Predictive validity of the Static-99R and its dimensions

The aims of the current paper were to identify dimensions of the Static-99R, the most commonly used sex-offender actuarial scale, and to test their predictive validity. Exploratory factor analysis extracted three dimensions from the scale: sexual criminality, general criminality and detachment. Cox regression analyses revealed that different actuarial dimensions predicted different recidivism outcomes; sexual criminality predicted only sexual recidivism, while general criminality predicted only non-sexual violent recidivism and non-sexual non-violent recidivism. Detachment had an inconsistent predictive validity. Receiver operating characteristic curve analyses indicated that the latest Static-99 revision, the Static-99R, outperformed its predecessor in virtually every way. These same analyses revealed that the Static-99R was as effective with sexual aggressors of women as with sexual aggressors of children.