OMB questions SSV

OMB Questions on the Survey of Sexual Violence (SSV)

1. The Terms of Clearance for the preceding OMB clearance of this collection included the following language:

BJS shall submit to OMB all cover letters and other informational materials that will be provided to respondents prior to fielding. BJS shall also brief OMB on the results of the first year of data collection including an evaluation of data quality and item non-response. OMB recognizes the need to initiate data collection quickly in order to meet the legislative requirements for a Report to Congress. While a three year clearance is given in this approval, BJS should make every effort to improve the instrument during the three years if evaluation of first year data collection warrants changes.

1. Please provide all cover letters and other information materials planned for the 2008 collection.  

Please see attachment: SSV2 2007 mailout_letter. There are no other information materials provided to the respondent other than the approved SSV form(s).

2. Please provide evaluation results of data quality and item non-response.

Please see attached table 1 that summarizes reporting capabilities of correctional authorities, by type of reported assault for collection years 2004, 2005 and 2006. This table tracks significant improved capacity to report data using uniform definitions and survey categories among State and Federal prison systems. However, local jail authorities show lower levels of conformity to data standards. Due to sampling, approximately a two-thirds of jails selected each year are receiving the summary form containing the definitions and reporting criteria for the first time. Reporting capabilities are expected to improve as the survey includes an increasing percentage of jurisdictions that have previously been sampled.

Overall, the data demonstrate high capacity to provide data. Measures of sexual harassments have the highest rates of partial reporting (5% in jails) or inability to report (2%). Such levels do not impair the ability to draw accurate estimates of sexual violence by type.

See attached table 2 that summarizes item non-response on the SSV incident report. Overall, these data show low non-response rates for critical items. Notable exceptions are observed for (1) time of occurrence (item 6), especially for staff sexual misconduct (20.4%); (2) housing changes (item 15) for inmate victims (12.0%); and (3) length of employment for perpetrators (item 36) for staff (5.9%). Although these item-nonresponse rates are higher than desired, they reflect limits of information contained in the investigative reports (rather than ambiguity of item content or selective reporting from jurisdictions). The need to track these characteristics of substantiated incidents outweighs deletion of these items from the surveys.

3. Please provide any other evaluation results or plans.

BJS intends to continue to track data coverage, reporting capacities and item non-response annually. To date, the survey submissions have required minimal call backs. BJS attributes this high level of data quality to the fact that incident-level data are based on completed investigations in which allegations have been substantiated.

2. Data comparability -- Publications on the first several years of data indicate that results should not be compared at a system or facility level. 

1. Does BJS have a target year by which it expects that results will be comparable?  What indicators or criteria are being/will be used to make that determination?  

Prior to BJS collecting data, there was no system or set of definitions for collecting data on sexual assault in correctional facilities. The field has been extremely responsive in adapting to this new collection. There has been an improvement of the State prison systems each year since the 2004 collection in their ability to provide the information according to our definitions. The ability of States to fully report on nonconsensual sexual acts using BJS definitions went from 67% in 2004, to 71% in 2005, to 80% in 2006. Similarly, full reporting on staff sexual misconduct among State systems went from 69% in 2004, to 86% in 2005, to 90% in 2006.

Jail jurisdictions are a greater challenge, as we annually sample individual facilities from over 3,000 jails, many of which are different from year to year. Every year there are some jails to which this collection is entirely new. In 2004 BJS went to mainly large jail systems with greater reporting abilities (sampling the largest jail in each state and probability proportionate to size thereafter) and 78% of facilities were able to fully report on nonconsensual sexual acts; 88% could fully report on staff sexual misconduct. Subsequent years sampling changed to increase the likelihood of smaller jail facilities being selected, resulting in fewer facilities able to fully report by BJS definitions. In the 2006 collection 61% of facilities able to report fully on nonconsensual sexual acts and 74% reporting full on staff sexual misconduct.

Direct comparability between jurisdictions and facilities will remain a challenge, as different States define sexual assault and staff sexual misconduct differently under their statutes and track and record the data in different ways. For example, some jurisdictions automatically pass investigations of sexual assault to law enforcement and prosecution and thus have limited information on the outcome of such incidents. BJS will continue to strive for complete comparability of data; in the meantime, differences in reporting and definitions are carefully noted in the reports.

