Table 3.
Associations between quality criteria and effect sizes when compared to passive control conditions at posttreatment
| k | Hedges' g | I2 | p | ||
|---|---|---|---|---|---|
| Trials not meeting criterion | Trials meeting criterion | ||||
| Q1: Diagnosis | 77 | 0.75 | 1.15 | 80.71*** | 0.074 |
| (outlier-adjusted) | 73 | 0.75 | 1.04 | 73.45*** | 0.125 |
| Q2: Manual | 77 | 1.68 | 1.04 | 80.84*** | 0.046* |
| (outlier-adjusted) | 73 | 1.33 | 0.97 | 74.02*** | 0.224 |
| Q3: Training | 77 | 0.88 | 1.14 | 80.95*** | 0.193 |
| (outlier-adjusted) | 73 | 0.74 | 1.05 | 73.07*** | 0.075 |
| Q4: Integrity | 77 | 0.89 | 1.14 | 80.97*** | 0.196 |
| (outlier-adjusted) | 73 | 0.76 | 1.05 | 73.14*** | 0.078 |
| Q5: ITT | 77 | 1.10 | 1.08 | 81.54*** | 0.904 |
| (outlier-adjusted) | 73 | 0.98 | 1.00 | 74.27*** | 0.857 |
| Q6: N ⩾ 50 | 77 | 1.21 | 0.94 | 80.99*** | 0.095 |
| (outlier-adjusted) | 73 | 1.05 | 0.94 | 74.07*** | 0.419 |
| Q7: Randomization | 77 | 1.20 | 1.02 | 81.13*** | 0.285 |
| (outlier-adjusted) | 73 | 1.07 | 0.95 | 73.89*** | 0.406 |
| Q8: Blinding | 77 | 1.14 | 1.08 | 81.47*** | 0.791 |
| (outlier-adjusted) | 73 | 0.98 | 1.00 | 74.20** | 0.915 |
| Quality sum score (outlier-adjusted) | k | intercept | b | I2 | p |
| 77 | 1.15 | −0.01 | 81.52*** | 0.837 | |
| 73 | 0.82 | 0.03 | 74.11*** | 0.520 | |
ITT, analyses conducted on an intent-to-treat basis (as opposed to completer basis); k, number of trials included in the analysis for the given comparison; p values refer to the comparison of trials meeting v. not meeting the quality criterion or the continuous moderator analyzed.
*p < 0.05, **p < 0.01, ***p < 0.001.