Practice Profile: Juvenile Intensive Supervision Programs

Sarver, Molloy, and Butters (2012) conducted a meta-analysis to examine the extent to which intensive supervision programs (ISPs) decrease recidivism rates. A systematic review was conducted to identify eligible studies published in English between 1987 and 2011. Studies were eligible if 1) both the treatment and comparison groups consisted of juveniles (ages 12 to 22 or processed through the juvenile justice system), where the treatment group was on some form of intensive supervision (either probation or parole), and the comparison group was on some form of traditional supervision (e.g., probation, parole, or residential placement); 2) they evaluated a juvenile justice intensive supervision intervention (defined as increased surveillance that could include smaller caseloads for probation/parole officers, more frequent contact with those who committed the offense, home confinement, day reporting centers, and electronic monitoring); 3) they included a measure of recidivism (which could be arrest, conviction, return to residential placement, supervision failure, or other measures of delinquency); 4) the design was experimental or quasi-experimental (using a matching or statistical method to demonstrate equivalence between treatment and comparison groups). The systematic review included searches of databases and portals, including EBSCO, Criminal Justice Abstracts, ERIC, National Criminal Justice Reference Service, and others.

A total of 16 studies, which included 19 independent comparisons, were identified. Across the 16 studies, the total sample size was 4,511 juveniles in the treatment group and 3,456 juveniles in the comparison group. Of these 19 comparisons,13 were from randomized controlled trials, 1 was from a high-quality quasi-experimental design, and 5 were from quasi-experimental studies that used testing or matching. Further, of the 19 independent comparisons, 13 looked at the differences between ISPs and regular supervision, 5 looked at the differences between ISPs and secure placements, and 1 looked at the different forms of ISPs (the CrimeSolutions review of this meta-analysis focused on the comparisons between ISPs and regular supervision). Four of the comparisons were from evaluations of juveniles on parole, and the remaining 15 comparisons were from evaluations of juveniles on probation. In terms of study settings, 17 comparisons were from studies conducted in the United States, and 2 were from studies conducted outside the United States. Information on participants’ race/ethnicity and was not provided.

The data were coded and transformed into odds ratios, with values above 1 indicating a negative treatment effect (i.e., an increase in recidivism for the ISP treatment group) and values below 1 indicating a positive treatment effect (i.e., a decrease in recidivism for the ISP treatment group). A random effects model was used to analyze the data.

Bouchard and Wong (2018) conducted a meta-analysis to examine the effects of ISPs and aftercare programs on juvenile recidivism (the CrimeSolutions review of this meta-analysis focused on the examination of ISPs). A systematic review was conducted to identify eligible studies for inclusion. To be included, studies had to 1) examine ISPs that targeted juveniles between the ages of 12 and 18; 2) be delivered (at least partially) in a non-closed setting in the community; 3) be published in English and between January 1, 1990, and April 21, 2015; 4) use rigorous or moderately rigorous control group designs (such as randomized controlled trials and quasi-experimental designs in which participants were matched on at least one variable); 5) report on at least one individual-level measure of crime; 6) provide sufficient numerical or graphical data to allow for calculation of an effect size; 7) include a minimum sample of 20 subjects in both the treatment and control groups; and 8) be conducted in Australia, Canada, New Zealand, the United States, or a Western European country. Studies were excluded if the primary intervention was traditional supervision (i.e., standard probation, or if the program targeted a specific group, including those with serious mental health problems, substance users, people committed of sex offenses, known gang members, or status offenders. A systematic search was conducted across 20 electronic databases and by hand-searching the grey literature.

The search resulted in the inclusion of 26 individual program sites from 15 studies, which contributed 31 effect sizes. Recidivism was measured as either alleged offenses or convicted offenses; 12 sites reported on alleged offenses, and 19 sites reported on convicted offenses (the CrimeSolutions review of this meta-analysis focused on the outcome of convicted offenses from 19 individual program sites). Of the 19 program sites, 15 sites came from randomized controlled trial studies, 3 sites came from quasi-experimental design (matched comparison) studies, and 1 site came from a quasi-experimental design (weakly matched). Seven sites included a mix of male and female juvenile participants (12 sites did not report that information). With regard to race/ethnicity, five sites comprised white or mixed-race participants, and two sites comprised predominately minority participants (12 sites did not report that information). Sixteen sites were in the United States, and three sites were in England.

Because most of the included studies used dichotomous data to report outcome measures, effect sizes were calculated as odds ratios. The standardized mean difference was used to calculate effect sizes for continuous data. Due to small sample sizes, the primary analyses focused on findings from the fixed effects models.

The Washington State Institute for Public Policy (2019) conducted a meta-analysis of evaluations that compared crime outcomes for adjudicated youth on intensive supervision with outcomes for adjudicated youth on traditional probation supervision. Evaluations were excluded if they compared intensive supervision for youth placed directly on supervision with confined youth or compared intensive supervision for youth released from confinement with youth released from confinement and placed on traditional supervision.

Studies were identified from four primary sources: 1) bibliographies of systematic and narrative reviews of the research literature in various topic areas, 2) citations in individual studies that were located, 3) independent literature searches of research databases (using search engines such as Google, ProQuest, EBSCO, ERIC, PubMed, and SAGE), and 4) contacting authors of primary research to learn about ongoing or unpublished evaluation work. To be included, evaluation studies had to have a control or comparison group (i.e., study designs used random assignment or were quasi-experiments). Studies were not included if the treatment group was composed solely of program completers. To be included in the meta-analysis, studies also had to include enough information to calculate the effect size.

A total of 11 studies, providing 18 effect sizes, were identified and included in the meta-analysis. Of the 11 studies, 7 were randomized controlled trials, and 4 were quasi-experimental designs. Participants in the included studies were at high risk of recidivism (based on categorization from a validated risk assessment tool). In studies that reported demographic information, 60 percent of participants were youth of color, 12 percent were female, and the average age was 16 years (no additional information on race/ethnicity was provided). The length of intensive supervision ranged from 3 to 14 months (the average was 9 months). The studies were conducted in various regions across the United States, including Washington State, California, and Maryland; one study was conducted in the United Kingdom.

The calculated effect size was the standardized mean difference for continuous data. A random effects model was used to analyze the data.