Study 1
Rossman and colleagues (1999) conducted a randomized controlled trial to measure the effectiveness of the Opportunity to Succeed (OPTS). The study was conducted at three of the original demonstration sites: St Louis, Mo.; Kansas City, Mo.; and, Tampa, Fla. The target population for all three sites were individuals who had been convicted of felonies and who had alcohol- and drug-related histories, although the methods of recruitment and selection differed by site. At all three sites, individuals convicted of violent offenses were initially excluded from participation, but selection of individuals who were convicted of violent offenses was ultimately left to the sites. Identification of potential participants and eligibility occurred either prior to release from jail or from prison treatment programs, or upon return to the community depending on each site’s need to reach the desired number of participants.
At the Tampa site, the primary target population comprised individuals who were either on probation or parole supervision released from the Hillsborough County Jail Substance Abuse Program. The target population was then expanded partway through the study, due to the low numbers of participants, to include probationers and parolees from 1) Drug Abuse Comprehensive Coordinating Office’s (DACCO) residential drug treatment facilities, or 2) the Crossroads facilities for men and women.
The Kansas City and St. Louis sites targeted probationers and parolees from the Missouri state prison system incarcerated in Institutional Treatment Centers (ITCs), specifically, the Farmington ITC. St. Louis expanded its target population to include the Cramer facility, which included women. Kansas City expanded its sample to include the St. Joseph’s ITC, which was closer than Farmington.
Overall, the 343 study participants were 85 percent male, and approximately 75 percent were Black. There were 95 participants at the Tampa site, 159 at the St. Louis site, and 89 at the Kansas City site. The treatment group consisted of 175 participants, and the comparison group included 168 participants. Participants were randomly assigned by the researchers to the treatment or comparison group. The treatment group received the five case-management core services offered by the OPTS program. The comparison group received none of the enhanced core services of the OPTS program.
Data for recidivism was gathered through official requests from the central records of the Missouri Department of Probation and Parole in Kansas City for both Missouri sites. Recidivism data for the Tampa participants was available from the Department of Corrections central office in Tallahassee and the Tampa Central Field Office. Substance use relapse and urinalysis testing was conducted by offsite laboratories. Follow-up interviews were conducted roughly 1 year after individuals began participating in the intervention or, for comparison group participants, 1 year after they began their routine probation/parole sentences.
Simple linear regression modeling was used to determine the relationship between outcome behaviors and the main effects of substance use, crime, and employment. Specifically, ordinary least squares (OLS) was used to estimate the interval-level dependent measures, while OLS and logistic regression were used to estimate the dichotomous dependent measures.