Fig. 3.

Genetic and environmental contributions to face evaluation. (A) Correlations between MZ and DZ twin pairs on face impression scores for trustworthiness, dominance, and attractiveness, measured using intraclass correlation coefficients (ICC[1,1]) (32). Bars show 95% CIs; the large overlap between MZ and DZ correlations is indicative of a nongenetic contribution to individual differences in facial impressions. CIs were calculated using a bias-corrected and accelerated bootstrap with 2,000 bootstrap samples. (B) Plot of MZ vs. DZ intraclass correlations (ICCs) for the traits tested in the present study, as well as attractiveness face impressions (18) and Cambridge Face Memory Test face identity recognition ability (12, 13) reported elsewhere for twins recruited from the same registry. The dashed line represents the extreme case where all covariation within twin pairs is caused by shared environmental factors, resulting in equal ICCs for MZ and DZ pairs. The solid line represents the extreme case where all covariation within twin pairs is caused by additive genetic factors, such that MZ ICC is twice that for DZ pairs. Critically, in this case, the influence of both shared environmental factors and additive genetic factors was small relative to personal environmental factors. The further left and/or lower in the graph, the more idiosyncratic the judgments and the higher the contribution of personal environmental factors. All face evaluation results are well to the left of the face identity recognition result. Bars show 95% CIs.