Supporting Statement - SSV - Part B

SUPPORTING STATEMENT

Survey of Sexual Victimization, 2017, 2018, and 2019

B. Statistical Methods

1. Universe and Respondent Selection

The SSV collects information on allegations and substantiated incidents of sexual victimization that occur in adult and juvenile correctional facilities. BJS uses a series of sampling frames to identify the universe of facilities covered by PREA. This universe is fluid due to changes in the operational status of facilities, including openings, closings, new contracts (i.e., privately operated), and ended contracts (i.e., no longer privately operated).

State and federal prison systems

The state and federal prison systems have centralized reporting. About 2 months prior to launching the survey each year, BJS receives an update of the listing of state and federal prison administrators from the Association of State Correctional Administrators.

Private prisons

The private prison sample is drawn from the Census of State and Federal Adult Correctional Facilities (CSFACF), which is conducted every 5-7 years by BJS. The most recent CSFACF was conducted in 2012 and will be used for the 2017, 2018, and 2019 samples. Between censuses, BJS updates the frame one or more times each year as changes are identified (for example, via internet searches of private correctional companies or as reported by past SSV respondents). The next prison census is scheduled to be conducted in 2019.

Public jails

The public jail sample is drawn from the Mortality in Correctional Institutions (MCI, formerly the Deaths in Custody Reporting Program (DCRP)). SSV uses the Annual Summary on Inmates under Jail Supervision portion of this frame. The MCI/DCRP file for the year preceding the SSV collection has been used as the sampling frame for SSV since 2007, and will continue to be used as the public jail sampling frame for SSV for 2017-19. For example, MCI 2016 will be used for SSV 2017.

Private jails

The private jail sample is also drawn from the MCI frame. MCI includes an indicator which identifies privately operated facilities, and the frame will be split into public and private facilities prior to drawing the samples each of the three years.

Other prisons and jails

There are four military prison systems (Army, Navy, Marines, and Air Force).¹ Each one has centralized reporting, maintains their own facility list, and provides BJS with a point of contact a couple months before fielding SSV each year.

A list of facilities operated by the Department of Homeland Security Immigration and Customs Enforcement (ICE), called dedicated ICE facilities, is maintained by ICE. BJS receives an update of these dedicated facilities from ICE each year, about two months before fielding the survey.

The Indian country jail sample is drawn from the Annual Survey of Jails in Indian Country (ASJIC). BJS maintains the ASJIC, which is updated annually and includes all known Indian country correctional facilities (adult and juvenile) operated by tribal authorities or by the U.S. Department of the Interior, Bureau of Indian Affairs. Facilities in Indian country that hold only juveniles are excluded here and included below in Local and private juvenile correctional facilities. The prior year of ASJIC has been used for the SSV sample each year. BJS will continue this for the next three years of SSV.

State juvenile correctional systems

The state juvenile correctional systems (including the District of Columbia) have centralized reporting. A couple of months before fielding the SSV each year, BJS receives an updated listing of state administrators from the Council of Juvenile Correctional Administrators.

Local and private juvenile correctional facilities

The local and private juvenile facility sample is drawn from three different frames, two of which are used in alternating years. The Census of Juveniles in Residential Placement (CJRP) is conducted by the Census Bureau for the Office of Juvenile Justice and Delinquency Prevention (OJJDP) in even-numbered years. The Juvenile Residential Facility Census (JRFC) is conducted, again by the Census Bureau for OJJDP in odd-numbered years. Again, the most recent frame is from the year preceding the SSV collection. That is, BJS will use CJRP 2016 for SSV 2017; JRFC 2017 for SSV 2018; and CJRP 2018 for SSV 2019. The juvenile correctional facilities in Indian country sample is drawn from the Annual Survey of Jails in Indian Country.

Approximately 7,050 facilities are covered by PREA. This includes about 5,200 adult prisons and jails and 1,850 juvenile facilities. PREA requires that BJS collect data from a sample of at least 10% of these correctional facilities. (See table 2 for the number of facilities and reporting units covered by PREA and sampled in 2016.)

Because of the low numbers of reported sexual victimizations to correctional authorities and the centralized authority at the system level that governs responses to the BJS surveys, the SSV elected to conduct a complete enumeration at the system level – including the Federal Bureau of Prisons, all state departments of correction, all state juvenile justice systems, and each branch of the U.S. military. In each of these cases, both the information systems and the authority to report are centralized. Moreover, this annual enumeration minimizes burden on the respondents (rather than selecting a sample from the more than 1,800 facilities operated by these systems).

BJS also conducts a complete enumeration of the facilities operated by or under contract with the United States Immigration and Customs Enforcement (ICE), called dedicated ICE facilities. 

Finally, private prisons, local jail jurisdictions, private jails, Indian country jails, and local and private juvenile facilities are completely decentralized. Each of these are sampled using the most current versions of the frames listed above. Sample designs have been relatively stable from year to year. For each type of sampled facility or jurisdiction, a detailed description of the sample design used for SSV 2016 has been provided, including sampling frames, coefficients of variation, and variance estimates. (See Attachment 6.)

