ABSTRACT
Many studies have validated the psychometric properties of the Static-99, the most widely used measure of sexual offender recidivism risk. However much of this research relied on instrument coding completed by well-trained researchers. This study is the first to examine the interrater reliability (IRR) of the Static-99 between practitioners in the field and researchers. Using archival data from a sample of 1,973 formerly incarcerated sex offenders, field raters' scores on the Static-99 were compared with those of researchers. Overall, clinicians and researchers had excellent IRR on Static-99 total scores, with IRR coefficients ranging from "substantial" to "outstanding" for the individual 10 items of the scale. The most common causes of discrepancies were coding manual errors, followed by item subjectivity, inaccurate item scoring, and calculation errors. These results offer important data with regard to the frequency and perceived nature of scoring errors.
