Using a retrospective quasi-experimental design, this study evaluated the effectiveness of prison-based treatment by examining recidivism outcomes among 2,040 sex offenders released from Minnesota prisons between 1990 and 2003 (average follow-up period of 9.3 years).
To reduce observed selection bias, the authors used propensity score matching to create a comparison group of 1,020 untreated sex offenders who were not significantly different from the 1,020 treated offenders. In addition, intent-to-treat analyses and the Rosenbaum bounds method were used to test the sensitivity of the findings to treatment refuser and unobserved selection bias. Results from the Cox regression analyses revealed that participating in treatment significantly reduced the hazard ratio for rearrest by 27 percent for sexual recidivism, 18 percent for violent recidivism, and 12 percent for general recidivism. These findings are consistent with the growing body of research that supports the effectiveness of cognitivebehavioral treatment for sex offenders. 31 references (publisher abstract modified)