Planning to Evaluate an RSAT Program? What Are Some Challenges?

Randomized controlled trials are frequently used to assess the effects of RSAT programs.

RSAT evaluation studies using experimental designs have provided the strongest evidence of program effectiveness. Programs using the experimental method have randomly assigned offenders to receive RSAT programming and compared their outcomes with those of offenders randomly assigned to receive no treatment or some other form of treatment. Use of random assignment to treatment enables evaluators to attribute differences in recidivism and other outcomes to the RSAT program. When random assignment has not been used, studies have used quasi-experimental designs, such as comparing a group of volunteers who enrolled in treatment to a no-treatment comparison group of eligible volunteers who never attended treatment or offenders receiving other types of treatment.

Selection bias issues can be challenging for evaluators.

RSAT evaluations unable to use random assignment must establish that the treatment and control groups are equivalent. The equivalence of groups can be compromised when participation is voluntary, when treatment providers determine who is eligible for services, or when offenders drop out or are not located for follow-up. Dropout is a particular problem for evaluations of RSAT programs since the programs are designed as long-term interventions. Evaluations of RSAT programs, especially voluntary programs, must insure that selection bias issues are controlled for when determining the impact of the program(s). When evaluators have been unable to randomly assign offenders to treatment, they have compensated by using statistical techniques to control for differences between the groups. (For example, Pelissier et al. (n.d.) in Publications)

Program design and implementation information must be collected.

Information on how an RSAT program was designed and actually carried out allows the evaluator to determine how these factors affected program outcomes. RSAT programs should address implementation issues such as the role of referral systems, aftercare services, and segregation of RSAT offenders from the general prison population.

| | |
| | |