, all used during the
Wave 1 weight raking process. The sampled-based controls reflected any tobacco use and e-cigarette
use reported at Wave 1 cross-classified with some of the same characteristics (e.g., sex and age). The
tobacco use variable had three levels: current established user, ever user but not current established
user, and never user; those with missing data for tobacco use were pooled for raking purposes with
those associated with the category “ever user but not current established user.” 12 The e-cigarette
variable had two levels: never user and not never user; those with missing data for e-cigarette use
were pooled with those associated with the category “not never user.”
For continuing youth respondents and aged-up adult respondents, the raking was done using
Wave 1 census region, single-year of age, race/ethnicity, and sex (all used during the Wave 1 weight
raking process), as well as any tobacco use (for respondents who were age 16 or 17 at Wave 1) and
e-cigarette use (for respondents who were age 17 at Wave 1); 13 the tobacco use and e-cigarette use
variables were both defined using two levels: never user and not never user; those with missing data
were pooled with those associated with the category “not never user.”
For aged-up youth respondents, the raking was done using Wave 1 census region, single-year of age,
race/ethnicity, and sex.

For the purpose of sample-based raking, cases with missing values were pooled with other cases in such a way that
the categories thought to be of most analytic interest were unaffected.

The ages for which the e-cigarette use and any tobacco use dimensions were used for raking were determined by the
need to have sufficient sample sizes for stability of the weighting adjustments.

Population Assessment of Tobacco and Health Study

A trimming procedure was implemented as called for, after the raking was completed. Whenever
trimming was implemented, the corresponding set of weights was re-raked. After the raking and
trimming process, the final weight, denoted as 𝑊𝑊2𝑅𝑅𝑅𝑅𝑅𝑅𝑇𝑇𝑖𝑖 , was computed for each participant as
𝑊𝑊2𝑅𝑅𝑅𝑅𝑅𝑅𝑇𝑇𝑖𝑖 = 𝑊𝑊2𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑖𝑖 × 𝑅𝑅𝑅𝑅𝑖𝑖 ,

where 𝑅𝑅𝑅𝑅𝑖𝑖 is the combined raking and trimming adjustment factor associated with respondent i.

Replicate weights were created for variance estimation purposes using balanced repeated replication
methodology, reflecting the sample design and weighting process including raking to control totals
that included some PATH Study sample estimates, as described above.
Tables 2-19 through 2-25 repeat the analyses in Tables 2-8 through 2-14; however, the Wave 2
nonrespondent estimates are replaced by estimates for the Wave 1 respondents who remained
eligible at Wave 2, and the Wave 2 respondent estimates use the Wave 2 final weights (rather than
the Wave 1 final weights). Differences between the weighted estimates for Wave 2 and Wave 1
respondents therefore reflect the extent to which the Wave 2 weighting process corrected for
potential nonresponse bias. Across all seven tables, the biggest difference found was 0.8 percent (for
non-Hispanic White aged-up adults, see Table 2-24) and this was not statistically significant at the
.05 level. These results suggest that the Wave 2 weighting nonresponse adjustments were highly
effective.
Table 2-19 shows that the Wave 2 weighting process corrected for potential nonresponse bias
associated with Wave 1 demographic characteristics and health insurance coverage among
continuing adults at Wave 2. All the differences between the point estimates are essentially zero. The
astute reader will notice that the 95 percent confidence intervals around the difference estimates
barely exclude zero for a small number of subgroups (males and 18-44 year-olds), however this is
primarily due to the large sample sizes. 14

Part of the explanation also lies in technical subtleties associated with the Wave 2 weighting process. Persons
ineligible at Wave 2 were included with the Wave 2 respondents when raking to Wave 1 demographics and tobacco
use estimates but are excluded from the estimates in the table. Also, cases with missing values for the tobacco use
estimates used in the Wave 2 raking process were pooled with cases with non-missing values (as described above)
during raking whereas they are excluded from the estimates in the table.

Population Assessment of Tobacco and Health Study

Table 2-19.

Comparison of Wave 2 continuing Adult Interview respondents with Wave 1 Adult Interview respondents

Characteristic at Wave 1a
Sex

Wave 2 continuing adult respondents
Weighted percentage, using
Unweighted
adult Wave 2 final weights
count
[95% confidence interval]

Male

13,080

Female

13,356

Age group
18-24

7,330

25-44

9,233

45-64

7,375

65+

2,516

Race/ethnicity
Non-Hispanic White alone

15,767

Other

10,252

Health insurance
Yes

21,269

4,924

Education
< HS or GED

5,279

6,056

Some college, no degree

9,308

Bachelor degree +

5,684

Wave 1 adult respondentsc
Weighted percentage, using
Unweighted
adult Wave 1 final weights
count
[95% confidence interval]

47.9%
[47.3%, 48.5%]
52.1%
[51.5%, 52.7%]

16,147
15,881

13.0%
[12.6%, 13.5%]
34.4%
[33.8%, 35.0%]
34.6%
[34.0%, 35.2%]
18.0%
[17.5%, 18.5%]

9072
11,216
8,731
3,027

65.9%
[65.4%, 66.5%]
34.1%
[33.5%, 34.6%]

19,138
12,368

85.4%
[84.8%, 86.1%]
14.6%
[13.9%, 15.2%]

25,550
6,110

16.6%
[16.2%, 17.1%]
24.3%
[23.8%, 24.8%]
31.1%
[30.6%, 31.7%]
28.0%
[27.4%, 28.5%]

6,376
7,477
11,231
6,771

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]

47.9%
[47.4%, 48.5%]
52.1%
[51.5%, 52.6%]

-0.0%
[-0.0%, -0.0%]
0.0%
[0.0%, 0.0%]

13.1%
[12.7%, 13.4%]
34.4%
[33.9%, 34.9%]
34.6%
[34.0%, 35.1%]
18.0%
[17.6%, 18.4%]

-0.0%
[-0.0%, -0.0%]
-0.0%
[-0.0%, -0.0%]
0.0%
[-0.0%, 0.0%]
0.0%
[-0.0%, 0.1%]

66.0%
[65.5%, 66.5%]
34.0%
[33.5%, 34.5%]

-0.0%
[-0.1%, 0.0%]
0.0%
[-0.0%, 0.1%]

85.5%
[84.9%, 86.0%]
14.5%
[14.0%, 15.1%]

-0.0%
[-0.2%, 0.2%]
0.0%
[-0.2%, 0.2%]

16.6%
[16.2%, 17.0%]
24.3%
[23.8%, 24.7%]
31.1%
[30.6%, 31.6%]
28.0%
[27.5%, 28.5%]

-0.0%
[-0.0%, 0.0%]
0.0%
[-0.0%, 0.1%]
0.0%
[-0.0%, 0.1%]
-0.0%
[-0.1%, 0.0%]

Population Assessment of Tobacco and Health Study

Table 2-19.

Comparison of Wave 2 continuing Adult Interview respondents with Wave 1 Adult Interview respondents (continued)

Characteristic at Wave 1a
Tobacco use statusb

Wave 2 continuing adult respondents
Weighted percentage, using
Unweighted
adult Wave 2 final weights
count
[95% confidence interval]

Current established user

11,693

Not current established
user

14,039

Wave 1 adult respondentsc
Weighted percentage, using
Unweighted
adult Wave 1 final weights
count
[95% confidence interval]

23.8%
[23.2%, 24.4%]
76.2%
[75.6%, 76.8%]

14,234
16,903

23.9%
[23.3%, 24.5%]
76.1%
[75.5%, 76.7%]

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]
-0.1%
[-0.1%, -0.0%]
0.1%
[0.0%, 0.1%]

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due to
missing values.

A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A ‘current established user’ of a given tobacco product is
someone who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used
that product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, ecigarettes, and dissolvable tobacco.

Excludes those who were ineligible at Wave 2.

Population Assessment of Tobacco and Health Study

Table 2-20.

Comparison of Wave 1 tobacco use* rates for Wave 2 continuing Adult Interview respondents with Wave 1 Adult Interview
respondents

Characteristic at Wave 1a
Overall

Wave 2 continuing adult respondents
Weighted percentage, using
Unweighted
adult Wave 2 final weights
count
[95% confidence interval]
23.8%
25,732
[23.2%, 24.4%]

Sex
Male

12,738

Female

12,973

Age group
18-24

7,214

25-44

9,052

45-64

7,113

65+

2,348

Race/ethnicity
Non-Hispanic White alone
Other

15,428
9,923

29.6%
[28.8%, 30.5%]
18.4%
[17.7%, 19.1%]
29.0%
[27.6%, 30.4%]
28.8%
[27.8%, 29.8%]
23.2%
[22.2%, 24.2%]
10.8%
[9.6%, 12.1%]
25.2%
[24.3%, 26.1%]
21.0%
[20.2%, 21.8%]

Wave 1 adult respondentsb
Weighted percentage, using
Unweighted
adult Wave 1 final weights
count
[95% confidence interval]
23.9%
31,137
[23.3%, 24.5%]
15,699
15,412
8,922
10,990
8,406
2,808
18,700
11,962

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]
-0.1%
[-0.1%, -0.0%]

29.8%
[28.9%, 30.6%]
18.4%
[17.7%, 19.1%]

-0.2%
[-0.3%, -0.1%]
-0.0%
[-0.1%, 0.0%]

29.0%
[27.6%, 30.4%]
29.0%
[28.1%, 29.9%]
23.2%
[22.3%, 24.1%]
10.7%
[9.6%, 11.9%]

-0.0%
[-0.1%, 0.0%]
-0.2%
[-0.5%, 0.1%]
-0.1%
[-0.4%, 0.2%]
0.1%
[-0.3%, 0.5%]

25.3%
[24.4%, 26.2%]
21.1%
[20.4%, 21.9%]

-0.1%
[-0.2%, -0.0%]
-0.1%
[-0.2%, 0.0%]

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due to
missing values.

Excludes those who were ineligible at Wave 2.

* Defined as current established use of tobacco. A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A ‘current
established user’ of a given tobacco product is someone who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any
other tobacco product, has reported they ever used that product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars,
pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 2-21.

Comparison of Wave 2 continuing Youth Interview respondents with Wave 1 youth respondents who were continuing youth at
Wave 2

Characteristic at Wave 1a
Sex

Wave 2 continuing youth respondents
Weighted percentage, using
Unweighted
youth Wave 2 final weights
count
[95% confidence interval]

Male

5,164

Female

4,892

Age group
12-13

4,164

14-17

5,917

Race/ethnicity
Non-Hispanic White alone

4,731

Other

5,141

Tobacco use statusb
Ever user

1,665

Never user

8,006

Wave 1 youth respondentsc
Weighted percentage, using
Unweighted
youth Wave 1 final weights
count
[95% confidence interval]

51.4%
[50.4%, 52.4%]
48.6%
[47.6%, 49.6%]

5,829
5,533

40.7%
[39.8%, 41.7%]
59.3%
[58.3%, 60.2%]

4,678
6,718

54.5%
[53.5%, 55.5%]
45.5%
[44.5%, 46.5%]

5,347
5,808

17.1%
[16.3%, 17.9%]
82.9%
[82.1%, 83.7%]

1,926
8,988

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]

51.3%
[50.4%, 52.3%]
48.7%
[47.7%, 49.6%]

0.0%
[-0.0%, 0.1%]
-0.0%
[-0.1%, 0.0%]

40.9%
[40.0%, 41.8%]
59.1%
[58.2%, 60.0%]

-0.2%
[-0.2%, -0.1%]
0.2%
[0.1%, 0.2%]

54.5%
[53.5%, 55.4%]
45.5%
[44.6%, 46.5%]

0.0%
[-0.1%, 0.1%]
-0.0%
[-0.1%, 0.1%]

17.3%
[16.4%, 18.3%]
82.7%
[81.7%, 83.6%]

-0.2%
[-0.7%, 0.3%]
0.2%
[-0.3%, 0.7%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due
to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A ‘never user’ is someone who has never used any of those tobacco
products. Ever use of a tobacco product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional
cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Excludes those who were ineligible at Wave 2 and also excludes Wave 1 youth who were aged-up adults at Wave 2.

Population Assessment of Tobacco and Health Study

Table 2-22.

Comparison of Wave 1 tobacco use* rates for Wave 2 continuing Youth Interview respondents with Wave 1 youth respondents who
were continuing youth at Wave 2

Characteristic at Wave 1a
Overall

Wave 2 continuing youth respondents
Weighted percentage, using
Unweighted
youth Wave 2 final weights
count
[95% confidence interval]
17.1%
9,671
[16.3%, 17.9%]

Sex
Male

4,922

Female

4,727

Age group
12-13

3,923

14-17

5,748

Race/ethnicity
Non-Hispanic White alone

4,555

Other

4,930

Wave 1 youth respondentsb
Weighted percentage, using
Unweighted
youth Wave 1 final weights
count
[95% confidence interval]
17.3%
10,914
[16.4%, 18.3%]

18.1%
[17.1%, 19.2%]
16.1%
[15.1%, 17.2%]

5,541
5,343

7.8%
[6.8%, 8.9%]
23.3%
[22.3%, 24.5%]

4,396
6,517

18.0%
[16.9%, 19.1%]
16.3%
[15.3%, 17.4%]

5,138
5,563

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]
-0.2%
[-0.7%, 0.3%]

18.4%
[17.3%, 19.6%]
16.2%
[15.2%, 17.4%]

-0.3%
[-1.0%, 0.3%]
-0.1%
[-0.7%, 0.4%]

7.8%
[6.9%, 8.9%]
23.7%
[22.5%, 24.9%]

-0.1%
[-0.4%, 0.2%]
-0.3%
[-1.1%, 0.5%]

18.4%
[17.1%, 19.8%]
16.3%
[15.2%, 17.4%]

-0.4%
[-1.0%, 0.2%]
0.0%
[-0.6%, 0.6%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due
to missing values.

Excludes those who were ineligible at Wave 2 and also excludes Wave 1 youth who were aged-up adults at Wave 2.

* Defined as ever use of tobacco. An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. Ever use of a tobacco product is
defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes,
smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 2-23.

