| Description |
Very few studies have investigated the construct validity of measures of masculinity. This study analyzed multitrait multimethod matrices of male and female subscale scores from the Male Role Norms Inventory-Revised (MRNI-R), the Conformity to Masculine Norms Inventory (CMNI), and the Gender Role Conflict Scale (GRCS). Six subscales were of interest. Specifically, measures of dominance, homophobia, restrictive emotionality, nonrelational sexuality, aggression, and self reliance were given to 176 undergraduates from a large university in Utah. Subscale correlations were compared to indicate convergent and discriminant validity, as well as method effects, for each masculinity measure. Generally, the measures performed slightly better among males than females and the multidimensional structure of the CMNI was supported. Surprisingly, results indicated modest or poor validity among many of the subscales. These results were explained in terms of a construct confound on the GRCS, a very high method effect for the MRNI-R, and inconsistency between constructs and definitions among the measures. Finally, a preliminary content analysis of masculinity scales generated hypotheses about convergence between masculinity science and other social science disciplines. Limitations of this study and recommendations for future research were noted. |
