Table 4.
Results of Study-Level Moderation Analyses.
| Moderator | Summary effect and 95% CI | Test of moderation | Egger’s test | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| g | LL | UL | t | df | p | m | Statistic | p | I 2 | p | |
| Subtlety of norm salience | t(50.0) = 2.87 | .006 | 55.5% | ||||||||
| Explicit | 0.42 | 0.34 | 0.51 | 10.54 | 33.8 | <.001 | 39 | .036 | |||
| Subtle | 0.20 | 0.07 | 0.34 | 3.09 | 23.0 | .005 | 25 | .015 | |||
| Norm category | F(2, 14.8) = 5.41 | .017 | 57.2% | ||||||||
| Injunctive norm | 0.25 | 0.16 | 0.35 | 5.61 | 37.7 | <.001 | 42 | .002 | |||
| Descriptive norm | 0.56 | 0.31 | 0.80 | 5.63 | 5.8 | .001 | 7 | — | |||
| Personal norm | 0.47 | 0.31 | 0.63 | 6.27 | 12.6 | <.001 | 15 | — | |||
| Order of manipulations | t(50.2) = 0.00 | .997 | 63.8% | ||||||||
| MS first | 0.33 | 0.22 | 0.45 | 5.96 | 32.3 | <.001 | 35 | .015 | |||
| Norm salience first | 0.33 | 0.22 | 0.45 | 5.89 | 23.2 | <.001 | 27 | .011 | |||
| Data collection | F(2, 8.4) = 4.52 | .047 | 61.1% | ||||||||
| Laboratory | 0.40 | 0.29 | 0.50 | 7.60 | 36.0 | <.001 | 40 | .013 | |||
| Internet | 0.21 | 0.08 | 0.34 | 3.41 | 17.6 | .003 | 20 | .151 | |||
| Field | 0.45 | 0.32 | 0.58 | 10.92 | 3.0 | .002 | 4 | — | |||
| MS control topic | t(31.9) = 2.77 | .009 | 57.7% | ||||||||
| Neutral | 0.18 | 0.04 | 0.32 | 2.68 | 16.5 | .016 | 18 | — | |||
| Aversive | 0.40 | 0.31 | 0.49 | 9.11 | 39.5 | <.001 | 46 | .002 | |||
| Sample origin | F(4, 8.9) = 1.52 | .276 | 60.5% | ||||||||
| Europe | 0.30 | 0.17 | 0.42 | 4.90 | 18.6 | <.001 | 21 | .061 | |||
| North America | 0.32 | 0.20 | 0.43 | 5.77 | 29.8 | <.001 | 33 | .002 | |||
| Asian | 0.37 | — | — | 3.46 | 4.0 | — | 6 | — | |||
| Arabian | 0.72 | — | — | 2.97 | 2.0 | — | 3 | — | |||
| Israel | 0.41 | — | — | — | — | — | 1 | — | |||
| Number of delay tasks | F(2, 17.3) = 0.53 | .600 | 62.7% | ||||||||
| Zero | 0.39 | 0.26 | 0.53 | 6.54 | 8.9 | <.001 | 10 | — | |||
| One | 0.31 | 0.21 | 0.41 | 6.24 | 37.3 | <.001 | 41 | .012 | |||
| Two | 0.34 | 0.07 | 0.61 | 2.90 | 8.4 | .019 | 11 | — | |||
| Research team | F(3, 34.1) = 8.22 | .007 | 52.4 | ||||||||
| Schindler/Reinhard | 0.11 | −0.02 | 0.25 | 1.77 | 16.6 | .095 | 19 | — | |||
| Fritsche/Jonas | 0.43 | 0.30 | 0.55 | 7.30 | 12.7 | <.001 | 15 | — | |||
| Pyszczynski | 0.55 | 0.35 | 0.76 | 6.34 | 7.8 | <.001 | 9 | — | |||
| others | 0.37 | 0.24 | 0.51 | 5.84 | 17.7 | <.001 | 21 | .021 | |||
| Opposed norms salient | t(36.4) = 1.37 | .180 | 62.5% | ||||||||
| No | 0.30 | 0.21 | 0.39 | 6.88 | 39.1 | <.001 | 44 | .007 | |||
| Yes | 0.42 | 0.26 | 0.58 | 5.40 | 18.3 | <.001 | 20 | .104 | |||
Note. g = effect size; CI = confidence interval; LL = lower limit of the 95% CI; UL = upper limit of the 95% CI; t = t value associated with the g value in the same row testing statistical significance in the respective moderator subgroup; df = associated small-sample-corrected degrees of freedom; p = p value associated with the t value and df in the same row; m = number of effect sizes in the respective moderator subgroup. Statistic (test of moderation): t value for single parameter tests or F value for multiple parameter tests and according degrees of freedom. Significant test statistics indicate the significance of the overall model. I2 reflects the proportion of true variance in the total observed variance of effect sizes after accounting for the respective moderator. Note that for two subgroups in the sample origin analysis, degrees of freedom fell below 4. Significance tests for the summary effects should thus not be interpreted. Accordingly, we did not report LL, UL, and p values for the respective subgroups. For one subgroup in the sample origin analysis, there was only one study; therefore, significance tests for the summary effects were not conducted. Egger’s test refers to the positive relationship between the effect size and the standard error. A significant p value indicates a significant relationship suggesting small-study effects and an overestimation of the unadjusted effect size. Egger’s test was only conducted for moderator subgroups including 20 studies or more. All tested relationships were positive in direction; nonsignificant effects should be interpreted with caution due to low test power.