Comparison of Wave 2 aged-up Adult Interview respondents with Wave 1 youth respondents who were aged-up adults at Wave 2

Characteristic at Wave 1a
Sex

Wave 2 aged-up adult respondents
Weighted percentage, using
Unweighted
adult Wave 2 final weights
count
[95% confidence interval]

Male

957

Female

956

Race/ethnicity
Non-Hispanic White alone

954

Other

945

Tobacco use statusb
Ever user
Never user

811
1,061

Wave 1 youth respondentsc
Weighted percentage, using
Unweighted
youth Wave 1 final weights
count
[95% confidence interval]

50.7%
[48.5%, 53.0%]
49.3%
[47.0%, 51.5%]

1,126
1,102

55.1%
[52.8%, 57.3%]
44.9%
[42.7%, 47.2%]

1,127
1,086

43.2%
[40.9%, 45.4%]
56.8%
[54.6%, 59.1%]

937
1,241

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]

51.0%
[48.9%, 53.0%]
49.0%
[47.0%, 51.1%]

-0.2%
[-0.6%, 0.2%]
0.2%
[-0.2%, 0.6%]

55.3%
[53.2%, 57.4%]
44.7%
[42.6%, 46.8%]

-0.2%
[-0.6%, 0.2%]
0.2%
[-0.2%, 0.6%]

42.6%
[40.3%, 44.9%]
57.4%
[55.1%, 59.7%]

0.6%
[-1.7%, 2.8%]
-0.6%
[-2.8%, 1.7%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due
to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A ‘never user’ is someone who has never used any of those tobacco
products. Ever use of a tobacco product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional
cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Excludes those who were ineligible at Wave 2 and also excludes Wave 1 youth who were continuing youth at Wave 2.

Population Assessment of Tobacco and Health Study

Table 2-24.

Comparison of Wave 1 tobacco use* rates for Wave 2 aged-up Adult Interview respondents with Wave 1 youth respondents who
were aged-up adults at Wave 2

Characteristic at Wave 1a
Overall

Wave 2 aged-up adult respondents
Weighted percentage, using
Unweighted
adult Wave 2 final weights
count
[95% confidence interval]
43.2%
1,872
[40.9%, 45.4%]

Sex
Male

933

Female

936

Race/ethnicity
Non-Hispanic White alone

936

Other

919

Wave 1 youth respondentsb
Weighted percentage, using
Unweighted
youth Wave 1 final weights
count
[95% confidence interval]
42.6%
2,178
[40.3%, 44.9%]

46.6%
[43.4%, 49.8%]
39.7%
[36.6%, 42.9%]

1,097
1,078

47.0%
[43.8%, 50.2%]
38.2%
[35.1%, 41.4%]

1,105
1,055

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]
0.6%
[-1.7%, 2.8%]

46.1%
[42.9%, 49.4%]
39.0%
[36.0%, 42.0%]

0.5%
[-2.6%, 3.5%]
0.7%
[-2.2%, 3.6%]

46.2%
[42.8%, 49.7%]
37.9%
[34.4%, 41.4%]

0.8%
[-2.3%, 3.8%]
0.3%
[-2.8%, 3.5%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due
to missing values.

Excludes those who were ineligible at Wave 2 and also excludes Wave 1 youth who were continuing youth at Wave 2.

* Defined as ever use of tobacco. An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. Ever use of a tobacco product is
defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes,
smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 2-25.

Comparison of Wave 2 aged-up Youth Interview respondents with Wave 1 shadow youth who were aged-up youth at Wave 2
Wave 2 aged-up youth respondents

Characteristic at Wave 1a
Sex

Unweighted
count

Male

1,055

Female

1,036

Race/ethnicity
Non-Hispanic White alone

1,008

Other

1,078

Weighted percentage, using
youth Wave 2 final weights
[95% confidence interval]

Wave 1 shadow youthb
Weighted percentage, using
shadow youth Wave 1 final
Unweighted
weights
count
[95% confidence interval]

50.3%
[48.1%, 52.4%]
49.7%
[47.6%, 51.9%]

1,292
1,258

52.1%
[50.0%, 54.3%]
47.9%
[45.7%, 50.0%]

1,223
1,321

Difference in weighted
percentages
[Wave 2 – Wave 1]
[95% confidence interval]

50.2%
[48.3%, 52.1%]
49.8%
[47.9%, 51.7%]

0.1%
[-0.3%, 0.4%]
-0.1%
[-0.4%, 0.3%]

52.2%
[50.2%, 54.1%]
47.8%
[45.9%, 49.8%]

-0.0%
[-0.4%, 0.3%]
0.0%
[-0.3%, 0.4%]

The characteristics are as reported in the Household Screener at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due to
missing values.

Excludes those who were ineligible at Wave 2 and also excludes Wave 1 shadow youth who were continuing shadow youth at Wave 2.

Population Assessment of Tobacco and Health Study

Table 2-20 demonstrates the effectiveness of the Wave 2 weighting process on Wave 1 tobacco use
estimates for continuing adults. The differences between the point estimates are not substantive,
despite the 95 percent confidence intervals indicating that estimates overall and for males and nonHispanic Whites are marginally lower among Wave 2 respondents. The confidence intervals around
the estimates of difference between Wave 1 and Wave 2 respondents are narrow reflecting high
correlation between the two groups over time and the use of sample-based raking to Wave 1
tobacco use estimates.
Among continuing youth, using the Wave 2 final weights, Table 2-21 shows that 12-13 year-olds are
slightly underrepresented among Wave 2 respondents compared to Wave 1 respondents. However,
the magnitude of the difference is not practically meaningful. Table 2-22 shows no evidence of
nonresponse bias for Wave 1 estimates of ever tobacco use among continuing youth at Wave 2.
Similarly, no evidence of potential nonresponse bias was found for aged-up adults or aged-up youth
based on the estimates in Tables 2-23, 2-24, and 2-25.
Estimates of Wave 2 adult cigarette smoking prevalence in Table 2-26 using the Wave 2 final
weights are similar to the estimates using the Wave 1 IPS weights; both are in the range of values
obtained by other surveys. The use of the Wave 2 weights resulted in a slight decrease in estimated
cigarette smoking prevalence for females and non-White non-Hispanics.
Table 2-27 examines the effect of the Wave 2 final weights on estimates calculated for youth.
Cigarette smoking prevalence estimates with the Wave 1 IPS weights and with the Wave 2 weights
were generally lower than estimates from other surveys although, as noted above, the surveys took
place in different time periods.

Population Assessment of Tobacco and Health Study

Table 2-26.

Characteristic at
Wave 2
Current smoker
Current smoker,
male
Current smoker,
female
Current smoker,
age 18-24
Current smoker,
age 25-44
Current smoker,
age 45-64
Current smoker,
age 65+
Current smoker,
Hispanic
Current smoker,
White nonHispanic
Current smoker,
other nonHispanic
Current every-day
smoker
Current some-days
smoker

Current cigarette smoking based on Wave 2 Adult Interview respondents

Unweighted
count

PATH
Study:
Unweighted
percentage

PATH Study:
Weighted
percentage,
using adult IPS
weights
[95%
confidence
interval]

PATH Study:
Weighted
percentage,
using adult
Wave 2 final
weights
[95%
confidence
interval]

19.0%
[18.1%, 19.9%]
21.2%
[20.2%, 22.3%]
17.2%
[16.2%, 18.1%]
18.3%
[17.2%, 19.4%]
23.5%
[22.3%, 24.8%]
20.1%
[19.0%, 21.2%]
8.2%
[7.2%, 9.3%]
14.4%
[13.3%, 15.5%]

18.6%
[18.1%, 19.2%]
21.3%
[20.6%, 22.0%]
16.2%
[15.6%, 16.8%]
18.4%
[17.4%, 19.4%]
23.6%
[22.6%, 24.6%]
19.7%
[18.8%, 20.6%]
8.0%
[7.0%, 9.1%]
14.5%
[13.6%, 15.5%]

37.1%

19.4%
[18.2%, 20.6%]

6,143

33.7%

28,337

26.3%

28,337

7.9%

28,337

34.2%

14,014

35.3%

14,297

33.3%

8,173

22.9%

9,872

40.3%

7,525

42.4%

2,763

23.9%

5,033

25.0%

16,707

Percentage
from 20102011 TUS-CPS
[95%
confidence
interval]

Percentage
from 2014
NHIS
[95%
confidence
interval]

Percentage
from 20132014 NHANES
[95%
confidence
interval]
20.0%
[17.8%, 22.3%]
21.6%
[19.4%, 24.1%]
18.4%
[15.6%, 21.6%]
23.4%
[19.2%, 28.3%]
23.1%
[20.7%, 25.7%]
21.6%
[17.6%, 26.2%]
8.4%
[6.9%, 10.1%]
13.7%
[11.4%, 16.3%]

Percentage
from 2014
NSDUH, original
definitiona
[95% confidence
interval]

Percentage from
2014 NSDUH,
modified
definitiona
[95% confidence
interval]

NAb

18.5%
[17.0%, 20.1%]

15.2%
[13.8%, 16.6%]

16.1%
[15.8%, 16.3%]
18.0%
[17.7%, 18.4%
14.2%
[13.9%, 14.5%]
17.1%
[16.4%, 17.8%]
17.9%
[17.5%, 18.4%]
17.8%
[17.4%, 18.2%]
7.8%
[7.5%, 8.2%]
10.9%
[10.4%, 11.5%]

16.7%
[16.1%, 17.4%]
18.8%
[18.0%, 19.7%]
14.8%
[13.9%, 15.7%]
16.7%
[14.2%, 19.5%]
20.0%
[19.0%, 21.0%]
18.0%
[17.0%, 19.0%]
8.5%
[7.7%, 9.4%]
11.2%
[10.2%, 12.2%]

22.7%
[22.1%, 23.2%]
25.5%
[24.7%, 26.3%]
20.1%
[19.3%, 20.8%]

20.9%
[20.3%, 21.4%]
23.6%
[22.8%, 24.4%]
18.3%
[17.6%, 19.2%]

19.4%
[18.6%, 20.1%]

17.5%
[17.2%, 17.8%]

18.2%
[17.3%, 19.1%]

20.9%
[17.7%, 24.5%]

23.9%
[23.2%, 24.7%]

22.7%
[21.9%, 23.4%]

21.8%
[20.6%, 23.1%]

19.3%
[18.3%, 20.3%]

16.2%
[15.2%, 17.2%]

21.7%
[18.9%, 24.8%]

21.8%
[20.4%, 23.3%]

19.4%
[18.0%, 20.9%]

14.5%
[13.7%, 15.4%]
4.4%
[4.2%, 4.7%]

14.2%
[13.7%, 14.6%]
4.5%
[4.2%, 4.7%]

12.7%
[12.4%, 12.9%]
3.4%
[3.3%, 3.5%]

12.8%
[12.3%, 13.4%]
3.9%
[3.6%, 4.2%]

16.1%
[14.2%, 18.2%]
3.9%
[3.2%, 4.7%]

Detailed age information was not available in the public use file for NSDUH 2014.

Population Assessment of Tobacco and Health Study

Table 2-27.

Cigarette smoking* based on Wave 2 Youth Interview respondents

Unweighted
count

PATH Study:
Unweighted
percentage

12,148

11.8%

Ever tried smoking, male

6,208

11.8%

Ever tried smoking, female

5,911

11.7%

4,150

3.9%

7,998

15.8%

12,068

4.0%

Characteristic at Wave 2
Ever tried cigarette
smoking, even one or
two puffs

Ever tried smoking,
age 12-13
Ever tried smoking,
age 14-17
Have smoked in past
30 days

PATH Study:
Weighted percentage,
using youth IPS
weights
[95% confidence
interval]

PATH Study:
Weighted percentage,
using youth Wave 2
final weights
[95% confidence
interval]

Percentage from
2013-2014 NHANES
[95% confidence
interval]

Percentage from
2014 NSDUH
[95% confidence
interval]

Percentage from
2014 NYTS
[95% confidence
interval]

11.9%
[11.1%, 12.8%]

11.7%
[11.0%, 12.3%]

18.7%
[15.3%, 22.6%]

14.4%
[13.6%, 15.3%]

21.7%
[20.2%, 23.3%]

12.1%
[11.0%, 13.2%]
11.7%
[10.8%, 12.8%]
4.0%
[3.4%, 4.7%]
16.0%
[14.9%, 17.1%]
4.0%
[3.6%, 4.5%]

11.8%
[10.9%, 12.7%]
11.5%
[10.7%, 12.4%]
3.8%
[3.2%, 4.4%]
15.7%
[14.9%, 16.5%]
4.0%
[3.6%, 4.4%]

19.2%
[14.3%, 25.4%]
18.0%
[13.0%, 24.4%]
5.1%
[3.0%, 8.6%]
25.2%
[20.8%, 30.2%]
3.9%
[2.7%, 5.5%]

14.6%
[13.5%, 15.7%]
14.2%
[13.1%, 15.4%]
3.8%
[3.1%, 4.6%]
19.4%
[18.2%, 20.6%]
5.0%
[4.6%, 5.5%]

22.5%
[20.8%, 24.2%]
21.0%
[19.2%, 22.9%]
10.3%
[8.6%, 12.3%]
27.4%
[25.4%, 29.4%]
5.8%
[5.2%, 6.5%]

Population Assessment of Tobacco and Health Study

2.3.2

Biospecimens

The results in Tables 2-15 and 2-16 show estimates from the aged-up adults who provided urine or
blood specimens, respectively, at Wave 2, but not all specimens collected will be analyzed in the
laboratory. The biospecimens chosen initially for laboratory analysis are likely to come from a
probability sample of aged-up adults who are in specified tobacco use categories. 15 In other
categories of tobacco use, no samples may be selected for laboratory analysis. Consequently, the
samples of biospecimens from Wave 2 aged-up adults that are analyzed may not be representative of
the population of adults age 18 as a whole. If desired, an additional set of nonresponse-adjusted
weights can be developed for these adults, following procedures similar to those described in the
memo on Wave 1 biospecimen weighting procedures submitted to OMB (and approved on October
9, 2015). These weighting adjustments would address issues such as the overrepresentation of
current established tobacco users among aged-up adults who provided urine and/or blood
specimens, and the selection of adults whose biospecimens are sent for laboratory analysis.

2.4

Summary of Findings
Response Rates

As reported in Section 2.1, the response rates 16 for the PATH Study Wave 2 interviews were lower
than projected for continuing adults and for continuing and aged-up youth, but higher than
projected for aged-up adults (see Table 2-28). The unweighted response rates for the biospecimen
collections in Wave 2 were all higher than projected.
Table 2-28.

Summary of PATH Study Wave 2 response rates

Group
Continuing adults, Adult Interview
Continuing youth, Youth Interview
Aged-up adults, Adult Interview
Aged-up youth, Youth Interview
Continuing adults, urine collection
Aged-up adults, urine collection
Aged-up adults, blood collection

Unweighted
response rate
82.6%
88.5%
85.9%
82.0%
96.4%
82.8%
47.4%

Weighted
response rate
83.1%
88.4%
85.7%
82.1%
-

Projected
response rate*
86%
90%
85%
88%
80%
69%
45%

* Provided in the Revision Request to OMB for Wave 2 data and biospecimen collections.

Urine specimens from some continuing adults at Wave 2 will also be chosen for laboratory analysis.

The weighted response rates were computed using inverse probability of selection weights.