In late 2008 the Prison Rape Elimination Commission is expected to release standards related to policies and procedures for eliminating sexual violence. These standards are expected to include uniform definitions and counting rules related to data. Variations from State to State and jurisdiction to jurisdiction are expected to decline following promulgation of these standards.

2. Please describe what ranking, if any, occurs or will occur across the 5 collections that comprise the National Prison Rape Statistics Program.  

Results from the inmate self-report surveys - the National Inmate Survey (NIS) and National Survey of Youth in Custody (NSYC) - will be used to rank facilities as required under the Act. The administrative data collections are intended to provide officially reported data, but not to rank facilities specifically. The administrative data collected in the SSV are assumed to be an undercount given fear/shame of reporting, but yield information about what administrators know (consider the difference in number of  sexual assaults reported to police in the Uniform Crime Reports and the higher rate reported informally within the National Crime Victimization Survey).

The NIS and NSYC facility-level rates are based on inmate self-report and BJS has gone through great lengths to ensure confidentiality and comfort in reporting on this sensitive topic. Samples within facilities are designed to represent the entire population whenever possible, and standard errors will be provided by facility. These surveys overcome the jurisdiction-based variations in policy and procedures in reporting that limit comparability.

The Former Prisoner Survey (FPS) cannot report data at the facility level, as parolees are asked about the experience over the entirety of their last incarceration, including jail time and any prison facility in which they resided. Such small numbers cannot be applied to a facility as a whole.

The Medical Surveillance project is still under development, but is anticipated to be a smaller scale project that will focus on gathering additional information in those facilities identified through the NIS as having a high prevalence of sexual assault. Such facilities should exhibit other measurable indicators correlated with high rates of sexual assault, such as increased STDs, reports of victimization to medical staff, etc.

3. What analysis across the 5 collections has occurred to facilitate their use complementarily?  

The different collections are designed to compliment one another, but are not directly comparable. The SSV collects what administrators know about reported sexual assaults. The NIS in 2007 collected allegations of sexual assault from inmates, including whether or not they reported any incident to staff. The 2007 SSV collection is an important comparison during a similar snapshot in time as to what administrators say what was reported and what inmates claim they reported. The 2007 reference period will be the first time we can compare what is reported to administrators (SSV) to what inmates report to the NIS.

The NSYC will collect similar data in 2008, and the 2008 SSV for juvenile facilities will used as a benchmark for comparison purposes. The FPS is intended as an independent measure of the average experience of inmates over the entirety of their last incarceration, including jail, prison, and community corrections. Thus, the data from the SSV is not applicable to this collection. Finally, the Medical Surveillance project is being designed to yield additional data on measures correlated with sexual assault within a small number of facilities identified as having rates of sexual assault in the NIS.

2. Payments or gift to respondents – Part A should contain a discussion of this topic, including if no payment is anticipated.  

There is no payment or gift to staff in State or Federal systems, local jail jurisdictions, private facilities or juvenile facilities for responding to the survey.

2. Cost estimates for project – those provided indicate they are for 2006.  Please update as needed to reflect costs for the 3 year period for which approval is requested.  

BJS estimates a 4% increase in costs each year, per agreements in the 2007 and 2008 Inter-Agency Agreements (IAAs) with the U.S. Census Bureau. BJS costs are expected to remain stable, subject to Cost of Living Adjustments (COLA).

5. Survey Frame – please describe the frame from which each sample is drawn and what is known about its properties (e.g., completeness), as required in Part B, item 1.  

The sampling frame source varies for each type of correctional facility. For each year, the State Department of Correction and the Bureau of Prisons submit information based on reports of sexual assault among all state/federal operated facilities, thus, a complete enumeration without the need for sampling. The private prison sample draws from the most recent Census of State Correctional Facilities, conducted every 5-6 years by BJS (the 2000 Census for 2004 and 2005 samples, and the 2005 Census for the 2006 sample). The 2005 Census of State Correctional Facilities will be used as the frame for the 2007, 2008, and 2009 sample, adjusted for any openings and closings as we learn of them. The private and public jail sample is drawn from the most recent Census of Jails, conducted every 5-6 years by BJS (the 1999 Census for the 2004 and 2005 samples and the 2005 Census for the 2006 sample). The 2005 Jail Census will also be used for the 2207, 2008, and 2009 sampling frame, adjusted for any openings and closings as we learn of them.

