Responses to OMB questions

Question 1. The abstract in ROCIS says that this collection has two forms ("The project includes two forms: the Census of Juvenile Probation Supervision Offices and the Census of Juveniles on Probation."); yet only one form was submitted in ROCIS. I'm thinking the abstract needs to be fixed?

Angela is correcting this, it is just one form (the CJP) for this application.

Question 2. The form submitted in ROCIS says it's a draft; we need a final copy.

Attached is a final copy of the form, which includes the missing definitions. We also corrected the language in Box 1.

1. Page 18 is missing definitions for the terms.

Definitions are inserted

2. Box 1 on page 4 of the form says "Enter an identifying number of first name and last initial..." I think "of" should be "or." This is the case throughout the form in Box 1.

This has been corrected

Question 3. Can you explain how, specifically, respondents are supposed to rank the seriousness of offenses, as is requested in Box 6 of the form? What if two crimes are roughly equal in their seriousness? Is there a way to assist respondents with this?

Universally, offense seriousness has a high degree of consensus, with person offenses ranking highest, and with some potential, though small debate among which person offense is most serious (for example, armed robbery versus serious assault). The majority of states rank their charges according to statute. There may be slight variability within states, and perhaps slightly more variability across states. However, in terms of respondent burden and confusion, this self-ranking system has not proven to be an issue by respondents participating in our focus groups or earlier rounds of pretesting. Moreover, these same instructions and codes are used in the Census of Juveniles in Residential Facilities (CJRP) without any reports of difficulties by respondents in interpreting and deciding which offense is most serious. We are available to assist respondents make this determination on an individual basis, and this should reflect the very small but natural variation across states.

On a different but related note, codes that the form supplies have been extensively pretested both for this survey and the Census of Juveniles in Residential Placement. Respondents easily map their state statute onto these more universal codes. To further reduce burden and to assist in interpretation, the project team is developing a state-by-state cross walk of these census offense codes and each state’s statutes.

Question 4. The Summary of Burden Table doesn't appear to be right. For example, there's a 1 in the PRA Violation column, and there's a -789 in the change due to agency discretion column. Can you double check this?

Angela is correcting this.

Question 5. Was the focus group and any field testing submitted through ROCIS?

The focus group was not submitted through ROCIS/OMB as it took place under exploratory parameters and did not have a defined protocol. The prior administration of the CJP was a “field test” of the instrument on a sample of juvenile probation supervision offices. That was submitted through ROCIS/OMB. This was the prior approval of this form/collection

Question 6. On what specifically did OJJDP consult with BJS on and who were the relevant BJS staff members?

While OJJDP consults with BJS periodically on issues related to our statistics programs and data collection activities, in the case of the CJP, OJJDP consulted with BJS early in the development of the collection (2003), but has not consulted with any specific BJS staff members in the current collection process. We do periodically reach out to BJS in our custody collections (Census of Juveniles in Residential Placement and Juvenile Residential Facility Census). Specific staff include Paige Harrison, Mona Rantala, and Alan Beck. Howard Snyder (BJS’ Director of Special Projects) also has a long history of working with OJJDP on the custody collections.

Question 7. We understand that some focus group testing was done. Was there any post-focus group work operational testing done as well? For example, page 20 of Supporting Statement A says there was some field testing done - can you describe this? Also, was any mode effects testing done? For example, was it considered that providing the range of options for completing the survey (online, pre-formatted Excel doc, hard copy, etc) may reduce the response rate? Can you share your thoughts on this?

The OJJDP CJP Work Group (led by George Mason University) convened two testing sessions, February 7^th and 8^th, during the American Probation and Parole Association (APPA) annual winter meeting. At each testing session there were 2-3 volunteer responders (“ambassadors”) and 3 Work Group representatives. At each of the sessions ambassadors were asked to explore the online response system for the CJP and provide reactions to the system as well as census content. Ambassadors shared a computer, entered plausible but fictitious data, and considered possible problems based on their experience. Overall, ambassadors were delighted with the online system. For responders with only a few juveniles, ambassadors found it to be an easy to use, intuitive approach to data submission. For responders with hundreds or even thousands of juveniles, ambassadors were glad to have the alternative data submission options. All ambassadors indicated that the multiple options would increase participation.