Population Assessment of Tobacco and Health Study

The differential weighted response rates to the Wave 2 interview were modest for tobacco use status
and demographic subgroups (see Tables 2-2 to 2-5). Unweighted response rates to the urine
collection were consistently high among subgroups of continuing adults (see Table 2-6). The largest
differential unweighted response rates were for the current established tobacco use status of aged-up
adults asked to provide urine and blood specimens: response rates for current established users were
about eight percentage points higher for both types of biospecimen than for other aged-up adults
(see Table 2-7), which suggests a heightened potential for nonresponse bias.

Nonresponse Bias Analysis
Nonresponse bias analysis indicates that estimates of many key demographic and Wave 1 tobacco
use variables calculated using the Wave 1 final weights are comparable for Wave 2 respondents and
nonrespondents. However, males, 18-24 year-olds, and those with high school education are
underrepresented among continuing adult respondents; and 45-64 year-olds, persons with health
insurance, and those with at least a bachelor’s degree are overrepresented among respondents (see
Table 2-8). Estimates of current established tobacco use are lower overall, and for males, 18-44 yearolds, and non-Hispanic Whites among continuing adult respondents (see Table 2-9). Estimates of
ever use of tobacco are lower overall, and for females, 14-17 year-olds, and non-Hispanic Whites
among continuing youth respondents compared to nonrespondents (see Table 2-11). No evidence
of potential nonresponse bias was found for aged-up adults or aged-up youth (see Tables 2-12 to 214). Based on these results, for some subgroups, the PATH Study may experience attrition patterns
that are similar to those in other longitudinal surveys. Cunradi et al. (2005) and Young et al. (2006)
have found that smokers were less likely to be retained in subsequent waves of surveys than
nonsmokers.
Results of the Wave 2 biospecimen nonresponse bias analysis for aged-up adults found no evidence
of nonresponse bias with respect to sex and race/ethnicity; however, current established tobacco
users were overrepresented among both urine and blood specimen providers. No nonresponse bias
analysis was necessary for urine collection from continuing adults due to the high response rate of
96 percent (see Table 2-6).
When compared to national cross-sectional surveys that measure tobacco use (TUS-CPS, NHIS,
NHANES, and NSDUH), estimates of adult cigarette smoking from the PATH Study Wave 2
sample are roughly in the middle of the range of estimates on smoking. There is no indication of
nonresponse bias with respect to this measure.
58

Population Assessment of Tobacco and Health Study

PATH Study estimates of the selected youth cigarette smoking measure from the full Wave 2 sample
are at the low end of estimates in comparison with national cross-sectional surveys that measure
tobacco use (NHANES, NSDUH, and NYTS). However, estimates from the comparison surveys
are from 2013 through 2014 while those from the PATH Study are from October 2014 through
October 2015, and evidence suggests the use of traditional cigarettes is declining among youth. The
difference among surveys on time period alone is not large enough to account for the different
estimates; as indicated in Section 2.2.2, time period is one of a number of factors that may explain
the different estimates.

Statistical Approach for Addressing Nonresponse
The approach used to reduce potential nonresponse bias in Wave 2 of the PATH Study was to
adjust the Wave 1 final weights of respondents at the adult and youth levels to account for
nonrespondents. Results of applying this approach to the full Wave 2 sample indicate the
nonresponse adjustments essentially eliminated discrepancies between estimates based on Wave 1
respondents and estimates based on Wave 2 respondents with respect to demographic
characteristics and Wave 1 tobacco use, for all age groups.
Procedures similar to those described in the memo on Wave 1 biospecimen weighting procedures
submitted to OMB (and approved on October 9, 2015) can be used to address nonresponse among
adults asked to provide biospecimens at Wave 2.
Estimates of adult cigarette smoking at Wave 2 using the Wave 1 IPS weights (before any
nonresponse adjustments) are in line with estimates from other surveys; agreement in these
estimates was preserved using the Wave 2 final weights. Weighting adjustments for youth had little
effect on the Wave 2 estimates of youth cigarette smoking.

Population Assessment of Tobacco and Health Study

Wave 3

Wave 3 of the PATH Study is at the approximate mid-point of data and biospecimen collections.
This section discusses the predicted response rates for Wave 3, an interim nonresponse analysis, and
the study’s planned statistical approach for addressing nonresponse in Wave 3. All such analyses are
based on data collected from the cases in replicate group 1 that had been released by April 29, 2016.
All study participants who completed an adult interview in Wave 1 or Wave 2 are continuing adults
and, if eligible, are asked to complete an Adult Interview in Wave 3. Study participants whose last
completed interview prior to Wave 3 was a Youth Interview are aged-up adults in Wave 3 if they
completed a Wave 3 Adult Interview, and are continuing youth if they completed a Wave 3 Youth
Interview. Wave 2 shadow youth who completed a Wave 3 Youth Interview are aged-up youth.
Nonrespondents and interim cases for Wave 3, however, do not have a Wave 3 interview date, so
the following procedure was used to determine their ages and participant types for this report. 17 A
Wave 2 youth who has not responded in Wave 3 is classified as a continuing youth if his/her age
was determined to be 17 or younger on the age classification date; otherwise he/she is classified as
an aged-up adult. 18 A similar classification rule was used for persons who were shadow youth at
Wave 2. A Wave 2 shadow youth who has not responded in Wave 3 is classified as an aged-up youth
if he/she was determined to have attained age 12 on or before the age classification date.
As stated in Section 1, the PATH Study Wave 1 sample was divided among four replicate groups.
Replicate group 1, which consisted of the addresses that were released to the field in September
2013, obtained Wave 1 responses from 5,951 adults and 2,698 youth, and parental consent for 1,414
shadow youth. Approximately 96 percent of those cases had been released to the field for Wave 3 as
of April 29, 2016. Replicate group 1 roughly corresponds to the earliest set of follow-ups in Wave 3.

Please refer to Section 2.1 for a general explanation of how the PATH Study sets an age classification date to
determine the age of a nonrespondent in a follow-up wave for reporting purposes.

Under this rule, for retention rate calculation purposes, nonresponding and interim youth in Wave 3 who were youth
respondents in Wave 2 are assigned to the category (continuing youth or aged-up adult) that would result if they
completed an interview on their Wave 3 age classification date.

Population Assessment of Tobacco and Health Study

Table 3-1 displays the status of cases from replicate group 1 that had been released to the field as of
April 29, 2016 for the four categories of continuing adults, continuing youth, aged-up adults, and
aged-up youth. The finalized cases include respondents and finalized nonrespondents of all types
(eligible, ineligible, and unknown eligibility).
Table 3-1.

Status of Wave 3 released cases from replicate group 1, as of April 29, 2016

Group
Continuing adults
Continuing youth
Aged-up adults
Aged-up youth

3.1

n
5,154
1,832
405
440

Finalized

Case status
Interim
n
934
297
94
89

%
84.7
86.0
81.2
83.2

%
15.3
14.0
18.8
16.8

n
6,088
2,129
499
529

Total

%
100.0
100.0
100.0
100.0

Predicted Response Rates

This section summarizes the three types of predicted response rate calculations used for Wave 3 of
the PATH Study: the interim retention and recruitment rates for the interviews and the interim
response rates for the biospecimen collections. In Section 3.1.1, retention rates for Wave 3 apply to
persons who completed the Adult Interview in Wave 2 (i.e., continuing adults), and persons who
completed the Youth Interview in Wave 2 and who are age 17 or younger at Wave 3 (i.e., continuing
youth). In Section 3.1.2, recruitment rates for Wave 3 apply to those Wave 2 respondents who have
aged up, either as Wave 2 shadow youth who have turned age 12 and are eligible to participate in the
Wave 3 Youth Interview (i.e., aged-up youth), or as Wave 2 youth who have turned age 18 and are
eligible to participate in the Wave 3 Adult Interview (i.e., aged-up adults). The Wave 3 predicted
recruitment rates and the Wave 3 predicted retention rates reported in these sections are conditional
on Wave 2 response. For example, the denominator for calculating the Wave 3 predicted retention
rate for continuing adults is the count of Wave 2 Adult Interview respondents who are eligible for
Wave 3 Adult Interview. Section 3.1.3 presents the predicted recruitment and retention rates among
all Wave 1 respondents that are obtained by combining the prediction results for Wave 2
respondents with predicted numbers of completed Wave 3 interviews among the Wave 2
nonrespondents. Response rates for biospecimen collections appear in Section 3.1.4 and refer to the
percentages of persons providing biospecimens among those who are asked to provide
biospecimens.

Population Assessment of Tobacco and Health Study

3.1.1

Predicted Retention Rates among Wave 2 Respondents for
Continuing Adults and Continuing Youth

This section reports retention rates for continuing adults who completed the Adult Interview at
Wave 2, and continuing youth who completed the Youth Interview at Wave 2 and remained eligible
for the Youth Interview at Wave 3.

Method
Consistent with the response rate calculation guidelines specified by the Office of Management and
Budget (2006), final retention rates for Wave 3 will be calculated for adults as the ratio of the
number of Wave 3 Adult Interview completed cases (or sufficient partials) to the number of cases
eligible for the Wave 3 Adult Interview. A simplified formulation will be used for this report because
the eligibility status of some interim cases is unknown. The simplified formulation corresponds to
AAPOR RR1 (AAPOR, 2015), which treats all completed cases from Wave 2 as eligible for Wave 3.
The predicted retention rates are therefore slightly conservative because some ineligible persons are
included in the denominator.
If all the Wave 3 cases were finalized, the RR1 retention rate for continuing adults would be
calculated as (number of completes or sufficient partials at Wave 3)/(number of Wave 2 completed
cases minus number of persons who died, were in a correctional facility, or left the country), where
the denominator can equivalently be expressed as the sum of the respondents and finalized eligible
nonrespondents. Because the PATH Study Wave 3 data collection is ongoing, however, the formula
must consider “nonfinalized” or interim status cases as well as finalized cases; in this sense, the
retention and recruitment rates presented in this Interim Report are “predicted.” Furthermore, the
cases in replicate group 1 that had not been released by April 29, 2016 are not included in these
analyses – these cases represent study participants who took relatively longer to complete a Wave 2
interview (which in some instances was because they were reluctant to respond), and who therefore
may be less inclined to respond at Wave 3. To help offset this selection bias among the analyzed
cases, for prediction purposes, the interim refusals 19 and persons who are difficult to locate were

Interim refusals for adults are cases that initially declined to participate in the Wave 3 interview but are still being
followed for refusal conversion attempts. Interim refusals for youth are those for whom the parent initially declined
permission for the youth to participate in the Wave 3 interview, but whose parents are still being followed for refusal
conversion attempts.

Population Assessment of Tobacco and Health Study

considered to be finalized nonrespondents. In this report, the unweighted retention rate for
continuing adults is calculated as
𝐼𝐼

𝐶𝐶𝐶𝐶
RRUCA = (CCA+∑𝑖𝑖=1
𝑝𝑝̂𝑖𝑖,𝐶𝐶𝐶𝐶 )/(CCA+NCA+ICA),

where

CCA = number of Wave 3 completed cases or sufficient partials among Wave 2 adult
respondents;
NCA = number of Wave 3 finalized nonrespondents among Wave 2 adult respondents;
ICA = number of Wave 3 interim cases among Wave 2 adult respondents; and
𝑝𝑝̂ 𝑖𝑖,𝐶𝐶𝐶𝐶 = predicted probability of interim continuing adult i becoming a respondent.

For continuing youth, the denominator of the response rate is defined using the age classification
date described earlier, and the unweighted retention rate is calculated as
𝐼𝐼

𝐶𝐶𝐶𝐶
𝑝𝑝̂ 𝑖𝑖,𝐶𝐶𝐶𝐶 )/(CCY+NCY+ICY),
RRUCY = (CCY+∑𝑖𝑖=1

where

CCY = number of Wave 3 completed Youth Interviews or sufficient partials among
Wave 2 youth respondents;
NCY = number of Wave 3 finalized nonrespondents among Wave 2 youth respondents
who were age 17 or younger on the age classification date;
ICY = number of Wave 3 interim cases among Wave 2 youth respondents who were age
17 or younger on the age classification date; and
𝑝𝑝̂ 𝑖𝑖,𝐶𝐶𝐶𝐶 = predicted probability of interim continuing youth i becoming a respondent.

The weighted response rates are computed similarly, with each count of finalized cases replaced by
the sum of the Wave 1 IPS weights (AIPSWT or YIPSWT) for individuals in that category. The
𝐼𝐼𝐶𝐶𝐶𝐶
sums of the predicted probabilities for interim cases are replaced by ∑𝑖𝑖=1
𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑇𝑇𝑖𝑖 ∗ 𝑝𝑝̂ 𝑖𝑖,𝐶𝐶𝐶𝐶 in the
𝐼𝐼𝐶𝐶𝐶𝐶
formula for RRUCA and by ∑𝑖𝑖=1 𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑇𝑇𝑖𝑖 ∗ 𝑝𝑝̂ 𝑖𝑖,𝐶𝐶𝐶𝐶 ,in the formula for RRUCY.

The probability that an interim case will become a Wave 3 respondent is estimated using logistic
regression, which is commonly used to predict response propensities (Groves et al., 2008; Wagner,
2010). Models were fit to the sets of Wave 3 finalized and interim cases, separately for Wave 2
responding adults and youth, to predict the probability of an interim case becoming a respondent as
a function of respondent characteristics from earlier waves such as age, sex, race/ethnicity, tobacco
use status, education and general health condition (both for continuing adults only), number of

Population Assessment of Tobacco and Health Study

adults in the household, number of contact attempts before a completed interview was obtained,
and ever having refused participation at the person level. During the model fitting process, a
conservative approach was taken by setting the response variable equal to one for respondents and
zero for both finalized nonrespondents and interim cases. When calculating the predicted response
rate, the predicted value from the logistic regression model was used for the interim cases, with the
exception of interim refusals and persons who are difficult to locate who were treated as finalized
nonrespondents.

Results
Tables 3-2 and 3-3 provide predicted retention rates for continuing adults and continuing youth who
responded in Wave 2. In addition to the overall row, each table includes rows on tobacco use status,
age, sex, race, and ethnicity subgroups based on Wave 2 data for characteristics where information
was updated (age and tobacco use status), and Wave 1 data otherwise. Persons with missing values
for these characteristics were excluded from the response rate calculation for that characteristic.
The weighted predicted retention rates among Wave 2 respondents are approximately 92 percent for
continuing adults and 93 percent for continuing youth. The unweighted predicted retention rates are
approximately 91 percent for continuing adults and 93 percent for continuing youth. The predicted
retention rate for continuing adults is higher than the projected retention rate of 86 percent provided
in the Revision Request to OMB for Wave 3; the predicted retention rate for continuing youth is
slightly higher than the projected retention rate of 91 percent. The predicted retention rates are
sensitive to the models used for predicting the response propensities among the interim cases, and
the estimates of retention rates will be more accurate as more information accrues.
As shown in Tables 3-2 and 3-3, the variability among predicted retention rates for subgroups is
small. For continuing adults, females appear to have a slightly higher predicted retention rate than
males, and persons of ‘other’ race have a lower predicted retention rate than Whites and Blacks.
Current established users of tobacco at Wave 2 also appear to have a lower predicted retention rate.
However, these apparent differences depend largely on the disposition of the interim cases and no

Population Assessment of Tobacco and Health Study

Table 3-2.

