Supp_Statement_NIS3[1]

SUPPORTING STATEMENT

A. Justification

1. Necessity of Information

The Prison Rape Elimination Act of 2003 (known hereafter as the PREA, see Public Law 108-79.pdf), requires the Bureau of Justice Statistics (BJS) to “carry out, for each calendar year, a comprehensive statistical review and analysis of the incidence and effects of prison rape.” The PREA further instructs BJS to collect survey data, “…the Bureau shall…use surveys and other statistical studies of current and former inmates…” The law was passed in part to overcome a shortage of available research on the incidence and prevalence of sexual violence within correctional facilities.

To implement its responsibilities under the PREA, BJS developed the National Prison Rape Statistics Program, which includes five separate data collection efforts: the Survey on Sexual Violence (SSV, OMB 1121-0292), the National Survey of Youth in Custody (NSYC, OMB 1121-0319), the Former Prisoner Survey (FPS, OMB 1121-0316), and Clinical Indicators of Sexual Violence in Custody (CISVC, OMB 1121-0324), and the National Inmate Survey (NIS, OMB 1121-0311), the collection currently under review in this package.

Each of these collections is independent and provides various measures of the prevalence and characteristics of sexual assault in correctional facilities. The SSV series, ongoing since 2004, reports what incidents of sexual violence are reported to and substantiated by correctional authorities. The NSYC, first fielded in 2008 to 2009, collects allegations of sexual victimization from youth in residential placement, and the FPS, fielded in 2008, collects allegations of sexual assault experienced by inmates during their last incarceration from persons on active parole supervision. The CISVC, currently underway, will collect information from medical staff on persons seeking treatment for conditions consistent with sexual victimization. Finally, the NIS, fielded in 2007 (Year 1), and 2008 to 2009 (Year 2), collected allegations of sexual victimization reported by current inmates held in a large sample of prisons and jail facilities across the nation.

The Year 1 NIS (2007) yielded information about sexual victimization at the facility and national level that had never before been gathered (see Section 2, Needs and Uses). With each iteration, BJS has tried to improve the survey to address correlates of sexual victimization while balancing the effect on burden. For Year 2 (2008-09), with OMB approval, we simplified the offense data coding, and added questions on the first and most recent victimization, number of times, whether an incident occurred in the course of a pat down/strip search, as well as a follow-up question about the specific position of staff involved in allegations of sexual misconduct. BJS staff is currently working on the report from the NIS Year 2 (2008-09) rankings for prisons and jails, which will incorporate some of these new findings.

This submission is to renew approval for the NIS with instrument, sampling, and burden changes. BJS and the Research Triangle Institute (RTI), with whom BJS has a cooperative agreement to collect data, have re-designed and tested the survey (with OMB approval) to refine the collection as well as address an additional congressional mandate for data collection on the prevalence of mental illness among inmates required under the Mentally Ill Offender Treatment and Crime Reduction Reauthorization and Improvement Act of 2008 (hereafter known as the MI Act, see Public Law 110-416.pdf).

a) General Survey Information – The first part of the survey will be a brief Computer-Assisted Personal-Interview (CAPI) interaction wherein an interviewer will read a series of questions to an inmate and enter the answers directly into a laptop computer. All respondents will then receive a brief tutorial in how to use the ACASI instrument which involves inmates responding to a computer questionnaire using a touch-screen, following audio instructions delivered via headphones. After answering some basic demographic questions, the instrument will randomly assign each respondent to either the sexual assault and mental health screener questionnaire or the mental health/medical screeners and drug/alcohol use questionnaire.

Due to the sensitive nature of the topic, the survey has several design features to maximize confidentiality. First, the survey is administered one respondent at a time in a private setting with an interviewer to conduct the consent procedures, computer tutorial, and assist with any technical issues. Once the Audio Computer Assisted Self-Interview (ACASI) begins, the interviewer is on the opposite side of the computer screen and is not privy to the questions being asked. Second, there are two possible questionnaires the respondent could receive. Ninety percent will receive a questionnaire on focusing on sexual victimization and basic mental health and disability screeners (see A_NISquestionnaire.doc). Ten percent of respondents will be randomly assigned to an alternative questionnaire containing mental health, medical, and disability screeners, and questions on drug and alcohol use and treatment (see B_NIS_AltQuestionnaire.doc). This design offers a layer of protection to the respondents, as correctional staff, other inmates, and field staff will not know which modules the respondent received.

