NCJ Number
207968
Date Published
July 2004
Length
6 pages
Annotation
This study identified factors that predict whether Indigenous adult male violent and sexual offenders in Western Australia will reoffend violently or sexually and to use these findings to construct a risk-assessment instrument specific for this population.
Abstract
A retrospective analysis was conducted on the file data of 1,838 adult male Western Australian Indigenous offenders who had been identified as requiring either a violent or sexual offender program since 1987, including those who did not participate in or complete a program. The data for 525 offenders were used to identify possible predictors, and the data for 380 were used to develop the various instruments. The postrelease records of 258 offenders were examined to determine the accuracy of the risk-prediction instruments developed in the study. "Reoffending" was defined as "being found guilty in a court of a similar subsequent offense." Time at risk was calculated from time of release into the community or the beginning of a community order. Separate predictors were found for violent offenders who had no record of sex offenses and for sex offenders who were violent and nonviolent. Further, within the group of violent offenders a subgroup of family-violence offenders was identified. The study found that a risk assessment instrument developed specifically for Indigenous sex offenders can be more accurate than instruments developed overseas or for non-Indigenous populations in Australia; however, as more subgroups of sex offenders are distinguished in larger samples, instrument modifications will be necessary for each subgroup. The instrument developed to assess the risk of violent reoffending, including family violence, was disappointing, since the false positive rate was 55 percent. These findings suggest that instrument accuracy increases as subgroups within broader groups of offenders are more narrowly defined. 26 references