Briere J, Ward R, Hartsough W R
Psychological Service Center, University of Manitoba, Winnipeg, Manitoba, Canada.
J Pers Assess. 1983 Jun;47(3):300-2. doi: 10.1207/s15327752jpa4703_12.
Bem Sex-Role Inventory protocols from 848 subjects were categorized as androgynous or nonandrogynous according to each of three current scoring systems. Androgynous protocols by each scoring system were then examined for sex differences in trait item endorsements. The t-score method generated androgynous subjects who were cross-sex-typed, while androgynous subjects defined by the median split method tended to be same-sex-typed. Androgynous subjects by the intersect method presented the fewest number of sex differences. These method-specific versions of androgyny appear linked to the use of summated Masculinity and Femininity subscales of the BSRI.
