Contextual Influences on Nonresponse to Health Survey Questions About Sexual Orientation and Gender Identity.

Abstract

Purpose: We examined the relationship between state context and survey nonresponse to sexual orientation (SO) and gender identity (GI) questions. Methods: We obtained data from the 2014-2020 Behavioral Risk Factor Surveillance System (BRFSS) surveys. Item nonresponse was defined as selecting "don't know/not sure" or "refused" for each of two questions about SO and GI. Nonresponse patterns included responding to both SO and GI questions; responding only to the SO question (nonresponse to GI); responding only to the GI question (nonresponse to SO); and responding to neither question. State-level contextual measures included legal protections for lesbian, gay, bisexual, transgender, or other sexual or gender minority (LGBT+) people, LGBT+ social movement strength, and public opinion regarding LGBT+ issues. Results: The analysis included 1,459,525 respondents from 44 states (190 state-years). On weighted analysis, 96.5% of adults answered both SO/GI questions, 2.4% responded only to GI, 0.4% responded only to SO, and 0.7% responded to neither. The demographic profile of individuals with GI-only nonresponse differed markedly from the profile of adults with SO-only nonresponse. An increasingly favorable legal climate for LGBT+ people was associated with greater rates of response to SO and GI questions. However, a more LGBT+ friendly state climate measured by social movement strength or public opinion was not consistently associated with reduced SO and GI question nonresponse. Conclusion: Contextual factors have mixed association with nonresponse to SO and GI question on BRFSS surveys. Our results warrant continued development of health survey questionnaires to elicit accurate information on respondents' SO and GI.