Development and validation of the Comfort and Conformity of Gender Expression Scale (CAGES)

Abstract

Gender roles and stereotypes have been shown to be deeply entrenched across race, class, age and location (Bem, 1993). Gender has been a focus of psychological measurement since the early 20th century (Terman & Miles, 1936). The Comfort and Conformity of Gender Expression Scale (CAGES) was developed in order to assess an area not yet explored, comfort and discomfort with conformity and nonconformity of gender expression in a broad sample of adults. After initial item development, three studies were completed to assess the validity of the CAGES as a measure of comfort and conformity of gender expression. Study 1 consisted of the initial exploratory factor analyses (EFA) to determine factor structure and item relevance. A sample of 356 participants completed the initial CAGES. The final EFA resulted in a 24-item, 4-factor scale. The four factors were named; Discomfort with Nonconformity, Resentful Conformity, Active-Physical Comfort Conformity and Comfort with Conformity-Appearance. The CAGES was hypothesized to measure cognitive, affective and behavioral components of gender expression. The hypothesized scale was partially supported in that the final 4-factor CAGES measures behavioral and affective components of comfort-discomfort with conformity-nonconformity in gender expression. Study 2 examined reliability and convergent and divergent validity estimates for the CAGES by correlating the revised CAGES with two established scales, the Conformity to Masculine Norms Inventory (CMNI) and the Conformity to Feminine Norms Inventory (CFNI). A total of 176 participants completed the revised CAGES and the CFNI and CMNI. Moderate to strong correlations were found between all four of the CAGES subscales and six subscales on the CMNI and five subscales on the CFNI, and no correlation with a measure not related to gender, the Balanced Inventory of Desirable Responding-Impression Management. Study 3 established test-retest reliability through administering the CAGES to 56 participants who completed the revised CAGES twice over a two-week period. Significant correlations were found on three of four CAGES subscales, evidencing test-retest reliability for three CAGES subscales as a measure of gender expression. Thus, the CAGES contributes to the existing literature in gender measurement by introducing a 24-item, 4-factor scale that measures comfort and conformity of gender expression.