Third, all respondents will spend about the same amount of time completing the survey, regardless of which questionnaire they are assigned. Those respondents experiencing no victimization, about 95% based on the Year 1 and 2 collections, will complete the sexual violence and mental health/disability screeners more quickly. The remainder of the survey will be padded out using the medical health screener and drug and alcohol modules. In the pretest, conducted in January 2010 in two prison facilities and one jail facility with 377 inmates, respondents took about 35 minutes to complete the survey, whether they were victims or not, and whether they received the sexual assault questionnaire, the mental health, medical health, and disability screeners, drug and alcohol questions, or some combination thereof.

There are a portion of inmates who may be unable to come to the interviewing room, for medical or disciplinary reasons. It is not possible in most facilities to bring laptops into the housing units for safety and security reasons. In order to include these inmates in the survey, a representative will visit these inmates in their living area equipped with a consent form (see C_NIS_PAPIConsent.doc) and an abbreviated paper-and-pencil-instrument (see D_NIS_PAPIF12.pdf; D_NIS_PAPIFArrive.pdf; D_NIS _PAPIM12.pdf; D_NIS_PAPIMArrive.pdf ). Such interviews take about 15 minutes each and made up 2.3% of prison inmate interviews and 0.6% of jail inmate interviews in the 2007 data collection.

All allegations of sexual violence are important to include in the estimate of sexual assault to ensure accuracy. Thus, data from the PAPI are included in the prevalence estimates generated for facilities. BJS will analyze and note any differences between allegations reported using PAPI and ACASI methodologies, as well as any selection bias that may occur in administering either the PAPI or ACASI to a respondent.

b) Changes to the NIS - The NIS design necessitates some alternative questions for randomization and time-padding to provide additional protections for maintaining respondent confidentiality. The drug and alcohol use and treatment questions used in Years 1 and 2 and the proposed mental health, medical, and disability screeners for Year 3 are included based on the concern that inmates with such problems may be at additional risk to sexual assault while incarcerated. These questions were reviewed by experts in the field and tested in January 2010. For ease of review, all changes to the survey have been highlighted (see A_NISquestionnaire.doc and B_NIS-AltQuestionnaire.doc).

BJS is currently combining data from the first two years of data collection on drug and alcohol use and treatment for a special report. Data from the NIS 3 will also be used for a special report on the prevalence of serious mental illness among prison and jail inmates, fulfilling, in part, requirements mandated under the MI Act previously mentioned.

A short series of questions regarding veteran status have been added to instrument in an effort to gather information on rates of incarceration among persons returning from recent conflicts abroad; there is also interest in whether such persons may be at risk to victimization, particularly if there are associated mental health or drug/alcohol issues. Also included in the latest instrument are a series of facility climate questions to measure general conditions and safety and attitudes towards staff. Many of these were taken directly from the Bureau of Prisons’ Social Climate Survey, which have been validated over time. Similar questions were fielded in the first year of the NSYC work, revealing that facilities with higher rates of sexual victimization were more likely to have inmates reporting concerns with general conditions and safety. Such measures are of great use to researchers for correlation purposes and to administrators in understanding the conditions underlying sexual victimization.

There is no utility in information submitted by a person not able or willing to consent to the interview. BJS and RTI have appropriate protocols in place to ensure we obtain data from consenting and competent persons. The general consent process has been revised to include follow-up questions for all persons not subject to mandatory reporting procedures to confirm comprehension of consent (see E_NIS Consent Form.doc and F_NIS-C Consent Card.doc). After successfully implementing data collection among adults in correctional facilities for the first two iterations of the NIS, we now seek to incorporate special populations who may be at greater risk for sexual victimization in custody – youth held in prisons and jails and persons with mental health problems. BJS proposes to include all youth ages 16 years and older held in sampled prisons and jails for whom parental consent is not required.

