Primary Submission Category: Methodological approaches to studying public health
“I know who I am attracted to so I just clicked the options I am”: Improving assessment of attraction using data from a large survey of U.S. adolescents and young adults
Authors: Allegra Gordon Lynsie Ranker Kimberly Nelson R. Korkodilos Jennifer Conti Ziming Xuan Sabra Katz-Wise
Presenting Author: Allegra Gordon*
Background. Sexual and/or romantic attractions are a key feature of adolescent and young adult (AYA) health and development. Accurate, inclusive assessment of attractions is critical to population health research. Yet little is known about how to best assess attractions for AYA. Methods. We use cross-sectional quantitative and qualitative data from an online panel of 14,985 US-based AYA (14-25 years). We asked participants to report current attractions (check all that apply: girls/women, boys/men, nonbinary people, people of another gender identity [write-in]), and how they decided which of those answers to choose (open-ended). We assessed selection frequency and conducted: (i) content analysis of “another gender identity” write-ins (n=253); (ii) inductive thematic analysis of a stratified random sample of decision process responses (n=120; 20 across 6 gender groups). Results. Most respondents (n=14,675; 98.3%) endorsed one or more of the three gender categories offered. Among the 253 (1.7%) who selected “another gender” there were two common write-ins: (1) Any/all/no preference (n=116); (2) No attraction (n=74). Some wrote-in specific gender modalities (n=19; e.g., “trans men”) or expressions (n=10; e.g., “masculinity”). Decision-making question responses suggest the question was interpretable and easy to answer. Thematic analysis identified three main decision-making patterns: (a) considering gender(s) of past/current/possible partners, (b) stating one’s sexual orientation identity, and (c) naming other attraction facets (e.g., personality). Conclusions. Nearly all AYA in a large online sample answered an attraction question using three check-all-that-apply gender categories, suggesting this approach serves most AYA. To increase accuracy and inclusivity, population health surveys of AYA should include additional options in attraction questions (e.g., “no attractions” and “any gender”). Research is needed to examine attraction patterns across population subgroups.