PATH Study Wave 3 predicted retention rates by respondent characteristics: Adult Interview (continuing adults)

Characteristica
Overall
Tobacco use statusd
Current
established user
Not current
established user
Age
18-24
25-44
45-64
65+
Sex
Male
Female
Race
White alone
Black alone or in
combination
Other
Ethnicity
Hispanic
Non-Hispanic

A:
Adult Interviews
completed
(n)
4,074

B:
Interim likely to be
completedb
(n)
393

C:
Finalized
nonresponse
(n)
179

D:
Interim cases
(n)
660

Unweighted predicted
retention rate for
Wave 3c
(%)
90.9

Weighted predicted
retention rate for
Wave 3c
(%)
91.8

1,740

153

284

89.7

89.9

2,280

236

369

91.9

92.4

1,182
1,383
1,097
411

143
143
86
21

40
52
45
42

248
229
143
40

90.1
91.7
92.0
87.7

90.7
93.4
92.2
88.6

1,994
2,077

190
202

93
86

335
324

90.2
91.6

90.3
93.0

2,915

272

139

448

91.0

92.1

676

107

91.8

92.2

360

88.4

88.2

722
3,293

81
308

20
155

137
519

91.4
90.8

93.2
91.5

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Adult Extended Interview. The age information and tobacco use status are as reported in the Wave 2 Adult
Interview. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due to missing values.

Interim likely to be completed is the sum of predicted probabilities of an interim case becoming a respondent over all interim cases.

Predicted retention rate = (A+B)/(A+C+D).

A tobacco user is defined as someone who uses one or more of the following tobacco products covered by the Wave 2 Adult Interview: cigarettes, traditional cigars, cigarillos, little filtered
cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco. A ‘current established user’ of a given tobacco product is someone who currently uses the product
every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used that product regularly.

Population Assessment of Tobacco and Health Study

Table 3-3.

PATH Study Wave 3 predicted retention rates by respondent characteristics: Youth
Interview (continuing youth)

Characteristica
Overall
Tobacco use statusd
Ever user
Never user
Age
12-13
14-17
Sex
Male
Female
Race
White alone
Black alone or in
combination
Other
Ethnicity
Hispanic
Non-Hispanic

A:
Youth
Interviews
completed
(n)
1,620

Unweighted
predicted
retention
rate for
Wave 3c
(%)
92.5

Weighted
predicted
retention
rate for
Wave 3c
(%)
92.5

B:
Interim
likely to be
completedb
(n)
153

C:
Finalized
nonresponse
(n)
47

D:
Interim
cases
(n)
250

289
1,260

36
113

13
31

71
174

87.2
93.7

87.5
93.7

682
938

54
99

20
27

88
162

93.2
92.0

92.8
92.3

836
781

67
85

21
26

118
131

92.7
92.3

92.8
92.2

1,095

140

93.0

273

91.8

91.9

176

92.5

92.4

441
1,148

56
95

11
36

95
152

90.9
93.0

91.0
93.1

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Youth Extended Interview. The age information and tobacco
use status are as reported in the Wave 2 Youth Interview. For each characteristic, the sum of the counts in all the categories may not
be equal to the count in the overall row due to missing values.

Interim likely to be completed is the sum of predicted probabilities of an interim case becoming a respondent over all interim cases.

Predicted retention rate = (A+B)/(A+C+D).

An ‘ever user’ is someone who has ever used one or more of the following tobacco products covered by the Wave 2 Youth Interview:
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco,
bidis, and kreteks. A ‘never user’ is someone who has never used any of those tobacco products. Ever use of a tobacco product is
defined as having ever used the product, even one or two times.

definitive conclusions can be made. The predicted retention rates are similar for most of the
subgroups of continuing youth; however, the predicted retention rate for ever users of tobacco at
Wave 2 appears to be slightly lower than that for never users.

3.1.2

Predicted Recruitment Rates among Wave 2 Respondents for Agedup Adults and Aged-up Youth

This section reports recruitment rates for aged-up adults who completed the Wave 2 Youth
Interview and are eligible for the Adult Interview at Wave 3, and aged-up youth who were

Population Assessment of Tobacco and Health Study

participating shadow youth at Wave 2 and are eligible for the Youth Interview at Wave 3. The
Wave 3 Youth Interview is the first interview for responding aged-up youth, and aged-up adult
respondents complete the Adult Interview for the first time. Please refer to Table 3-1 (above) for the
status of the aged-up adults and aged-up youth from replicate group 1.

Method
The methods described in Section 3.1.1 for estimating the retention rates were also used to estimate
the recruitment rates for aged-up adults and aged-up youth. For aged-up adults, the unweighted
recruitment rate is calculated as
𝐼𝐼

𝐴𝐴𝐴𝐴𝐴𝐴
RRUAUA = (CAUA+∑𝑖𝑖=1
𝑝𝑝̂𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 )/(CAUA+NAUA+IAUA),

where

CAUA = number of Wave 3 completed cases or sufficient partials among persons who
completed the Youth Interview at Wave 2 and were administered the Adult
Interview at Wave 3;
NAUA = number of Wave 3 finalized nonrespondents among Wave 2 youth respondents
who were age 18 by the age classification date;
IAUA = number of Wave 3 interim cases among Wave 2 youth respondents who were age
18 by the age classification date; and
𝑝𝑝̂𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 = predicted probability of interim aged-up adult i becoming a respondent.

The unweighted recruitment rate for aged-up youth is calculated as
𝐼𝐼

𝐴𝐴𝐴𝐴𝐴𝐴
RRUAUY = (CAUY+∑𝑖𝑖=1
𝑝𝑝̂ 𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 )/(CAUY+NAUY+IAUY),

where

CAUY = number of Wave 3 completed cases or sufficient partials among persons who
were participating shadow youth at Wave 2 and were administered the Youth
Interview at Wave 3;
NAUY = number of Wave 3 finalized nonrespondents among Wave 2 participating shadow
youth who were age 12 by the age classification date;
IAUY = number of Wave 3 interim cases among Wave 2 participating shadow youth who
were age 12 by the age classification date; and
𝑝𝑝̂ 𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 = predicted probability of interim aged-up youth i becoming a respondent.
67

Population Assessment of Tobacco and Health Study

The weighted recruitment rates were calculated by substituting the sum of Wave 1 IPS weights
(YIPSWT or SIPSWT 20) for the counts of finalized cases in each category, and replacing
𝐼𝐼𝐴𝐴𝐴𝐴𝐴𝐴
𝐼𝐼𝐴𝐴𝐴𝐴𝐴𝐴
𝐴𝐴𝐴𝐴𝐴𝐴
∑𝐼𝐼𝑖𝑖=1
𝑝𝑝̂𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 in the RRUAUA formula by ∑𝑖𝑖=1
𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑇𝑇𝑖𝑖 ∗ 𝑝𝑝̂ 𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 , and ∑𝑖𝑖=1
𝑝𝑝̂ 𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 in the RRUAUY
𝐼𝐼𝐴𝐴𝐴𝐴𝐴𝐴
formula by ∑𝑖𝑖=1 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑇𝑇𝑖𝑖 ∗ 𝑝𝑝̂ 𝑖𝑖,𝐴𝐴𝐴𝐴𝐴𝐴 .

Results
Table 3-4 provides predicted recruitment rates for the Adult Interview for aged-up adults, and Table
3-5 provides predicted recruitment rates for the Youth Interview for aged-up youth. In addition to
the overall row, each table includes rows on sex, race, and ethnicity subgroups; Table 3-4 also
includes rows on tobacco use status. There are no rows corresponding to age subgroups in Table
3-4 or Table 3-5 because most of the aged-up adults are 18 years old and most of the aged-up youth
are 12 years old; there are no rows for tobacco use status in Table 3-5 because no information was
collected about the tobacco use of shadow youth at Wave 2. Information from the Wave 1 Youth
Interview was used to define the demographic characteristics for the aged-up adults, and
information from the Wave 1 Household Screener was used to define the demographic
characteristics for the aged-up youth. Information from the Wave 2 Youth Interview was used to
define tobacco use status for aged-up adults. Persons with missing values for these characteristics
were excluded from the response rate calculation for that characteristic.
The predicted recruitment rate is approximately 94 percent for aged-up adults, which is higher than
the projected recruitment rate of 87 percent in the Revision Request to OMB for Wave 3. The
predicted recruitment rate for aged-up youth of 88 percent is slightly lower than the projected rate
of 89 percent.
Subgroup recruitment rates in Tables 3-4 and 3-5 for aged-up adults and for aged-up youth,
respectively, are similar with the exception of sex. Among aged-up adults, females appear to have a
higher recruitment rate than males; the reverse pattern is observed among aged-up youth.

SIPSWT is the name of the Wave 1 IPS weight for shadow youth. The construction of these weights is not described
in the 2015 Interim Report; however, they were created analogously to the youth IPS weights, YIPSWT.

Population Assessment of Tobacco and Health Study

Table 3-4.

PATH Study Wave 3 predicted recruitment rates by respondent characteristics:
Adult Interview (aged-up adults)

Characteristica
Overall
Tobacco use statusd
Ever user
Never user
Sex
Male
Female
Race/ethnicity
Non-Hispanic
White alone
Other

D:
Interim
cases
(n)
53

Unweighted
predicted
recruitment
rate for
Wave 3c
(%)
94.4

Weighted
predicted
recruitment
rate for
Wave 3c
(%)
94.1

1
6

25
27

94.8
94.5

94.3
94.3

20
21

7
1

27
26

92.6
96.3

92.0
96.3

145

94.6

94.7

153

94.1

93.3

A:
Adult
Interviews
completed
(n)
301

B:
Interim
likely to be
completedb
(n)
41

C:
Finalized
nonresponse
(n)
8

133
161

18
22

155
145

Interim likely to be completed is the sum of predicted probabilities of an interim case becoming a respondent over all interim cases.

Predicted recruitment rate = (A+B)/(A+C+D).

Table 3-5.

PATH Study Wave 3 predicted recruitment rates by respondent characteristics:
Youth Interview (aged-up youth)

Characteristica
Overall
Sex
Male
Female
Race/ethnicity
Non-Hispanic
White alone
Other

D:
Interim
cases
(n)
52

Unweighted
predicted
recruitment
rate for Wave 3c
(%)
88.0

Weighted
predicted
recruitment
rate for
Wave 3c
(%)
88.1

12
16

23
29

90.6
85.1

90.6
85.2

88.4

87.7

87.8

A:
Youth
Interviews
completed
(n)
311

B:
Interim
likely to be
completedb
(n)
33

C:
Finalized
nonresponse
(n)
28

174
137

15
18

148
163

Interim likely to be completes is the sum of predicted probabilities of an interim case becoming a respondent overall all interim cases.

Predicted recruitment rate = (A+B)/(A+C+D).

Population Assessment of Tobacco and Health Study

3.1.3

Predicted Retention and Recruitment Rates among Wave 1
Respondents

Sections 3.1.1 and 3.1.2 describe the predicted retention and recruitment rates, respectively, among
Wave 2 respondents. Most Wave 2 nonrespondents are also eligible for Wave 3 data collection (see
Section 1.2.2), and experience in the field to date suggests that the PATH Study will be successful in
obtaining completed interviews from about 10 percent of these cases. Table 3-6 presents the counts
of Wave 2 respondents and nonrespondents by participant type at Wave 3 among the replicate
group 1 cases released by April 29, 2016. The table also shows the predicted recruitment and
retention rates among all Wave 1 respondents, obtained by combining predicted results for Wave 2
respondents with predicted numbers of completed Wave 3 interviews among the Wave 2
nonrespondents. Due to small sample sizes, results for subgroups of Wave 2 nonrespondents among
each participant type are not presented. 21
Table 3-6.

PATH Study Wave 3 predicted retention and recruitment rates among Wave 1
respondents

Group
Continuing adults
Continuing youth

Wave 2 respondents
B:
A:
Unweighted
Unweighted
predicted
count
response rate
(n)
(%)
4,913
90.9

Wave 2 nonrespondents
C:
Unweighted
count
(n)
1,175

D:
Assumed
response rate
(%)
10

Wave 1 respondents
Unweighted predicted
response ratea
(%)
75

1,917

92.5

212

Aged-up adults

362

94.4

137

Aged-up youth

391

88.0

138

Predicted response rate = (A*B+C*D)/(A+C).

Predicted retention and recruitment rates for Wave 1 respondents in Table 3-6 provide perspective
on the Wave 3 data collection efforts. However, these predicted rates depend on the assumed degree
of success in obtaining completed Wave 3 interviews from Wave 2 nonrespondents (10 percent),
which is based on small sample sizes and may change by the end of the field period. Section 2.1
discusses the Wave 2 response rates; the remainder of Section 3 addresses retention and recruitment
from Wave 2 to Wave 3 and therefore focuses on the results for Wave 2 respondents, presented in
Sections 3.1.1 and 3.1.2.
21

The assumed response rate of 10 percent for the Wave 2 nonrespondents does not have a weighted counterpart;
therefore, only unweighted predicted response rates were computed for the Wave 1 respondents.

Population Assessment of Tobacco and Health Study

3.1.4

Biospecimen Collections

This section addresses response rates for the collection of urine and blood specimens from
continuing adults and aged-up adults who completed a Wave 3 Adult Interview and were asked to
provide a specimen. The PATH Study requests a urine specimen from a subsample of continuing
adults in Wave 3 who provided urine specimens at a previous wave; it also requests urine and blood
specimens at Wave 3 from all aged-up adults.

Method
The response rates were calculated using the following formula:
RRU = (Number of adults who provided a specimen)/(Number of adults from whom a
specimen was requested)
The urine response rate for continuing adults is based on the 2,140 adults who, as of April 29, 2016,
were asked to provide a urine specimen following their Wave 3 Adult Interview. Similarly, the urine
and blood response rates for aged-up adults are based on the 326 aged-up adults who completed the
Wave 3 Adult Interview as of this date.
Tables 3-7 and 3-8 provide unweighted response rates 22 for the biospecimen collections. In addition
to the overall row, each table includes rows on tobacco use status, sex, race, and ethnicity subgroups.
All persons asked to provide a biospecimen completed the Wave 3 Adult Interview so the subgroup
definitions use Wave 3 data for characteristics where information was updated (age and tobacco use
status), and Wave 1 data otherwise. Table 3-7 includes rows on age subgroups; this is not necessary
for Table 3-8 because the age range among Wave 3 aged-up adults is narrow. Adults with missing
values for such characteristics were excluded from the response rate calculation for that
characteristic.