BJS is cognizant of the sensitivities associated with interviewing youth and persons with mental health issues. Facility liaisons will be asked about these populations during logistics planning to address special needs or concerns. The consent process has been adapted to include follow-up questions to be sure the youth comprehends issues of mandatory reporting (if applicable in the state, see G_NIS_J Consent Card.doc). Protocols for reporting will be established as dictated by each jurisdiction. BJS has experience with interviewing youth from the NSYC. About .002% of youth failed to demonstrate understanding of the consent procedure and were not interviewed. Another 1% of youth made a statement to field staff that triggered the reporting of mandatory abuse and neglect. We anticipate sampling approximately 1,000 youth; assuming an 80% response rate, we can expect that up to 8 youth may make a statement subject to mandatory reporting.

Interviewers, while not able to screen for mental competency, will receive special training to identify comprehension issues (see H_MCverification.doc). BJS has experience interviewing persons with mental health challenges in the first two collections of the NIS. The table below illustrates how often a case was coded as ineligible due to mental incompetence by facility type. The rate was slightly higher in jail facilities, where respondents are more likely to be newly admitted and awaiting evaluation and treatment for mental health or substance abuse issues.

Facility type	NIS Year 1 (2007)			NIS Year 2 (2008-09)
	Number of cases	Percent of sampled inmates		Number of cases	Percent of sampled inmates
Jails	552	.74	%	334	.42	%
Prisons	88	.23		138	.30
Prisons with MH function	36	.63		50	.57

The changes in the instrument and sampling procedure have resulted in a small increase in burden to facilities and respondents (see Section 12). We believe the increase in burden is not only justified, but necessary to better understand incidents of sexual victimization and to better identify persons at risk of sexual victimization in the future.

BJS requests approval for all data collection activities related to the NIS. It is anticipated that these activities will span the 12-month period from approximately September 2010 to September 2011. BJS requests authorization for 3 years of data collection. As mandated by the PREA, BJS will produce system-level and facility-level estimates of sexual assault within correctional facilities and report these findings to Congress each year.

The IRB response and approval for the Year 3 instrument and protocols is attached (I_IRBResponse.doc; J_IRB Approval Notice.pdf). RTI will be meeting with the IRB for Year 3 full implementation approval in April 2010. The full IRB package, response, and final approval will be submitted to OMB as each are received from RTI. Any additional IRB approvals required from sampled jurisdictions will be obtained prior to conducting any data collection under this clearance.

Data collection for the NIS project is authorized under the Prison Rape Elimination Act of 2003 (Public Law 108-79), a copy of which is attached. The Omnibus Crime Control and Safe Street Act of 1968 (see attachment) as amended (42 U.S.C. 3732), authorizes BJS to collect and disseminate statistical data on all aspects of criminal justice, including criminal victimization, occurring in the United States.

2. Needs and Uses

This clearance request is to obtain approval to conduct national data collection required under the PREA. Data collection is necessary to measure the incidence and prevalence of sexual assault within correctional institutions, at a facility-level, as required under the PREA. The purposes of the PREA include: “to develop and implement national standards for the detection, prevention, reduction, and punishment of prison rape,” and “increase the available data and information on the incidence of prison rape, consequently improving the management and administration of correctional facilities.”

The BJS special reports Sexual Victimization in State and Federal Prisons Reported by Inmates, 2007 (December 2007) and Sexual Victimization in Local Jails Reported by Inmates, 2007 (June 2008), based on the data collected in the NIS, received a considerable amount of national media attention. BJS found an estimated 60,500 prison inmates (4.5%) reported some type of sexual victimization in the 12 months prior to the interview, or since admission, if shorter. An estimated 24,700 jail inmates (3.2%) reported some type of sexual victimization within the 6 months prior to the interview, or since admission, if shorter. BJS also provided individual facility rankings in each report.

Several facilities listed in the reports as having “high rates” and “low rates” were selected by the Prison Rape Review Panel to send representatives to Washington, DC to participate in separate hearings for prison and jail administrators to discuss these results. Transcripts and reports from these hearings are available at http://www.ojp.usdoj.gov/reviewpanel/reviewpanel.htm. Data were also cited extensively by the National Prison Rape Elimination Commission in its recommended standards, currently under review at the Department of Justice. These standards can be found at http://www.cybercemetery.unt.edu/archive/nprec/20090820154816/http://nprec.us/publication/.

