Skip to main content
. 2014 Jul 2;9(7):e99629. doi: 10.1371/journal.pone.0099629

Table 2. Results from regression analyses for African and Scottish observers' attractiveness judgements of African faces.

β F/t P Effect size 95% CI
Lower Upper
African ratings of African female faces
Model 1.642 0.131 0.389
Colour component 0.410 2.530 0.017 0.414 0.023 0.440
Shape component 9 −0.269 −1.812 0.080 −0.309 −0.036 0.008
Scottish ratings of African female faces
Model 2.921 0.008 0.531
Shape component 9 −0.472 −3.627 0.001 −0.546 −0.056 −0.012
Shape component 8 −0.324 −2.460 0.020 −0.404 −0.054 −0.003
Shape component 4 −0.259 −2.083 0.046 −0.350 −0.024 0.001
African ratings of African male faces
Model 2.374 0.031 0.474
Colour comp 1 0.472 2.854 0.008 0.473 0.037 0.512
Scottish ratings of African male faces
Model 0.792 0.647 0.231

Results obtained using simultaneous regression method. F statistics are indicated for the overall model and t statistics for individual coefficients. Effect size: R2 (model); Partial eta squared (variables). β indicates the standardized beta coefficient and CI the confidence interval based on 1000 bootstrap samples. Only significant and marginally significant (p≤0.08) coefficients are indicated. Higher values for the colour components indicate a lighter, yellower and redder complexion. High values for the shape components seem to indicate: higher facial adiposity and/or robustness (shape component 9); lower femininity and/or neoteny (shape component 8 and 4).