On the Design, Development, and Testing of Sexual Identity Questions: A Discussion and Analysis of Kristen Miller and J. Michael Ryan's Work for the National Health Interview Survey

A remarkable article by Kristen Miller and J. Michael Ryan, working at the Centers for Disease Control and Prevention, deserves wider attention. They developed two sexual identity-related questions for the (U.S.) National Health Interview Survey: one pertaining to gender identity and one to sexual orientation. They conducted 386 in-depth cognitive interviews and concluded that there are substantial problems with existing sexual identity questions. Words such as ‘heterosexual’ and ‘bisexual’ are often grossly misunderstood (especially by those with less education), triggering substantial measurement errors. Revised questions substantially reduced missing-value responses and increased validity, even though responses used phrases such as ‘straight, that is, not gay’ and avoided the words ‘heterosexual’ and ‘homosexual.’ Spanish translation was neither straightforward nor intuitive. These questions should generate improved estimates of the lesbian, gay, and bisexual population and raise important issues that must be considered by anyone designing sexual orientation questions in the future.