In its second report, Sexual Victimization in Local Jails Reported by Inmates, 2007, BJS provided additional detail on the characteristics of victims and incidents of sexual victimization. Jail inmates ages 18 to 24 were assaulted at a higher rate (4.6%) than jail inmates ages 25 and older. Inmates with a sexual orientation of homosexual had the highest rate of victimization (18.5%) followed by bi-sexual inmates (9.8%) and heterosexual inmates (2.7%). Inmates with a prior sexual assault were 6 times more likely to report a sexual victimization in jail (11.8%), compared to those with no sexual assault history.

Inmate-on-inmate reports were more than twice as likely to involve physical force (34%) as compared to staff sexual misconduct (15%). Inmate-on-inmate incidents most often occurred in the victim’s cell or room (64%), while staff sexual misconduct occurred most often in a closet, office or other locked room (47%). Nearly 62% of all reported incidents in local jails of staff sexual misconduct involved female staff with inmates; 8% involved male staff with female inmates. Female staff were involved in 48% of incidents reported by male inmates who said there were unwilling and in 79% of incidents with male inmates who said they were willing.

BJS continues to strive to improve the quality and usefulness of the data gathered in this most important area. This proposal furthers this work in adapting the sampling measures to make estimates of victimization for youth held in adult facilities and persons with serious mental health illness, both widely believed to be high-risk for sexual assault. While sexual victimization continues to be the focus of the collection, we have added measures we believe to be correlated with victimization, including facility safety and staff measures, and mental health/ medical/disability screeners.

Users of these data include the following:

U.S. Congress – Each year Congress will receive a report on data collected under the PREA. The report will include information about the prevalence of sexual assault at each facility in the sample.

U.S. Department of Justice – The Review Panel on Prison Rape will solicit testimony from correctional administrators in facilities with the highest and lowest rates of sexual violence as identified in the annual reports.

National Institute of Corrections (NIC) – is responsible for establishing a “national clearinghouse for the provision of information and assistance to Federal, State, and local authorities responsible for the prevention, investigation, and punishment of instances of prison rape.” NIC will also develop periodic training and educational programs for “…authorities responsible for the prevention, investigation, and punishment of instances of prison rape.”

National Institute of Justice and the Bureau of Justice Assistance – are responsible for studying characteristics of victims and perpetrators and identifying trends in sexual violence within correctional settings. Findings from the NIS activities disclosed in the Congressional reports may be used to inform research proposals for grant funding opportunities provided in the PREA.

Civil Rights Division, U.S. Department of Justice – may use data from the Congressional reports to understand the magnitude and scope of sexual violence within correctional facilities as they relate to the violation of inmate civil rights.

Federal, State, and local corrections and juvenile officials and administrators – use data from the Congressional reports to assess and compare facility rates and characteristics of inmate-on-inmate and staff-on-inmate sexual victimization.

3. Use of Technology

Using the latest technology in survey methodology, RTI interviewers will conduct interviews using laptop computers. Respondents will enter the answers themselves using ACASI technology which allows them to hear the question being read over headphones as it appears on the screen. In addition, the ACASI methodology allows even respondents with low literacy levels to participate because the audio component provides clear instruction for how to enter answers and is highlighted as the question and corresponding answers are read. The survey will be offered in both English and Spanish. About 5% of interviews of the Year 1 NIS (2007) were conducted using the Spanish instrument.

CAPI and ACASI technology improves the flow of the interview through built-in skip patterns and filled-in reference periods that tailor specific questions to individual inmates. This allows for the instrument to be tailored by gender, type of facility (prison or jail), and length of stay (“in the last 12 months” or “since you arrived at this facility”). As previously stated, the technology also allows for modules to be administered in a manner that results in an average of 35 minutes per interview, regardless of whether a report of victimization is alleged. This technology also produces more accurate data through built in edit checks (for example, answers that are out of range can trigger a follow-up question that asks the respondent about an inconsistent or out-of-range response). Finally, due to the sensitive nature of the questions, the NIS is programmed to end the sexual victimization questions upon receiving three “refuse” or “don’t knows” responses to minimize risk of trauma (these inmates skip ahead to alternative modules to pad out the time).

Research with ACASI suggests respondents provide more honest reporting of sensitive behaviors when the questions are administered via ACASI as opposed to traditional face-to face interviews and other methodologies.¹ Due to residual concerns among stakeholders about the validity of the data reported, the questionnaire incorporates latent class modeling (LCM) to detect false positives and false negatives of sexual assault reports among respondents. This involves using several different questions to measure the same phenomenon. Data collection from the Year 1 NIS indicates that LCM revealed low levels of both error types.

