Stereotypes and crossed-category evaluations: the case of gender and science education.

A sample of UK adolescents (n = 1140), grouped by sex and liking of science, evaluated themselves, and girl and boy targets who did or did not like science, on masculine, feminine and gender non-specific traits. Contrary to sociological concerns about the masculine image and appeal of science, those who liked science more rated themselves more positively on feminine and gender non-specific-but not masculine-traits. The girl target was rated lower on feminine traits if she liked science, but the boy was rated higher on feminine traits if he liked science. Target ratings also showed in-group enhancement based on liking of science, and a 'black sheep' effect: those who liked science less discriminated against the same-sex target who liked science, especially on gender in-group relevant traits. We argue that gender differences in science education should be attributed partly to subjective group dynamics and not solely to images of science.