267 Supplementary material | Appendix A A Table 5. Results of publication bias analyses. Outcome category Studies (s) Effect sizes (k) Funnel Plot test Egger regression test Trim-and-fill analyses k imputed left k imputed right k imputed left > cutoff Social network 20 164 z =-0.475; p = 0.635 z = 2.869; p = 0.004** 0 0 no General functioning 14 32 z =-0.392; p = 0.695 z = 1.880; p = 0.060 0 0 no Mental health treatment adherence 18 70 z = 1.741; p = 0.082 z =-0.228; p = 0.819 0 6 no Psychiatric functioning 20 80 z =-1.375; p = 0.169 z = 0.304; p = 0.761 0 1 no Psychological functioning 16 59 z = 1.056; p = 0.291 z =-0.339; p = 0.735 0 0 no Substance use 9 120 z =-0.136; p = 0.892 z = 0.235; p = 0.814 9 1 yes Criminal behavior 6 28 z = 0.890; p = 0.374 z = 0.750; p = 0.453 0 0 no Note: s = number of studies. k = number of effect sizes. Imputed left = number of effect sizes imputed on the left side of the funnel plot. Imputed right = number of effect sizes imputed on the right side of the funnel plot. k imputed left > cutoff = cutoffs for the estimator of the Trim and Fill method as proposed by Fernández-Castilla et al. (2021b). *p < 0.05. **p < 0.01. ***p < 0.001.
RkJQdWJsaXNoZXIy MjY0ODMw