Sex offender risk assessment: A reexamination of the coffee can study

Abstract

There exists a valid concern among forensic psychology scholars that measures intended for sexual offenders have plateaued in their predictive accuracy. The current study examined this concern using the “coffee can” methodology of Kroner, Mills, and Reddon (2005). The Iowa Sex Offender Risk Assessment (ISORA8), Level of Service Inventory — Revised (LSI-R), and Static-99R were used to predict various methods of detected recidivism (general, sexual, and violent recidivism) as compared to three randomly generated measures and a weighted generated measure. Independent correlation and receiver operating characteristic comparisons found that in almost all cases, generated measures outperformed established measures. Analyses indicated that the current conceptualization of risk for sexual offenders has not been thoroughly integrated into established measures. Current measures appear incomplete, but have a clear and empirically-noted ability to improve.