A Two-Question Method for Assessing Gender Categories in the Social and Medical Sciences

Abstract

Three studies (N = 990) assessed the statistical reliability of two methods of determining gender identity that can capture transgender spectrum identities (i.e., current gender identities different from birth-assigned gender categories). Study 1 evaluated a single question with four response options (female, male, transgender, other) on university students. The missing data rate was higher than the valid response rates for transgender and other options using this method. Study 2 evaluated a method of asking two separate questions (i.e., one for current identity and another for birth-assigned category), with response options specific to each. Results showed no missing data and two times the transgender spectrum response rate compared to Study 1. Study 3 showed that the two-question method also worked in community samples, producing near-zero missing data. The two-question method also identified cisgender identities (same birth-assigned and current gender identity), making it a dynamic and desirable measurement tool for the social and medical sciences.

Acknowledgments

No funding source aided the completion of this article. The data reported in this article are tabulated in Table 1. The authors declare no conflicts of interest. We thank Avi Ben-Zeev and the members of the Social Perception, Attitudes, Mental Simulation (SPAMS) Lab for helpful comments on drafts of this article.

Notes

Note: Studies 1 and 2 are college student samples from San Francisco State University. Study 3 is a community sample from the San Francisco Bay Area. All percentages are the number of people in the category divided by the total sample size. Percentages within a row can be directly compared. Cells with “—” indicate that the value could not be determined by the method used. Cis is a short form of cisgender; trans is a short form of transgender.

^a Single-question method.

^b Two-question method.

^c Missing data percentages reflect individual respondents who provided a nonresponse to at least one of the questions.