final jrfc supporting statements B

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS

The JRFC universe includes all public, private, territory and tribal juvenile facilities in the United States. In the best of all possible scenarios, statistical estimation would not be required. However, given the inevitable facility non-response and item non-response, OJJDP (as in previous years) will work with the Census Bureau to ensure valid and reliable procedures to estimate the population characteristics.

1. Respondent Universe and Sampling Methods

For this census, OJJDP has defined the universe to include all facilities that hold juveniles as offenders. An offender is defined as a youth who has committed a crime or status offense¹ and who is being held because of that offense. It is important that the collection include facilities that hold juveniles for an offense and not for some other problem behavior such as alcohol or drug abuse. While most States define juveniles to be any person under 18 years of age, some states define the age of majority (i.e., a juvenile) differently (e.g., 15 in New York, 16 in Wisconsin).

OJJDP intends to survey all public and private facilities in the United States that fulfill these requirements. These facilities run the gamut of environments from open facilities in which the youth reside in a home environment to the high security training schools that house upwards of 400 youth.

Based on expert consultation, OJJDP determined that a census is most efficient and effective in meeting the government’s interest because a nationally representative sample of facilities would not allow for between state comparisons, hampering states’ ability to develop informed juvenile justice policy at the State level.

2. Information Collection Procedures

To maintain an accurate and complete list of all facilities of interest, OJJDP annually funds the Census Bureau to maintain a list that includes the facilities’ names, addresses, locations, phone numbers, e-mail addresses, and classification information. Although the JRFC collection occurs, biennially, the universe of juvenile residential facilities is supported and maintained on an annual basis because it is used for both the JRFC and CJRP, which occur in alternating years.

To maintain this list, the Census Bureau regularly receives resource materials from OJJDP and other professional and state juvenile justice organizations, and periodically contacts OJJDP grantees, juvenile justice stakeholders, and state juvenile justice agency personnel to gather information on new facilities (births), facility closings (deaths), and changes in facility characteristics. The following individuals and organizations are queried annually to provide updates to the juvenile justice facility universe:

• OJJDP Core Protections Division (compliance data submitted by states);

• Council of Juvenile Correctional Administrators;

• National Center for Juvenile Justice;

• Center for Coordinated Assistance to States (a training and technical assistance provider that assist states in complying with the four core protections of the Juvenile Justice and Delinquency Prevention Act);

• State Juvenile Justice Specialists (these individuals oversee the management of OJJDP’s State Formula and Block grant funds); and

• State Compliance Monitors (these individuals oversee the monitoring of juvenile justice and adult facilities within the state for compliance with requirements of the Juvenile Justice and Delinquency Prevention Act).

Since the initiation of the CJRP and JRFC in 1997 and 2000 respectively, positive, long-term relationships have developed among the data collection agents at the Census Bureau, OJJDP, and the CJRP/JRFC respondents. Many of the updates result from direct respondent contact with the Census Bureau. Often, new or changed information is on submitted CJRP or JRFC survey forms, prompting follow-up as necessary.  Additionally, a non-deliverable form returned via the United States Postal Service or a non-response will prompt the Census Bureau to initiate research on a facility to gain updated information.  This outreach is often as simple as contacting the respondent on file or a state agency. Depending on the outcome, a more in-depth search may be implemented, at times with OJJDP assistance.

Since FY2013, OJJDP and the Census Bureau have agreed to add a separate and distinct task to their Interagency Agreement to conduct comprehensive research, development, and maintenance of the juvenile facility frame used for both the JRFC and the CJRP. Work to be performed by the Census/Governments Division includes:

• Maintain an accurate and up-to-date contact list of the survey respondents. Verify and update the names, titles, addresses, and contact information of the respondents, including the phone and fax numbers, and email addresses prior to each data collection cycle.

• Engage key stakeholders to obtain input frame maintenance and development. Network and participate in meetings of state juvenile corrections administrators to seek advice on developing new methods for universe development and maintenance. Establish partnerships at the state and local level to build and maintain up-to-date lists of all facilities.

• Research facilities online, through publications and associations, and work with State data providers and private juvenile residential companies to develop new ways to obtain data and implement new central reporting arrangements.

