Abstract
The recent surge in empirical work on androgyny requires a comparison among the different instruments and different methods by which masculinity, femininity, and androgyny are measured. Subjects (N = 237) completed five instruments, which were scored for androgyny, by both Bem's (1974) t-ratio and Spence, Helmreich, and Stapp's (1975) median-split methods, and for masculinity and femininity. Confirmatory factor analysis of the total scores failed to yield the predicted two-factor structure, and correlations among masculinity and femininity scores varied widely. Despite moderate to high correlations among androgyny scores derived by different instruments, there was at most only 65% agreement between any two instruments in classifying subjects into the five Bem or four Spence et al. sex-role categories.