Gender bias in the abstractness of verbs and adjectives

Participants of both sexes were asked to write responses to sentences about members of both sexes exhibiting socially desirable or undesirable behaviors. Using Semin and Fiedler's (1988) coding scheme, I analyzed the sentences for the abstractness of the verbs and adjectives. As predicted, when the participants wrote about socially desirable behaviors performed by someone of their own gender, they used more abstract words than when they wrote about undesirable behaviors. The opposite effect was not found when the participants wrote about the other gender; possibly there was no true out-group in this study. The bias occurred even though gender was not a salient issue at the time; therefore, a more subtle bias than has been shown previously may exist.