August 19, 2015
Stanford sociologist urges rethinking of sex and gender in surveys
New research reveals that most social surveys are not measuring what surveyors think is being measured when it comes to sex and gender. To better reflect today's diversity, Stanford sociologist Aliya Saperstein says survey designers should ask themselves if using only one question with two possible answers is enough.
By Christina Dong
Stanford sociologist Aliya Saperstein says survey questions involving sex and gender need to change as society's perceptions change. (Shutterstock
The International Olympic Committee is revisiting its standards for deciding which athletes are eligible to compete in men’s or women’s events. And in Washington, D.C., the signs on some public restrooms are changing to allow access for all genders.
Stanford sociologist Aliya Saperstein says she hopes the major social surveys in the United States will be next in line to reconsider how they classify Americans into “males” and “females” or “men” and “women.”
“If the world is changing and [surveyors] are not changing the measures, it’s not clear that we’re getting the information we think we’re getting, even if we ask the same question we always have,” Saperstein said in an interview.
Traditional understandings of sex and gender found in social surveys – such as only allowing people to check one box when asked "male" or "female? – reflect neither academic theories about the difference between sex and gender nor how a growing number of people prefer to identify, Saperstein argues in a study she coauthored with Grand Valley State University sociology professor Laurel Westbrook.
In their analysis of four of the largest and longest-running social surveys in the U.S., the sociologists found that the surveys not only used answer options that were binary and static, but also conflated sex and gender. These practices changed very little over the 60 years of surveys they examined.
“Beliefs about the world shape how surveys are designed and data are collected,” they wrote. “Survey research findings, in turn, shape beliefs about the world, and the cycle repeats.”
“We have been taking gender categories for granted for too long,” Westbrook added. “It doesn’t help us better understand health disparities or income gaps or voting patterns to always divide the population into he's and she's.”
Questioning the options
Contemporary gender theory, according to the researchers, approaches sex and gender as related but distinct concepts – sex is a system of classification based on biological criteria such as chromosomes and genitalia, while gender reflects normative expectations for behavior based on one’s sex category.
There are more than two sexes and more than two genders, the researchers say, and how people identify in terms of gender may differ from how others perceive and classify them. Gender identities and classifications can also change over the course of a person’s life.
Saperstein and Westbrook point out that sex and gender categories in social surveys and official statistics do not align with this understanding. For example, the researchers found that answer options to questions about a person’s sex or gender were limited to “male” or “female” only. Though this finding was not surprising to them, they noted the practice stands in contrast to other demographic questions that allow for more diversity.
“Characteristics from race to political affiliation are no longer counted as binary distinctions, and possible responses often include the category ‘other’ to acknowledge the difficulty of creating a preset list of survey responses,” they wrote.
“By always providing the same two-answer options, surveys imply that the categories 'male' and 'female' cover all possible ways of being," they noted.
The surveys also treat sex and gender as synonymous, often using “sex” and “gender” interchangeably or using the sex terms “male” and “female” to measure gender, rather than gender terms such as “man” or “woman;” “cisgender” or “transgender.”
Suggestions for reform
The researchers suggest the following changes to social surveys:
• Surveys must consistently distinguish between sex and gender
• Surveys should rethink binary categories
• Surveys need to incorporate self-identified gender and acknowledge it can change over time.
Finally, Saperstein and Westbrook suggest measuring gender as a spectrum. One possibility would be to ask respondents to describe themselves on scales of masculinity and femininity. “You get to rate yourself on both scales, so that way you’re not either/or; you can be some of both,” Saperstein said.
To better distinguish between sex and gender, the researchers also recommend asking two separate questions in which respondents indicate their sex at birth, followed by the gender they currently present as or identify with.
“Then, if you’re interested in studying inequality, you have both pieces of information,” Westbrook said. “So you can determine whether one or the other, both or neither, makes a difference in how people experience the world.”
Saperstein and Westbrook noted in their study that they intentionally have not outlined specific guidelines for improving surveys because they aim to inspire both creativity and rigorous testing of alternative measures.
“We encourage an acknowledgement of sex and gender fluidity and diversity within, as well as beyond, traditional categories,” they wrote.
Also, they recognize there is no easy solution. “We don’t think that there’s one ideal way of measuring sex and gender that will fit for every purpose,” Saperstein said in an interview. “The best way is going to vary depending on the goal of the research and who is being studied.”
Above all, surveys have to be flexible, Saperstein said, noting that in 10 or 20 years, as ideas about sex and gender continue to evolve, surveys could warrant even more changes.
Christina Dong is an intern at the Stanford News Service.
For more Stanford experts on sociology and other topics, visit Stanford Experts.