Table 2. Number of facilities and reporting units in 2016 covered by the Prison Rape Elimination Act

		Facilities					Reporting units
Facility type	Form	Universe		Sampled	Percent		Universe		Sampled	Percent
Total		7,045		3,277	48.1%		5,126		1,579	49.1%
Prisons
Public – federal system	SSV-1	98		98	100%		1		1	100%
Public – state systems	SSV-2	1,293		1,293	100%		50		50	100%
Private	SSV-4	530		155	29.2%		530		155	29.2%
Jails
Public – local jurisdictions	SSV-3	3,100		~700	22.6%		2,884		700	24.3%
Private	SSV-4	39		15	38.5%		39		15	38.5%
Other adult facilities
Indian country jails	SSV-4	60		25	41.7%		60		25	41.7%
Military systems	SSV-4	44		44	100%		4		4	100%
ICE	SSV-4	29		29	100%		29		29	100%
Juvenile facilities
Public – state/DC systems	SSV-5	369		369	100%		51		51	100%
Public – local	SSV-6	661		266	40.2%		661		266	40.2%
Private	SSV-6	798		264	33.1%		798		264	33.1%
Indian country	SSV-6	19		19	100%		19		19	100%

Sample designs

A brief description of each sample design and stratum is provided below. These are the sample designs that were used for SSV 2016 and will be used for SSV 2017. However, BJS will review and evaluate the precision of each of these designs. If changes are needed, BJS will submit a request for a substantive change for SSV 2018 and 2019.

Federal and state prisons

The Bureau of Prisons and the 50 state departments of correction submit annual information on sexual victimization for all state and federal facilities resulting in a complete enumeration of all allegations reported within publicly operated prisons.

Private prisons

A sample of 155 (29%) privately operated state and federal prison facilities is drawn from the private prisons identified in the most recent version of the Census of State and Federal Adult Correctional Facilities frame. (The updated 2012 CSFACF frame contains 530 private prisons). The facilities are first sorted by average daily population (ADP). To determine the size cut-off for certainty facilities, the total ADP for all eligible facilities is divided by the number of facilities in the sample. Facilities with an ADP greater than or equal to this cut-off are sampled with certainty. The remaining total ADP is then divided by the remaining number of facilities in the sample to determine whether any facilities meet the subsequent certainty cut-off size. This process is repeated until no remaining facilities have an ADP greater than or equal to the cut-off. The facilities remaining in the frame at this point are sorted by region (i.e., Northeast, Midwest, South, or West), state, and ADP, and sampled systematically with probability proportional to size. (See Attachment 6 for details on the SSV 2016 private prison sample design.)

Public jails

A sample of 700 (24%) publicly operated jail jurisdictions is selected from the most recent Mortality in Correctional Institutions (MCI, formerly the Deaths in Custody Reporting Program) file.

The largest jail jurisdictions in 45 states and the District of Columbia are selected to meet the PREA requirement that at least one jail per state is selected each year.² All jail jurisdictions with ADPs greater than or equal to 1,000 inmates are selected with certainty. The remaining sample of jail jurisdictions are selected using a stratified systematic random sample as follows. The remaining jail jurisdictions on the frame are grouped into three strata. The cumulative sqrt(f(y)) method is used to determine the boundaries of these three strata based on ADP as the measure of size (Cochran, Sampling Techniques, 1997 edition, p. 129). The jail jurisdictions in each of these three strata are then sorted by region, state, and ADP, and selected systematically with probability proportional to their size. (See Attachment 6 for details for the SSV 2016 sample.)

Private jails

A sample of 15 (38%) privately operated jails is selected from the private jails in the most recent MCI file. These private facilities are sorted by region, state, and ADP, and 15 jails are systematically sampled with probability proportional to size.

Given the large standard errors, estimates for private jails are combined with public jails. The separate sample is used to ensure inclusion of private jails in the SSV. (See Attachment 6 for details for the SSV 2016 sample.)

Other prisons and jails

Each of the four military prison systems submits an SSV report each year, covering all facilities run by the Army, Navy, Marines, and Air Force.

All dedicated ICE facilities are included each year. Most years the forms are mailed directly to each ICE facility. However, for SSV 2015, officials from ICE headquarters opted to coordinate the data collection. That year all packets were placed in a large envelope and mailed to headquarters, and ICE headquarters mailed out the individual packets. BJS intends to use these procedures for 2017-2019.

For jails in Indian country, 25 (42%) of the approximately 60 jails in Indian country housing adult inmates are selected. Using the same method to determine certainty size cut-offs as used for private prisons, n facilities are sampled with certainty due to their size. The remaining facilities on the frame are sorted by state and ADP, and 25 - n facilities are sampled with probability proportionate to size. Facilities in Indian country housing exclusively juveniles are excluded from the adult sample. See Private and local juvenile facilities below for a description of sampling procedures for juvenile facilities in Indian country. (See Attachment 6 for details for the SSV 2016 sample.)

