2 44 CHAPTER 2 Identification Screening Eligibility Included 854,033 records through database searches 42 records identified through other sources 854,019 records after removal of duplicates 264 records screened 116 records excluded 148 records assessed for eligibility 90 records excluded (failing to meet eligibility criteria) 6 records effect sizes could not be traced 52 records included in meta-analysis 46 records included for research question 1 22 records included for research question 2 12 records included for research question 3 Figure 2.1. Selection steps for meta-analysis (van den Berg et al., 2018, p.181). Included dynamic risk assessment instruments contain risk factors within at least one of the following domains (Thornton, 2013): sexual interests (e.g., sexualized coping, multiple paraphilias, sexual preoccupation, sexual interest in children), distorted attitudes (e.g., pro-offending attitudes, supportive attitudes towards child sexual abuse and/or rape), relational style (e.g., emotional congruence with children, lack of sustained marital type relations, callousness), or self-management (non-compliance with supervision, violation of conditional release, impulsivity/recklessness, poor coping). A complete listing of all static and dynamic instruments used in the included studies can be found in Appendix B, which is available in the online supplemental materials. 2.3.3.2 Overall predictive properties To evaluate the overall predictive properties of dynamic risk assessment instruments (Research Question 1), 46 papers were eligible for inclusion. Twentynine of these papers consisted of studies with (partly) overlapping samples. These overlapping
RkJQdWJsaXNoZXIy MjY0ODMw