Table 1.
Descriptive Statistics for Each Study Design
| Design | Control prior intelligence | Policy change | School age cutoff |
|---|---|---|---|
| k studies | 7 | 11 | 10 |
| k data sets | 10 | 12 | 20 |
| k effect sizes | 26 | 30 | 86 |
| N participants | 51,645 | 456,963 | 107,204 |
| Mean age at early test in years (SD) | 12.35 (2.90) | — | — |
| Mean time lag between tests in years (SD) | 53.17 (15.47) | — | — |
| Mean age at policy change in years (SD) | — | 14.80 (2.59) | — |
| Mean age at outcome test in years (SD) | 63.48 (18.80) | 47.92 (19.39) | 10.36 (1.60) |
| n outcome test category (composite:fluid:crystallized) | 5:20:1 | 2:23:5 | 3:67:16 |
| n achievement tests (achievement:other) | 1:25 | 7:23 | 38:48 |
| Male-only estimates (male only:mixed sex) | 2:24 | 8:22 | 0:86 |
| Publication status (published:unpublished) | 22:4 | 21:9 | 64:22 |
Note: To estimate N from studies with multiple effect sizes with different ns, we averaged sample sizes across effect sizes within each data set and rounded to the nearest integer. “Unpublished” refers to any study not published in a peer-reviewed journal.