Weighted response rates are not provided because the subset of continuing adults asked to provide a urine specimen
at Wave 3 does not represent a readily interpretable portion of the population. Similarly, while most aged-up adults
are age 18 at Wave 3, some are older due to the actual time elapsed between the Wave 2 and Wave 3 interviews, or
due to nonresponse at Wave 2.

Population Assessment of Tobacco and Health Study

Table 3-7.

PATH Study Wave 3 response rates by respondent characteristics: Urine collection
(continuing adults)

Characteristica
Overall
Tobacco use statusc
Current established user
Not current established user
Age
18-24
25-44
45-64
65+
Sex
Male
Female
Race
White alone
Black alone or in combination
Other
Ethnicity
Hispanic
Non-Hispanic

Urine
Unweighted response
rate for Wave 3b
(%)
96.8

A:
Adults requested to
provide urine
(n)
2,140

B:
Urine collected
(n)
2,072

1,158
929

1,135
888

98.0
95.6

596
838
548
158

573
824
528
147

96.1
98.3
96.4
93.0

1,093
1,046

1,050
1,021

96.1
97.6

1,514
355
203

1,463
348
195

96.6
98.0
96.1

410
1,698

391
1,650

95.4
97.2

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Adult Extended Interview. The age information and tobacco
use status are as reported in the Wave 3 Adult Interview. For each characteristic, the sum of the counts in all the categories may not
be equal to the count in the overall row due to missing values.

Response rate = B/A.

A tobacco user is defined as someone who uses one or more of the following tobacco products covered by the Wave 3 Adult Interview:
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable
tobacco. A ‘current established user’ of a given tobacco product is someone who currently uses the product every day or some days
and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever
used that product regularly.

Population Assessment of Tobacco and Health Study

Table 3-8.

PATH Study Wave 3 response rates by respondent characteristics: Biospecimen
collections (aged-up adults)

Characteristica
Overall
Tobacco use statusc
Current established user
Not current established user
Sex
Male
Female
Race
White alone
Black alone or in combination
Other
Ethnicity
Hispanic
Non-Hispanic

A:
Adult
Interviews
completed
(n)
326

Urine
Unweighted
B:
response rate
Collected
for Wave 3b
(n)
(%)
284
87.1

Blood
Unweighted
B:
response rate
Collected
for Wave 3b
(n)
(%)
156
47.9

47
265

44
227

93.6
85.7

26
122

55.3
46.0

167
158

147
137

88.0
86.7

78
78

46.7
49.4

212
55
41

182
50
34

85.8
90.9
82.9

99
27
19

46.7
49.1
46.3

96
227

82
200

85.4
88.1

52
101

54.2
44.5

The sex, race, and ethnicity characteristics are as reported in the Wave 1 Youth Extended Interview. The tobacco use status is as
reported in the Wave 3 Adult Interview. For each characteristic, the sum of the counts in all the categories may not be equal to the
count in the overall row due to missing values.

Response rate = B/A.

Population Assessment of Tobacco and Health Study

Results
The projected response rates for biospecimen collections in the Revision Request for Wave 3 were
97 percent for urine collection among continuing adults and, among aged-up adults, 83 percent for
the collection of urine and 43 percent for the collection of blood. To date, 97 percent of the
continuing adults asked to provide urine specimens have done so, meeting the projected response
rate. Among the aged-up adults, the response rates for urine and blood collection are 87 percent and
48 percent, respectively; the response rates to date for both biospecimen collections exceed the
projected response rates. Subgroup differences in response rates may be exaggerated by small
sample sizes; however, current established tobacco users at Wave 3 appear to have higher response
rates to the urine and blood collections. Hispanics appear to be more willing to provide blood
specimens than other aged-up adults and Blacks appear to be more willing to provide urine
specimens.

3.2

Nonresponse Bias Analysis

The standard approach for an analysis of nonresponse bias in a longitudinal cohort study such as the
PATH Study would be to compare Wave 3 respondents with Wave 3 nonrespondents with respect
to characteristics from a previous wave (Bose and West, 2002; Javitz and Wagner, 2005; Brownstein
et al., 2009). Because some of the cases fielded at Wave 3 were nonrespondents at Wave 2, Wave 1
characteristics are compared. At the mid-point of Wave 3 data collection, there are a number of
interim cases yet to be finalized as either respondents or nonrespondents. The number of finalized
nonrespondents is small and does not include interim cases that will ultimately be nonrespondents.
For this analysis, Wave 3 respondents are compared with the finalized nonrespondents. To explore
the sensitivity of results to the disposition of the interim cases, Wave 3 respondents are also
compared with provisional nonrespondents, defined to be the set of finalized nonrespondents plus
interim refusals and persons who are difficult to locate. Some of the interim cases among provisional
nonrespondents are expected to complete the Wave 3 interview; however, they are more likely to
require intensive contact tracing and follow-up efforts than are other interim cases. For this reason,
they are considered to be more similar to finalized nonrespondents than are other interim cases.
Sections 2.2 and 2.3 of the 2015 Interim Report describe the weight construction for Wave 1 of the
PATH Study. The final raked weights from Wave 1 were designed to reduce the potential
nonresponse bias from Wave 1. For Wave 3, the nonresponse bias analysis uses the Wave 1 final

Population Assessment of Tobacco and Health Study

weights and presents results for demographic and tobacco use subgroups defined using Wave 1
characteristics. This is necessary because Wave 2 final weights and characteristics are not available
for Wave 2 nonrespondents. Differences between the weighted estimates of Wave 1 characteristics
for Wave 3 respondents and nonrespondents therefore identify characteristics that might be
associated with nonresponse bias due to attrition between Wave 1 and Wave 3 of the study, after
compensating for Wave 1 nonresponse and possible undercoverage.
The statistical method used for detecting differences between the characteristics of the Wave 3
respondents and nonrespondents is the same as was used for comparing Wave 2 respondents and
nonrespondents in Section 2. Point estimates were calculated using the Wave 1 final weights as
described above. The corresponding replicate weights were used to calculate variances, and account
for the complex sampling features of stratification and clustering. Precisions for the estimates are
reported using 95 percent confidence intervals based on the modified Wilson confidence interval
approach. SAS software version 9.4 was used to calculate all point estimates and confidence
intervals. An estimated difference between Wave 3 respondents and nonrespondents is considered
statistically significant if the confidence interval for the difference excludes zero.
Tables 3-9 and 3-10 compare Wave 1 demographic characteristics, health insurance coverage, and
tobacco use rates for Wave 3 adult respondents who were adults at Wave 1 with the finalized
nonrespondents and with the provisional nonrespondents (who were adults at Wave 1). Tables 3-11
and 3-12 present similar comparisons for the persons who were interviewed as youth at Wave 1
(these cases are a mixture of Wave 3 continuing youth, continuing adults who were Wave 2 aged-up
adults, and Wave 3 aged-up adults); the cases are combined for this analysis because the number of
finalized nonrespondents among aged-up adults is too small (less than 90) to allow for meaningful
comparisons.

Population Assessment of Tobacco and Health Study

Table 3-9.

Characteristic at
Wave 1a
Sex

Comparison of Wave 3 Adult Interview respondents with finalized and provisional nonrespondents (Wave 1 adults)
Wave 3 respondents to Adult
Interview
Weighted
percentage, using
adult Wave 1 final
Unweights
weighted
[95% confidence
count
interval]

Male

1,911

Female

1,974

Age group
18-24

1,043

25-44

1,347

45-64

1,130

65+
Race/ethnicity
Non-Hispanic
White alone
Other

368
2,328
1,494

Health insurance
Yes
No

3,066
801

Education
< HS or GED

801

826

Some college,
no degree
Bachelor
degree +

1,391
859

47.1%
[45.1%, 49.2%]
52.9%
[50.8%, 54.9%]
11.7%
[10.7%, 12.8%]
35.0%
[33.1%, 36.9%]
35.6%
[33.6%, 37.8%]
17.7%
[15.9%, 19.6%]
66.3%
[63.4%, 69.1%]
33.7%
[30.9%, 36.6%]
85.4%
[83.7%, 86.8%]
14.6%
[13.2%, 16.3%]
17.1%
[15.3%, 19.0%]
21.7%
[20.1%, 23.5%]
31.6%
[29.6%, 33.6%]
29.6%
[27.2%, 32.1%]

Wave 3 finalized nonrespondents
to Adult Interview
Weighted
percentage, using
adult Wave 1 final
Unweights
weighted
[95% confidence
count
interval]
515
416
240
295
259
137
634
272
723
192
169
240
326
188

Difference in
weighted
percentages
[respondents –
finalized
nonrespondents]
[95% confidence
interval]

53.6%
[49.3%, 57.9%]
46.4%
[42.1%, 50.7%]

-6.4%
[-11.3%, -1.6%]
6.4%
[1.6%, 11.3%]

10.4%
[8.6%, 12.5%]
30.6%
[26.9%, 34.6%]
33.3%
[29.3%, 37.5%]
25.7%
[22.1%, 29.6%]

1.3%
[-0.5%, 3.1%]
4.4%
[0.0%, 8.7%]
2.3%
[-2.7%, 7.4%]
-8.0%
[-12.4%, -3.6%]

73.1%
[68.5%, 77.3%]
26.9%
[22.7%, 31.5%]

-6.8%
[-11.1%, -2.4%]
6.8%
[2.4%, 11.1%]

85.7%
[82.8%, 88.1%]
14.3%
[11.9%, 17.2%]

-0.3%
[-3.1%, 2.5%]
0.3%
[-2.5%, 3.1%]

17.9%
[14.6%, 21.9%]
26.2%
[22.5%, 30.1%]
31.3%
[27.4%, 35.4%]
24.6%
[21.3%, 28.3%]

-0.9%
[-5.3%, 3.6%]
-4.4%
[-8.6%, -0.3%]
0.3%
[-4.4%, 5.0%]
5.0%
[0.7%, 9.3%]

Wave 3 provisional nonrespondents
to Adult Interview

Unweighted
count
680
522
339
396
315
152
781
393
929
253
256
296
410
230

Weighted percentage,
using adult Wave 1
final weights
[95% confidence
interval]

Difference in
weighted
percentages
[respondents –
provisional
nonrespondents]
[95% confidence
interval]

54.3%
[50.2%, 58.4%]
45.7%
[41.6%, 49.8%]

-7.2%
[-11.9%, -2.5%]
7.2%
[2.5%, 11.9%]

11.6%
[9.9%, 13.6%]
31.9%
[28.6%, 35.5%]
32.6%
[28.8%, 36.5%]
23.9%
[20.8%, 27.3%]

0.1%
[-1.6%, 1.9%]
3.0%
[-1.1%, 7.1%]
3.1%
[-1.6%, 7.8%]
-6.2%
[-10.1%, -2.3%]

69.8%
[65.2%, 74.0%]
30.2%
[26.0%, 34.8%]

-3.5%
[-7.8%, 0.8%]
3.5%
[-0.8%, 7.8%]

85.2%
[82.4%, 87.6%]
14.8%
[12.4%, 17.6%]

0.2%
[-2.7%, 3.0%]
-0.2%
[-3.0%, 2.7%]

19.6%
[16.6%, 23.1%]
25.1%
[21.7%, 28.8%]
31.6%
[27.9%, 35.4%]
23.7%
[20.6%, 27.1%]

-2.5%
[-6.6%, 1.5%]
-3.3%
[-7.4%, 0.7%]
0.0%
[-4.4%, 4.4%]
5.9%
[1.8%, 9.9%]

Population Assessment of Tobacco and Health Study

Table 3-9.

Characteristic at
Wave 1a
Tobacco use
statusb
Current
established
use
Not current
established
user

Comparison of Wave 3 Adult Interview respondents with finalized and provisional nonrespondents (Wave 1 adults) (continued)
Wave 3 respondents to Adult
Interview
Weighted
percentage, using
adult Wave 1 final
Unweights
weighted
[95% confidence
count
interval]

Wave 3 finalized nonrespondents
to Adult Interview
Weighted
percentage, using
adult Wave 1 final
Unweights
weighted
[95% confidence
count
interval]

Difference in
weighted
percentages
[respondents –
finalized
nonrespondents]
[95% confidence
interval]

Wave 3 provisional nonrespondents
to Adult Interview

Unweighted
count

Weighted percentage,
using adult Wave 1
final weights
[95% confidence
interval]

Difference in
weighted
percentages
[respondents –
provisional
nonrespondents]
[95% confidence
interval]

1,717

22.7%
[21.3%, 24.1%]

450

25.4%
[22.3%, 28.7%]

-2.7%
[-6.2%, 0.8%]

592

27.0%
[24.0%, 30.3%]

-4.4%
[-7.8%, -0.9%]

2,066

77.3%
[75.9%, 78.7%]

449

74.6%
[71.3%, 77.7%]

2.7%
[-0.8%, 6.2%]

566

73.0%
[69.7%, 76.0%]

4.4%
[0.9%, 7.8%]

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall row due to
missing values.

A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A ‘current established user’ of a given tobacco product is someone
who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for any other tobacco product, has reported they ever used that product
regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and
dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 3-10.