All 50 State juvenile justice jurisdictions report on incidents of sexual violence within all state-operated juvenile facilities, thus, a complete enumeration with no need for sampling. For private and local juvenile facilities, BJS utilized the 2003 Census of Juvenile in Residential Placement (CJRP) for the 2004 sample, and the 2004 Juvenile Residential Facility Census (JRFC) for 2005 and 2006 samples. Both frames are collected by the Office of Juvenile Justice and Delinquency Prevention every 2-3 years. For the 2007, 2008, and 2009 samples, BJS will use the most recent frame available.

BJS collects a report from the main branches of the military each year to cover all facilities run by the Army, Navy, Coast Guard, Marines, and Air Force. Similarly, all facilities operated by the Bureau of Immigration and Customs Enforcement (ICE) are included each year. Jails in Indian Country are sampled each year based on the number of facilities known to the Bureau of Indian Affairs (BIA).

5. Please describe precision requirements and justify specific sample sizes proposed, as required in Part B, item 1.  Please note that OMB is aware of the sampling requirements in the statute, but those do not appear to require sampling at the specific levels proposed for each facility type, e.g., local prisons. 

The Act requires a 10% sample of correctional facilities. In the case of systems operated by the State or Federal governments we take all facilities, as information tends to be centralized, and one report minimizes burden on the respondents (Department of Correction, Bureau of Prisons, etc.) The same is true for military facilities - a form goes to each independent branch to report on all facilities for which it is responsible. In the case of the Bureau of Immigration and Customs Enforcement (ICE), a form is sent to each facility operated by ICE, as the system is largely decentralized. Finally, for local jails and private prisons, jails, juvenile facilities, and facilities in Indian Country, which are completely decentralized, a 10% sample of each is drawn based on the most current frame available. BJS provides a detailed description of the sampling methods in the back of each published report (see attached reports).

7. Please provide anticipated response rates, as well as the actual response rate in the most recent collections, as required in Part B, item 1. 

The response rate varies according to the type of correctional jurisdiction. All 50 State Departments of Correction and the Federal Bureau of Prisons have participated in the 2004 - 2006 collections. For the remaining entities, responses have been as follows: 4 non-responses in 2004 (one in Indian country, and 3 juvenile facilities); 0 non-respondents in 2005 (adult facility report only); 5 non-respondents in 2006 (adult facility report only, all non-respondents were local jails). Overall response rates were 99.6% in 2004 (adult and juvenile facilities), 100 in 2005 (adult facilities only), and 99.2% in 2006 (adult facilities only).

8. Survey Methods –

1. What advance contacts, reminders, calls or other contacts are planned?  Please describe, as required in Part B item 2. 

Each system/facility will receive a letter (attached) and a copy of the appropriate SSV form(s) shortly following approval of this collection (they usually go out at the end of the calendar/reference year). Our data collection agent, the U.S. Census Bureau, will follow up with phone calls and emails after the expiration of the return date on the form. Typically, Census is successful in obtaining the majority of the responses. The remaining non-respondents are forwarded to BJS and we make personal calls to obtain the information. BJS has been successful in this endeavor (see response rate above).  

2. Also, please provide associated materials.

See Methodology section in the attached copies of our last three reports.

Questionnaire – Please note that question 8 (and other similar questions) on SSV-IJ does not meet OMB’s statistical standards for the collection of information on race and ethnicity.  Please revise to conform.  See the email exchange below with your colleagues at BJS for how we resolved this issue for a different administrative records collection that provides a useful model.  

Rather than "Other," BJS will add the phrase "Other racial/ethnic origin" to satisfy the OMB standard.  BJS will specify in publication what is specifically included in the “Other racial/ethnic origin” category as reported by the data respondents.