State juvenile correctional systems

All state operated juvenile residential placement facilities used to house juveniles and youthful offenders are included in SSV, regardless of age or reason for placement. As defined in the CJRP, residential placement facilities include detention centers, training schools, long-term secure facilities; reception or diagnostic centers; group homes or halfway houses; boot camps; ranches; forestry camps, wilderness or marine programs, or farms; runaway or homeless shelters; and residential treatment centers for juveniles. All juvenile correctional facilities operated by the 50 states and the District of Columbia are included in SSV (a total of 369 juvenile facilities in 2016). Data for these facilities are collected from 51 central reporters. BJS intends to employ the same procedures for 2017-2019.

Local and private juvenile correctional facilities

The total sample size for juvenile correctional facilities is 600. A sample of 530 local and private juvenile facilities (600 minus 51 state systems and 19 Indian country facilities) is drawn as follows –

1. First, the facilities in the frame are separated into three groups: local, private, and detention facilities. The local and private facility groups are sorted by state and size. The measure of size for juvenile facilities is the number of persons assigned to beds.

   1. The largest locally operated facility in each state is sampled with certainty.

   2. The largest privately operated facility in each state is sampled with certainty.

2. Next, the remaining facilities in each of the three groups are sorted by size. Using the same method to determine certainty size cut-offs as used for private prisons,

   1. large detention facilities are sampled with certainty due to their size,

   2. large local facilities are sampled with certainty due to their size, and

   3. large private facilities are sampled with certainty due to their size.

3. The remaining facilities are serpentine sorted by geographic region (Midwest, Northeast, South, and West), state, facility type (detention, local, or private), and size.

   1. The remaining detention facilities are sampled with probabilities proportionate to size from 4 strata based on geographic region.

   2. The remaining local facilities are divided into 2 strata based on commitment status (commitment and non-commitment), and sampled with probabilities proportionate to size.

   3. The remaining private facilities are sorted by region and state, then sampled with probabilities proportionate to size.

Finally, the juvenile sample includes all facilities in Indian country identified as housing exclusively juveniles (19 in 2016, based on the Annual Survey of Jails in Indian Country). (See Attachment 6 for details for the 2016 sample.)

The total sample size over all types of systems and facilities is approximately 1,580 sampling units per year.

2. Procedures for Collecting Information

Statistical methodology for stratification and sample selection are described above in 1. Universe and Respondent Selection. As mentioned above, the precision of SSV data is being evaluated. BJS will follow the same estimation procedures used for SSV 2016. Estimates are produced using weighted data. Weights are calculated at the unit level and are equal to the reciprocal of the sampling probability for each facility, jurisdiction, or system.

3. Methods to Maximize Response Rates

At the start of each collection, selected facilities receive a survey packet containing a letter signed by the BJS Principal Deputy Director describing PREA, the importance of the survey, and information about how to respond (see Attachment 7). For those that have no history of responding by Web, paper copies of the SSV forms will be included in the packet. BJS expects the packets for SSV 2017 to be mailed in early fall 2018, following OMB approval.

As the data collection agent, the U.S. Census Bureau will conduct the mailing, data collection, and follow-up. The Census Bureau will send out two reminder emails, one 4 weeks prior and one 2 weeks prior to the survey due date. Two weeks after the due date, they will follow up with non-respondents first by email and then by phone. Follow-up will be repeated about every two weeks until the data collection closes. (See Attachment 8 for non-response follow-up protocols.)

SSV is a multi-mode survey. Respondents can provide their data using a paper survey, fillable PDF forms, or the Web-based survey. A data extraction methodology is also being developed, whereby selected state systems or very large jail jurisdictions will have the ability to provide files from their records management systems (i.e., electronic datasets), from which the Census Bureau will extract the information to populate the annual SSV dataset.

Overall, participation is expected to exceed 95% for the SSV 2017, 2018, and 2019 collection years based on prior year response rates. In 2016, 95% of sampled systems and facilities responded. For adult facilities, all 50 state departments of correction, the Federal Bureau of Prisons, and all military and dedicated ICE facilities responded. Among the 698 public jail jurisdictions sampled in 2016, 27 did not respond to the survey. (See table 3.) Six (out of 25) sampled Indian country jails and 4 (out of 143) private adult prisons did not respond. For juvenile facilities, all state juvenile systems responded. Twelve (out of 257) local facilities, 24 (out of 249) private facilities, and 2 (out of 19) juvenile facilities in Indian country did not respond. (Counts exclude facilities that closed or no longer housed inmates or juvenile offenders.)

