Chung Y B
Department of Educational Psychology, University of Illinois at Urbana-Champaign 61820, USA.
J Homosex. 1995;30(2):87-97. doi: 10.1300/j082v30n02_05.
This study examined the construct validity of the Bem Sex-Role Inventory (BSRI; Bem, 1978) for heterosexual and gay men. Sixty heterosexual and 63 gay male participants were recruited through networking and advertisements. These two groups were of equivalent age, socioeconomic background, race, student status, and educational level. They completed the Lifestyle Questionnaire assessing sexual orientation and the BSRI assessing sex-role orientation. The internal consistency and discriminant validity of the BSRI scales were examined by corrected item-total correlations, coefficient alphas, inter-scale correlations, and factor analysis. Results suggested that the BSRI was equally valid for heterosexual and gay men, and the psychometric data reported in the BSRI Manual (Bem, 1981) were essentially replicated. However, the short-form BSRI is recommended for use with male respondents because of the problematic non-short-form Femininity items.
本研究考察了贝姆性别角色量表（BSRI；贝姆，1978）对异性恋男性和男同性恋者的结构效度。通过社交网络和广告招募了60名异性恋男性和63名男同性恋参与者。这两组在年龄、社会经济背景、种族、学生身份和教育水平方面相当。他们完成了评估性取向的生活方式问卷和评估性别角色取向的BSRI。通过校正后的项目总分相关性、α系数、量表间相关性和因子分析来检验BSRI各量表的内部一致性和区分效度。结果表明，BSRI对异性恋男性和男同性恋者同样有效，并且基本上重复了BSRI手册（贝姆，1981）中报告的心理测量数据。然而，由于非简版女性气质项目存在问题，建议对男性受访者使用简版BSRI。
