Categorical and gradational: alternative survey measures of sex and gender

Abstract

Demands for recognition of gender diversity and transgender people are growing. We tested non-binary sex and gender measures using nationally representative samples of US adults to assess feasibility for general population surveys. We find more support for a two-step categorical approach, with separate questions about natal sex and gender identity, than for a single question assessing transgender status as the latter was less reliable within our online surveys and over time. We also consider the challenge of determining measurement reliability for fluid characteristics and argue that using categorical and gradational gender measures in combination should become standard practice.