Finally, use of the computer allows for random assignment of inmates to one of two questionnaires, as described earlier, and pads out response time with additional questions when necessary to assure a consistent amount time spent taking the survey across respondents, which averaged 35 minutes in the pretest.

For inmates who cannot leave their cell or living area, a representative will be taken to the inmate with a PAPI consent form and questionnaire (See D_NIS_PAPIConsent.doc; C_NIS_PAPIF12.pdf; C_NIS_PAPIFArrive.pdf; C_NIS _PAPIM12.pdf; C_NIS_PAPIMArrive.pdf).

4. Efforts to Identify Duplication

This research does not duplicate any other data collection being conducted done by BJS or any other Federal agencies. BJS will be the only government agency that collects national data on the incidence and prevalence of sexual violence within correctional settings.

5. Impact on Small Businesses

This research does not involve small businesses or other small entities. The respondents are inmates held in adult correctional institutions.

6. Consequences of Less Frequent Collection

The Bureau of Justice Statistics is required by law to collect these data annually.

7. Special Circumstances Influencing Collection

This data will be collected in a manner consistent with the guidelines in 5 CFR 1320.6.

8. Federal Register Publication and Outside Consultation

The research under this clearance is consistent with the guidelines in 5 CFR 1320.6. The 60 and 30-day notices for public commentary will be published in the Federal Register.

In developing the original survey for the NIS, BJS staff consulted with Federal, State, and local corrections administrators as well as representatives from their professional organizations, prisoner rights advocates, former inmates, specialists in prison rape research, practitioners, and survey methodologists. These individuals have and will continue to provide valuable input regarding the development of the questionnaires, definitions and counting rules, anticipated data analysis, and data presentation.

For the addition of mental health and medical screeners, BJS solicited input from a series of highly regarded experts in the field:

Ronald C. Kessler

Professor

Department of Health Care Policy

Harvard Medical School

180 Longwood Avenue

Boston MA 02115

Robert L Trestman

Executive Director, Correctional Managed Health Care

Interim Chief, Division of Public Health and Population Sciences

Interim Director, UCHC Signature Program of Public Health

Interim Co-Director, UConn Center for Public Health and Health Policy

Professor of Medicine and Psychiatry

University of Connecticut Health Center

MC-5386

263 Farmington Avenue

Farmington CT 06030

Lisa Colpe

Statistician

Division of Population Surveys

Office of Applied Studies

Substance Abuse and Mental Health
Services Administration

1 Choke Cherry Road

Rockville, MD 20857

Denise Juliano-Bult

Chief, Systems Research Program

Chief, Disablement and Functioning Research Program

Division of Services and Intervention Research

National Institute of Mental Health

6001 Executive Blvd., RM 7137 - MSC 9631

Bethesda, MD  20892-9631

Michael Schoenbaum

Senior Advisor for Mental Health Services, Epidemiology, and Economics

Division of Services and Intervention Research

National Institute of Mental Health

6001 Executive Blvd, Room 7142 MSC 9629

Bethesda, MD 20892-9669

Jeff Buck

Chief, Survey, Analysis and Financing Branch

Center for Mental Health Services

Substance Abuse and Mental Health
Services Administration

1 Choke Cherry, Room 2-1089

Rockville, MD 20857

Ingrid Goldstrom

Social Science Analyst

Survey, Analysis and Financing Branch

Center for Mental Health Services

Substance Abuse and Mental Health
Services Administration

1 Choke Cherry, Room 2-1081

Rockville, MD 20857

Henry Steadman

President

Policy Research Associates

345 Delaware Avenue

Delmar, NY 12054

9. Payment or Gift to Respondents

A non-substantive change was requested for the second year of NIS data collection to incorporate the use of non-monetary incentives (a packet of cookies) for respondents in approving jurisdictions. The table below illustrates that response rates were significantly higher in facilities allowing the cookies (p <.0001). BJS proposes that we continue to use the incentives where permitted.

