2 157 REVIEW OF APPLIED & CASUAL GAMES FOR MENTAL HEALTH Figure A.6 Forest plot of standardised effect sizes of studies examining social skills in youth with autism or social skill challenges Beaumont et al. 2015 Beaumont et al. 2021 Einfeld et al. 2018 Griffin et al. 2021 Murphy et al. 2021 Beaumont & Sofronoff, 2008 Faja et al. 2021 Sanchez, Brown et al. 2017 Applied game vs Active condition Applied game vs Passive condition Cohen's d [95% CI] 0.19 [-0.29, 0.68] 1.00 [0.51, 1.50] 0.47 [-0.08, 1.02] -0.60 [-1.23, 0.04] -0.06 [-0.67, 0.56] 1.41 [0.79, 2.04] 0.19 [-0.31, 0.69] 1.02 [0.52, 1.52] -1.25 -1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 Favours active / passive condition Favours applied game Social skills Notes. The study of Einfeld et al. (2018) used cluster randomisation and a crossover design. The adjusted effect size was calculated based on an average cluster size of 5.60 and an ICC of 0.05 (i.e., design effect = 1.23), which gave a Cohen’s d of 0.47, with 95% CI [-0.14, 1.08]. The study of Beaumont et al. (2015) used cluster randomisation. The adjusted effect size was calculated based on an average cluster size of 4.06 and an ICC of 0.05 (i.e., design effect = 1.15), which gave a Cohen’s d of 0.19, with 95% CI [-0.33, 0.71]. The study of Murphy et al. (2021) used cluster randomisation as well. The adjusted effect size was calculated based on an average cluster size of 4.17 and an ICC of 0.05 (i.e., design effect = 1.16), which gave a Cohen’s d of -0.06, with 95% CI [-0.72, 0.61]. Faja et al. (2021) report a Cohen’s d of 0.19. CI = confidence interval, ICC = intracluster/intraclass correlation coefficient.
RkJQdWJsaXNoZXIy MTk4NDMw