Seminar Series: Brady T. West

This research sought to assess how differences in response options for survey questions about sexual identity affect population estimates of the associations of sexual identity with substance use and misuse. A randomized experiment was implemented in the 2015-2017 National Survey of Family Growth. In this experiment, two random half-samples of a large, nationally representative sample of 10,048 individuals aged 15-49 years responded to two different versions of a sexual identity question (a three-option version, including heterosexual, gay/lesbian, and bisexual, or a four-option version, adding “something else”). We examined changes in population estimates of the associations of sexual identity with substance use and misuse across the two half-samples.We will present evidence of dramatic differences in estimates of the associations of sexual identity with substance use and misuse across the half-samples that remained robust after adjusting for sociodemographics. This study demonstrates that when survey respondents choose their sexual identity from a more limited number of response options, the heterogeneity of the sexual identity subgroups increases, weakening estimated associations of sexual identity with substance use and misuse. The findings from this research suggest that open-ended survey questions may be necessary to measure sexual identity accurately and understand its associations with substance use and misuse (along with other survey outcomes that may be of interest to researchers in other domains).

Presentation by Brady West, Institute for Social Research.