Characteristic
at Wave 1a
Overall
Sex

Comparison of Wave 1 tobacco use* rates for Wave 3 Adult Interview respondents with finalized and provisional
nonrespondents (Wave 1 adults)
Wave 3 respondents to Adult
Interview
Weighted
percentage, using
adult Wave 1 final
weights
Unweighted
[95% confidence
count
interval]
22.7%
3,783
[21.3%, 24.1%]

Male

1,859

Female

1,921

Age group
18-24

1,026

25-44

1,315

45-64

1,089

65+
Race/ethnicity
Non-Hispanic
White alone
Other
a

353
2,272
1,452

27.1%
[25.0%, 29.3%]
18.7%
[17.0%, 20.5%]
29.4%
[26.1%, 32.9%]
26.0%
[23.3%, 28.9%]
23.4%
[20.9%, 26.0%]
9.9%
[7.2%, 13.5%]
23.1%
[21.3%, 25.1%]
21.4%
[19.1%, 23.9%]

Wave 3 finalized nonrespondents
to Adult Interview
Weighted
percentage, using
adult Wave 1 final
weights
Unweighted
[95% confidence
count
interval]
25.4%
899
[22.3%, 28.7%]
497
401
238
289
244
127
613
263

Difference in
weighted
percentages
[respondents –
finalized
nonrespondents]
[95% confidence
interval]
-2.7%
[-6.2%, 0.8%]

29.2%
[24.9%, 34.0%]
20.8%
[17.1%, 25.2%]

-2.2%
[-7.5%, 3.2%]
-2.1%
[-6.2%, 2.0%]

39.7%
[33.2%, 46.7%]
29.5%
[24.3%, 35.4%]
26.8%
[21.5%, 32.8%]
12.1%
[7.5%, 18.9%]

-10.4%
[-18.0%, -2.7%]
-3.5%
[-10.3%, 3.2%]
-3.4%
[-9.6%, 2.7%]
-2.2%
[-7.1%, 2.8%]

28.0%
[24.1%, 32.3%]
18.9%
[14.6%, 24.1%]

-4.9%
[-9.4%, -0.4%]
2.5%
[-2.9%, 7.9%]

Wave 3 provisional
nonrespondents to Adult
Interview
Weighted
percentage, using
adult Wave 1 final
weights
Unweighted
[95% confidence
count
interval]
27.0%
1,158
[24.0%, 30.3%]
655
502
335
383
297
142
757
376

Difference in
weighted
percentages
[respondents –
provisional
nonrespondents]
[95% confidence
interval]
-4.4%
[-7.8%, -0.9%]

31.6%
[27.3%, 36.2%]
21.6%
[17.9%, 25.8%]

-4.5%
[-9.7%, 0.7%]
-2.9%
[-6.9%, 1.1%]

41.0%
[35.2%, 47.0%]
32.3%
[27.4%, 37.6%]
27.5%
[22.5%, 33.2%]
12.0%
[7.7%, 18.4%]

-11.6%
[-18.3%, -4.9%]
-6.3%
[-12.5%, -0.0%]
-4.2%
[-10.1%, 1.8%]
-2.1%
[-6.8%, 2.6%]

30.2%
[26.2%, 34.4%]
20.4%
[16.6%, 24.7%]

-7.0%
[-11.5%, -2.6%]
1.0%
[-3.6%, 5.7%]

The characteristics are as reported in the Adult Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall
row due to missing values.

* Defined as current established use of tobacco. A tobacco user is defined as someone who uses one or more of the tobacco products covered by the Wave 1 Adult Extended Interview. A ‘current
established user’ of a given tobacco product is someone who currently uses the product every day or some days and: for cigarettes, has smoked at least 100 cigarettes in their lifetime and, for
any other tobacco product, has reported they ever used that product regularly. The products covered by the Wave 1 Adult Extended Interview are cigarettes, traditional cigars, cigarillos, little
filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, and dissolvable tobacco.

Population Assessment of Tobacco and Health Study

Table 3-11.

Comparison of Wave 3 interview respondents with finalized and provisional nonrespondents (Wave 1 youth)

Characteristic at Wave 1a
Sex
Male
Female

Wave 3 respondents
Weighted
percentage,
using youth
Wave 1 final
weights
Unweighted [95% confidence
count
interval]
1,005
918

Age group
12-13

691

14-17

1,236

Race/ethnicity
Non-Hispanic White alone

936

Other

949

Tobacco use statusb
Ever user
Never user

396
1,467

52.1%
[49.8%, 54.4%]
47.9%
[45.6%, 50.2%]
35.2%
[33.1%, 37.3%]
64.8%
[62.7%, 66.9%]
55.9%
[52.2%, 59.6%]
44.1%
[40.4%, 47.8%]
21.0%
[18.6%, 23.6%]
79.0%
[76.4%, 81.4%]

Wave 3 finalized
nonrespondents
Weighted
percentage,
using youth
Wave 1 final
weights
Unweighted [95% confidence
count
interval]
117
134

89
163

139
110

56
191

Difference in
weighted
percentages
[respondents –
finalized
nonrespondents]
[95% confidence
interval]

46.4%
[40.2%, 52.8%]
53.6%
[47.2%, 59.8%]

5.7%
[-1.1%, 12.6%]
-5.7%
[-12.6%, 1.1%]

34.5%
[28.3%, 41.2%]
65.5%
[58.8%, 71.7%]

0.7%
[-6.2%, 7.6%]
-0.7%
[-7.6%, 6.2%]

61.4%
[53.9%, 68.4%]
38.6%
[31.6%, 46.1%]

-5.4%
[-13.4%, 2.5%]
5.4%
[-2.5%, 13.4%]

22.2%
[16.4%, 29.3%]
77.8%
[70.7%, 83.6%]

-1.2%
[-8.1%, 5.8%]
1.2%
[-5.8%, 8.1%]

Wave 3 provisional
nonrespondents
Weighted
percentage,
using youth
Wave 1 final
weights
Unweighted [95% confidence
count
interval]
166
180

122
226

180
164

87
254

Difference in
weighted
percentages
[respondents –
provisional
nonrespondents]
[95% confidence
interval]

47.8%
[42.4%, 53.3%]
52.2%
[46.7%, 57.6%]

4.3%
[-1.7%, 10.3%]
-4.3%
[-10.3%, 1.7%]

34.2%
[29.1%, 39.6%]
65.8%
[60.4%, 70.9%]

1.0%
[-4.9%, 6.8%]
-1.0%
[-6.8%, 4.9%]

58.2%
[51.3%, 64.8%]
41.8%
[35.2%, 48.7%]

-2.3%
[-9.4%, 4.8%]
2.3%
[-4.8%, 9.4%]

25.3%
[20.4%, 30.9%]
74.7%
[69.1%, 79.6%]

-4.3%
[-10.2%, 1.6%]
4.3%
[-1.6%, 10.2%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

An ‘ever user’ is someone who has ever used one or more of the tobacco products covered by the Wave 1 Youth Extended Interview. A ‘never user’ is someone who has never used any of
those tobacco products. Ever use of a tobacco product is defined as having ever used the product, even one or two times. The products covered by the Wave 1 Youth Extended Interview are
cigarettes, traditional cigars, cigarillos, little filtered cigars, pipes, smokeless tobacco, snus, hookah, e-cigarettes, dissolvable tobacco, bidis, and kreteks.

Population Assessment of Tobacco and Health Study

Table 3-12.

Characteristic at
Wave 1a
Overall
Sex

Comparison of Wave 1 tobacco use* rates for Wave 3 respondents with finalized and provisional nonrespondents (Wave 1
youth)
Wave 3 respondents
Weighted
percentage, using
youth Wave 1
final weights
Unweighted
[95% confidence
count
interval]
21.0%
1,863
[18.6%, 23.6%]

Male

967

Female

892

Age group
12-13

653

14-17

1,210

Race/ethnicity
Non-Hispanic
White alone
Other
a

907
917

21.5%
[18.6%, 24.8%]
20.4%
[17.3%, 23.9%]
7.1%
[5.1%, 9.9%]
28.3%
[25.2%, 31.6%]
21.1%
[17.9%, 24.6%]
21.2%
[18.0%, 24.8%]

Wave 3 finalized nonrespondents
Weighted
percentage, using
youth Wave 1 final
weights
Unweighted
[95% confidence
count
interval]
22.2%
247
[16.4%, 29.3%]
115
131
86
160
136
108

Difference in
weighted
percentages
[respondents –
finalized
nonrespondents]
[95% confidence
interval]
-1.2%
[-8.1%, 5.8%]

26.2%
[18.0%, 36.5%]
18.8%
[12.3%, 27.5%]

-4.7%
[-14.9%, 5.5%]
1.6%
[-6.6%, 9.9%]

8.0%
[3.8%, 15.8%]
29.6%
[21.3%, 39.6%]

-0.9%
[-7.2%, 5.4%]
-1.3%
[-11.2%, 8.5%]

26.4%
[18.2%, 36.8%]
16.1%
[10.2%, 24.6%]

-5.4%
[-15.1%, 4.4%]
5.1%
[-3.4%, 13.6%]

Wave 3 provisional
nonrespondents
Weighted
percentage, using
youth Wave 1 final
weights
Unweighted
[95% confidence
count
interval]
25.3%
341
[20.4%, 30.9%]
163
176
119
221
177
160

Difference in
weighted
percentages
[respondents –
provisional
nonrespondents]
[95% confidence
interval]
-4.3%
[-10.2%, 1.6%]

28.7%
[20.8%, 38.1%]
22.4%
[16.9%, 29.1%]

-7.2%
[-16.7%, 2.3%]
-2.0%
[-8.7%, 4.7%]

10.2%
[5.8%, 17.3%]
33.2%
[26.4%, 40.8%]

-3.1%
[-9.1%, 3.0%]
-4.9%
[-12.9%, 3.0%]

29.6%
[22.3%, 38.2%]
19.0%
[13.5%, 26.1%]

-8.5%
[-17.2%, 0.1%]
2.2%
[-5.6%, 10.0%]

The characteristics are as reported in the Youth Extended Interview at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the
overall row due to missing values.

Population Assessment of Tobacco and Health Study

Table 3-13 compares Wave 1 demographic characteristics for Wave 3 youth respondents who were
shadow youth at Wave 1 with the provisional nonrespondents (these cases are a mixture of
continuing youth who were Wave 2 aged-up youth and Wave 3 aged-up youth). The numbers of
finalized nonrespondents are too small (less than 120 overall) to permit comparisons of respondents
and finalized nonrespondents for the persons who were Wave 1 shadow youth. Table 3-13 does not
include subgroups for tobacco use status because shadow youth did not self-report information in
Wave 1.
Among Wave 1 adults, some trends differ for the comparisons of respondents with finalized
nonrespondents and with provisional nonrespondents (see Table 3-9). The estimated percentage of
adults who are non-Hispanic Whites is lower for respondents than for finalized nonrespondents but
the difference is not statistically significant when the respondents and provisional nonrespondents
are compared. A similar pattern is observed for the percentage of adults with high school education.
The estimated percentage of Wave 1 adults ages 25 to 44 is higher for respondents than for finalized
nonrespondents. The estimated percentage of adults with at least a bachelor’s degree is higher for
respondents than for both nonrespondent groups. The estimated percentages of males and persons
age 65 or older are both lower for respondents than for finalized and provisional nonrespondents.
Respondents have lower rates of current established tobacco use compared to finalized and
provisional nonrespondents among non-Hispanic White and young adults (ages 18 to 24) at Wave 1
(see Table 3-10). Overall and among Wave 1 adults ages 25 to 44, the respondents have a lower rate
of current established tobacco use than the provisional nonrespondents, but there is no statistically
significant difference in the rate of current established tobacco use between the respondents and
finalized nonrespondents.
The results in this report are based on preliminary data, and may change as more cases are finalized.
If the trends seen among provisional nonrespondents continue as the interim cases are finalized,
however, the PATH Study may experience attrition patterns that are similar to those in other
longitudinal surveys. Thompson (2015) noted that younger persons and persons with lower
educational levels are more difficult to retain in longitudinal surveys. Cunradi et al. (2005) and
Young et al. (2006) have found that smokers were less likely to be retained in subsequent waves of
surveys than nonsmokers.
Among Wave 1 youth and shadow youth, there is no evidence of nonresponse bias (all confidence
intervals for estimates of differences between Wave 3 respondents and finalized or provisional
nonrespondents include zero, see Tables 3-11 to 3-13).
81

Population Assessment of Tobacco and Health Study

Table 3-13.

Comparison of Wave 3 Youth Interview respondents with provisional nonrespondents (Wave 1 shadow youth)

Characteristic at Wave 1a
Sex

Wave 3 respondents to Youth Interview
Weighted percentage,
using shadow youth Wave 1
Unweighted
final weights
count
[95% confidence interval]

Male

362

Female

351

Age group
9-10

353

360

Race/ethnicity

Non-Hispanic White alone

364

Other

348

Wave 3 provisional nonrespondents to
Youth Interview
Weighted percentage,
using shadow youth Wave 1
Unweighted
final weights
count
[95% confidence interval]

49.2%
[45.5%, 52.9%]
50.8%
[47.1%, 54.5%]

74
86

49.9%
[46.3%, 53.6%]
50.1%
[46.4%, 53.7%]

83
77

55.6%
[51.1%, 59.9%]
44.4%
[40.1%, 48.9%]

77
83

Difference in weighted
percentages [respondents –
provisional nonrespondents]
[95% confidence interval]

47.8%
[39.0%, 56.6%]
52.2%
[43.4%, 61.0%]

1.4%
[-8.4%, 11.2%]
-1.4%
[-11.2%, 8.4%]

52.4%
[44.7%, 60.0%]
47.6%
[40.0%, 55.3%]

-2.5%
[-11.1%, 6.1%]
2.5%
[-6.1%, 11.1%]

51.6%
[43.7%, 59.4%]
48.4%
[40.6%, 56.3%]

4.0%
[-4.1%, 12.0%]
-4.0%
[-12.0%, 4.1%]

The characteristics are as reported in the Household Screener at Wave 1. For each characteristic, the sum of the counts in all the categories may not be equal to the count in the overall
row due to missing values.

Population Assessment of Tobacco and Health Study

3.3

Statistical Approach for Addressing Nonresponse

Initial weights for Wave 3 respondents will be adjusted to address nonresponse at Wave 3. Similar to
the Wave 2 weighting process, the nonresponse adjustment process will be done in two stages. The
first stage is to account for the fact that some of the nonrespondents may not have been eligible for
Wave 3 but this could not be formally ascertained because, for example, they could not be located at
the time of Wave 3 data collection. Doing so is routine and amounts to adjusting the sample weights
of all those whose eligibility status is known for Wave 3 (e.g., Wave 3 respondents, Wave 3 refusals,
those who died) to account for nonresponse among those whose eligibility status could not be
ascertained. After this first stage of nonresponse adjustment is completed, the second stage is carried
out only among those respondents and nonrespondents known to be eligible for Wave 3. This is
accomplished by simply removing from the weighting process those people who died, were in a
correctional facility, or left the country prior to Wave 3.
Weight adjustments will be computed within cells formed from the cross-classification of variables
with the potential for reducing nonresponse bias. Such variables from Wave 1 include age, race,
ethnicity, sex, employment status, education level, tobacco use status, household composition,
census block characteristics, and the type of interview completed (adult, youth, or none for shadow).
Wave 2 data available for both Wave 2 respondents and nonrespondents (e.g., Wave 2 paradata) may
also be included. 23 A tree-based classification method will be employed to identify cells that
distinguish between subgroups with different propensities to respond to Wave 3 the PATH Study
(see Roth et al., 2006 and Schouten and deNooij, 2005). SAS macros will then be used to compute
and apply the weighting adjustment factors and identify potential sources of concern in the
adjustment process, such as small cell sizes and large adjustment factors.
Procedures similar to those described in the memo on Wave 1 biospecimen weighting procedures
submitted to OMB (and approved on October 9, 2015) can be used to address nonresponse among
adults asked to provide biospecimens at Wave 3.

The handling of Wave 2 nonrespondents in Wave 3 presents a challenge in the development of nonresponse
adjustments for Wave 3 weighting. Because the formation of nonresponse adjustment cells requires information for
both Wave 3 respondents and nonrespondents, the only information uniformly available for both groups is from
Wave 1. Response status in Wave 2 would serve as a key discriminator in terms of differentiating between
propensities to respond at Wave 3 but would produce very large adjustment factors, resulting in an inordinate
contribution to the variance.