Table 3. Sampled reporting units by type of facility and response status, 2016

				Reporting units
	Facility type		Form	Sampled		Active	Re-sponded		Did not respond		Response rate
	Total			1,574		1,540	1,465		75		95%
	Prisons
	Public – federal system		SSV-1	1		1	1		0		100%
	Public – state systems		SSV-2	50		50	50		0		100
	Private		SSV-4	155		143	139		4		97
	Jails
	Public – local jurisdictions		SSV-3	700		698	671		27		96%
	Private		SSV-4	15		15	15		0		100
	Other adult facilities
	Indian country jails		SSV-4	25		25	19		6		76%
	Military systems		SSV-4	4		4	4		0		100
	ICE		SSV-4	29		29	29		0		100
	Juvenile facilities
	Public – state/DC systems*		SSV-5	51		50	50		0		100%
	Public – local		SSV-6	266		257	245		12		95
	Private		SSV-6	264		249	225		24		90
	Indian country		SSV-6	19		19	17		2		89
	* In Arkansas, all state juvenile facilities were privately operated in 2016.

Significant effort is being made to make the survey materials clear and straightforward. The SSV questionnaires have been designed to make collection of the data as concise and easy for the respondents as possible. Uniform definitions of terms and concepts as well as counting rules for items to be reported are included on the forms. The SSV also uses some questions that have been used previously in other surveys and are easily reported by most respondents.

Table 4 (page 7) summarizes the reporting capabilities of correctional authorities, by type of reported sexual victimization for collection year 2016.³ This table demonstrates significant capacity to report data using uniform definitions and survey categories among state and federal prison systems and state juvenile systems. However, local jails and local and private juvenile facilities show somewhat lower levels of conformity to data standards.

In 2016, all federal and state adult prison system authorities were able to report on each type of allegation; 4% could not separate allegations of abusive sexual contact from nonconsensual sexual acts; and 2% could not separate allegations of staff sexual harassment from staff sexual misconduct. All sampled jail jurisdiction authorities were able to report on nonconsensual sexual acts (76% had full reporting, 20% could not separate abusive sexual contact from nonconsensual sexual acts, 2% reported only allegations that were substantiated or acts that were completed, and 1% had partial reporting of nonconsensual sexual acts combined with abusive sexual contact); at least 98% of jails were able to report on each of the other types of allegations: 2% were unable to report on abusive sexual contact; 1% were unable to report on inmate-on-inmate or staff-on-inmate sexual harassment, and less than 0.5% were unable to report on staff sexual misconduct. Reporting of nonconsensual sexual acts separate from abusive sexual contacts remains difficult for jails, with 21% unable able to report separately in 2016.

All state juvenile systems were able to report data on nonconsensual sexual acts in 2016, and 98% were able to report on each of the remaining types of allegations. Local and private juvenile facilities were also able to report on most types of allegations. All were able to report data on nonconsensual sexual acts and staff sexual misconduct, more than 99.5% were able to report on abusive sexual contact, 99% were able to report on staff sexual harassment, and 98% were able to report on youth-on-youth sexual harassment.

Table 4. Reporting capabilities of correctional authorities, by type of sexual victimization, 2016
		Adult						Juvenile
Type of victimization		Prisons		Jails			State		Local/Private
Nonconsensual sexual acts
	Full reporting	96	%	76	%		90		%	75			%
	Includes abusive sexual contact	4		20			10			22
	Partial reporting*	0		2			0			1
	Partial and includes abusive sexual contact	0		1			0			1
	Unable to report	0		0			0			0
Abusive sexual contact
	Full reporting	96	%	77	%		88		%	76			%
	Combined with nonconsensual sexual acts	4		21			10			24
	Unable to report	0		2			2			<0.5
Inmate-on-inmate sexual harassment
	Full reporting	100	%	96	%		98		%	96			%
	Partial reporting*	0		3			0			2
	Unable to report	0		1			2			2
Staff sexual misconduct
	Full reporting	98	%	85	%		98		%	84			%
	Includes sexual harassment	2		13			2			15
	Partial reporting*	0		1			0			1
	Partial and includes sexual harassment	0		< 0.5			0			<0.5
	Unable to report	0		< 0.5			2			0
Staff sexual harassment
	Full reporting	98	%	86	%		96		%	84			%
	Combined with sexual misconduct	2		14			2			15
	Unable to report	0		1			2			1
* Systems or facilities which report only allegations that were substantiated or acts that were completed.

Table 5 presents non-response rates for selected items on the SSV adult and juvenile Substantiated Incident Forms. Overall, these data show high response rates for critical items, ranging from 0% non-response for several items to 6.6% nonresponse for location of staff-on-inmate incidents.