During 2014 (the most recent period for which data are available--through the JRFC), the Census Bureau identified 251 facilities that reported changes affecting the JRFC/CJRP universe. These changes include information updates (changes to respondent contact, facility address and facility name) and status changes (facility births, deaths, merges, and temporary closures).

During the 2015 CJRP, the Census Bureau identified 208 facility closures, 122 point of contact changes, 258 address changes, and 4 facility name changes. In addition, verification calls identified an additional 82 new facilities and 74 closed facilities. Research, including review of state compliance documents identified 61 new facilities.

As for the actual collection of the information through the Census Bureau, OJJDP will pursue schedule below. This schedule was developed based on experience with other censuses and experience in testing and administering the JRFC in previous years. Should circumstances require changes (most likely to move forward the telephone follow-ups), the schedule will be changed accordingly. After the initial mailout and due date have passed, Census Bureau staff prepare and execute a second mailout of the form to those facilities that have not responded. Subsequent follow-up involve staff contacting the remaining respondents via phone to secure participation and/or to gather critical item data. Additional details on nonresponse contact protocols can be found in Attachment N.

Typically, OJJDP has been able to achieve a high response rate (85 to 95 percent) for its facility-based censuses. Such a level of response has proven sufficient for designated analysis purposes. The Office expects to continue this high response rate in future administrations of the JRFC.

TASK NAME	FINISH DATE
PROJECT PLANNING
OMB clearance
Develop project schedule	01/16
Prepare NPC cost estimate	03/16
Submit ASD Rollover cost estimate	05/16
Questionnaire design	03/16
JAL/Universe Update	continuous
CENTURION (ASD)
Kickoff Meeting	06/16
Bi-weekly meetings	continuous
Requirements Document (EIS Document)	05/16
Changes to the Centurion Site document	05/16
Create Input File	09/16
Design and Submit Fillable PDF	09/16
Test Online Data Collection Tool	09/16
Re-test Online Data Collection Tool	09/16
Centurion User Guide Document	09/16
Upload Fillable PDF to Instrument	09/16
Instrument moved to production	10/16
GPS (ISB)
Kickoff Meeting	07/16
Bi-weekly meetings.	continuous
Specifications for mandatory updates, enhancements and corrections submitted to ISB (GPS)	09/16
Updates to Application: includes all updates, presentation, and edits that were documented on specifications	08/16
Test Processing application (GPS) includes testing usability, functionality and edits	09/16
Requirements: Submit Crosswalk to ISB and ASD	08/16
Test Output transfer from Centurion to GPS	09/16
Test transfer from Processing App to Database	09/16
Data Review, Analysis & Editing	08/17
DATA COLLECTION (ACSD/NPC)
Questionnaire printing	10/16
Mailout and Preparation
Requirements: Submit to CENDOCs	04/16
Submit Draft of CJ-15 form to ACSD	05/16
Receive updated mailout letter from OJJDP	06/16
Print CJ-15 Form	09/16
Order mailout materials (envelopes, supplies, etc)	07/16
Receive letter approval from sponsor	08/16
Print Letters and webflyers	09/16
Establish and transmit mail file to NPC	09/16
Mail CJ-15 Questionnaires Spec to NPC	10/16
Mailout(s)
JRFC Questionnaires/letter/webflyer mailed to respondents	10/16
Reference date: October 26, 2016	10/26/16*
Data collection due	12/02/16*
Second request mailed	01/11/17*
NPC Keying Requirements
Develop keying specifications	08/16
Develop non-response specifications	08/16
Conduct Keying Operations	06/17
Nonresponse follow-up
Develop procedures	01/17
Conduct training	02/17
Conduct nonresponse follow-up	06/17
Closeout	06/17
TABULATION/ANALYSIS (ESMPD)
Imputations
Develop specifications	05/17
Program specifications	05/17
Generate and review imputations	05/17
Tables/estimates
Develop specifications	06/17
Program specifications	06/17
Generate and review tables	07/17
Data file
Review/correct data	06/17
Prepare preliminary file	06/17
Deliver file to ESMPD	07/17
Prepare documentation package	08/17
Deliver final file to OJJDP	09/17
Archive Data	12/17
EVALUATION
Evaluate survey operations	continuous
Develop recommendations	continuous

* Subject to OMB approval. If approval is not obtained in time to meet the planned reference date, OJJDP will plan to administer the data collection in early 2017.

