Measuring sexual identity in US health surveys

Virginia S. Cain, National Center for Health Statistics (NCHS), CDC

Sexual orientation has been linked to a number of negative physical and mental health outcomes and access health care (Institute of Medicine, 2011). However, these results have come mainly from small scale non-representative samples. While some major US health surveys have asked questions about sexual orientation, identifying the lesbian, gay, and bisexual (LGB) population has been challenging. Problems with questions on sexual orientation have led to results with substantial rates of missing data which are not randomly distributed across subgroups within the population. Highest rates of missing data have been found among racial and ethnic minority groups and lower education groups. Efforts to improve questions using cognitive interviewing techniques have resulted in questions that address the conceptual complexity of the broad term sexual orientation by disaggregating sexual identity from sexual attraction and sexual behavior, concepts that are related but not identical. The improvements in question wording and response categories have substantially reduced the rates of missing data across population subgroups reducing the bias in the estimates. The present paper discusses the process of incorporating sexual identity questions into the largest US health survey, the National Health Interview Survey (NHIS), Newly designed questions will be field tested in a 5,000 case nationally representative sample in 2012 with a planned full implementation in the 2013 NHIS, a survey of approximately 40,000 households representative of the US population. The present paper examines the validity of questions used to measure sexual orientation or identity, presents results of cognitive interviewing which developed improved questions, discusses issues around question placement and the plans for incorporating the questions on sexual identity into the NHIS. The paper discusses the pilot test which includes a test of two modes of interviewing. e.g. interviewer administered and Audio Computer Assisted Self Interviewing (ACASI).