DCRP Supporting Statement Part B

9

B. Statistical Methods

1. Universe and Respondents Selection

The universe for the Deaths in Custody Reporting Program (DCRP) is all deaths occurring during the process of arrest by, or in the custody, of state or local law enforcement personnel; while persons are in custody in local jails; and while offenders are incarcerated in state prisons. Respondents are persons in organizations having custody over the decedent. The potential respondent universe consists of (1) all state and local law enforcement agencies in the United States (about 18,000 agencies); (2) all local jails in the United States (about 3,000 such jails); and (3) each of the 50 state departments of corrections. For (1), BJS uses State Reporting Coordinators (SRCs) who, as agents of BJS, use various methods to collect data about arrest-related deaths occurring in state and local law enforcement agencies. For (2) and (3), BJS collects data from persons in the organizations directly.

For each of the three collections under the DCRP (as identified in the introduction to Part A), BJS collects data on the entire universe of deaths that fall under the scope of the DCRP, rather than sampling deaths or agencies. The original Death in Custody Reporting Act (DICRA) legislation (referenced in Part A, Item 1) required all state prisons, local jails, and state and local law enforcement agencies to report data on deaths in custody to the Attorney General. BJS established reporting procedures for the DCRP in response to the DICRA legislation and has maintained these reporting procedures since. In 2000 and 2001, BJS established procedures for collecting data directly from the roughly 3,000 local jails and 50 state departments of corrections. In doing this, BJS built upon established contacts and procedures for collecting data from these entities for other BJS collections.

Due to cost constraints, BJS developed a different approach to identifying respondents for the Arrest-Related Deaths (ARD), as contacting the 18,000 state and local law enforcement agencies was beyond BJS’ means. In establishing the ARD portion of DCRP, BJS identified and relied upon centralized state-level coordinators (the State Reporting Coordinators or SRCs) to identify deaths that were within the scope of ARD and then collect data directly from the local law enforcement agencies in their respective states in which the deaths occurred. The SRCs would then transmit the data to BJS. The SRCs are typically located within state criminal justice commissions (commonly administered by the governor’s office) as well as in state attorney general offices and state police departments. (See Part B, Item 2, “Procedures for Information Collection” for information about how the SRCs obtain the data.) BJS chose the SRCs because the offices in which they were located were established and recognized state-level, criminal justice offices that had legitimacy with local police departments, because they had connections with local law enforcement agencies in their states, and in many cases the SRCs had pre-existing contacts with local law enforcement agencies.

BJS obtains the universe of deaths rather than samples for a number of reasons. One goal of the DCRP is to track changes in mortality in state prisons, local jails, and in the process of arrest, and sampling would complicate this task immensely as well as preclude making important comparisons over time among causes of death and by demographic groups. Sampling deaths to obtain information about the circumstances leading to DCRP-related deaths would significantly compromise BJS’ ability to accurately report on mortality in prisons, jails and in the process of arrest. Sampling respondents would not appreciably reduce burden but would greatly diminish the quality and utility of the data.

For the prison portion of the DCRP, one goal of the program is to track state-level changes in mortality in prisons, not simply to generate a national estimate of prison mortality. The only option to achieve this goal by sampling would be to sample death records within a state. Given the relative rarity of some important manners of death (e.g., suicide and homicide), sampling strategies, by design, would be very complicated and potentially burdensome on the respondents.

In the local jail portion of the DCRP, BJS tracks changes in mortality and makes temporal and geographic comparisons over time, while also producing national data on mortality in jails. Sampling local jails would not appreciably reduce burden, and would diminish the quality of the data, since a sampling strategy producing national estimates of jail deaths would necessarily include the largest jails in the country, plus regional jails (those that serve multiple counties). These 300 jails consistently provide the majority of deaths, so a sampling plan that yielded precise national estimates of mortality would not reduce their burden.

Sampling smaller jails would decrease the reliability of estimates. During a given year, approximately 80% of jails report no deaths, so their reporting burden is quite small—i.e., they only fill out the annual summary form and participate in the verification calls. However, jails do not report zero deaths from year to year on a consistent or predictable basis. Sampling to capture variation would require either a fairly large number of small jails or different cross-sections from year to year. A large panel would not appreciably reduce burden and would reduce accuracy. A cross-sectional design could allow for greater accuracy in estimates of deaths in jails with small numbers of deaths, but would involve greater administrative burden and possibly increase respondent burden for respondents that fall in and out of the sample.