The February meeting resulted in an invitation to attend a Juvenile Justice Association of Texas meeting in April. The same representatives and an additional OJJDP representative provided a presentation on the Benefits of Data Sharing. This presentation was attended by 15 Juvenile Probation Officers. Attendees were provided with an overview of the CJP and its content, an overview of data submission methods, and encouraged to find ways to use the collection for their benefit. Once again the various methods for submitting data was welcomed and suggestions to reduce responder burden were volunteered from the audience. Suggestions included creating a cross walk between state offense code and federal offense categories. At no time were there enough ambassadors or responders involved for testing of mode effects to occur.

Currently pre-registration for the data collection is very high. Interest in participation is highly encouraging. Once OBM approval has been finalized several of the non-responders have indicated that they will register and participate. (see U.S. map below)

Question 8. Have DOJ attorneys cleared the "peer pal program" discussed on page 11 of Supporting Statement A (see paragraph below)? Specifically, do they believe that OJJDP can ask respondents to waive confidentiality for some items? Also, we're having trouble understanding the highlighted sentence below. Can you re-convey?

Details regarding the Peer Pal program have been provided to Kathi Grasso, OJJP Legal Policy Advisor and Peter Brien, OJP Attorney Advisor, for their review and to the CSR, Inc., Institutional Review Board for its approval of the 2009 Census of Juveniles on Probation.

The following two paragraphs, which summarize the rationale for the Peer Pal Program and describe how it would operate, can replace the paragraph in the original submission that raised the question about the Peer Pal Program. Also, a justification for a very limited waiver of confidentiality is provided below, along with limitations on the scope of release and permissible use of the data. References to appropriate sections of OJP confidentiality regulations are provided.

In planning the 2009 Census of Juveniles on Probation (CJP), the Office of Juvenile Justice and Delinquency Prevention (OJJJDP) conducted formal focus groups and sought informal feed back from the field to encourage participation. Results indicated a strong desire among the respondents to have products and tools developed from the CJP that would be useful to them. In response, George Mason University (GMU) has developed data sharing tools and will manage a Peer Pals networking program. The first data sharing tool constitutes a “give back” to all respondents their own data presented in the form of charts and graphs to depict the distribution of their caseloads by selected variables of interest—hereinafter, the Juvenile Population Profile. In addition to providing respondents with their own results to view, early feedback to all respondents will assist the process of data editing and verification.

The second level of data sharing involves only registered Peer Pal participants who choose to volunteer to participate. Informed consent (see below) will be required to participate in the Peer Pal program. For those jurisdictions that choose to participate in the program through the secure web portal, GMU will obtain authorization from participants to disclose their identity to other Peer Pal participants. This facilitated linkage will enable them to exchange contact information (which is protected by federal law) with other participants who have also agreed to share the identity of their jurisdiction. Along these lines, focus group participants and others who attended training outreach sessions were very open to the creation of an e-mail listserve that would allow cross-jurisdictional discussions of problems of mutual interest and potential solutions that participants have found promising among Peer Pal Participants. GMU would construct the exchange process to allow participants to define the types of jurisdictions that they would like to either contact individually or receive summaries of 2009 CJP Juvenile Probation Profiles from a grouping of selected jurisdictions. Shared results will not reveal the identity of individual juvenile probationers nor the jurisdiction from which the data are provided, without prior approval by the comparison jurisdiction.

The protections contained in the Informed Consent, below, and the Privacy Certificate, have been developed with the purposes of the regulations in mind, especially (Sec. 22.1 (e)) of the DOJ Confidentiality of Identifiable Research and Statistical Information, which is, [to] “Insure appropriate balance between individual privacy and essential needs of the research community for data to advance the state of knowledge in the area of criminal justice.” Over the last decade OJJDP has promoted the use of best practices by practitioners, encouraged collaboration among practitioners and researchers to utilize data in crafting effective policies and practices. Under the Peer Pal Program, OJJDP plans to facilitate this exchange among practitioners who are participants in this data collection, to share contacts, Juvenile Population Profiles and operational information for systems improvements that are data driven.

