Table 3|.
Intervention group comparisons on pre-registered secondary and exploratory outcome measures.
| Measure (sample size) | Pre-training |
Post-training |
Active vs. Control post- vs. Pre-training t-test |
Repeated measures ANOVA |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Active |
Control |
Active |
Control |
Group |
Time |
Group × Time |
|||||||||||||
| m | sd | m | sd | m | sd | m | sd | diff | se | p | 95% CI | d | F | p | F | p | F | p | |
| BDI (PRE:28, POST:24) | 4.7 | 4.4 | 8.1 | 8.1 | 4.7 | 4.2 | 7.2 | 7.5 | 0.58 | 1.05 | 0.59 | −1.60 to 2.76 | 0.23 | 1.20 | 0.29 | 2.27 | 0.15 | 0.31 | 0.59 |
| IGQ-Self-hate (PRE:24, POST:21) | 27.4 | 4.9 | 30.9 | 7.4 | 27.9 | 4.8 | 29.9 | 6.9 | 0.43 | 2.17 | 0.79 | −4.12 to 4.98 | 0.09 | 0.83 | 0.37 | 0.22 | 0.64 | 0.04 | 0.85 |
| Rosenberg Self-esteem (PRE:27, POST:24) | 22.1 | 4.0 | 22.4 | 3.8 | 24.2 | 2.9 | 21.8 | 5.7 | 3.34 | 1.44 | 0.03* | 0.35 to 6.33 | 0.98+ | 0.70 | 0.41 | 0.65 | 0.43 | 5.40 | 0.03* |
| Negative Affect (PANAS) (PRE:28, POST:28) | 3.8 | 4.0 | 3.9 | 4.6 | 4.1 | 4.2 | 3.5 | 5.1 | 0.64 | 1.52 | 0.68 | −2.49 to 3.77 | 0.16 | 0.03 | 0.87 | 0.002 | 0.96 | 0.18 | 0.68 |
| Positive Affect (PANAS) (PRE:28, POST:28) | 25.0 | 7.9 | 23.1 | 9.2 | 18.6 | 8.7 | 17.9 | 10.7 | −1.29 | 2.77 | 0.65 | −6.99 to 4.42 | 0.18 | 0.16 | 0.69 | 17.39 | 0.0003* | 0.21 | 0.65 |
| Guilt vs. Indignation intensity ratings (PRE:28, POST:28) | −0.07 | 1.33 | 0.07 | 1.33 | −0.50 | 0.65 | 0.07 | 1.33 | −0.43 | 0.40 | 0.30 | −1.26 to 0.40 | 0.40 | 0.78 | 0.39 | 1.14 | 0.30 | 1.14 | 0.30 |
*Significant at p = .05, 2-sided.Between-group Cohen's d scores were computed from the means and pooled standard deviations (Cohen, 1988) of the differences between Post vs. Pre-training. Mean differences and 95% confidence intervals were taken from independent samples t-test for differences between Post vs. Pre-training. CI = confidence interval, m = mean, sd = standard deviation, se = standard error, diff = difference of means, d = Cohen's d. +After excluding one outlier in the active group (outside of ±2.5 standard deviations from the mean), the intervention's effect on self-esteem remained large: d = 0.91. The Positive and Negative Affect Schedule(Watson et al., 1988b) was used to explore affective changes of how participants felt at the moment. Ratings of guilt and indignation-evoking events were obtained prior and after the scan. In the scanner, after each run the participants were asked over the intercom to rate on a scale from 0 to 9 how much they felt guilt or indignation. Analyses of Covariance on post-intervention outcomes with pre-intervention measures as covariates confirmed the repeated measures ANOVA results (Supplementary Table 1).