Further, although local jails are typically administered at a county level, BJS generates state-level total counts of jail inmate deaths. On average, there are 60 jails per state, so sampling to produce state-level estimates would require a nearly full enumeration of jails, eliminating any reduction in burden. Consequently, sampling jails to obtain mortality data does not provide cost or burden savings and would likely reduce the accuracy of information. Finally, jails are now accustomed to the collection and have adapted their information systems to readily provide the data we seek.

Sampling law enforcement agencies to collect data on arrest-related deaths would present the same obstacles as sampling local jails. Much like local jails, arrest-related deaths are rare events and a sampling scheme would be costly, burdensome, and unlikely to produce accurate results.

2. Procedures for Information Collection

Procedures for the DCRP-ARD collection. Since 2003, BJS has used State Reporting Coordinators (SRC) to compile information about deaths that occur during the process of arrest by, or in the custody of, state and local law enforcement personnel. The role of the SRC is to actively identify deaths that are reportable to the ARD program, compile information about these deaths, and submit data on these incidents to BJS. Some SRCs use BJS funding through BJS’ State Justice Statistics Program for State Statistical Analysis Centers to support staff time for ARD-related work.

In January of each year, BJS notifies the SRCs by email (see Appendix C, Attachment G) about the initiation of a new year of ARD data collection and provides the SRCs with key program materials, including (a) the CJ-11 Summary of Incidents form; (b) the CJ-11A Incident Report, and (c) a Question-by-Question Guide to aid in completing the CJ-11A. In addition, the email provides general program information, such as key dates and methods for submitting data. SRCs may submit data on an ongoing basis or they may submit records according to a predetermined submission schedule. The SCRs can transmit data to BJS via mail, encrypted email, or secure fax.

The SRCs two major responsibilities are (1) to identify all ARD reportable deaths in their state and (2) to complete a CJ-11A (incident report) for each death. SRCs use a variety of methods to identify ARD-reportable deaths, including relying upon centralized reporting systems on violent deaths that occurred in their states, using their contacts with law enforcement agencies, medical examiners, or coroners’ offices in their respective states, and/or conducting systematic open-source searches. Once SRCs identify ARD-reportable deaths, they follow-up with local law enforcement agencies to obtain the data necessary to complete a CJ-11A form for each death. In instances where the involved law enforcement agency does not complete the CJ-11A form, the SRC completes the form using information from official source documents (such as a police report or death certificate).

Independently of the SRC efforts to identify ARD-reportable deaths, BJS’ ARD program staff continuously monitor open sources for reportable deaths, using a set of Google Alerts to identify deaths that occur during an interaction with law enforcement personnel. The names of the decedent and the law enforcement agency involved with the death are identified and compiled to be compared to records that the SRCs submit to BJS. The comparisons occur during the production of the status report (see below) that occurs at the end of a collection year.

SRCs are to submit to BJS their final set of records within 60 days following the end of the calendar year. BJS’ ARD staff review these data and assess them for completeness, consistency of responses, and make a determination as to whether a reported death falls within the scope of ARD. BJS prepares a “status report” (see Appendix C, Attachment H) that identifies incomplete records that BJS sends back to the SRCs for follow-up on missing items. The status report also notifies SRCs of any changes made to submitted records or any records that were determined outside the scope of the program. Lastly, the status report notifies SRCs of reportable deaths identified by ARD program staff that were not included in the SRC’s submission, and the SRCs follow-up with local law enforcement agencies to verify these deaths. The SRCs obtain data from the local agencies to complete a CJ-11A form for ARD reportable deaths.

Follow-up contact for non-responding SRCs begins 60 days after the end of the year. These respondents are contacted via email and asked to submit their data from the prior calendar year as soon as possible. Concentrated data quality follow-up with SRCs occurs during the summer, in which SRCs are reminded to submit responses to issues identified on the status report and to submit any additional cases. All follow-up contact with SRCs occurs via email. SRCs are permitted to respond to status reports in whatever format is easiest for them and ARD program staff modifies the previously submitted CJ-11A forms.

Procedures for the DCRP-Prisons and DCRP-Jails collections. During November of the year prior to the upcoming calendar year collection, BJS’ data collection agent (RTI International) notifies both jail and prison respondents of the annual collection cycle. The notification occurs during annual verification calls. In the annual verification calls, RTI verifies that the current collection year respondent will be the respondent for the upcoming collection year and if not, will obtain the new respondent’s contact information. For the DCRP-Jails collection, RTI also verifies that the jail facilities listed on the current roster within a jail jurisdiction is complete and accurate. Annually, about 20 of the DCRP-Prison or DCRP-Jail respondents change.