3. Methods to Maximize Response Rates

OJJDP and the Census Bureau are committed to very high response rates and high quality data. Using the number of in-scope facilities as a base, the JRFC facility response rate is typically between 85 and 95 percent. There were 2,223 in-scope facilities on reference day in 2014: 1,942 of the 2,223 facilities were counted as respondents for a unit response rate of 87.4 percent. The reported unit response rates do not count critical item facilities as respondents.

A small proportion of in-scope facilities neither return JRFC forms nor do they respond to telephone calls from Census Bureau personnel requesting that the institution participate in the census. In total 186 facilities refused to participate in the 2014 JRFC. The Census Bureau imputed records for the nontribal facilities in the 50 states and the District of Columbia that were refusals. Some facilities were not able to provide all of the information requested. For example, the item non-response rate for key items in 2014 was 7.8 percent. In such cases, data are imputed from complete records to fill in incomplete records. Nonresponse adjustment procedures are summarized below and described in the attached methodology report (see Attachment O).

OJJDP will use the following techniques to maximize response:

• Electronic reporting in a manner acceptable to the respondent

• Streamlined forms and clarify response instructions

• Continued support at the Census Bureau through a toll free number to answer any questions that arise

• Continuous contact with respondents through email and U.S mail (see the schedule for mailout and reminder notices above)

• Call-back procedures that continue until data close-out in May/June of the following year.

Notably, based on feedback from respondents and analysis of JRFC paradata, the Census Bureau implemented enhancements to the functionality of its online data collection system to improve respondent experiences, including: adding comment boxes to each section, adding a filter to the review pages, and adding totals to the edit message to help respondents correct those responses that do not sum as expected.

Additionally, staff at the Census Bureau have reported that facilities may be more likely to respond if they are able to see the importance of their data in understanding national trends, and those respondents with confidentiality concerns may be reassured that their data are only reported in the aggregate. So in 2015, in an effort to demonstrate to facilities how their data are used by policymakers and the public, the OJJDP-published bulletin: Juveniles in Residential Placement, 2011, was included in the October mailout of the CJRP instrument. This technique is also planned for the 2016 JRFC data collection which will include copies of the OJJDP bulletin: Juvenile Residential Facility Census, 2014: Selected Findings.

OJJDP anticipates this ongoing effort to engage respondents will continue to yield positive outcomes.

Nonresponse Adjustments

The 2016 JRFC is a census of all the juvenile residential facilities in the United States and its territories. No sample weights are used in the data collection. As noted above, it is expected that a small proportion of in-scope facilities will not respond to the census and some facilities will not respond to all items. Overall, unit and item-level response rates are expected to remain high. Nevertheless, it is expected that missing values will be imputed for both unit and item nonresponse. The imputation methods include:

• Prior year data,

• Growth rate applied to prior year data,

• Median value reported in imputation cell,

• Most commonly reported value in imputation cell,

• Value derived based on other fields on the facility record,

• Ratio estimation.

If the number of adults, juvenile offenders, or juveniles assigned to beds for reasons other than offense are missing, the Census Bureau applies a growth rate to the prior year value. If no prior year value is available, the Census Bureau uses the median value reported within the imputation cell.

If the number of juveniles assigned to beds is missing, the Census Bureau derives the value by adding juvenile offenders and juveniles assigned to beds for reasons other than offense. If persons assigned to beds is missing, the Census Bureau derives the value by adding juveniles assigned to beds and adults assigned to beds.

If total standard beds is missing, the Census Bureau pulls forward prior year data. If no prior year data is available, The Census Bureau multiplies the median value of (total standard beds)/(persons assigned to beds ) by persons assigned to beds.

If occupied makeshift beds is missing, the Census Bureau pulls forward prior year data. If no prior year data is available, we use the median value reported within the imputation cell.

