Gender identity questions
The recommended approach is two steps: ask the sex someone was assigned at birth, then their current gender. Comparing the two answers identifies transgender and gender-diverse respondents, including people who wouldn't describe themselves as transgender. Ask the pair together, with sex at birth first.
Two-step (recommended)
From NASEM 2022:
1. What sex were you assigned at birth, on your original birth certificate?
- Female
- Male
- (Don't know)
- (Prefer not to answer)
2. What is your current gender? [Select one]
- Female
- Male
- Transgender
- Two-Spirit [show only to American Indian or Alaska Native respondents]
- I use a different term: [free text]
- (Don't know)
- (Prefer not to answer)
Federal survey variant
OMB's 2023 recommendations use the same first question (Female or Male), with a different second question and a confirmation step:
2. How do you currently describe yourself (mark all that apply)? Female; Male; Transgender; I use a different term: [free text]
3. Just to confirm, you were assigned {FILL} at birth and now you describe yourself as {FILL}. Is that correct? Yes; No [return to questions 1 and 2]
Ask the confirmation only when the first two answers differ. It catches entry errors, which matter because a small error rate in a large sample can produce many false gender minority classifications.
The two versions aren't interchangeable. Cite the one you used.
Coding
OMB 2023 suggests four reporting groups from the two-step answers:
| Group | Answers |
|---|---|
| Cisgender female | Assigned female, currently female only |
| Cisgender male | Assigned male, currently male only |
| Gender minority | Current gender differs from sex assigned at birth, or includes transgender |
| Another gender identity | Used a different term (after coding write-ins) |
Single-item alternatives
When only one item fits, OMB 2023 offers: "Are you: Female / Male / Transgender, non-binary, or another gender" (mark all that apply). Both NASEM and OMB present the two-step as the main approach, so use the single item only when there's no room for two.
Evidence
- Reisner et al. found two-step items valid in a US national cohort of young adults (BMC Public Health, 2014, doi).
- Bauer et al. evaluated transgender-inclusive sex and gender measures for population surveys (PLoS One, 2017, doi).
Sources: NASEM 2022 recommended questions (NIH summary); OMB 2023