Using the contact information obtained from the verification calls, RTI mails to each DCRP-Prisons and DCRP-Jails respondent an information packet announcing the start of the current year’s collection cycle and requesting annual summary data for the previous year (see Appendix C, Attachment I). In addition to the announcement letter, the packets contain year-specific envelopes that distinguish the current year’s surveys (CJ-9 or CJ-10 for jails; NPS-4A for prisons) from the annual summary forms (CJ-9A or CJ-10A for jails; NPS-4 for prisons).

The packets include a reporting instructions sheet and a DCRP fact sheet. The instruction sheet directs the respondents to the DCRP Web-site and provides them with login information to use in submitting data via the Web-reporting tool. In addition, respondents are reminded to submit an annual summary form and any outstanding death record forms from the preceding year by March 1^st of the current year. The primary mode of collection, the Web-reporting tool, allows respondents to submit individual death record forms on an ongoing basis. Secondary modes of submission of records include email, fax or U.S.P.S. A postage-paid business reply envelope is provided for respondents who choose to submit their data via U.S. mail.

All data for a calendar year are due within 60 days following the end of the year. Follow-up contacts begin in the spring following the collection year to solicit late responses. RTI contacts nonrespondents via phone, fax, email, and mail (for respondents for whom RTI does not have an email address). Concentrated data quality follow-up with respondents occurs during the summer (see Appendix C, Attachment J).

RTI monitors the DCRP data and its respondents throughout the active data collection period. This includes tracking the time to complete, number of contacts made to each agency, and other information (e.g., which form(s) have and have not been returned to date). This information is then used for tailored follow-up prompts to partial or nonrespondents. A similar approach is employed when there is missing or conflicting data within the returned forms. For such outreach activities, agencies are assigned and managed by project team members, each of whom serves as that agency’s liaison for DCRP.

3. Methods to Maximize Response

BJS has consistently achieved high rates of response for the state prison and local jail inmate death collections. In each year since collection began in 2001, all 50 state departments of corrections have participated in the DCRP-Prisons collection in each year since it began in 2001. Since BJS initiated the DCRP-Jails collection in 2000, it has obtained responses from at least 97% of the roughly 3,000 local jail jurisdictions (table 4).

Table 4. Response rates for eligible jails in the DCRP frame, 2000-2010

	2000	2001	2002	2003	2004	2005	2006	2007	2008	2009	2010
# jails responding	3,072	3,052	3,036	3,019	3,003	2,977	2,946	2,924	2,908	2,826	2,813
# of jails in the frame	3,095	3,068	3,059	3,046	3,025	3,019	2,972	2,953	2,944	2,918	2,892
% of jails responding	99.2%	99.4%	99.2%	99.1%	99.2%	98.6%	99.1%	99.0%	98.9%	96.9%	97.3%

To maximize response rates with state prison and local jail respondents, BJS has provided multiple modes of data submission, with web as the primary mode. As some local jail respondents either have no or limited access to the web, the data collection plan allows for local jail respondents to submit data by mailing their reply in a postage-paid envelope, via secure fax, encrypted e-mail, or telephone, in addition to web-based reporting. BJS developed a bulk file template in which state prison administrators can submit their prison inmate death records electronically due to the volume of deaths associated with prison submissions. During 2010, 92% of prison respondents submitted their records electronically (82% via the web, 10% via the bulk file).

Throughout the history of the ARD collection, BJS has established SRCs in up to 48 states (including the District of Columbia). During the 2013 through 2015 collection cycle, BJS will work to establish SRCs in the remaining states—Georgia, Arkansas, and Wisconsin—that currently do not have an SRC for ARD. In the interim, RTI, BJS’ contractor for ARD collection and analysis, will serve as the SRC in these states.

