| This study examined the function of dynamic risk assessment in the prediction of female offenders' post release adjustment in the community. The predictive validity of seven dynamic risk factors from the Community Intervention Scale were assessed with 497 federally sentenced women released into the community. The dynamic risk predictors were assessed four times at 6-month intervals in the community. A within-subjects repeated measure design was performed to determine if authentic dynamic characteristics could be attributed to the predictor variables in the study. Cox regression survival analyses with time dependent covariates and Receiver Operator Characteristics were then utilized to determine the ability of the seven needs variables to predict conditional release failure.;Overall, the results were positive and several of the asserted predictions were validated. Six of the seven domains, excluding substance abuse, demonstrated genuine change throughout the study period, and all seven variables were significantly related to outcome measures. The employment and associates variables were empirically deemed the strongest predictors of failure, while the remaining variables displayed moderate to weak predictive relationships with recidivism and revocations. Finally, antisocial associates, attitudes and substance abuse all predicted violent recidivism. Theoretical, practical and statistical implications of this work are discussed. |