[NOTE: PLAN B: However, should the information-sharing functions of the Peer Pal Program be deemed non-research in nature, the regulations do allow for the sharing of the information for administrative purposes as long as the proper consent has been obtained. Sec. 22.28 Use of Data Identifiable to a Private Person for Judicial, Legislative or Administrative purposes, specifies that “(a) Research or statistical information identifiable to a private person … shall only be … used for any purpose in any action, suit, or other judicial, legislative or administrative proceeding with the written consent of the individual to whom the data pertains; (b) where consent is obtained, such consent shall: (1) be obtained at the time that information is sought for use in judicial, legislative or administrative proceedings; (2) set out specific purposes in connection with which information will be used; (3) limit, where appropriate, the scope of the information subject to such consent.” The Notification of Consent, below, complies with the above requirements.”]

Sec. 22.2 (b) Revelation of Identifiable Information, specifically authorizes the type of release as proposed under the Peer pal Program under the condition that “Information may be revealed in identifiable form where prior consent is obtained from an individual…” In this instance, permission is being provided to share protected information, i.e., the identity contact information of an employee who has participated in the 2009 CJP, with the specific authorization of the agency CEO.

Further, provisions of the Peer Pal Informed Consent to Waive Confidentiality are consistent with Sec. 22.21 Use of Identifiable Data, which specifies that “Research or statistical information identifiable to a private person may be used only for research or statistical purposes.” While the relevant regulations relate primarily to protected information such as the identity of employees participating in the research, because of the OJJDP policy to not identify specific jurisdictions, authorization has been requested for that as well. Under no circumstances will the Juvenile Population Profiles, which contain aggregate information on age, race, offense, etc., enable the identification of individual youth, which is the other principal element protected by the regulations.

The following will appear on the web site with the appropriate mechanisms to authorize their participation in the program.

Description of the CJP 2009 Peer Pal Program and Statement of Informed Consent to Participate

To increase the utility of data for improving juvenile probation practices nationwide, OJJDP wishes to facilitate networking among Juvenile Probation Officials who participate in the 2009 Census of Juveniles on Probation (CJP 2009). Participation will enable exchange of contact information and certain juvenile probation population profiles and operational practices with other jurisdictions that have also chosen to participate in the program. Participation in this component of the program is also entirely voluntary and nonparticipation will not affect your access to other data resources and support provided to all CJP 2009 participants.

Federal laws and regulations require GMU and OJJDP to preserve the confidentiality of individual subjects of federally-sponsored studies, as well as the privacy of employees who have participated in such studies. Furthermore, it is OJJDP’s policy to not identify individual participating jurisdictions. Therefore, in order to participate in the Peer Pal Program, it is necessary for you to authorize the release of any protected information. Note that any such release excludes the disclosure of any information about the characteristics of juveniles that may reveal their identity. Civil and criminal penalties for the unauthorized disclosure of protected information that you cannot or have not waived are still applicable.

Please carefully read and enter your approval of the provisions for waiving confidentiality of selected information for specific purposes and agreement to abide by the terms of use of information provided by through the CJP 2009 Peer Pal Program.

I agree to the release of the following information for the purposes of participating in the CJP 2009 Peer Pal Program. I understand that my participation in this program is entirely voluntary and I may withdraw my permission at any time. I understand that any information provided to me through the Peer Pal Program, including, but not limited to the following: the identity of other Peer Pal Participants, their contact information, their population data, and selected agency operations is confidential and shall be used only for research or statistical purposes and shall not be divulged or shared with anyone without the prior consent of the other participant/s.

I agree to the terms above and am authorizing the release of the following information to others Peer Pal participants:

a) Participant (your) name

b) Job title

c) E-mail address/ CJP Userid –for GMU verification purposes only

d) Office mailing address

e) Name of JPSO jurisdiction covered in CJP 2009

f) Authorization of CEO authorizing your waiver, including his/her Name and CJP Userid, if you are not the CEO of the participating CJPSO

g) Selected data for an aggregate Juvenile Population Profile (identified on the portal)

h) Selected CJPSO profile characteristics (drawn from external factors, including jurisdictional crime rates, poverty rates, and other crime/social data at the jurisdiction level)

I further understand that anyone who requests the above information from me or my jurisdiction will be under the same obligations to protect its confidentiality use it only for the purpose establish for the Peer Pal Program.

I also understand that while I may withdraw at any time, and GMU will be required to immediately remove my jurisdiction’s data from the pool of other Peer Pal participant who may be interested in it, they may not be able to retract information that has already been shared with others up to that date.

Question 9. What is the statistical justification for conducting a census as opposed to a sample?