BJS acknowledges that the concept of maximizing response rates in the context of the ARD collection has limited meaning. Even if BJS were to meet its goal of establishing an SRC in each state, that would not guarantee that BJS would obtain data on all arrest related deaths. For example, results from a preliminary analysis of ARD data on law enforcement homicides and SHR data on justifiable homicides by law enforcement agencies—the categories of arrest-related deaths that are most comparable between the two programs—BJS found SHR cases that were not reported in ARD and ARD cases that were not reported in SHR. Across four years, from 2006 through 2009, there were about 500 SHR homicides that appeared not to be included within the roughly 1,800 ARD law enforcement homicides over this period. In addition, there were at least 700 ARD law enforcement homicides that did not appear in the roughly 1,600 SHR justifiable homicides over this period. Both data systems will improve once a deeper understanding of the incidents found only in ARD or SHR is developed. In addition, the SRCs ability to identify arrest-related law-enforcement homicides would be improved if more current SHR data were made available for them to use in their searches.

Under the current collection method, SRCs first identify deaths then contact the local agencies in which the deaths occurred to get the information to complete a CJ-11A for each death. By this method, BJS can only affirm that an identified death occurred and met the criteria for inclusion within ARD but cannot confirm that there were in fact no other arrest-related deaths. This is because BJS does not survey each of the 18,000 law enforcement agencies to determine whether, and if so, how many arrest-related deaths occur. The two SRCs with state-mandated collection efforts actively survey all law enforcement agencies in their respective states to confirm that, in fact, no deaths or some deaths occurred.

Despite the limitations of the current method, it is a method that has generated useful information that contributes to the understanding of arrest-related deaths. For example, despite coverage error and incomplete reporting, at least 40% of the ARD deaths that BJS has collected were not law enforcement homicides (i.e., outside the scope of SHR). Rather, these arose from other causes such as heart attacks, intoxication, accidents and suicides (other than suicide by cop).

Given that BJS resource constraints still preclude it from developing a system of surveying 18,000 law enforcement agencies about arrest-related deaths on a recurring basis, BJS aims to undertake a set of studies of past ARD data and of the ARD data to be collected under this clearance to 1) test the methods and procedures for data collection; 2) determine the portions of the ARD collection that result in reliable data on arrest related deaths; and 3) for those procedures that do not result in reliable data, and develop alternative approaches to the collection information about arrest-related deaths.

These planned studies include the following:

1. SHR-ARD comparison: A potential core of the ARD program is the set of arrest-related deaths occurring as the result of a homicide by a law enforcement officer. These types of deaths receive the greatest attention by policy makers and the media, and they were the types of death that gave rise to the original DICRA legislation. They are the set of deaths that ARD shares most commonly with SHR, although most importantly, the ARD collection obtains information about these deaths that is not covered by the SHR.

For this assessment, BJS will examine law enforcement homicides in ARD and compare them with justifiable homicides in SHR using data from the entire ARD program (2003-present). The assessment will consist primarily of matching records from the two programs. This presents the greatest challenge because common identifiers (such as police department ORI numbers), event identification information (such as date of death), and decedent information (such as age) may differ slightly; combined, the measurement error in identification variables will affect matches, but BJS will explore both direct and probabilistic match methods. The purpose of this test is to confirm, for matched cases, that they refer to the same event.

BJS will work collaboratively with the FBI in conducting the comparison and share the result of its report on the analysis with the FBI, given that BJS’ findings may have implications for the SHR. But more directly for the ARD program, BJS will share the findings of the report with the SRCs for the purpose of assisting them in developing methods to address reported coverage issues.

2. Coverage of agencies within states: Within the current collection method, SRCs typically do not contact local law enforcement agencies until they have identified an arrest-related death for which they need information. In addition, SRCs may have routine contact with agencies for other purposes but use these contacts to obtain information about arrest-related deaths. BJS will undertake a study of the SRCs to determine the scope of routine coverage of local law enforcement agencies within their states and to determine if the SRCs could use their routine operational procedures and contacts with local law enforcement agencies to obtain information for ARD, including confirmation that no arrest related deaths occurred. This study will provide BJS with information that could be used to establish a low-cost method for surveying local law enforcement agencies within states. Conducting this study would require that BJS obtain clearance to survey the SRCs for this purpose. As BJS is currently seeking a Generic Clearance for testing methods, BJS will seek clearance under its Generic Clearance after that has been approved.

3. Census of Law Enforcement Agencies: Periodically, BJS conducts a census of law enforcement agencies for the purpose of developing a sampling frame of agencies that it uses for other surveys. BJS will, in its next census, add a minimum number of questions about arrest related deaths to help to determine its coverage under ARD. For example, if BJS next Census of Law Enforcement Agencies has a reference date in 2013, BJS would ask for agencies to report a count of ARD-related deaths for the past two or three years. BJS could use these data to determine if the ARD program had captured these deaths.