NIS Year 2 (2008-09) Response Rates by Facility Type and Use of Incentive
Facility Type	Incentive Received	Number of Facilities	Sample Size	Response Rate
Prison	No	132	37,275	70.2 %
	Yes	35	8,914	76.3
Jail	No	105	32,516	61.9
	Yes	181	47,790	71.9

10. Assurance of Confidentiality

BJS and RTI hold in confidence any information that could identify an individual according to Title 42, United States Code, Sections 3735 and 3789g. All respondents as well as correctional facility administrators who participate will be given written assurance that the identity of all participants, victims, and perpetrators will be protected as required under Title 42 (see Attachments D and F). Rates of sexual victimization at the facility level will be published, as required under the PREA.

All interviews will be conducted in a private room, and names and other personal identifiers will not be linked to the questionnaire data, such that if someone were to somehow obtain the survey data, they could not associate any data with a particular individual. As required under Title 42 USC, section 3879g, BJS and its data collection agents will take all necessary steps to mask the identity of survey respondents, including suppression of demographic characteristics and other potentially identifying information, especially in situations in which cell sizes are small.

Further, BJS has masked and padded the survey to ensure that no correctional official, inmate, or the field representative will know which questionnaire is administered, or make assumptions based on the time taking the survey (See Section 1).

BJS and RTI have received the requisite approvals from the appropriate Institutional Review Boards (IRB) to ensure that the data collection procedures are in compliance with human subjects protection protocols and confidentiality regulations (see I_IRBpackage.doc; J_IRBresponse.doc; K_IRB Approval Notice.pdf).

11. Justification for Sensitive Questions

The PREA requires BJS to collect highly sensitive information (see Section 4, Attachment A). As previously mentioned in Section 3, research with ACASI suggests respondents provide more honest reporting of sensitive behaviors when the questions are administered via ACASI as opposed to traditional face-to face interviews and other methodologies.

According to extant research, it is beneficial to begin broadly and narrow down when asking questions about sensitive topics. BJS has employed this approach by asking first about the respondent’s sexual activities. This serves two main purposes. First, a global binomial (yes/no) question leaves the instrument with limited ability to define what is meant by sex and sexual assault and leaves interpretation largely in the hands of the respondent. Further, if the response is negative, the interview is essentially over. Second, the literature in this area notes that sexual assault, particularly in correctional facilities, occurs on a continuum of coercion from no coercion at all to serious physical violence. The lesser kinds of coercion may be easily overlooked as consensual unless the general (sexual activity) to specific (coerced, pressured, or forced sexual activity) approach is utilized. Findings from the Year 1 and 2 NIS collections confirm a substantial portion of sexual victimization is staff sexual misconduct, and often is reported by the inmate as “willing.” These incidents would likely go unreported if not for beginning with questions about general sexual activity.

To protect against undue trauma or distress, BJS has also implemented several safeguards to protect inmates. All respondents are told in the consent process that they will receive a questionnaire about either their sexual experiences in the facility or a series of questions on mental health, medical, and substance use issues. Respondents are also reminded that participation is voluntary and they may quit the survey at any time. The sexual assault module of the questionnaire is also set up to redirect the respondent to the beginning of the next module if the respondent hits the “refuse” or “don’t know” button provided on the screen three times within the sexual victimization section in an effort to minimize distress. Finally, BJS and RTI will work with each facility to determine the point of contact either within or outside the facility should a respondent experience distress or trauma as a result of participating in the survey.

12. Estimate of Hour Burden

Based on our experience with the Years 1 and 2 NIS, we estimate that each facility staff member will take two hours to provide a roster of inmates. This includes working with a member of RTI’s Logistics Team to determine the most efficient manner for submitting the roster to RTI. It is also expected that each ACASI inmate will spend 35 minutes in the interview and each PAPI inmate will spend approximately 15 minutes. The total respondent burden, including both staff and inmates, is summarized in the following chart:

Annual Respondent Burden for the Personal Interviews
Description of service	Burden hours per response	Number of responses	Total expected burden hours
Provide roster	1.0 hour	558	558 hours
Verify roster	1.0 hour	558	558 hours
Inmates, Interview time -- ACASI	.6 hour	103,400	62,040 hours
Inmates, Interview time -- PAPI	.25 hour	2,100	525 hours
TOTAL BURDEN	2.60 hours*	105,500	63,681 hours
*Inmates will receive either the ACASI or the PAPI, thus total burden is between 2.25 and 2.6 hours per response. For the total expected burden hours the 2.6 is used.

