Asking about gender

By Kathleen Frankovic

Market and opinion researchers must be increasingly aware that gender identity is not binary.  The standard question — Male or female? – may no longer be a question that some respondents can answer.  While there is no one standard (and ESOMAR has not adopted such a question or issued a best practice in this area), many organizations in many places have.

This is a review of guidance found in a few of those proposals.

Perhaps the first question is “Does it really matter?” That answer is yes.  According to Wikipedia, some type of legal recognition of a third sex, or of indeterminate or transitioning sex, exists in Australia, Canada, Germany, India, Nepal, New Zealand, Pakistan, Thailand, the United Kingdom and the United States.

The second question might be: “What should we be asking – sex or gender?” The relevant literature seems to suggest that the better option is to ask about gender, which is about identity.

The Market Research Society in the UK has issued a guideline that may be applicable to market research in other countries.  It notes the possible difference between anatomy and identity, as well as the existence of gender change after birth.  The United Kingdom protects transgender individuals from discrimination, and legally recognizes the adopted gender.  Since the MRS code requires that participants be permitted to express themselves as they wish and to incur no harm due to research, this affects how to ask about gender.

It also reminds researchers that at the very least, an “other” response to a question about sex or gender identity is necessary.  At times, interviewers may also accept a “prefer not to say” option.

Guidelines from Australia also suggest a possible introduction telling the respondent why data on sex and/or gender is being collected.  Here is that example: “In order to avoid making assumptions, I am required to ask you to report your sex.  Like information on age, sex is a primary means of measuring and analyzing many aspects of the population, for example population counts and projections, health, etc., and stringent confidentiality measures are applied to all statistical data sets to ensure the privacy of individual responses.”

A final question to ask is “What does your research require?” There may be times when asking about sex or gender is irrelevant.  There may also be times when research requires learning whether a person is transgender.  For that there are additional questions.  The Williams Institute in the United States, which studies human sexuality and gender identity, offers guidance for questions on this subject.

ESOMAR itself has not issued a standard.  It urges researchers, as always, to observe local regulations, and to monitor any guidance from local associations.  This is becoming a sensitive area; consequently, this is an appropriate time to examine your standard questions and consider potential revisions.

Kathleen Frankovic is a member of the ESOMAR Professional Standards Committee