Moving Beyond Sex: Measuring Gender Identity in Telephone Surveys

Abstract

While researchers have long understood that sex and gender are theoretically distinct concepts, much survey research continues to rely on the sex of respondents as a proxy for their gender. Recent work on political and social attitudes has shown that gender has far-reaching consequences for attitudes that are not neatly summed up by measures of sex, but measures of gender identity have generally been used either in forms that are too lengthy for most survey research or have only been used in online modalities. Longer sets of gender identity items may be theoretically desirable, but a gender identity item embedded in a live-caller random digit dialing survey in New Jersey shows that just one gender identity item can account for a great deal of within-sex variance in political and social attitudes without leading to an overly high rate of refusals.