Previous research on the survey measurement of sexual orientation, gender identity, and gender expression (SOGIE) often focuses on the measurement of identity, with comparably little research focused on gender expression as a key feature of how gender is lived and experienced. This study examines the reliability and validity of survey questions about gender expression in a 2-by-5-by-2 factorial experiment that varies the question order, type of response scale, and the order of gender presentation in the response scale.

The results indicate that the effect of which (side of the) scale is presented first on gender expression varies by gender for each of the unipolar items and one of the bipolar items (behavior). In addition, the unipolar items also show distinctions among the gender minority population in ratings of gender expression as well as more nuance with respect to concurrent validity in predicting health outcomes among cisgender respondents. The results of this study have implications for researchers who are interested in accounting for gender holistically in survey and health disparities research.