Table 1. The Results of 40,000 Simulated Comparative Cognition Studies by Power.
Power | Proportion Published | Unstandardized Effect Size | ||
---|---|---|---|---|
All Samples | Published | Mean overestimation in “published” findings | ||
80 | 0.796 | 5.77 | 6.53 | 13% |
50 | 0.494 | 3.82 | 5.56 | 45% |
20 | 0.205 | 1.89 | 4.88 | 158% |
5 | 0.053 | 0.007 | 4.45 | 64486% |
Note. The proportion published represents the proportion of studies producing p < .05, and the unstandardized effect sizes are the mean differences between the groups.