2 56 CHAPTER 2 studies was not significant (Q = 3.5, p = .48). The I2 (<0.00%) indicated that there was no variability among studies that could not be explained by chance. There was not enough information to run any moderator analysis. Table 2.4. Effect sizes research for question 3: Predictive validity of change scores corrected for static and initial dynamic scores (van den Berg et al., 2018, p.186) Recidivism Type Fixed-Effect Random-Effect Q I2 (%) N K L Hazard ratio 95% CI Hazard ratio 95% CI Sexual 0.91 [0.87, 0.95] 0.91 [0.87, 0.95] 2.4 <0.00 1,980 6 6 Violent 0.93 [0.90, 0.97] 0.93 [0.90, 0.97] 3.5 <0.00 4,168 5 5 Any 0.95 [0.93, 0.98] 0.95 [0.93, 0.98] 1.5 <0.00 1,172 3 3 K = Number of studies; L = Number of unique samples 2.4.4.3 Predictive validity of controlled change scores for any recidivism For the analysis of any recidivism the fixed-effect weights of the individual studies (the inverse of the variance) varied between 3,460.21 and 452.69 with a median value of 1,111.11. The study producing the largest weight had a sample size of 392 (recidivism rate of 21.2%) and more than twice the weight of the next largest study (1,111) and more than 7 times more weight than the smallest study weight. To reduce the influence of this study, its variance was increased artificially to ensure its weight did not exceed the next largest study weight by factor 2. The fixed-effect weighted hazard ratio on any recidivism for corrected change scores on dynamic risk assessment instruments was 0.95, 95% CI [0.93, 0.98], N = 1,172, based on 3 studies representing 3 unique samples. As such, changes on the dynamic risk factors added to the prediction of any recidivism above and beyond static and initial dynamic risk scores. The Q indicated that the variability across studies was not significant (Q = 1.5, p = .47). The I2 (<0.00%) indicated that there was no variability among studies that could not be explained by chance. Not enough information was available to run any moderator analysis. 2.5 DISCUSSION This meta-analysis is the first to examine the predictive properties of dynamic risk assessment instruments designed to assess recidivism risk in adult men with a history of sexual offenses. We evaluated their value in predicting sexual, violent (including sexual), and any (general) recidivism in terms of overall predictive properties, incremental predictive validity over static risk assessments, and the predictive validity of change
RkJQdWJsaXNoZXIy MjY0ODMw