4. Open-source methods and capture-recapture methods: BJS has explored the use of open-source methods to obtain ARD data but has not systematically assessed the results of this effort or the quality of these data. At the same time, BJS has been capturing meta-data on ARD death records that identify the source or sources that have identified and reported on a death. For example, if both an SRC and a BJS’ Google searches identify the same record and report the same information about the death, that would give some indication that open source methods could be useful. While we doubt that sole reliance on open source methods would produce wholly reliable ARD data, an examination of the information that can be obtained from open source records could help in assessing the scope of ARD’s current coverage of arrest-related deaths. For example, given that BJS has started and will continue to collect paradata about the source of information on an ARD record, BJS can use this information to identify both open source and law enforcement agencies records to assess the quality of the information between the two sources. Or, tests can be constructed based on varying search terms to assess the quantity and quality of information obtained from open source records. More generally, the extent to which the same records are returned from different searches or different sources can be used with capture-recapture models to estimate the size of the potential universe of ARD records. These estimates can provide a basis for determining the completeness of the collected records.

Our plan is to move forward with studies (A) and (D) first, as these do not require collecting additional information other than what is currently being collected. We will use the information from (A) to help to develop the scope of effort in (B), and during the planning phase for the next census of law enforcement agencies, we will develop and test the ARD-related questions prior to fielding the census.

In addition, to improve the SRCs understanding of ARD, BJS will develop online training that will provide concise guidelines on how to complete the CJ-11A death report form, the qualifying criteria for arrest-related deaths, and the definition, purpose, and scope of the ARD program. All training programs will be short (10-15 minutes) and can be administered online with no additional software or cost. Also during the upcoming collection cycle, BJS will work with SRCs to identify SHR justifiable homicide cases that do not appear among the ARD cases that they submit to BJS. In some situations, such as where the SRC also reports to the UCR, this can be done more quickly and in real time; in other settings, BJS will provide the SRCs with information about SHR justifiable homicides not included in ARD reports after SHR data are released. As reported above in Section A. 4. “Efforts to Identify Duplication,” there is some overlap in the universe of the cases eligible to be included in both the SHR and ARD. However, this effort would not assist in maximizing cases that are reportable to ARD, but outside of the scope of the SHR.

4. Test of Procedures or Methods

In 2010, BJS held a meeting of experts to review of the ARD collection and solicit feedback on the data collection instrument (CJ-11A) from. Participants included law enforcement personnel, medical examiners and coroners, SRCs, and researchers in the law enforcement and public health fields. The ARD incident reporting form (CJ-11A) was revised in 2012 in light of comments received from the meeting. The revised 2013 CJ-11A instrument was pilot tested in June 2012 with nine of the program’s current SRCs (see Appendix C, Attachment F).

In June 2012, BJS asked nine of the program’s current SRCs to assess the revised instrument and related instructions in terms of substance, breadth, and clarity. Specifically, BJS asked the SRCs to evaluate whether the questionnaire items enhanced the quality, utility, and precision of the information to be collected and to note any items that were unclear, concerns with respect to the availability of the information requested, and response burden. Lastly, SRCs were asked to complete the revised version of the CJ-11A form for a death record that was previously submitted to the ARD program.

BJS removed items that required respondents to speculate about outcomes or that relied on respondents’ opinions or perceptions. For example, BJS removed items related to alleged criminal involvement in the events leading up to the death, removed an item that had asked respondents to indicate whether the decedent “appeared intoxicated” or “exhibited mental health problems.” BJS removed other items that were repetitive or could be determined from responses to other items. For example, Item 13 in the old instrument (“Did the deceased die from a medical condition, injuries sustained during the arrest process, or alcohol/drug intoxication?”) was omitted because this information could be determined from information provided in response to Item 9 (“What was the cause of death?”). In addition, four questions about deaths occurring at a booking center/police lockup were omitted because the items either asked for speculative information or were redundant.

