I’m a qualitative researcher, and I regularly write screening questionnaires. The basic demographic questions we have asked for decades are: age, gender, marital status, education, employment and sometimes ethnicity. We do this for two reasons: 1) to ensure we include a mix of people, and 2) to exclude people who have no experience or interest in the topic. So, for example, you would not recruit men to discuss tampons, or women to discuss jock itch. Thanks to Christopher Ireland for that vivid example. Christopher, by the way, is a woman. And that brings me to the impetus for writing about this.
What’s in a Name?
I recently conducted a study about people and their pets. I decided to leave the gender question off the screening questionnaire. It seemed superfluous and possibly “old school” in light of new gender paradigms. But as recruiting progressed, I noticed the sample was skewing female so I began to place a priority on selecting men to keep it balanced. I was happy when “Joey” replied to my email saying yes. But when I called Joey I discovered she is a woman. That sent me back to the list of names. I began to question my assumptions…was it correct to assume Shelby and Zarrin were women? That led me to question my own, and my peers’ way, of handling the gender question. Should we take a fresh look at if, how, when, and why we ask it?
What Are the Learning Objectives?
Let’s take “if” we should ask about gender first. Carla Sarett, seasoned researcher and deep thinker, takes the position that our ultimate responsibility as researchers is to find answers for our clients.
“Marketers need to sell pedestrian things, like TV shows and cookware; and most buying patterns are predicted by the pedestrian categories of sex and age. If market researchers forget that, they’re neglecting their duty to the client– which is to sell products, not further political freedoms. Furthermore, the Census collects data by sex and age, and it’s impossible to weight data sets without those variables.”
Qualitative research samples are small, and do not require weighting, but Carla’s point is still valid: the decision about whether to collect gender data should be based on the learning objectives of the study—not to further political freedoms. So, the answer is YES, you must ask the gender question if you are conducting quantitative research from which statistical inferences can be drawn. But if you are conducting qualitative research where the goal is exploration, you may not need a gender screen. It depends on how important gender is to the topic under study.
How Do We Ask?
I’m not alone in my concern about HOW to ask the gender question given that our binary model is a thing of the past. Ryan Holeywell, Senior Editor at Rice University’s Kinder Institute for Urban Research notes in his blog that The US Census Bureau is currently grappling with this question. But they have no plans to test a transgender option in the near future. Further Holeywell notes:
“…research data shows even if there was a transgender option on the Census, it may only be used by a small segment of the transgender population. Those were the conclusions published earlier this year in the journal Transgender Studies Quarterly.”
I asked two of my male-to-female transgender friends to chime in: one said she would choose “female” because that’s how she sees herself. The other friend said it would depend on the context, but that she would lean toward “other” most of the time. It may be that some transgender people identify more with the label “male” or “female” than they do with the possibly stigmatizing label “transgender.” If given a choice NOT to be stigmatized, who wouldn’t choose that?
“Other” and its Alternatives
It is unclear whether the “other” option also stigmatizes people, but in my view it is a better alternative than “transgender.” The reason being that “other” is less defined, and therefore more inclusive. That way, a person who may not label themselves as male or female can appropriately be included in a study about cookware.
If tampons are the topic of discussion, and a transgender person chooses the “female” answer option, it is incumbent upon me, as the writer of the screening questionnaire, to filter out people who have no experience with tampons. I would do this by asking about past and present tampon use. If they have no experience using tampons for menstruation (whether they are transgender or not) they will appropriately be screened out of the study.
Another idea we could consider is “prefer not to say” which is a standard answer option for the income question. In my view, the best alternative to the binary model when asking about gender for qualitative research is to offer these four answer options: male, female, other, prefer not to say.
As always, we would love to hear from you on this or any other nerdy research topic in the comments.
Stay tuned for our next topic: Do we really need to ask people if they are widowed or divorced?