Date of Conferral







Sharon Xuereb


Recidivism within the sexually violent predator (SVP) population has gained worldwide attention because of the lack of protection offered to the victims that may lead to loss of life. Behavioral theory suggests that accuracy of predictive behaviors based on empirical judgement is more reliable than that based on clinical judgement. The purpose of this research was to see whether three actuarial assessment tools, Static-99, PCL-R, and MnSOST-R, could predict recidivism and whether the combination of the three-increased predictive value in the Texas SVP population. As yet, the literature provides no evidence. The Texas Open Record System provided assessment scores and violations of 90 SVPs committed during fiscal years 2009-2013. Texas had 58.9% violated commitment laws within the SVP population of the civil commitment program. The scores on these three assessment tools were analyzed along with the violations using bivariate logistic regression. According to the results, Static-99, PCL-R, and MnSOST-R can, in combination, predict recidivism better than any tool by itself in the Texas SVP. However, individually, only the PCL-R approached significance as a predictor. This study could lead to positive social change in both the targeted treatment of labeled SVP and in the accuracy of predicting recidivism among SVPs. Therapists should use the three actuarial assessment tools when developing treatment plans, intervention techniques, and when adjusting supervision requirements to assist in both targeted treatment and to reduce the number of victims.