The Census Bureau uses the most commonly reported value in the imputation cell for the questions on sleeping room arrangements, minutes per day of required large muscle activity, and days per week of required large muscle activity when the values are missing for 2016 and there are no prior year values.

If the values for the check box questions in the General Facility Information section are missing, the Census Bureau pulls forward prior year data if available. Those questions include the questions about the physical layout of the facility, whether or not on-site residential treatment, foster care, or independent living are provided, the types of sleeping units in the facility, who owns and operates the facility, and how juveniles are confined in specific areas.

The Juvenile Residential Facility Census and the Census of Juveniles in Residential Placement are sent to juvenile residential facilities in alternating years. Prior year data exist for most facilities, and researchers and policy makers are interested in longitudinal trends. The key items collected in the JRFC are:

• Persons assigned to beds,

• Persons assigned to beds age 21 or older,

• Persons under age 21 assigned to beds,

• Persons under age 21 assigned to beds for offenses,

• Persons under age 21 assigned to beds for reasons other than offenses,

• Total number of standard beds,

• Total occupied makeshift beds.

In addition to tracking the key item response rates over time, the Census Bureau also tracks the total quantity response rates over time. The total quantity response rates for 2012 and 2014 are more than 90 percent, with the exception of the occupied makeshift beds. Very few facilities report occupied makeshift beds, and very few facilities had nonzero values imputed for occupied makeshift beds.

The total quantity response rate shows how much of the quantity estimate was reported. The key item quantity response rates are usually higher than the response rates for the key items, with the exception of persons under age 21 assigned to beds for reasons other than offenses and total occupied makeshift beds. These results are expected, since nonresponse follow up places more emphasis on the larger facilities and central reporters.

Additional details on unit and item level nonresponse and nonresponse adjustments are included in an attached methodology report (see Attachment O).

4. Tests of Procedures

There are no changes to the forms or data collection methods at this time. However, in the future, any significant changes to the form would require that OJJDP obtains feedback from state, local, and private agencies to ensure new items, definitions, and counting rules are clear across jurisdictions. In addition, OMB approval would need to be obtained, which includes an outside review of the survey. Consistently high response rates in recent JRFC and CJRP data collection cycles suggest that that the planned procedures for the 2016 JRFC will yield similar or improved results.

The Census Bureau and OJJDP tested the data collection procedures in a field test that began in October 1998. This test included a sample of 500 facilities. It was not a random sample but rather a sampled based on facility type and facility size. The combination of these variables created a series of strata. The goal was to test the procedures on each stratum and to modify the data collection procedures as necessary. The results of this test resulted in several changes to the instrument and to the collection methods; however, these changes were more cosmetic rather than substantive. The content of the survey did not require changing, but the Census Bureau learned much about how to administer this type of survey to the facilities included.

Since then, OJJDP has worked with the Census Bureau and other consultants to periodically retest the instrument in the field, in particular at times when new questions or topics have been proposed.

5. Statistical Consultants

Presently, OJJDP funds an Interagency Agreement with the Economic Reimbursable Surveys Division at the Census Bureau to perform data collection, and to maintain the data file and address lists. This IAA also funds the imputation activity (ESMPD) related to the CJRP file.

Relevant Contacts:

Brecht Donoghue

Deputy Associate Administrator, Innovation and Research Division

Office of Juvenile Justice and Delinquency Prevention

(202) 305-1270

Benjamin Adams

Social Science Analyst, Innovation and Research Division

Office of Juvenile Justice and Delinquency Prevention

(202) 616-3687

Nicole Adolph

Chief, Criminal Justice Statistics Branch

Economic Reimbursable Surveys Division

US Census Bureau

(301) 763-1577

Alonzo Johnson

Supervisory Statistician, Criminal Justice Statistics Branch

Economic Reimbursable Surveys Division

US Census Bureau

(301) 763-4511

Suzanne Dorinski

Mathematical Statistician

Economic Statistical Methods Division

US Census Bureau

(301) 763-4869

1Status offenses are those offenses which are illegal for minors but not for adults. For example, truancy or running away may be a status offense depending on the state in which the juvenile resides. Other status offenses include ungovernability, underage drinking, or curfew violations.

OMB Submission Juvenile Residential Facility Census Page 10 of 26