Design, Sites, and Data Sources
The CHAMPS evaluation targeted individuals who were recently released from prison to parole supervision. The evaluation used a random assignment design to compare the efficacy of three approaches with parolees: (1) parole officers use standard supervision practices (the “business as usual” control group); (2) parole officers use core correctional practices with a focus on desistance (the Next Generation model); and (3) parole officers use the Next Generation model and parolees participate in Motivational Enhancement Therapy-Thinking for a Change (MET-T4C), a cognitive behavioral program for individuals formerly incarcerated offered by service providers.
The three sites participating in the evaluation were: Dallas, Texas; Des Moines, Iowa; and Denver, Colorado.
Outcome data for all study groups was collected from sample member interviews and administrative records. Administrative records data were collected from each of the state agencies to measure criminal history at baseline and key outcomes over time, including arrest, conviction, reincarceration, and parole violations. We collected data from each of the sites’ management information systems in order to measure treatment session attendance, employment, discharge status, parole contacts, drug testing, and other events and processes as available. The survey, administered up to one year after study enrollment, focused on key outcomes that cannot be measured with the administrative records, including receipt and dosage of reentry services, educational and employment services, and substance use and mental health services; housing stability and services; and other outcomes.