Prior to revision, the CJ-11A did not have reporting instructions or a national list of SRCs; both of which were only included on the CJ-11 Summary Report. The revised CJ-11A instrument includes a description of what deaths should be reported, contact information for assistance in completing the form, information about how completed forms should be submitted, and contact information for each SRC. The revised CJ-11A also includes a “data supplied by” section in order to assist SRCs track where information for each death is coming from. The order of items on the revised CJ-11A was changed to improve the flow of the instrument. Items were reorganized into groups that reflect characteristics of the incident, characteristics of the decedent and actions the decedent took during the interaction, actions of law enforcement personnel during the interaction, and characteristics of the death. Lastly, the question-structure of items changed from a “mark all that apply” format to a “Yes/No” format. This change was done to improve data quality by forcing data providers to reply to each response category, thereby decreasing the risk of skipping over responses that should have been marked.

SRCs were asked to complete the form with information that was previously submitted in order to compare the reliability of the items across both versions of the form. As a result of this test, a proposed item that asked about whether specialized units responded to the incident was omitted from the final form because of concerns about burden and reliability of responses. In addition, an item regarding whether the decedent had a history of mental illness was removed following the pilot test out of concern that the information obtained would not be reliable.

5. Consultation Information

The Corrections Statistics Unit of BJS takes responsibility for the overall design and management of the activities described in this submission, including data collection procedures, development of the questionnaires, and analysis of the data. BJS contacts include the following:

Margaret Noonan, Program Manager

Deaths in Custody Reporting Program

Bureau of Justice Statistics

810 Seventh Street, NW

Washington, DC 20531

(202) 353-2060

E. Ann Carson, Ph.D., Statistician

Deaths in Custody Reporting Program

Bureau of Justice Statistics

810 Seventh Street, NW

Washington, DC 20531

(202) 616-3496

William J. Sabol, Ph.D.

Principal Deputy Director

Statistical Collections & Analysis
Bureau of Justice Statistics
810 Seventh Street, NW
Washington, DC 20531
202-514-1062

The Project Director at RTI International for the DCRP data collection is:

Christopher Ellis, Project Director

Deaths in Custody Reporting Program

RTI International

3040 Cornwallis Road

Research Triangle Park, NC 27709

(919) 541-6480

The Law Enforcement Statistics Unit of BJS takes responsibility for the overall design and management of the activities described for the deaths occurring during the process of arrest within this submission, including data collection procedures, development of the questionnaires, and analysis of the data. BJS contacts for the Arrest-Related Deaths Program include the following:

Andrea M. Burch, Program Manager
Arrest-Related Deaths Program

Bureau of Justice Statistics

U. S. Department of Justice

810 7th St. NW

Washington, DC 20531

202-307-1138

Howard N. Snyder, Ph.D.
Deputy Director

Law Enforcement, Prosecution, Courts, Recidivism and Special Projects

Bureau of Justice Statistics
U.S. Department of Justice
810 7th Street NW, Room 2326
Washington, DC 20531
Office Phone: 202-616-8305
Cell Phone: 202-305-5592
Fax: 202-616-1351

The Project Director at RTI International for the Arrest-Related Deaths program is:

Kevin J. Strom, Project Director

Arrest-Related Deaths Program

RTI International

3040 Cornwallis Road

Research Triangle Park, NC 27709

(919) 485-5729

C. Attachments

1. BJS authorizing legislation (42 USC 3732)

2. 2000 Death in Custody Reporting Act (P.L. 106-297)

3. 2011 Death in Custody Reporting Act (H.R. 2189)

4. BJS confidentiality regulations (Title 42, USC, Sections 3789g and 3735)

5. 2012 CJ-9A (Annual Summary on Inmates Under Jail Jurisdiction)

2012 CJ-10A (Annual Summary on Inmates in Private and Multi-Jurisdictional Jails); 2012 NPS-4 (Annual Summary of Inmate Deaths in State Prisons);

2013 CJ-9 (Death Report on Inmates Under Jail Jurisdiction);

2013 CJ-10 (Death Report on Inmates in Private and Multi-Jurisdictional Jails);

2013 NPS-4A (State Prison Inmate Death Report)

6. 2013 CJ-11 (Arrest-Related Deaths Summary of Incidents);

2013 CJ-11A (Arrest-Related Deaths Incident Report);

2013 CJ-11A (Arrest-Related Deaths Incident Report (CJ-11A) Question-by-Question Guide;

7. Example of arrest-related death program launch email to state reporting coordinator;

8. Example of arrest-related death ‘status report’ to a state reporting coordinator;

9. Example of jail and prison DCRP program launch mailing to jail and prison respondents;

10. Example of letters, emails and telephone scripts for data quality and non-response follow-up for DCRP jail respondents

9