|| Checking for direct PDF access through Ovid
Questions about sexual orientation are increasingly utilized in population-based surveys. The New Zealand Attitudes and Values Study (NZAVS), a national probability longitudinal panel, used an open-ended measure which asked participants to describe their sexual orientation in their own words. The resulting data allowed us to explore 3 waves of missing responses in the NZAVS (Ns = 13,944–18,261). We examined the demographics, attitudes, and sexual orientation identities of those who provided an answer outside of the scope of the question (8.9%), or did not answer the question (10.8%). Encouragingly, we found that missing data decreased 5.5% over 2 years. Those who were coded as missing at 1 wave but provided an answer on others tended to identify as heterosexual at higher rates than the rest of the sample. Additionally, the vast majority of those with missing data at all 3 time points reported that they were in relationships with someone of a different gender. Finally, we found that gender, ethnicity, nation of birth, response mode, education, parental status, socioeconomic status, and support for marriage equality predicted differences in response type. Our results suggest that there are systematic differences between those willing to answer sexual orientation questions and those who cannot or will not.