Table 5. Non-response for selected items by type of incident, 2016
		Adult (SSV-IA)			Juvenile (SSV-IJ)
		Percent			Percent
Item		non-response			non-response
1. Date	Month	0.1		%	0.0	%
	Day	0.5			0.2
	Year	0.4			0.2
2. Facility		0.0			0.0
3. Location	Inmate/youth	2.1		% n = 1,639	2.8	% n = 507
of incident	Staff	6.6		n = 624	5.8	n = 86
6. # of victims
	Inmate/youth	0.7		%	2.0	%
	Staff	0.3			1.2
7. Victim #1 gender
	Inmate/youth	1.3		%	3.4	%
	Staff	0.6			0.0
8. Victim #1 age
	Inmate/youth	1.8		%	4.5	%
	Staff	1.0			1.2
9. Victim #1 race
	Inmate/youth	1.8		%	3.4	%
	Staff	0.6			3.5
13. Victim injury
	Inmate/youth		0.0	%	2.0	%
	Staff		0.0		0.0
15. Medical follow up
	Inmate/youth		0.8	%	0.4	%
	Staff		1.0		0.0
16. Victim housing change
	Inmate/youth		0.9	%	2.4	%
	Staff		0.8		1.2
18. # of inmate/youth perps			1.8	%	2.0	%
19. Perp #1 gender			1.5		1.4
20. Perp #1 age			2.6		2.4
21. Perp #1 race			1.9		2.8
26. Use of force (inmate/youth)			1.3		1.8
28. Use of force (staff)			1.6	%	1.2	%
29. # of staff perps			0.3		0.0
30. Perp #1 gender			0.2		0.0
31. Perp #1 age			3.0		1.2
32. Perp #1 race			1.0		1.2

A few items related to the details of incidents have non-response rates greater than 5%. On the adult Substantiated Incident forms for SSV 2016, item non-response rates were greater than 5% for the following items (n’s are the same as in table 5 unless otherwise noted):

• item 3, where incident occurred (6.6% for staff-on-inmate victimization);

• item 4, video monitoring of location of incident (12.9% inmate-on-inmate and 20.8% staff-on-inmate);

• item 5, time of occurrence (21.3% inmate-on-inmate, 33.2% staff-on-inmate);

• item 13a, type of injury to victim (20.0% staff-on-inmate, n = 5);

• item 13b, treatment received by injured victim (12.5% inmate-on-inmate, n = 96);

• item 23, age of second inmate perpetrator (9.7% inmate-on-inmate, n = 103); and

• item 34, age of second staff perpetrator (13.0% staff-on-inmate, n = 23).

On the juvenile Substantiated Incident forms for SSV 2016, item non-response rates were greater than 5% for the following items (n’s are the same as in table 5 unless otherwise noted):

• item 3, where incident occurred (5.8% for staff-on-youth victimization);

• item 4, video monitoring of location of incident (10.3% youth-on youth and 9.3% staff-on-youth);

• item 5, time of occurrence (9.3% youth-on-youth, 30.2% staff-on youth);

• item 10, sex/gender identity of second victim (6.8% youth-on-youth, n = 59);

• item 11, age of second victim (11.9% youth-on-youth, n = 59, 9.1% staff-on-youth, n = 11);

• item 12, race of second victim (6.8% youth-on-youth, n = 59);

• item 13a, type of injury to victim (20.0% youth-on-youth, n = 5);

• item 13b, treatment of injured victim (60.0% youth-on-youth (n = 5));

• item 34, age of second staff perpetrator (33% staff-on youth, n = 3).

Some of these rates may be considered high, however, it is important to track these characteristics. In many cases, the item specific non-response rates reflect an absence of information, even after an investigation is complete.

Three items relating to inmate-on-inmate sexual harassment (items 7, 8, and 9 in SSV-1) were added to the summary forms in 2013. (See Attachment 2 for corresponding items in forms SSV-2, SSV-3, SSV-4, SSV-5, and SSV-6.) These items and related definitions were added to align the SSV with the PREA standards. In 2016, 100% of adult correctional systems and 96% of local and private jails and other adult correctional facilities were able to answer these items (see table 4 above). During the same period, 98% of juvenile correctional systems and 96% of local and private juvenile correctional facilities were able to answer these items.

The national standards issued in 2012 (28 C.F.R. Part 115) have extensive requirements related to the treatment of inmates who are transgender or intersex. To be compliant, correctional authorities need to know the sexual status of inmates. The PREA standards specific to transgender and intersex prisoners include 115.15 (e) and (f), limits to cross-gender viewing; 115.31 (9), employee training; 115.41 (7), screening for risk; 115.42 (c), use of screening information; and 115.86, sexual abuse reviews. As such, correctional authorities must track both victims and perpetrators who may be considered transgender or intersex.

To help ensure compliance, the SSV 2013 survey added two additional response categories related to the sex/gender identity of victims and inmate/youth perpetrators. (See forms SSV-IA and SSV-IJ, items 7 and 10 for victims and items 19 and 22 for inmate/youth perpetrators.) These response categories include “transgender” and “intersex.” In 2016, the question itself was reworded at OMB’s request, changing from “What was the sex” to “What was the sex or gender identity.” The PREA-specified definitions are provided on the forms (see page 5 of SSV-IA and SSV-IJ.)

