RESUMO
Dynamic risk variables (or criminogenic needs; Andrews & Bonta, 2003) increase risk for criminal behavior and are conceptualized as changeable through intervention. Yet this assumption of changeability has been examined in only a few studies, and none of these studies have examined whether the measurement properties the risk assessments measure remain invariant over time--a necessary precursor to determining if actual change is occurring or if changes in measurement are producing differences. This study examines the dynamic needs index (DNI) and protective strengths index (PSI) of the inventory of offender risk, needs, and strengths (Miller, 2006b) pre- and post-treatment. Findings suggest the measurement properties of the DNI are acceptably invariant over time, although there is evidence that the intercept of the alcohol/drug Problems scale is higher after (opposed to before) treatment and the intercept of the Intra/Interpersonal Problems scale is higher before treatment. Subsequent latent difference score models suggest--as expected--that the dynamic needs latent variable decreased and the protective strengths latent variable increased through treatment. This study presents first evidence for invariance of measurement properties of a risk assessment measure at different points in treatment.
