The purpose of the study it to examine the issue of dosage and recidivism in therapeutically-oriented corrections for the juveniles in the state of Washington. The study aimed to assess whether time in placement was associated with the acquisition of social-emotional skills and subsequent felony recidivism.
Youth in the study were housed in one of four secure juvenile institutions or in one of two community-based group homes managed by Juvenile Justice Rehabilitation Administration (JJRA). In order to determine the effects that therapeutically-oriented corrections for the juveniles in the state of Washington have on recidivism rates for incarcerated juvenile offenders, data was collected from three different sources:
- The R-PACT was conducted within 30-45 days of intake for youth committed to a JJRA facility. The R-PACTs were completed by 10 different raters. Raters were not nested within facility or youth; youth often had different raters for admission versus release assessments.
- The JJRA administrative database maintains information on admission, release and placement changes while a youth is incarcerated.
- The Administrative Office of the Courts (AOC) contact database is the state data management system for all court contacts (juvenile and adult) across jurisdictions in Washington State and was used to measure recidivism post-release.
Additionally, a replication study was conducted to determine whether the relationships length of stay, skill acquisition and youth predictors of acquisition were reliable over multiple cohorts of youth. Recidivism was not measured because not enough time had elapsed to measure 12 month offending post release. The second cohort included youths admitted between April 2013 and February 2015 who had been released by May 2015.
The sample for this study includes one cohort of youth who were admitted to a Juvenile Justice Rehabilitation Administration (JJRA) in the state of Washington and were given an R-PACT within 30-45 days of the youth's admission. Youth admission dates ranged from December 5, 2008 through May 29, 2013 and 30-45 days of release. Youth release dates ranged from February 12, 2009 through August 1, 2013 (n=637).
A second cohort of youth were also given the R-PACT in a replication study. They were youths admitted between April 2013 and February 2015 who had been released by May 2015.(n=397)
The study used a convenience sampling method. The majority of youth were male (84 percent) with an average age of 16 years old. The racial/ethnic distribution included 38 percent White, 27 percent African American, 16 percent Hispanic, 3 percent Asian, 3 percent Native American, 2 percent Other, and 11 percent Mixed youth. Approximately 98 percent of youth identified English as their primary language.
Longitudinal: Cohort / Event-based
Youth aged 10-19 in juvenile offender facilities in Washington State 2008-2015.
The Residential Positive Achievement Change Tool (R-PACT).
Washington State Administrative Office of the Courts (AOC) court contact database.
Juvenile Justice and Rehabilitation Administration's administrative database.
administrative records data
cognitive assessment test
paper and pencil interview (PAPI)
There are two data sets in this study. The first data set is derived from data collected on cohort 1 (Integrated Treatment Assessment_Cohort 1 Data.sav). It contains 239 variables and 637 cases (n=637) and is comprised of data from three sources:
- The R-PACT, contains 12 categories related to criminal and social risk/protective factors: criminal history, school, use of free time, employment, relationships, family, living arrangements, alcohol and drugs, mental health, attitudes/beliefs, aggression, and social skills.
- JJRA Administration Database helped create variables based on length of stay of the incarcerated youth. This was calculated as the number of months from admission to release and then further categorized into dosage categories. Five dose categories in increments of three months were created to assess change in behavior and skill acquisition over time.
- Washington State Administrative Office of the Courts (AOC) records captured recidivism rates of Juvenile offenders. Data was requested for youth 12 months post release from JJRA's custody, and includes charges from juvenile and adult systems. Recidivism was operationalized as a count of all felony charges (violent and nonviolent) a youth obtained up to 12 months post release.
Propensity score variables were also created. Ten variables were significantly associated with the amount of time youth were in custody in order to balance youth histories and characteristics across dosage categories.
Data set 2 (Integrated Treatment Assessment_Cohort 2 Data.sav) contains 191 variables and 397 cases (n=397). It was designed as a replication study and used the same measurement tools as cohort 1. The main difference being that recidivism was not measured because not enough time had elapsed to measure 12 month offending post release.
Additionally, both datasets contain SPSS syntax files which create scales, creates propensity scores, creates recidivism data, creates length of stay indicators, and runs models.
Residential Positive Achievement Change Tool (R-PACT)
Integrated Treatment Assessment