RESOURCE B: RISK ASSESSMENT

In this resource, I identify actuarial risk scale developers and explain how risk scales can be obtained. I have no financial interest in any of these measures. Many of the scales can be obtained online; Web sites frequently change, however, so it is best to contact the scale developers for current information. Unless otherwise noted, all of the links provided in this resource were verified on November 9, 2006. Sex offender risk assessment is a fast-moving field, and new studies, scale revisions, and clarifications regularly appear. Users of these risk scales should make sure that they have the up-to-date versions of the scales and scoring rules.

Readers interested in using the scales should refer to the original sources for scoring instructions. Scale items are included here to illustrate the item content and to demonstrate the relative weights associated with items. Risk assessment training is available from the scale developers or through professional workshops; workshop providers can be identified by contacting the Association for the Treatment of Sexual Abusers (http://www.atsa.com ).

Useful resources are available to those who would like to learn more about the complexities of sex offender risk assessment. These include a bibliography compiled by Dennis Doren on September 2004 and available from the Association for the Treatment of Sexual Abusers Web site (see Doren, 2004a); Doren’s (2002) book on sex offender evaluation; and chapter 7 from Quinsey, Harris, Rice, and Cormier’s (2006) book, Violent Offenders: Appraising and Managing Risk .

Many of the key sex offender risk assessment studies have been completed by Canadian researchers. The risk assessment work of the prolific Penetanguishene research group (Grant Harris, Zoe Hilton, Vern Quinsey, Marnie Rice, Catherine Cormier, Carol Lang, and Terry Chaplin) is summarized on their Web site (see Mental Health Centre Penetanguishene Research Department, n.d.), and the contributions made by Karl Hanson and his colleagues are available from the Web site of Public Safety Canada, a government ministry involved in criminal justice (http://www.publicsafety.gc.ca/res/cor/rep/cprmindex-en.asp ).

The items, scoring guidelines, and sample reports for the Violence Risk Appraisal Guide (VRAG) and Sex Offender Risk Appraisal Guide (SORAG) are presented in the appendixes of Quinsey et al. (2006). Though the VRAG and SORAG can be used without any fees, both scales require the scoring of the Psychopathy Checklist—Revised (PCL–R: Hare, 2003), which is commercially available through Pearson Assessments (http://www.pearsonassessments.com/tests/hare.htm ) or Multi-Health Systems (http://www.mhs.com ).

The Rapid Risk Assessment of Sexual Offense Recidivism (RRASOR) and Static-99 scoring manuals are available online. For information on the RRASOR, see Hanson (1997), and for the Static-99, see A. Harris, Phenix, Hanson, and Thornton (2003). A revised version of the scale, the Static-2002, is available for research purposes but is not yet recommended by the developers for clinical use (Hanson & Thornton, 2003).

I focused on the VRAG, SORAG, RRASOR, and Static-99 in my review of actuarial risk scales because they have consistently shown good predictive validities in peer-reviewed studies conducted by investigators other than the original scale developers, they are widely used in sex offender risk assessments in Canada and the United States, and they are free and relatively easy to use (though care must be taken to follow the scoring instructions and to clarify any unusual or complicated scoring situations). All of these scales can be scored by a trained rater and do not require an advanced mental health or criminal justice degree. The PCL–R has certain user requirements, however. Other sex offender risk scales are available, and two of these are mentioned here for interested readers. For information on the Minnesota Sex Offender Screening Tool—Revised, see Epperson, Kaul, et al. (2005), and for the Sexual Violence Risk—20, which is commercially available through Psychological Assessment Resources, Inc., see Boer, Hart, Kropp, and Webster (1997).

Hanson, Morton, and Harris (2003) conducted a quantitative review of the sex offender literature on actuarial risk scale validity and found that the RRASOR (17) had the most number of replications, followed by the Static-99 (15), SORAG (5), and VRAG (5). The Static-99 had the highest average predictive accuracy for sexual recidivism (average area under the curve [AUC] = .76), followed by the SORAG (average AUC = .68), RRASOR (average AUC = .66), and VRAG (average AUC = .64). The confidence intervals around these average AUCs indicate there were no significant differences across the four scales in terms of their predictive validity for sexual recidivism.