A multimodel meta-analysis assessing moderators of sexual recidivism as an indicator of treatment effectiveness in persons with sexual offense histories

Abstract

Abstract Objectives This meta-analysis tested whether multimodel inference provides more conclusive evidence than traditional single-hypothesis testing regarding predictors that moderate sexual recidivism as an indicator of treatment effectiveness in persons with sexual offense histories. Methods A dataset including 35 studies equivalent to the meta-analysis by Holper et al. (Sex Abuse 2023; 0: 1–37) was used. Multimodel inference based on information theory tested 15 publication-, study-, treatment-, and individual-specific moderators. Results Only risk level was related to sexual recidivism. A greater posttreatment reduction in sexual recidivism was apparent in high- and medium- compared to low-risk individuals. This moderator explained 77% of the residual heterogeneity. Conclusions Compared to previous reports, the multimodel approach provided clearer evidence on which factors moderate sexual recidivism. Results corroborated the relevance of risk level, which relates to the risk-need-responsivity model. The findings may support treatment recommendations in persons with sexual offense histories in the criminal justice system.