Abstract
The structure of the short form Bem Sex Role Inventory (Bem, 1981) was examined in two UK samples of 214 women and 166 men, and 215 women and 166 men, respectively, aged 17-22 years. A confirmatory factor analysis of the masculinity and femininity items for the first sample produced a poor fit.
An exploratory factor analysis of the data from this sample produced a 3 factor solution which, together with the original 2 factor solution was examined in confirmatory factor analyses of the data from the second sample. Neither produced a good fit, although the 3 factor solution was better.
Gender differences in factor loadings were examined in a further multigroup CFA, which demonstrated that an unconstrained model produced a better fit than a constrained one. The findings suggest that when using the BSRI it would be prudent to examine underlying factor loadings.
Original language | English |
---|---|
Pages (from-to) | 384-387 |
Number of pages | 4 |
Journal | Personality and Individual Differences |
Volume | 46 |
Issue number | 3 |
DOIs | |
Publication status | Published - Feb 2009 |
Externally published | Yes |
Keywords
- Expressiveness
- Gender
- Gender roles
- Instrumentality