supporting Statement B

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS

This data collection will be a census of juveniles in custody on the reference data. 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 juvenile offenders 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 juvenile be held for the offense and not for some other problem behavior such as alcohol or drug abuse. Also, it is important that the facility specifically be holding the youth for the offense. OJJDP has defined juveniles to be any person under 18 years of age although many States define the age of majority differently (e.g., 16 in New York, 17 in Wisconsin).

OJJDP intends to survey all public and private facilities in the United States that fulfill these requirements. The 2006 CJRP included a total of 1,191 public and 1,878 private residential facilities. 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.

OJJDP has determined that a census will serve the government=s interest better than a sample survey. Based on the input from OJJDP=s outside consultants, the Office determined that a nationally representative sample of facilities would not suffice. States wish to make comparisons among themselves, and given that juvenile justice policy is made at the State level, a national sample would not serve their purposes. Most States have only a few facilities (some with one or two). In order to create a sample that would suffice for making adequate State level estimates, OJJDP would in effect conduct a census in many States. Thus, to create the State level estimates of residential placement would almost require a national census.

2. Information Collection Procedures

OJJDP must maintain an accurate and complete list of all facilities that are of interest. To accomplish this task, OJJDP annually funds a broad agreement with the Census Bureau to maintain a list of facilities that includes the name, address, location, phone numbers, and classification information. To maintain this list, the Census Bureau has developed procedures for assuring a complete list of facilities. These procedures include examining resource materials from professional organizations and State organizations, as well as periodically contacting agency personnel in the States able to provide information on new facilities, facility closings, and changes in facility characteristics.

As for the actual collection of the information, through the Census Bureau, OJJDP will pursue the following schedule:

Time frame	Action
4 weeks prior to reference date	mailout advance notice letter
2 weeks prior to reference date	mailout survey forms
1 week following reference date	mailout reminder letters (non-respondents only)
4 weeks following reference date	mailout a second-notice survey form (non-respondents only)
6 weeks following reference date	begin telephone follow-up

This schedule was developed based on experience with other censuses and experience in testing and administering the CJRP in previous years. Should circumstances require changes (most likely to move forward the telephone follow-ups), the schedule will be changed accordingly.

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

3. Methods to Maximize Response Rates

OJJDP and the Census Bureau are committed to very high response rates and high quality data. For many years, the CJRP collection resulted in high response rates. In 2007, there were a total of 2,911 “in scope” facilities in the universe. For the 2007 CJRP, the response rate was 76% (full response), representing 2,215 respondent facilities. An additional 687 facilities provided “critical item information” and 9 facilities did not respond. Over the past few collection periods, the CJRP has experienced a reduction in the percentage of full responders, and an increase in the percent of “critical item” responders, necessitating that we need to impute for an increasing proportion of the universe on several key variables. Specifically, this means imputation rates have increased significantly over the past few survey cycles for juvenile facilities that hold persons under age 21 that are charged with or court-adjudicated for an offense. The gradual trend has become a source of concern for subsequent survey cycles. If the trend continues, imputed data will make up such high percentage of the final data file that the quality will be compromised. The details of these concerns are outlined in the cover memo of the 2007 CJRP Imputation Report (see attachment C.). As a result of these concerns, OJJDP is in the process of launching a new “client development” effort to enhance retention and increase full respondent rates. This client development effort will utilize staff from OJJDP’s corrections data advisory group in exploring and possibly utilizing the following techniques::

• Electronic reporting in a manner acceptable to the respondent (Census Bureau is currently exploring a more user friend online reporting mechanism

• Streamlined forms and clear response instructions

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

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

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

These procedures have been developed and tested through the administration of the CIC and through four administrations of the CJRP. They have proven effective in achieving and maintaining a high response rate of which OJJDP and the Census Bureau are quite proud.

4. Tests of Procedures

The development of the CJRP followed a solid development design. The Center for Survey Methods Research at the Census Bureau began with semi-structured exploratory interviews of 20 respondents. The respondents were varied based on size and type of facility. These interviews were designed to learn how respondents think about the population in their facilities and how they understand various important concepts OJJDP wishes to report on (for example, delinquent versus status offense).

Results of these interviews informed the development of a test instrument which was reviewed and refined by OJJDP staff and a group of consultants. CSMR used the refined draft instrument to conduct multiple rounds cognitive interviews with respondents. Based on these interviews, CSMR and OJJDP produced an instrument for pre-testing. Using a reference date of October 30, 1996, the Governments Division of the Census Bureau conducted a pre-test of 400 facilities. This test included a small sample of facilities (96) that would receive the CIC form so that CSMR could compare the results of these two tests to more accurately judge how the CJRP form performed compared to the CIC. Statisticians at the Center for Survey Methods Research (CSMR) at the Census Bureau analyzed the data and provided a report to OJJDP.

Since the first full administration of the CJRP in 1997, OJJDP has worked with the Census Bureau to establish and maintain appropriate statistical procedures for the data files. As part of the normal procedures for each CJRP file, the Economic Statistical Methods and Programming Division (ESMPD) of the Census Bureau analyze the quality of the data and develop methods for imputing for facility non-response and item non-response. ESMPD provides a detailed report for each file. This report demonstrates the quality of the data collection efforts and the procedures the Governments Division uses to collect the CJRP data. The Imputation Report for the 2003 collection is included in Attachment D.

5. Statistical Consultants

Presently, OJJDP funds an Interagency Agreement with the Governments 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.

OJJDP also funds an IAA with George Mason University for field testing of questions and form design issues, as well as a grant with the National Center for Juvenile Justice to perform initial analysis of the CJRP data file, and to produce summary findings for publication and online dissemination.

Relevant Contacts:

Stephen Simoncini

Governments Division

U.S. Census Bureau

Washington, DC 20233

(301) 763-1549

Suzanne M. Dorinski and Carma Hogue

ESMPD

U.S. Census Bureau

Washington, DC 20233

(301) 763-4869

Catherine Gallagher

Assistant Professor

Department of Public and International Affairs

George Mason University

Fairfax, VA

(703) 993-8480

Melissa Sickmund and Charles Puzzanchera

National Center for Juvenile Justice

Pittsburgh, PA

(412

1Status offenses are those offenses which are illegal for minor 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 incorrigibility, underage drinking, or curfew violations.