2 42 CHAPTER 2 which included a mix of static and dynamic variables. Following this, Hanson and Harris (2000a) conducted a retrospective study with the specific goal of identifying dynamic factors that are associated with recidivism. For this purpose, they interviewed community supervision officers and reviewed the files of 208 men with a history of sexual offenses who recidivated and 201 who desist. The identified dynamic risk factors were subsequently tested in prospective research (Hanson et al., 2007). This work forms the basis for most dynamic risk assessment instruments that are currently being used in the evaluation of adult men with a history of sexual offenses. 2.2 CURRENT STUDY Recently, the number of studies on the predictive properties of dynamic risk assessment instruments, developed for use in adult men with a history of sexual offenses, has increased substantially. The current meta-analysis is the first to our knowledge focusing on the predictive properties of such risk assessment instruments and, as a secondary goal, examines whether dynamic risk assessment instruments have predictive validity above and beyond that of static risk assessment instruments. As a step toward elucidating possible causal links between dynamic risk factors and reoffending behavior, our third goal is to explore to what degree changes in dynamic risk factors over time are associated with recidivism risk. Summarized, the current meta-analysis will address the following three questions: 1) What are the predictive properties of dynamic risk assessment instruments developed for adult men with a history of sexual offenses? 2) What is the incremental predictive validity of these dynamic risk assessment instruments above and beyond that of static risk assessment instruments? 3) What is the predictive validity of change scores on these dynamic risk assessment instruments? 2.3 METHOD 2.3.1 SEARCH STRATEGY A literature search was performed on empirical papers, published up to the end of March 2014, that present the findings of studies on the predictive properties of dynamic risk assessment instruments (or risk assessment instruments containing a section measuring dynamic risk factors) developed for adult men with a history of sexual offeses. The search was conducted in PubMed, PsycINFO, Pro-Quest, SciVerse, and Web of Science databases using combinations of the following search terms: sexual offending behavio(u) r (e.g., child molesting, exhibitionism), predictive validity (e.g., accuracy, area under the curve [AUC]-value), recidivism (e.g., relapse, reoffense), and dynamic risk assessment
RkJQdWJsaXNoZXIy MjY0ODMw