13. Estimate of Respondent Cost Burden

The total respondent cost includes the facility staff time to prepare and verify the inmate selection roster and to escort 105,500 potential respondents to an interview location. It is estimated that the facility staff will be available for approximately 27,491 hours to complete the interview process. This includes 1,116 hours for providing and verifying the roster and 26,375 hours for escorting inmates to and from the interview site (.25 hours per respondent). At an estimate of $20 per hour for 27,491 hours, the estimated respondent cost burden for the entire national survey is $549,800.

14. Estimated Cost to Federal Government

The total estimated cost to the government for full survey implementation for the Year 3 NIS is $15,044,200.

RTI International NIS Cost Estimate

Survey and instrument planning, development, management, and processing	$5,746,100
Equipment and supplies	$96,200
Training, travel, data collection	$9,081,300
Total costs	$14,923,600

Bureau of Justice Statistics costs – $120,600

25% of GS-14, Statistician ($26,300)

25% of GS-14, Statistician ($26,300)

25% of SL-1530, Senior Statistician ($39,000)

10% of GS-13, Statistician ($8,900)

Benefits (@20% - $20,100)

15. Reasons for Change in Burden

The change in burden is a result of increase the number of facilities included in the sample, and the increase of the questionnaire time from an average of 25 minutes to 35 minutes per interview. The additional of facilities holding youth and persons with serious mental illness are of great importance to stakeholders of the PREA. BJS has successfully demonstrated the NIS can be conducted and is now in a position to address special populations. The additional time burden to complete the survey will yield information on measures we believe to be correlated with victimization, including facility safety and staff perception measures, as well as mental health/ medical/disability screeners. This information will allow facility administrators to determine additional characteristics by which to classify inmate risk upon intake and prevent sexual victimization in their facilities.

16. Plans for Publication

BJS issued two data reports as soon as possible following data collection, cleaning, and weighting. Sexual Victimization in State and Federal Prisons Reported by Inmates, 2007 (December 2007) was based on data collection April to August 2007. Sexual Victimization in Local Jails Reported by Inmates, 2007 (June 2008) was based on data collection between April and December 2007). Both special reports offered national estimates of sexual victimization, as well as specific rankings of facilities as required under the PREA.

Rankings of prison and jail facilities and characteristics of victims and incidents from the Year 2 NIS (conducted 2008-09) will be published by July 2010.

The 2007 NIS data are currently undergoing disclosure review at the National Archive of Criminal Justice Data at the University of Michigan. Identifiable information is being stripped, perturbed, and masked to preserve the confidentiality of respondents. When the review is complete, the data will be archived at the enclave at the University. Researchers seeking to access the data will submit applications with specific research questions and IRB approvals. Upon proposal approval, the data must be accessed at the University and is subject to censoring should analysis reveal small cell sizes. All output will be reviewed before leaving the premises. We anticipate a similar archive plan for the 2008-09 NIS data to begin following the publication of the associated ranking report.

17. Expiration Date Approval

The OMB Control Number and the expiration date will be published on all forms given to respondents.

18. Exceptions to the Certification Statement

There are no exceptions to the Certification Statement. The Collection is consistent with the guidelines in 5 CFR 1320.9.

1 Langhaug, Lisa F.; Sherr, Lorraine; Cowan, Frances M. (March 2010). How to improve the validity of sexual behaviour reporting: systematic review of questionnairedelivery modes in developing countries. Tropical Medicine & International Health, Vol. 15 (3): 362-381.

Mensch, Barbara S.; Hewett, Paul C.; Gregory, Richard; Helleringer, Stephane. (Dec 2008). Sexual Behavior and STI/HIV Status Among Adolescents in Rural Malawi: An Evaluation of the Effect of Interview Mode on Reporting. Studies in Family Planning, Vol. 39 (4): 321-334.

Mimiaga, Matthew J.; Reisner, Sari L.; Vanderwarker, Rodney; Gaucher, Michael J.; O'Connor, Catherine A.; Medeiros, M. Susana; Safren, Steven A. (Sept 2008). Polysubstance Use and HIV/STD Risk Behavior among Massachusetts Men Who Have Sex with Men Accessing Department of Public Health Mobile Van Services: Implications for Intervention Development. AIDS Patient Care & STDs, Vol. 22 (9): 745-751.

9