Bipolar-biasing effects of sex-role extremity on memory for traits

Two studies examined the relation between the extremity of sex-role self-ratings and recall of trait names. In each study, subjects were read a list of traits from Bem's Sex Role Inventory and asked to rate the degree of self-descriptiveness on a 7-point scale. Later they were asked to write down as many of the traits as possible in free-recall format. Results showed a predicted U-shaped relation between trait extremity and recall. That is, traits given extremely high or extremely low ratings were remembered better than those given moderate ratings. This pattern held for both idiographic and nomothetic scaling of traits. Subjects were also scored on Bem's Masculinity and Femininity dimensions. The same U-shaped relationship was observed between extremity on each sex-role dimension and recall of traits relevant to that dimension. Subjects scoring extremely high or low on a dimension recalled traits consistent or inconsistent with their overall level. No relation was found between scores on one sex-role dimension and recall on the other dimension. These results support the following conclusions: (1) Recall is enhanced for information related to a particular sex-role dimension, whether consistent or inconsistent. (2) The two sex-role dimensions are best viewed as functionally unrelated traits rather than as components of androgyny.