Some Methodological Comments on Labels, Traits, Interaction, and Types in the Study of "Femininity" and "Masculinity": Reply to Spence

We concur with Spence that the labels "femininity" and "masculinity" refer to a multidimensional domain and have been used inappropriately for scales that measure more circumscribed gender-related traits such as dominance and nurturance. It is nevertheless possible to investigate more encompassing constructs, such as androgyny and gender identity, through appropriate multivariate methods that incorporate, for example, interactions between dominance, nurturance, and other gender-related variables. In contrast, the common practice of exclusively relying on "feminine," "masculine," and other types or patterns, and treating these as purely nominal categories, hinders rather than aids recognition of underlying simple and complex (for example, interactive) relations. Furthermore, the study of interactions among measures of individual differences should be distinguished from a search for "real types." Interactions do not imply types, types do not imply interactions, although empirically the two phenomena may co-occur. The study of types does have a place in the study of gender-related characteristics and individual differences generally; it bears directly on questions of continuity and discontinuity important in personality and psychopathology and requires its own methodology.