Table 5.
Results of the multi-level analyses for RQ2A: mis-/matching rater and target ethnicity per stimulus type.
| Fixed | Random | |||||||
|---|---|---|---|---|---|---|---|---|
| Coeff | B | CI | stand. B | SE | t | Coeff | SD | |
| Whole face | ||||||||
| Intercept | β00 | 5.39 | 5.35–5.43 | 0.02 | 275.1*** | r0i | 1.08 | |
| Match ethnicity | β10 | 0.15 | 0.11–0.19 | 0.04 | 0.02 | 7.8*** | r1i | 0.84 |
| R2conditional = 28%, AIC = 437,792, BIC = 438,850, Ω2 = 31% | ||||||||
| Mouth part | ||||||||
| Intercept | β00 | 5.02 | 4.98–5.06 | 0.02 | 250.9*** | r0i | 1.12 | |
| Match ethnicity | β10 | 0.10 | 0.06–0.13 | 0.03 | 0.02 | 5.7*** | r1i | 0.77 |
| R2conditional = 37%, AIC = 403,475, BIC = 403,532, Ω2 = 40% | ||||||||
| Mid-face part | ||||||||
| Intercept | β00 | 5.00 | 4.96–5.04 | 0.02 | 250.2*** | r0i | 1.13 | |
| Match ethnicity | β10 | 0.09 | 0.06–0.12 | 0.03 | 0.01 | 6.4*** | r1i | 0.60 |
| R2conditional = 45%, AIC = 369,850, BIC = 369,907, Ω2 = 47% | ||||||||
| Eyes part | ||||||||
| Intercept | β00 | 5.38 | 5.34–5.42 | 0.02 | 268.6*** | r0i | 1.12 | |
| Match ethnicity | β10 | 0.17 | 0.13–0.21 | 0.05 | 0.02 | 9.0*** | r1i | 0.90 |
| R2conditional = 35%, AIC = 414,294, BIC = 414,352, Ω2 = 38% | ||||||||
Rater-target ethnicity match: 0 = mismatch; 1 = match. Nobservations = 107,712; Nraters = 3,366; whole face trust ratings served as reference category. *p < 0.05, **p < 0.01, ***p < 0.001.