The Prison Rape Elimination Act of 2003 (P.L. 108-79) requires the Bureau of Justice Statistics (BJS) to carry out a comprehensive statistical review and analysis of the incidence and effects of prison rape for each calendar year.
The National Survey of Youth in Custody, 2008-2009 provides the first facility-level estimates of youth reporting sexual victimization in juvenile facilities.
The 2008-2009 NSYC survey consisted of an audio computer-assisted self-interview (ACASI) in which youth, using a touch-screen, interacted with a computerized questionnaire and followed audio instructions delivered via headphones. The NSYC utilized self-administered procedures to ensure the confidentiality of the reporting youths and to encourage fuller reporting of victimization. The survey made use of audio technology to provide assistance to youth with varying levels of literacy and language skills. Approximately 98 percent of the interviews were conducted in English; 2 percent in Spanish.
Administrators in each state, county, and private facility determined the type of consent required for youths to be eligible for participation. Administrators provided in loco parentis (ILP) consent in 63 facilities. In loco parentis is when administrators provide consent "in the place of the parent" to contact youth. In the remaining 132 facilities, administrators required consent from the youths' parents or guardians (PGC). Youth in all facilities also had to assent to participate in the interview.
In each sampled PGC facility, administrators were asked eight weeks prior to data collection to provide a roster of all adjudicated youth assigned a bed; in ILP facilities a roster was provided four weeks prior to data collection.
All youth were sampled in ILP facilities and in PGC facilities with 240 or fewer youth on the roster. In larger PGC facilities, all females and a random sample of males were selected. In both PGC and ILP facilities all incoming youth were added to the sample up to four weeks prior to the survey. Youth who had been present in the facility at least four weeks prior to the survey and were present at the time of the survey were considered eligible.
The result of this process yielded a sample representing 26,551 adjudicated youth held nationwide in state operated and large locally or privately operated juvenile facilities. A total of 10,263 youth participated in the survey. Of these, 1,065 received an alternative survey on drug and alcohol use and treatment, and 9,198 youth participated in the survey of sexual victimization.
Sampling of facilities
A multistage stratified sample design was used.
At the first stage of selection, a total of 284 facilities
was selected from the over 500 eligible facilities
in the United States. Eligible juvenile facilities
included state-owned or operated juvenile
facilities and non-state facilities with 105 or
more adjudicated youth.
Selection of facilities was completed using the
2006 Census of Juveniles in Residential Placement
(CJRP), conducted by the Office of Juvenile
Justice and Delinquency Prevention. Based
on a complete enumeration, 548 facilities were
determined to be eligible for selection. A small
number of facilities were later determined to be
out-of-scope. Facilities were out-of-scope if the
sampled facility (1) had closed, (2) was a non-state
facility housing fewer than 105 youth, or (3) did
not house youth for more than 90 days. The
facility sampling rates ranged from a low of
about 1 in 10 for the smallest facilities to certainty
among the largest facilities.
For sample selection purposes, a measure of size
equal to the number of adjudicated youth
(according to the 2006 CJRP) was assigned to
each facility in the frame. Two hundred-and-one
facilities were included in the sample with certainty.
These certainty facilities were state facilities
with at least 90 youth or non-state facilities
with at least 105 youth.
Next, one state facility was randomly selected
from a designated substratum within each of the
nine states that did not contain a certainty facility.
The designated substratum consisted of the largest
facilities in the state. Within each of the designated
substrata, one facility was selected with
probability proportionate to the size of facility.
An additional 74 state facilities were selected for
the study from the remaining facilities. Facilities
were assigned to strata defined by gender of
youths held in the facility, percent of youths who
were female, facility size, region, and state.
Within each stratum, between two and five facilities
were selected with probabilities proportionate to
size of facility.
In the interest of completing data collection
activities by April 2009, the size criterion for the
non-state facilities was increased to 150. This
eliminated 32 facilities from the original sample.
Of the 252 selected juvenile facilities:
- 26 were determined to be ineligible due to an
average length of stay of less than 90 days or
some other constraint that precluded obtaining
consent of parent or guardian,
- 18 had closed,
- 6 housed pre-adjudicated youth only or too few
adjudicated youth to permit interviewing,
- 2 had merged with another participating facility,
- 2 participated but yielded no usable interviews
from the sexual victimization survey.
Of the remaining 198 eligible juvenile facilities, 3
privately operated facilities refused to participate
in the survey:
- Glen Mills School, Glen Mills, PA
- Northwestern Academy, Coal Township, PA
- Gulf Coast Trade Center, New Waverly, TX
Selection of youth
Rosters of adjudicated youth were provided by
facilities granting in loco parentis (ILP) four weeks
prior to data collection and by facilities requiring
parental/guardian consent (PGC) eight weeks prior
to data collection. All youth were sampled in ILP
facilities and in PGC facilities that had 240 or
fewer youth on the roster. In PGC facilities that
exceeded 240 youth, an initial sample of 240 was
selected. Additionally, all females among those
not selected were included with certainty.
