Table 3.
Overall effect sizes and results for the categorical moderator analysis.
| Category | All Measures | Compassion | ||||||
|---|---|---|---|---|---|---|---|---|
| ES | 95% CI | Fail-Safe | Heterogeneity | ES | 95% CI | Fail-Safe | Heterogeneity | |
| Main effect | 0.26 | 0.18, 0.34 | 396 | Q 25.71 | 0.36 | 0.25, 0.48 | 181 | Q 8.96 |
| Prosocial Type | ||||||||
| Compassion | 0.37 | 0.24, 0.49 | 580 | Qb 15.32 | N/A | |||
| Connectedness | 0.22 | −0.05, 0.49 | Qw 21.38 | |||||
| Empathy | 0.44 | 0.03, 0.84 | ||||||
| Aggression | 0.11 | −0.17, 0.48 | ||||||
| Prejudice | 0.11 | −0.09, 0.30 | ||||||
| Measure Type | ||||||||
| Questionnaire | 0.24 | 0.13, 0.35 | 387 | Qb 0.40 | 0.43 | 0.27, 0.59 | 221 | Qb 3.004 |
| Behavioural | 0.29 | 0.15, 0.43 | Qw 24.78 | 0.27 | 0.05, 0.49 | Qw 7.96 | ||
| Control Group | ||||||||
| Active | 0.27 | 0.12, 0.43 | 380 | Qb 0.08 | 0.37 | −0.04, 0.78 | 160 | Qb 0.006 |
| Passive | 0.25 | 0.14, 0.36 | Qw 24.67 | 0.36 | 0.20, 0.79 | Qw 7.92 | ||
| Teacher | ||||||||
| External/Audio | 0.22 | 0.09, 0.35 | 394 | Qb 0.97 | 0.26 | −0.07, 0.59 | 214 | Qb 2.65 |
| Author | 0.30 | 0.18, 0.41 | Qw 24.64 | 0.42 | 0.27, 0.57 | Qw 7.97 | ||
| Intervention Duration | ||||||||
| One session | 0.28 | 0.07, 0.48 | 379 | Qb 0.06 | Fewer than two cases per category available | |||
| Multiple sessions | 0.25 | 0.16, 0.35 | Qw 24.67 | |||||
Note: Significant results are highlighted in bold; Mean effect size (ES) reported as Pearson’s r with their corresponding 95% confidence interval (CI); Rosenthal’s fail-safe number, and Q heterogeneity statistic; Qb heterogeneity statistics for between-group and Qw for within-group effect size differences. The analysis contained 26 independent samples, exceeding the N of included studies because more than one outcome measure (e.g. compassion and empathy) was used in most studies.