Item 7 (from SSV-IA page 1)

7. Victim #1: What was the victim’s sex or gender

identity? (See definitions on page 5.)

01  Male 03  Transgender

02  Female 04  Intersex

Gender categories (from SSV-IA page 5)

TRANSGENDER: A person whose gender identity (i.e.,

internal sense of feeling male or female) is different from

the person’s assigned sex at birth.

INTERSEX: A person whose sexual or reproductive anatomy or

chromosomal pattern does not seem to fit typical definitions of

male or female. Intersex medical conditions are sometimes

referred to as disorders of sex development.

The SSV 2013 added two new items to the incident forms in response to requests from external users to enhance our understanding of the circumstances surrounding incidents of sexual victimization. The first (item 4) relates to video monitoring of the place where the incident occurred. While this item had a relatively higher non-response rate, the results are still very informative. BJS expects the response rate for this item to increase as systems and facilities begin to track the information. The second (item 39) relates to the length of time staff perpetrators worked at the facility, and respondents were able to answer this in 2016 (item response rate = 97%).

Also, in 2013 on both the summary and Substantiated Incident Forms, text was added to the definitions of sexual victimization to ensure that respondents understood that the SSV definitions and the PREA definitions are consistent. These definitions have remained unchanged.

OMB approved the SSV 2016 with a one-year conditional clearance. OMB requested more testing of items 7, 10, 19, and 22. The methods and results of this testing are described below in Section 4, Final Testing of Procedures.

4. Final Testing of Procedures

BJS proposes no changes to the Summary forms (SSV-1, 2, 3, 4, 5, 6). BJS proposes to remove one answer category from item 25 on the Substantiated Incident Forms (SSV-IA and SSV-IJ, see Attachment 2). Item 25 asks about the nature of inmate-on-inmate or youth-on-youth sexual abuse. The proposed response category for removal is “voluntary sexual contact between inmates/youth.” This response option was included initially when PREA reporting first began and facilities were reporting all sexual encounters that were investigated and found to have occurred, even if they did not meet the standard for sexual abuse. However, voluntary contact between inmates or between youth is not covered by PREA. In 2013, the “voluntary” response option was removed from item 17, but not from item 25. In 2016, juvenile facilities did not select this response for any substantiated incidents. Adult facilities selected this response for 18 out of 1,360 incidents [it was also checked in an additional 22 incidents, but other boxes were checked that indicated the contact was not voluntary (e.g., unwanted sexual contact)]. This answer category is no longer needed and can be removed.

Changes previously approved for collection in 2016 will not change and are described below. (See Attachment 2 for the 2016 forms.)

Minor wording changes were made to the instructions on the SSV 2016 Summary Forms related to the inclusion and exclusion statements for SSV-3 and SSV-4. Due to changes in how multi-jurisdictional facilities are handled in the sampling frame (MCI/DCRP), multi-jurisdictional facilities may receive either the SSV-3 or the SSV4. The inclusion and exclusion statements were adjusted on each of these forms to reflect this change.

Three changes were made to the Incident forms (SSV-IA and SSV-IJ) for SSV 2016. The first change was the title of the form. In prior years a few respondents completed the Incident Form for allegations that were unfounded, unsubstantiated, or still under investigation. In order to reduce respondent burden and improve data quality, BJS added the word Substantiated to the title (i.e., Substantiated Incident Form). The clarification was effective. For SSV 2016, all incident forms received were for substantiated incidents.

The second change related to respondent error in completing the perpetrator information. While incident and victim questions are the same regardless of the type of incident, the perpetrator questions for inmate-on-inmate victimizations (Section A) differ from those for staff-on-inmate victimizations (Section B). In prior years BJS received forms with perpetrator information entered into the wrong section. Again, to reduce respondent burden and improve data quality, BJS changed the instructions by directing respondents to the appropriate sections on inmate/youth and staff perpetrators. The section for inmate perpetrator information is presented first, so it was more common for the information on staff perpetrators to be entered into the inmate perpetrator fields. To resolve this in the web-based survey for SSV 2017, skip patterns will take the respondent to the correct section based on the type of incident marked in item 17.

The third change affected the wording of the question for items 7, 10, 19, and 22. For SSV 2013-15, these items asked about the sex of victims and inmate/youth perpetrators. Two answer categories were added in 2013, transgender and intersex. For SSV 2016, the wording of these items was modified to ask about the sex or gender identity of victims and inmate/youth perpetrators. The answer categories remained the same. BJS received clearance of these four items with the condition that further testing be conducted.

In response to the terms of clearance, BJS 1) monitored the number of reports and response rates for these items, and 2) conducted record-keeping interviews with 9 respondents. The results are summarized below.

Since the introduction of these two answer categories on the SSV, the number of reported intersex and transgender victim and perpetrators has increased each year from 2014 to 2016. Overall, 182 transgender and intersex victims and inmate/youth perpetrators of sexual victimization in correctional facilities have been reported through SSV (see table 6).

