Table 2.
Demographics of the exploration and validation samples
| Exploration sample | Validation sample | Statistics | |
|---|---|---|---|
| N = 204 | N = 304 | F/χ2 (p value), effect size | |
| Age | 22.85 (1.7) | 22.82 (1.73) | F = 0.02 (0.88); η2 = 4.9E10−5 |
| Sex | 105 M (51 %) | 155 M (51 %) | χ 2 = 0.01 (0.92); Φ = 0.005 (0.92) |
| Education (N) | χ 2 = 0.07 (0.97); Φ = 0.01 (0.97) | ||
| Low | 25 (12 %) | 35 (12 %) | |
| Middle | 87 (43 %) | 131 (43 %) | |
| High | 92 (45 %) | 138 (45 %) | |
| Handedness (N) | χ 2 = 0.71 (0.40); Φ = 0.04 (0.40) | ||
| Left | 22 (11 %) | 26 (9 %) | |
| Right | 182 (89 %) | 278 (91 %) | |
| AQa (Mean, SD, range) | |||
| Total | 55.63 (8.96) [33–80] | 57.05 (8.70) [32–91] | F = 3.18 (0.08); η2 = 0.006 |
| Broad factor social behavior | 45.69 (7.80) [27–70] | 47.31 (7.60) [25–75] | F = 5.43 (0.02); η2 = .01b |
| Social skills | 12.39 (3.65) [7–28] | 13.02 (3.67) [7–24] | F = 3.55 (0.06); η2 = 0.007 |
| Routine | 8.39 (1.84) [4–13] | 8.63 (1.73) [4–14] | F = 2.16 (0.14); η2 = 0.004 |
| Switching | 8.87 (1.85) [4–13] | 9.08 (2.00) [4–15] | F = 1.44 (0.23); η2 = 0.003 |
| Imagination | 16.03 (3.38) [8–23] | 16.58 (3.39) [8–31] | F = 3.19 (0.08); η2 = 0.006 |
| Numbers/patterns | 9.94 (3.10) [5–19] | 9.74 (3.21) [5–20] | F = 0.48 (0.49); η2 = 0.01 |
M males
a4-point scale
bA common side effect of working with large samples is the tendency that small between group differences become statistically significant, but may be unimportant. Here we report effect sizes to illustrate the magnitude of these effects. The difference in mean score on the broad factor social behavior subscale is significantly different between groups; however, the effect size shows that the magnitude is rather small (a value of η = 0.02 is considered small, here we report an even lower value: of η = 0.01). It should be noted that in all of our analyses we used the aggregated score of the AQ, i.e. we didn’t examine associations based on AQ subscales in any of our analyses