Criterion and Predictive Validity of the Static-99 for Adult Males Convicted of Sexual Offences Against Children

Abstract

Official demographic and offence history data and confidential self-report data obtained from adult m ales convicted of sexual offences against children were used to produce actuarial recidivism risk prediction (Stati c-99) scores, four sexual deviance and general criminological pre dictor variables, and four (official and unofficial) offen ce-related outcome variables. Static-99 scores applicable at t he time of the current sexual offence conviction (n = 172) pre dicted sexual and nonsexual offending outcomes. Regression models were improved with the addition of sexual deviance and criminological predictors not included in the Stati c-99. Static- 99 scores applicable at the time of the first sexua l offence conviction for known recidivist sexual offenders (n = 40) predicted official nonsexual offence convictions an d self- reported number of sexual offence victims, but not official sexual offence convictions or self-reported total p eriod of sexual offending. Again, regression models were imp roved with the addition of other sexual deviance and crim inological predictors. Results provide partial support for crit erion and predictive validity of the Static-99. Applied risk assessments may benefit from consideration of key self-report a nd official data not included in the Static-9