Table 6. Transgender or intersex victims and inmate/youth perpetrators reported in the Survey of Sexual Victimization, 2014-16
	Total	2016	2015	2014
All systems, jurisdictions, or facilities that--
Responded	6017	1469	1527	1519
Had one or more substantiated incidents	1160	405	375	380
Had a transgender or intersex victim or perpetrator	108	41	34	23
Number of transgender or intersex victims or perpetrators	182	85	55	30
Victims	132	60	42	20
Inmate/youth perpetrators	50	25	13	10
Adult systems, jurisdictions, or facilities that--
Responded	3808	979	952	939
Had substantiated incidents	1068	293	269	269
Had a transgender or intersex victim or perpetrator	91	31	>28	>17
Number of transgender or intersex victims or perpetrators	137	60	38	>24
Juvenile systems or facilities that--
Responded	2209	490	575	580
Had one or more substantiated incidents	419	112	106	111
Had a transgender or intersex victim or perpetrator	17	10	<5	<5
Number of transgender or intersex victims or perpetrators	45	25	17	<5

BJS developed a Record-Keeping Interview protocol to gain a better understanding of broader record-keeping practices among the various correctional systems and facilities. (See Attachment 9.) Nine correctional systems, jurisdictions, or facilities (herein, facilities) were selected (see table 7 for a list of facilities and their characteristics).

Table 7. Characteristics of sampled facilities for SSV Record-Keeping Interview
Name	State	Type	Popu-lation	ADP	Response of intersex or transgender
New York Dept. of Corrections and Community Supervision	NY	State	Adult	51,485	Transgender
George W. Hill Correctional Facility (GEO Group)	PA	Private	Adult	1,883	Transgender
Stutsman Co. Correctional Center	ND	Local	Adult	70	Intersex
Los Angeles Co. Jail	CA	Local	Adult	17,054	Neither
Louisiana Dept. of Public Safety and Corrections	LA	State	Adult	15,295	Neither
Wayne Co. Jail	MI	Local	Adult	2,069	Neither
Texas Juvenile Justice Department	TX	State	Juve	1,212	Neither
Oregon Youth Authority	OR	State	Juve	566	Both
Juvenile Detention Center	MO	Local	Juve	44	Transgender

BJS called each selected facility to request their participation and schedule the interview. The interviews lasted about 35 minutes. Results of the study are summarized below.

All nine facilities stored the data in a central location, but the types of storage varied. Most facilities used either a records management system (RMS) or spreadsheets or a combination of the two. A few referred to electronic or paper copies of investigation reports. The RMSs, spreadsheets, and investigation reports contained all of the information needed to respond to the SSV. Most contained additional information not covered by the SSV, such as names of victims.

All nine facilities said the information was easy to access. Some of the larger facilities mentioned that it was very time-consuming. Four facilities needed to go to a single source of information to complete the SSV Substantiated Incident Form (SSV-IA or SSV-IJ), and five needed to go to multiple sources, e.g., their RMS and their incident reports. Regardless of where the data are stored, all nine facilities review each incident and complete each SSV-IA or -IJ individually. Some of the smaller facilities complete the form on a flow basis after an incident has been substantiated. Two of the state systems require multiple levels of review before submitting their incident forms to BJS.

The items on sex and gender identity of victims and inmate/youth perpetrators were easy to answer for all nine facilities. Two states noted that while it was easy, it was time-consuming. All nine facilities use the PREA definitions of intersex and transgender. No facilities use other terms for the concepts of intersex or transgender. However, one juvenile facility uses whatever term a youth prefers when speaking to that youth. Another juvenile facility also collects data on gender non-conforming youth. Although none of their gender non-conforming youth have been the victim or perpetrator of a sexual assault while in their facility, if that occurred the facility said they would include gender non-conforming with transgender for SSV purposes. This facility also mentioned that they mark “Transgender” for transgender victims or perpetrators, even though they could also mark Male or Female.

Data on intersex and gender identity were relatively complete for all nine facilities. All nine facilities gather information on sex (male/female) for all inmates/youth at intake. Sex (male/female) is generally determined by observation or arrest or court records. All but one facility screen all inmates/youth for being intersex or transgender at intake, either through observation or through a series of questions. The facility that does not screen everyone for gender identity at intake is small and screens those who do not appear to be obviously male or female. Gender identity is usually self-reported by the inmate/youth. It is usually taken at face value and sometimes verified through family or friends (without revealing information the family or friends did not already know). Court, medical, and housing records are sometimes used as sources for information on gender identity.

Seven facilities collect information from everyone at intake on whether an inmate/youth is intersex. Two collect this information if sex (male/female) is not obvious. Again, this information is collected primarily through self-report. Mental health evaluations and medical records are sometimes used. Medical examinations are rarely used. Facilities that held inmates/youth who were incarcerated prior to the passage of the national standards in 2012 either went back and screened each inmate/youth or picked up the information during regular mental health evaluations or counseling sessions.