Population Assessment of Tobacco and Health Study

3.4

Summary of Findings
Response Rates

Because the PATH Study Wave 3 data collection is ongoing, response rates 24 for Wave 2
respondents were calculated using predicted response propensities for interim cases, as described in
Section 3.1. With one exception, the predicted responses rates exceed the projected response rates
provided in the Revision Request to OMB for Wave 3. Table 3-14 indicates that the predicted
retention rate for continuing adults is higher than the projected rate, and the predicted retention rate
for continuing youth is slightly higher than the projected rate. The predicted recruitment rate for
aged-up adults is also higher than the projection, and the predicted recruitment rate for aged-up
youth is slightly lower than the projected value.
The response rate for each biospecimen collection is calculated as the percentage of persons
requested to give a biospecimen who provided it. For both continuing adults and aged-up adults in
Wave 3, the response rates for the biospecimen collections meet or exceed those projected.
Table 3-14.

Summary of PATH Study predicted response rates for Wave 3

Unweighted predicted
response rate
90.9%
92.5%
94.4%
88.0%
96.8%
87.1%
47.9%

Weighted predicted
response rate
91.8%
92.5%
94.1%
88.1%
-

Projected
response rate*
86%
91%
87%
89%
97%
83%
43%

* Provided in the Revision Request to OMB for Wave 3 data and biospecimen collections.

The variability among predicted retention rates for subgroups is small (see Tables 3-2 and 3-3). For
continuing adults, females appear to have a slightly higher predicted retention rate than males, and
persons of ‘other’ race have a lower predicted retention rate than Whites and Blacks. Current
established users of tobacco at Wave 2 also appear to have a lower predicted retention rate. The
predicted retention rates are similar for most of the subgroups of continuing youth; however, the
predicted retention rate for ever users of tobacco at Wave 2 appears to be slightly lower than that
24

Response rates include retention rates for continuing adults and youth, recruitment rates for aged-up adults and
youth, and response rates for providing biospecimens.

Population Assessment of Tobacco and Health Study

for never users. Subgroup recruitment rates for aged-up adults and for aged-up youth are similar
with the exception of sex: among aged-up adults, females appear to have a higher recruitment rate
than males; the reverse pattern is observed among aged-up youth (see Tables 3-4 and 3-5). Response
rates for the Wave 3 biospecimen collections meet or exceed the projected rates for all categories
(see Tables 3-7 and 3-8).

Nonresponse Bias Analysis
The nonresponse bias analysis found differences between Wave 1 adult respondents and
nonrespondents for a few demographic and tobacco use characteristics (see Tables 3-9 and 3-10),
but no differences between those groups for Wave 1 youth or shadow youth (see Tables 3-11 to 313). For Wave 1 adults, the estimated percentages of males and persons age 65 or older tend to be
lower for the Wave 3 respondents than for the Wave 3 finalized and provisional nonrespondents;
whereas the estimated percentage of adults with at least a bachelor’s degree is higher for
respondents. Wave 1 adults ages 25 to 44 are overrepresented and those with a high school
education and non-Hispanic Whites are underrepresented among respondents compared to finalized
nonrespondents, but the differences between estimates for respondents and provisional
nonrespondents are not statistically significant. Among all adults, estimated rates of current
established tobacco use at Wave 1 are not significantly different between respondents and finalized
nonrespondents, although the provisional nonrespondents exhibit higher tobacco use rates than the
respondents (particularly among 18-44 year olds and non-Hispanic Whites).

Statistical Approach for Addressing Nonresponse
For Wave 3, the weights of respondents will be adjusted to account for the nonrespondents by
forming weighting adjustment cells using Wave 1 characteristics of respondents and
nonrespondents, as well as Wave 1 and Wave 2 paradata that are available for both Wave 3
respondents and nonrespondents. This weighting will compensate for differences between
respondents and nonrespondents with respect to sex, age, other demographic and geographic
variables, and selected Wave 1 tobacco use measures.

Population Assessment of Tobacco and Health Study

Conclusions and Implications for
Study Going Forward

This section summarizes the findings presented in this report on the PATH Study’s Wave 1, 2, and
Wave 3 response rates, nonresponse bias analyses, and approach to addressing nonresponse. Its
conclusions are based on the full sample for Waves 1 and 2, and on the data collected from replicate
group 1 during the first 6 months (out of 12) of Wave 3. The section closes with a discussion of the
implications of the conclusions for the study going forward.

Conclusions
Response rates in Wave 1 for the Household Screener and Adult Extended Interview were lower
than projected in the Non-substantive Change Request to OMB for Wave 1 of the PATH Study.
However, nonresponse bias analysis found that many characteristics of respondents in Wave 1
aligned with the 1-year estimates from the 2013 ACS. Exceptions were found for single-person
households, education, and ethnicity when comparing PATH Study estimates using IPS weights to
1-year 2013 ACS estimates. Estimates of cigarette smoking among adults in Wave 1 were within the
range of estimates found by other national health studies. Moreover, when full sample estimates
were adjusted for nonresponse using the raked weights, they more closely approximated the ACS
estimates, and adult cigarette smoking rates remained essentially the same.
The response rate for the Wave 1 Youth Interview was higher than projected. Nonresponse bias
analysis among youth found that many characteristics of respondents were consistent with the 1-year
estimates from the 2013 ACS, with the exception of ethnicity. When the full sample estimates were
adjusted for nonresponse among youth, they more closely approximated the 2013 ACS estimates,
but the ever-tried cigarette smoking rates for youth in Wave 1 remained lower than those found by
other national studies.
The response rates for urine and blood collections in Wave 1 were lower than initially projected.
Despite this, nonresponse bias analysis found that many of the characteristics of respondents were
generally aligned with estimates of these characteristics from the 1-year 2013 ACS. In addition, when
the sample estimates were adjusted for nonresponse, they were found to approximate the ACS
estimates more closely.

Population Assessment of Tobacco and Health Study

The retention rate for Wave 2 continuing adults was about three percentage points lower than
projected in the Revision Request to OMB for Wave 2; however, the retention rate for continuing
youth was almost the same as the projection (lower by about 0.5 percentage points). The recruitment
rate for aged-up adults was about one percentage point higher than the projected rate, and the
recruitment rate for aged-up youth was approximately six percentage points lower than the projected
rate. The response rates for the biospecimen collections in Wave 2 were all higher than projected.
The largest differential response rates were for the tobacco use status of aged-up adults asked to
provide urine and blood specimens (for each biospecimen, about eight percentage points higher for
current established tobacco users than for those who were not).
However, nonresponse bias analysis found that many characteristics of Wave 2 respondents aligned
with those of Wave 2 nonrespondents. Some exceptions were found when comparing estimates for
continuing adults (current established tobacco use was lower overall, and for males, 18-44 year-olds,
and non-Hispanic Whites among respondents) and for continuing youth (ever use of tobacco was
lower overall, and for females, 14-17 year-olds, and non-Hispanic Whites among respondents). For
continuing adults, males, 18-24 year-olds, and those with high school education were
underrepresented among respondents; and 45-64 year-olds, persons with health insurance, and those
with at least a bachelor’s degree were overrepresented among respondents. However, estimates of
cigarette smoking among adults in Wave 2 were within the range of estimates found by other
national health studies. Moreover, when the estimates of Wave 1 characteristics based on the full
Wave 2 sample were adjusted for nonresponse using the Wave 2 final weights, they were almost
identical to the estimates based on the Wave 1 sample and Wave 1 final weights. The Wave 2 adult
cigarette smoking rates remained essentially the same using the Wave 2 final weights (compared to
using the Wave 1 IPS weights), but the ever-tried cigarette smoking rates for youth remained lower
than those found by other national studies. Among aged-up adults, current established tobacco users
were more likely to provide urine and blood specimens; urine collection rates were very high among
continuing adults.
With one exception, the predicted Wave 3 responses rates exceed the projected response rates
provided in the Revision Request to OMB for Wave 3. The predicted Wave 3 retention rate for
continuing adults who responded at Wave 2 is about five percentage points higher than projected in
the Revision Request to OMB for Wave 3; the predicted retention rate for continuing youth who
responded at Wave 2 is about 1.5 percentage points higher than the projection. The estimated
recruitment rate for aged-up adults who completed a Youth Interview at Wave 2 is seven percentage
points higher than the projected rate, and the estimated recruitment rate for aged-up youth who
87

Population Assessment of Tobacco and Health Study

participated as shadow youth at Wave 2 is one percentage point higher than projected. The Wave 3
response rates for biospecimens also approximately equal or exceed the projected rates.
There is no evidence of nonresponse bias at Wave 3 for the Wave 1 youth or shadow youth. For
Wave 1 adults, the estimated percentage of persons with at least a college degree is higher when
calculated from the respondents than from the finalized or provisional nonrespondents. The
estimated percentages of males and persons age 65 or older tend to be lower for the Wave 3
respondents than for both nonrespondent groups. Current established use of tobacco is significantly
lower among respondents than among provisional nonrespondents, particularly for non-Hispanic
Whites and adults ages 18-24 at Wave 1 (for whom tobacco use rates are also significantly different
when respondents are compared to finalized nonrespondents). However, as noted, these findings are
preliminary pending finalization of interim cases and the remainder of data collection in Wave 3.

Implications for the Study Going Forward
Findings on the response rates, nonresponse bias analysis, and approach to addressing nonresponse
for Waves 1, 2, and 3 have important implications for the PATH Study. First, the study should
continue implementing new approaches to increase response rates for Wave 3 and subsequent
waves. The PATH Study is continually seeking ways to boost the response rates. For example,
beginning in Wave 2, it enhanced its efforts to communicate by text message and email with
participants who indicated they may be contacted in these ways. In each wave, the PATH Study
varies the appearance of materials it provides participants to enhance their interest and engagement.
In addition, the study provides a certificate of appreciation to participating youth, and it takes extra
steps to interview continuing adults who have relocated to group quarters facilities since their initial
interviews. Participants continue to access information about the study on the participant pages of
the PATH Study website; adult participants may update their contact information on this website, as
well as by returning update forms mailed to them and by calling a toll-free telephone number. The
study employs special interviewers with skills in refusal conversion and has a cadre of experienced
traveling interviewers that help to augment staffing in specific areas. For tracing difficult-to-locate
cases, the study has examined the effectiveness of various on-line search services; as needed, it
accesses multiple services for a given case. For Wave 4 and potential future waves, the PATH Study
plans to further expand its engagement activities to include the use of short videos, to be displayed
on field interviewer laptops (e.g., for participants who are reluctant to participate) and on the study
website. In these videos and other materials, the study will provide information on the scope of the

Population Assessment of Tobacco and Health Study

PATH Study, to underline the message participants are contributing to a large and important effort
that will increase understanding of tobacco use and health.
Second, data analyses will need to consider carefully the achieved sample sizes and adjustments may
be necessary, such as combining some subgroups in analyses. Future planning for longitudinal
analyses will need to account for the declining rates of cigarette smoking in the U.S. population in
general, notably among youth. For the PATH Study, the trends observed among youth mean smaller
sample sizes for youth cigarette smokers and therefore less statistical power for examining withinperson changes among this group. At the same time, the larger sample size of youth nonsmokers
provides more power for examining the initiation of cigarette smoking over time. The impact of the
smaller sample size for youth cigarette smokers must also be considered in the context of the recent
and rapid increase of youth use of alternative tobacco products such as e-cigarettes and hookah
(CDC, 2015d).
Third, on an ongoing and consistent basis, the PATH Study should examine the sample sizes
achieved, as well as those projected for future data collection. In this way, it would be possible to
detect differential rates of attrition among subgroups early and make extra efforts to retain persons
in subgroups of special analytic interest. The Wave 1 shadow sample, which serves as a reservoir for
aged-up youth in subsequent waves, will be exhausted after Wave 3. Therefore, with OMB’s
approval, the PATH Study plans to replenish the sample at Wave 4. The Wave 4 design includes the
selection of a new shadow youth sample, as well as new samples of youth and adults to replenish the
Wave 1 sample, taking into account the subgroup sample sizes in the continuing cohort.

Population Assessment of Tobacco and Health Study

References
American Association of Public Opinion Research (AAPOR). (2015). Standard Definitions: Final
Dispositions of Case Codes and Outcome Rates for Surveys, 8th ed.
http://www.aapor.org/AAPOR_Main/media/publications/StandardDefinitions2015_8theditionwithchanges_April2015_logo.pdf.
Bose, J., and West, J. (2002). Examining additional nonresponse bias introduced through attrition.
Proceedings of the Survey Research Methods Section, American Statistical Association, 278-283.
Brick, J.M., Lê, T., and West, J. (2003). Dealing with movers in a longitudinal study of children. In
Statistics Canada Symposium-challenges in survey taking for the next decade.
Brownstein, N., Kalsbeek, W.D., Tabor, J., Entzel, P., Daza, E., and Harris, K.M. (2009). NonResponse in Wave IV of the National Longitudinal Study of Adolescent Health.
http://www.cpc.unc.edu/projects/addhealth/data/guides/W4_nonresponse.pdf.
Center for Behavioral Health Statistics and Quality (CBHSQ). (2015a). National Survey on Drug Use
and Health, 2014. Rockville, MD: Substance Abuse and Mental Health Services Administration,
http://www.samhsa.gov/samhsa-data-outcomes-quality/major-data-collections/public-usefiles-2014-nsduh
Center for Behavioral Health Statistics and Quality (CBHSQ). (2015b). Results from the 2014 National
Survey on Drug Use and Health: Detailed Tables. Prevalence Estimates, Standard Errors, P Values, and
Sample Sizes. Rockville, MD: Substance Abuse and Mental Health Services Administration,
http://www.samhsa.gov/data/sites/default/files/NSDUH-DetTabs2014/NSDUHDetTabs2014.pdf.Centers for Disease Control and Prevention (CDC). (2013). Tobacco
product use among middle and high school students — United States, 2011 and 2012.
Morbidity and Mortality Weekly Report, 62, 893–897. Retrieved from
www.cdc.gov/mmwr/preview/mmwrhtml/mm6245a2.htm.
Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS).
(2015a). National Health and Nutrition Examination Survey Data, 2013-2014. Hyattsville, MD: U.S.
Department of Health and Human Services, Centers for Disease Control and Prevention,
http://wwwn.cdc.gov/Nchs/Nhanes/Search/DataPage.aspx?Component=Questionnaire&C
ycleBeginYear=2013Error! Hyperlink reference not valid..
Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS).
(2015b). National Health Interview Survey Data, 20144. Hyattsville, MD: U.S. Department of
Health and Human Services, Centers for Disease Control and Prevention,
http://www.cdc.gov/nchs/nhis/nhis_2014_data_release.htm.
Centers for Disease Control and Prevention (CDC). (2015c). National Youth Tobacco Survey Data,
2014. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Chronic
Disease Prevention and Health Promotion, Office on Smoking and Health,
http://www.cdc.gov/tobacco/data_statistics/surveys/nyts.

