Are Sexual Minorities Less Likely to Participate in Surveys? An Examination of Proxy Nonresponse Measures and Associated Biases with Sexual Orientation in a Population-Based Health Survey.

One of the implicit assumptions in survey research is lower response rates by sexual minorities than non-minorities. With rapidly changing public attitudes towards same-sex marriage, we reconsider this assumption. We used data from the 2013 and 2014 National Health Interview Survey (NHIS) that include contact history data for all sample families (n=117,589) as well as sexual orientation information about adults sampled from responding families (n=71,110). We created proxy nonresponse indicators based on contact efforts and reluctance from contact history data and linked them to sexual orientation of the sample adult and simulated nonresponse. The data did not support the assumption: straight adults were more difficult to get cooperation from than non-straights. With female sexual minorities showing higher nonresponse than the male counterpart, special considerations are required. Replication analyses may provide insights into what factors influence study participation decisions, which will inform how nonresponse may impact the accuracy of research findings.