Even though all inmates/youth have the opportunity at intake to disclose whether they are intersex or transgender, some do not. It is not uncommon for an inmate/youth to disclose their gender identity at a later time. This is done most frequently during a regular meeting with mental health staff or counselors (most facilities have these for all inmates/youth) or during a conversation with correctional or medical staff. Six facilities said information on sex or gender identity is sometimes uncovered or corrected during an investigation.

Data on intersex and gender identity were available for all nine facilities. Access was straightforward in most cases, but was more time-consuming for larger facilities with more substantiated incidents. Five of the nine facilities store information on sex and gender identity in the same place as the other information needed to complete a Substantiated Incident Form. Two facilities usually store all the information together, but sometimes it is not entered into the main RMS/spreadsheet, and two facilities always store the information on sex and gender identity separate from the other information.

Eight facilities keep the information in a centralized RMS or centralized spreadsheet. One keeps it in hard copies of investigation reports. Eight facilities need to refer to individual incident reports. All nine facilities complete each Substantiated Incident Form separately.

All nine facilities said the data on sex and gender identity were neither easier nor harder to access than the other required information. All nine keep data on intersex and gender identity in the same place as the information on sex (male/female).

All facilities thought the questions in the protocol covered everything that needed to be covered. Several shared an experience or observation:

• Having used paper, Web, and fillable PDFs to respond to SSV, the fillable PDFs were preferable because they’re the easiest to store and print.

• Sex and gender identity is an important issue, and recent training heightened their awareness and made them more proactive.

• It is helpful when their system’s Transgender, Gender Non-Conforming, Intersex Committee includes the PREA Coordinator in its membership.

• The SSV is a helpful annual review.

Two facilities had other issues they wanted to discuss. One said they would be sending comments during the 60-day notice (these were summarized and addressed in Part A). The other mentioned that PREA required specific staffing ratios for juvenile facilities, one during awake hours and another during sleeping hours (sleeping hours may differ by facility, theirs are 10 pm to 6 am). They noted that one time-frame response category on the SSV-IJ (6 pm to midnight) included both awake and sleeping hours, so it was not useful to them for planning or analytic purposes. They requested that we consider adding a question to the SSV-IJ on whether the incident occurred during awake or sleeping hours.

Based on the reported responses to SSV and the nine interviews, BJS found:

• The items on sex and gender identity are understood by respondents. The items are being answered consistently and how they were intended to be answered.

• All interviewed facilities collect and maintain information on sex, intersex, and gender identity for all inmates/youth to the best of their ability without being intrusive with the inmate/youth. Those that do not screen every inmate/youth for gender identity are small facilities, and they screen those that present visual or behavioral cues of being transgender. If an inmate/youth chooses not to disclose that they are transgender or intersex (sometimes intersex is chromosomal with no obvious physical or biological indicators), the inmate/youth has multiple opportunities to do so at a later time.

• The items on sex, intersex, and transgender are on par with the other items on the Substantiated Incident Form regarding ease or difficulty in accessing the information and answering the item.

• Issues of burden for larger systems/facilities are due to PREA requirements for compliance rather than time or effort to complete the SSV forms in general or answer the items on sex and gender identity in particular. One way BJS is addressing issues of burden related to data submission by developing the capability to upload PDFs. BJS is also planning on investigating the feasibility of developing electronic reporting whereby a system/facility can provide information in database, spreadsheet, or some other electronic format.

BJS proposes that the questions and answer categories for the items on sex and gender identity of victims and inmate/youth perpetrators (items 7, 10, 19, and 22) remain as they were on the SSV 2016 Substantiated Incident Forms.

5. Contacts for Statistical Aspects of Data Collection

BJS takes responsibility for the overall design and management of the survey, including sampling procedures, development of the questionnaires, and the analysis and publication of the data. The BJS contacts is—

Ramona Rantala, Statistician

Bureau of Justice Statistic

810 Seventh St., N.W.

Washington, DC 20531

(202) 307-6170

Ramona.Rantala@usdoj.gov

The Economic Reimbursable Surveys Division at the Census Bureau is the collection agent and is responsible for the collection of all data. The Economic Statistical Methods Division is responsible for drawing the samples. The Census Bureau contact is—

Greta Clark, Survey Statistician

Criminal Justice Statistics Branch

Economic Reimbursable Surveys Div.

U.S. Census Bureau

4600 Silver Hill Road

Washington, DC 20233-6800

(301) 763-2586

Greta.B.Clark@census.gov

1 The Coast Guard does not have its own facilities, relying instead on those operated by other branches of the Armed Forces, primarily the Navy.

2 Five states with combined jail/prison systems had no public jails: Connecticut, Delaware, Hawaii, Rhode Island, and Vermont.

3 Results for SSV 2012-2014 from adult correctional facilities are scheduled for publication in 2017.

28