Population Assessment of Tobacco and Health Study

Centers for Disease Control and Prevention (CDC). (2015d). E-cigarette use triples among middle and high
school students in just one year. Press Release (April 16). Atlanta, GA: Centers for Disease Control
and Prevention, National Center for Chronic Disease Prevention and Health Promotion,
Office on Smoking and Health, http://www.cdc.gov/media/releases/2015/p0416-e-cigaretteuse.html.
Cunradi, C. B., Moore, R., Killoran, M., and Ames, G. (2005). Survey nonresponse bias among
young adults: the role of alcohol, tobacco, and drugs. Substance Use & Misuse, 40, 171-185.
Currivan, D.B., Nyman, A.L., Turner, C.F., and Biener, L. (2004). Does telephone audio computerassisted self-interviewing improve the accuracy of prevalence estimates of youth smoking?
Evidence from the UMass Tobacco Study. Public Opinion Quarterly, 68, 542-564.
Fowler, F.J., and Stringfellow, V.L. (2001). Learning from experience: Estimating teen use of
alcohol, cigarettes, and marijuana from three survey protocols. Journal of Drug Issues, 31, 643664.
Groves, R.M., Brick, J.M., Couper, M., Kalsbeek, W., Harris-Kojetin, B., Kreuter, F., Pennell, B.,
Raghunathan, T., Schouten, B., Smith, T., Tourangeau, R., Bowers, A., Jans, M., Kennedy, C.,
Levenstein, R., Olson, K., Peytcheva, E., Ziniel, S., and Wagner J. (2008). Issues facing the
field: Alternative practical measures of representativeness of survey respondent pools. Survey
Practice, 1(3).
Javitz, H., and Wagner, M. (2005). Analysis of potential bias in the Wave 1 and Wave 2 respondents to the
National Longitudinal Transition Study-2 (NLTS-2). Menlo Park, CA: SRI International,
http://www.nlts2.org/studymeth/nlts2_analysis_bias_respondents.pdf.
Lundström, S., and Särndal, C-E (1999). Calibration as a standard method for treatment of
nonresponse. Journal of Official Statistics, 15 (2), 305-327.
Office of Management and Budget. (2006). Standards and Guidelines for Statistical Surveys, available at
http://www.whitehouse.gov/sites/default/files/omb/inforeg/statpolicy/standards_stat_surv
eys.pdf.
Roth, S., Montaquila, J., and Chapman, C. (2006). Nonresponse bias in the 2005 National
Household Education Surveys Program. Technical Report. (NCES 2007-016). U.S.
Department of Education, National Center for Education Statistics. Washington, DC: U.S.
Government Printing Office.
Ryan, H., Trosclair, A., and Gfroerer, J. (2012). Adult current smoking: Differences in definitions
and prevalence estimates—NHIS and NSDUH, 2008. Journal of Environmental and Public Health,
online.
Särndal, C.-E., and Lundström, S. (2005). Estimation in Surveys with Nonresponse. Hoboken, NJ: Wiley.
SAS Institute, Inc. (2013). SAS® 9.4 Help and Documentation. Cary, NC: SAS Institute, Inc.
Schenker, N., and Gentleman, J.F. (2001). On judging the significance of differences by examining
the overlap between confidence intervals. The American Statistician, 55, 182-186.

Population Assessment of Tobacco and Health Study

Schouten, B., and de Nooij, G. (2005). Nonresponse adjustment using classification trees. Discussion paper
05001, Statistics Netherlands. Available at www.cbs.nl.
Song, Y. (2013). Rotation group bias in smoking prevalence estimates using TUS-CPS. Paper presented at the
Federal Committee on Statistical Methodology Research Conference, Washington DC, paper
available at http://www.fcsm.gov/13papers/I3_Song_2013FCSM.pdf, and slides available at
http://www.copafs.org/UserFiles/file/fcsm/I3_Song_2013FCSM.pdf.
Thompson, M.E. (2015). Using longitudinal complex survey data. Annual Review of Statistics and its
Application, 2, 305-320.
United Nations. (2005). Designing Household Survey Samples: Practical Guidelines. United Nations
Publication ST/ESA/STAT/SER.F/98, New York: United Nations. Available at
http://unstats.un.org/unsd/demographic/sources/surveys/Handbook23June05.pdf.
United States Department of Commerce, Census Bureau. (2012). National Cancer Institute-sponsored
Tobacco Use Supplement to the Current Population Survey (2010-11):
http://appliedresearch.cancer.gov/studies/tus-cps/.
Technical documentation: http://www.census.gov/cps/methodology/techdocs.html.
Retrieved from http://appliedresearch.cancer.gov/studies/tuscps/results/data1011/table1.html.
Wagner, J. (2010). The fraction of missing information as a monitoring tool for survey data quality.
Public Opinion Quarterly, 74(2):223-243.
Wilson, E.B. (1927). Probable Inference, the Law of Succession, and Statistical Inference. Journal of
the American Statistical Association, 22, 209-212.
Young, A.F., Powers, J.R., and Bell, S.L. (2006). Attrition in longitudinal studies: Who do you lose?
Australian and New Zealand Journal of Public Health, 30, 353-361.

Population Assessment of Tobacco and Health Study

Appendix A
Cigarette Smoking Questions in the PATH Study and Other Surveys
Table A-1 lists the questions used to ask about current smoking status of adults in the PATH Study
and in the surveys used for comparison and describes the populations included in the estimates from
those surveys.
Note that although the questions used to define current cigarette smoking are similar among the
surveys, small differences could have an effect on the answers given. In the PATH Study, the
question used to establish whether a respondent has smoked at least 100 cigarettes in his or her
lifetime has closed response categories:
“How many cigarettes have you smoked in your entire life? A pack usually has 20 cigarettes in it.”
For adults:
1.

1 or more puffs but never a whole cigarette;

1 to 10 cigarettes (about ½ pack total);

11 to 20 cigarettes (about ½ pack to 1 pack);

21 to 50 cigarettes (more than 1 pack but less than 3 packs);

51 to 99 (more than 2 ½ packs but less than 5 packs); and

100 or more cigarettes (5 packs or more).

For youth:
1.

1 or more puffs but never a whole cigarette;

1 cigarette;

2 to 10 cigarettes (about ½ pack total);

11 to 20 cigarettes (about ½ pack to 1 pack);

21 to 50 cigarettes (more than 1 pack but less than 3 packs);

A-1

51 to 99 (more than 2 ½ packs but less than 5 packs); and

100 or more cigarettes (5 packs or more).

In TUS-CPS, NHIS, and NHANES, however, the question “Have you smoked at least 100
cigarettes in your entire life?” calls for a yes/no response.
The positioning of the questions also differs among the surveys. In the PATH Study, the cigarette
smoking questions are near the beginning of the adult questionnaire, and the respondent knows that
the questionnaire is about tobacco use behaviors. In TUS-CPS, the smoking questions are near the
beginning of the adult questionnaire on tobacco, but the survey is administered as part of the CPS.
In NHIS, the smoking questions follow a long series of questions on health problems (breathing
problems, diabetes, hernias, hemorrhoids, etc.). These question contexts may be associated with
differences in responses.
Table A-2 lists the questions used to define youth cigarette smoking in the PATH Study, NHANES,
NSDUH, and NYTS.

Population Assessment of Tobacco and Health Study

Table A-1.

Questions used to define adult current cigarette smoking in the PATH Study, TUS-CPS, NHIS, NHANES, and NSDUH

PATH Study
TUS-CPS*
NHIS
NHANES
Question to define current smoking (answers defining current smoking given in parentheses)
[“Have you ever smoked a
“Have you smoked at “Have you smoked at least
“{Have you/Has SP}
cigarette, even one or two
least 100 cigarettes 100 cigarettes in your
smoked at least
puffs?” (Wave 1 question, yes) in your entire life?”
ENTIRE LIFE?” (yes) and “Do 100 cigarettes in
or “In the past 12 months,
(yes) and “Do you
you NOW smoke cigarettes {your/his/her} entire
have you smoked a cigarette, now smoke
every day, some days or not life?” (yes) and “{Do
you/Does SP} now
even one or two puffs”
cigarettes every day, at all?” (every day or some
smoke cigarettes
(Wave 2 question, yes)] and
some days, or not at days)
every day, some
“Do you now smoke cigarettes all?” (every day or
days or not at all?”
every day, some days, or not
some days)
(every day or some
at all?” (every day or some
days)
days) and “How many
cigarettes have you smoked in
your entire life? A pack usually
has 20 cigarettes in it.” (100
or more cigarettes (5 packs or
more))
Age range included in estimate
18+
18+
18+
18+
Exclusions from population
The Wave 1 target population Includes only the U.S. Includes only the civilian
Includes only the
included only the U.S. civilian, civilian, nonnoninstitutionalized
U.S. civilian, nonnon-institutionalized
institutionalized
population residing in the
institutionalized
population.
population.
U.S. at the time of the
population.
interview. Several segments
The target population for
of the population are
Wave 2 was the Wave 1 target
excluded, such as: persons in
long-term care institutions;
population residing in the U.S.
persons on active duty with
at Wave 2, except for those
the Armed Forces; persons in
who were incarcerated at that
correctional facilities; and
time. Thus, it includes Wave 1
U.S. nationals living in
respondents who were on
foreign countries.
active duty or living in a health
care institution (e.g., a nursing
home) but not those in a
correctional facility at Wave 2.

A-3

NSDUH (original
definition)

NSDUH (modified
definition)**

“Have you ever
smoked part or all
of a cigarette?”
(yes) and “During
the past 30 days,
have you smoked
part or all of a
cigarette?” (yes)

“Have you ever
smoked part or all of
a cigarette?” (yes) and
“During the past 30
days, have you
smoked part or all of
a cigarette?” (yes) and
“Have you smoked at
least 100 cigarettes
in your entire life?”
(yes)

18+

Includes only the
U.S. civilian, noninstitutionalized
population.
Excludes homeless
persons who do not
use shelters,
military personnel
on active duty, and
residents of
institutional group
quarters, such as
jails and hospitals.

Includes only the U.S.
civilian, noninstitutionalized
population. Excludes
homeless persons
who do not use
shelters, military
personnel on active
duty, and residents of
institutional group
quarters, such as jails
and hospitals.

Table A-1.

Questions used to define “current smoking” in the PATH Study, TUS-CPS, NHIS, NHANES, and NSDUH (continued)

PATH Study
Proxy responses allowed
No

TUS-CPS*
Yes

NHIS
Yes, for individuals physically No
or mentally incapable of
responding.

NHANES

NSDUH (original
definition)
No

NSDUH (modified
definition)**
No

*Proxies are allowed if fourth callback, the person will not return before closeout, or the household is getting irritated. See http://appliedresearch.cancer.gov/studies/tuscps/surveys/tuscps_english_2010.pdf, p3.
**The modified definition is given in Ryan et al. (2012).

Population Assessment of Tobacco and Health Study

Table A-2.

Questions used to define youth cigarette smoking in the PATH Study, NHANES, NSDUH, and NYTS

PATH Study
NHANES
NSDUH
Question to define ever tried cigarette smoking (answers defining ever tried cigarette smoking given in parentheses)
“Have you ever tried cigarette
“About how many cigarettes have you smoked in
Have you ever smoked part or all
smoking, even one or two puffs?”
your entire life?” (1 or more puffs to 100 or more
of a cigarette?” (yes)
(Wave 1 question for all youth, and
cigarettes)
Wave 2 question for aged-up youth,
yes) or “In the past 12 months, have
you smoked a cigarette, even one or
two puffs?” (Wave 2 question for
continuing youth, yes)
Questions for determining whether have smoked in past 30 days
[“Have you ever tried cigarette
“On how many of the past 30 days did {you/SP}
“Have you ever tried cigarette
smoking, even one or two puffs?”
smoke a cigarette?” (1-30)
smoking, even one or two puffs?”
(Wave 1 question for all youth, and
(yes) and [“During the past 30
Wave 2 question for aged-up youth,
days, have you smoked part or
yes) or “In the past 12 months, have
all of a cigarette?” (yes) or
you smoked a cigarette, even one or
“During the past 30 days, that is
two puffs?” (Wave 2 question for
since [DATEFILL], on how many
continuing youth, yes)] and “When
days did you smoke part or all of
was the last time you smoked a
a cigarette?” (1-30)].
cigarette, even one or two puffs?”
(earlier today, not today but
sometime in the past 7 days, not in
the past 7 days but sometime in the
past 30 days)
Age range included in estimate
12-17
12-17
12-17

A-5

NYTS
“Have you ever tried
cigarette smoking, even
one or two puffs?” (yes)

“During the past 30 days,
on how many days did you
smoke cigarettes?” (1-30)

12-17

Table A-2.

Questions used to define youth cigarette smoking in the PATH Study, NHANES, NSDUH, and NYTS (continued)

PATH Study
Exclusions from population
The Wave 1 target population
included only the U.S. civilian, noninstitutionalized population.

NHANES
Includes only the U.S. civilian, noninstitutionalized
population.

The target population for Wave 2
was the Wave 1 target population
residing in the U.S. at Wave 2, except
for those who were incarcerated at
that time. Thus, it includes Wave 1
respondents who were on active duty
or living in a health care institution
(e.g., a nursing home) but not those
in a correctional facility at Wave 2.

Other comments

Youth with missing values for the response to the
question about number of lifetime cigarettes
smoked were excluded from the estimates of ever
tried cigarette smoking.
Youth with missing values for the response to the
question about number of cigarettes smoked in
the past 30 days were excluded from the
estimates of past 30 day cigarette use unless the
value was missing because the youth had never
smoked a cigarette in his/her lifetime. Youth who
had never smoked were treated as having smoked
zero cigarettes in the past 30 days.

NSDUH

NYTS

Includes only the U.S. civilian,
noninstitutionalized population.
Excludes homeless persons who
do not use shelters, military
personnel on active duty, and
residents of institutional group
quarters, such as jails and
hospitals.

Only includes youth who
are public and private
school students enrolled
in regular
middle schools and high
schools in grades 6
through 12 in the 50 U.S.
States and the District of
Columbia. Alternative
schools, special education
schools, Department of
Defense operated
schools, vocational
schools that serve only
pull-out populations, and
students enrolled in
regular
schools unable to
complete the
questionnaire without
special assistance, are
excluded.
Self-administered survey
in classroom.

File Type	application/pdf
File Modified	2016-06-24
File Created	2016-06-24