The initial sample was supplemented by youth
who were admitted to the facility between the 8th
and 4th weeks prior to data collection. In ILP
facilities and PGC facilities with at least 240
adjudicated youth, everyone was selected. In
PGC facilities with more than 240, incoming
youth were selected at the same rate as the initial
Prior to the start of data collection, interviewing
capacity at each facility was assessed based on
the number of available days, interviewing
rooms, and interviewers. In facilities in which
the NSYC team had the capacity to complete all
of the interviews, all youths for whom consent
had been given were selected. In other facilities,
youth were randomly sub-sampled so the number
of youth did not exceed interviewing capacity.
A total of 25,939 youth were selected. Among
these individuals, 7,175 left prior to the interviewing
team arriving at the facility. After
restricting the sample to those assigned to the
sexual assault interview, 54 percent of the youth
responded to the interview. Approximately 33 percent
of the youth did not participate because parental/guardian consent could not be obtained; 6 percent
refused to complete the interview; and 7 percent were
non-respondents for other reasons (e.g., they did
not complete the entire interview, they were not
at the facility at the time of visit, the facility
denied access, or they were excluded due to
extreme or inconsistent response patterns).
As a result of sampling and consent protocols,
youth who completed the NSYC were somewhat
older and had committed more serious offenses
than other youth in residential placement.
Nearly twice as many youth in the NSYC were
age 18 or older (26 percent), compared to adjudicated
and non-adjudicated youth who had been enumerated
in the 2006 CJRP (14 percent). Considerably
more youth in the NSYC had been placed
because of a violent offense (46 percent) than all youth
in residential placement (34 percent).
The survey data were weighted to provide facility-level and national-level estimates. To generate
facility estimates, an initial weight was
assigned to each youth corresponding to the
inverse of the probability of selection within each
facility. A series of adjustments were applied to
the initial weight to compensate for nonresponse.
These adjustments were completed in
1. Adjustment cells were constructed based on
the number of locked doors the youth had to
go through to leave the facility, offense, race/Hispanic origin, age, gender, and the number
of days the youth had been in facility.
2. An adjustment required a minimum nonresponse
cell size of 10 responding youth. In
many facilities, this resulted in no nonresponse
adjustment, as either the facility
had too few interviews (less than 20) to create
multiple cells or the differences between
respondents and non-respondents were not
significant. In facilities where significant differences
were observed, two to four non-response
cells were created.
3. After an initial non-response adjustment,
the weights within a facility were examined.
If the highest weight was four times greater
than the lowest weight in the facility, the
highest weights were trimmed and the difference
in weighted counts distributed to the
remaining youth, so that after trimming the
high-to-low ratio in the final weight would
be equal to four.
To generate national estimates, the facility
weights were adjusted to reflect each facility's
probability of selection into the sample and then
were adjusted for facility non-response. The
steps in creating the national weight adjustments
were the same as those described for facility-level
Mode of Data Collection:
audiovisual touch-screen computer-assisted self interview (AVT-CASI)
Calculating facility-level response rates
Within each facility a base weight was created for
each youth in the sexual victimization survey by
taking the inverse of each youth's probability of
selection. In most facilities youth selection probabilities
were the same. However, in some sampled
facilities (e.g. where females were oversampled
and where rosters contained duplicate
records) selection probabilities varied.
An initial facility response rate was calculated by
summing the base weights for all youth completing
the sexual victimization survey and dividing
it by the sum of the base weights for all sampled
youth (minus ineligible youth) in each facility.
A final response rate was calculated to account
for the deletion of interviews containing extreme
or inconsistent responses. This was achieved by multiplying the
initial facility response rate by an adjustment
ratio. In each facility this ratio represented the
sum of final weights for all interviewed youth
excluding those with extreme or inconsistent
responses divided by the sum of final weights for
all interviewed youth including those with
extreme or inconsistent responses. This final
adjusted response rate was then multiplied by
Calculations for Mt. Meigs Campus (Alabama)
illustrate the calculation of these weighted facility-level response rates. The facility listed 278
youth on its roster. Of these listed, 193 were subsampled.
Thirteen of the sampled youth were
roster errors (and were excluded from the sample).
Of the remaining 180 sampled youth, 32
were discharged prior to the visit, leaving 132
sampled for the sexual victimization survey and
16 for supplemental survey. Of the 132 eligible
youth, 121 completed the NSYC survey. After
adjusting for the probability of selection for each
youth, the 121 youth who completed the sexual
victimization survey represented 208 youth (or
91.7 percent of the 227 eligible youth in the facility).
Three of the youth provided extreme or three or
more inconsistent responses and were excluded.
After adjusting for the probability of selection
for each youth, a ratio adjustment of .974 was
applied to the initial response rate, resulting in
an overall facility response rate of 89.3 percent (.974
times .917 times 100 percent).
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of
disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major
statistical software formats as well as standard codebooks to accompany the data. In addition to
these procedures, ICPSR performed the following processing steps for this data collection:
Standardized missing values.
Checked for undocumented or out-of-range codes.