The purpose of this census is to provide a reasonable description of the juvenile probation population in terms of age, sex, race, most serious offense and status on probation at the national, regional, state, and (at the request of our respondents) local level. A statistical sample could reasonably be drawn to reflect the national characteristics, and each lower unit of generalizability would require an increasingly larger sample size and accurate consideration of all relevant covariates for sampling purposes. In conducting a cost-benefit review, we made two conclusions: 1. We were not confident that we had the external data points on which to accurately draw a representative sample, and 2. Sampling to report any meaningful data for local-level respondents were entail sampling from nearly every state. Pretesting reveals that in many cases, it is less of a burden for a central reporter to reporter to report on the full state than it is on a few jurisdictions within the state. Thus, it was determined that a more accurate and useful data collection would take the form a full census.

Question 10. Based on the survey literature, field testing, and experiences of similar collections, what response rate do you expect from this collection?

The 2006 pilot test of a statistical sample of jurisdictions (approximately 172) administered by the US Census Bureau yielded a disappointing response rate of 62%. This is especially disappointing since the companion Census of Juveniles in Residential Facilities has enjoyed a response rate of 90% or higher. The intense regrouping and the move to pre-registering and actively engaging respondents has led to the pre-registration of jurisdictions representing 85% of the juvenile population in the U.S.. These jurisdictions have agreed to participate in the October 2009 collection.

Question 11. What is OJJDP's nonresponse bias analysis plan for this collection?

These analyses are designed to determine whether nonresponse is random or systematic, and if the latter, systematic according to which theoretically important factors. Non-response analyses will examine statistical differences between full respondents, critical item respondent, pre-registered but non- or partial respondents, and respondents with whom we have had no successful communication. Bias analysis will include factors internal to the collections (eg, prior respondent status in CJPSO, office type, office size, respondent information collected in the preregistration process, including mode of response preferred) and external to the collections (eg, jurisdiction-level information on demographic and social characteristics of juveniles collected through national data program).

Question 12. Page 16 of supporting statement A says:

In addition, OJJDP expects to produce some publications that summarize the data findings (as either Fact Sheets or OJJDP Bulletins) for the juvenile justice field. The data will be archived and available to the field through an agreement with ICPSR.

Can you provide more information about the specific analysis plan and the specific kind of data that will be made available (e.g., what types of tabulations do you plan to publish?

In general, OJJDP produces summary data findings of our large data collections online through OJJDP’s Statisical Briefing Book, through OJJDP publications (fact sheets, bulletins) which are written for the juvenile justice field; and through numerous conference presentations. OJJDP maintains an ongoing grant with the National Center for Juvenile Justice to produce summary statistics. The grant, called the National Juvenile Justice Data Analysis Program, is for the ongoing maintenance and updating of the Statistical Briefing Book, and the production of data-related Fact Sheets and Bulletins.

OJJDP’s Statistical Briefing Book is located online at http://ojjdp.ncjrs.gov/ojstatbb/ As indicated under the left navigational bar, the briefing book provides statistical overviews of all key indicators and points in the system. Currently, the “Juveniles on Probation” section of the briefing book is rather sparse, but it is anticipated that the CJP data will enable OJJDP to provide Frequently Asked Questions and other resources under this category that will be similar to the type of information that is currently available under Juveniles in Corrections (http://ojjdp.ncjrs.gov/ojstatbb/corrections/index.html). With the CJP results, some of the FAQs we anticipate adding under Juveniles on Probation include:

• How many juveniles are on probation on a given day in the U.S.?

• What is the female proportion of juveniles on probation?

• How do probation rates vary by race?

• How old are most juveniles on probation?

• Does the race/ethnicity profile of juvenile offenders on probation vary by offense?

• Does the race/ethnicity profile of juveniles on probation vary by offense and gender?

• How do female probation rates vary by race/ethnicity and State?

• How do male probation rates vary by race/ethnicity and State?

• Does the offense profile of juveniles on probation vary by State?

• How does the type of offense resulting in probation vary by race/ethnicity?

Numerous other questions can be added; those listed above are an example only. In addition, OJJDP anticipates producing an overall “Juveniles on Probation” bulletin that would summarize the findings, as well as a series of online Fact Sheets that address some of the key issues outlined above. Regarding conference presentations, it is expected that OJJDP will present CJP findings at the next American Probation and Parole Annual Meeting, as well as at several juvenile justice related meetings and conferences (including a few sponsored by OJJDP) in 2010.