Key points are not available for this paper at this time.
The Staff Training at Reducing Rearrest (STARR) model is a training program designed to help federal probation officers incorporate effective rehabilitative practices into their client interactions. While previous research demonstrates that STARR training is associated with reductions in client recidivism, it remains unknown what level of service (or “dosage”) of STARR is necessary for optimizing its treatment effectiveness. This quasi-experimental study uses data collected from a federal probation office in a Southern state to conduct a dose-response analysis. Using a propensity score weighting for continuous treatments research design with double-robust framework, this study estimated the effects of the number of STARR skills received on four measures of recidivism. Study results indicated that greater doses of STARR skills were associated with lower rates of recidivism for higher risk clients, while the impacts were null (or iatrogenic) for low-risk clients. The policy and research implications of these findings are discussed.
Labrecque et al. (Sat,) studied this question.