The field validity of Static-99/R sex offender risk assessment tool in California.

Abstract

Policies that differentially apply to sexual offenders at different risk levels require defensible procedures for classifying offenders into risk categories. The current study examines the reliability and validity of Static-99 and Static-99R sexual offender risk assessment tools as implemented in the State of California. California is a valuable case study because it is a large jurisdiction that has devoted considerable resources to the implementation of risk tools. In Study 1, 55 corrections and probation officers scored Static-99R for 14 common cases. Overall, rater reliability was acceptable (intraclass correlation [ICC] .78), with higher reliability found for experienced scorers (ICC .85) than less experienced scorers (ICC .71). In the second study, the predictive ability of Static-99 and Static-99R was examined in a prospective study of 475 randomly selected adult males released in 2006 –2007 and followed for 5 years. Static-99/R scores were strongly related to subsequent sexual recidivism (areas under the curve .80). As well, there was acceptable fit between the expected and observed recidivism rates. These results suggest that it is possible to implement empirically derived risk assessment tools in large jurisdictions without degrading predictive accuracy.