How to write a trans- and nonbinary-inclusive gender demographic question on a survey in one picture (taken from a dissertation by M.L. Haupert, see following):
Further notes on gender demographic questions
This image is taken from a 2016 survey on Perceptions of Gender Questions performed as part of the doctoral research of Margaret L. Haupert at the University of Indiana, a survey I participated in and was deeply impressed by. Quoting from Dr. Haupert's 2019 dissertation discussing the results of this survey:
====
The short free-text question [see attached - pb] was best-liked and most identity-safe, but as described in Chapter 3, the complexity of analyzing its results make it unwieldy for general use. I recommend the multiple-choice gender question for most research purposes, as it is better liked by participants across contexts in both studies than the binary + other question. In Study 2, I experimentally demonstrated that participants who saw the multiple-choice question had higher expectations of belonging, higher comfort with research, and lower negative emotions toward researchers than those who saw the binary. In contrast, those who saw the binary + other question did not differ from those who saw the binary question on those measures.
----
Haupert, M. L. (2019). "Considerations for the development and implementation of transgender–inclusive gender demographic questions" (Doctoral dissertation). Retrieved from IUScholarWorks. http://hdl.handle.net/2022/22925
====
Whenever you perform a survey that asks for demographic data, you are asking people to trust you with some very personal information about who they are. When you ask a question like this, get it right.
P.S. "Transgender" describes medical history, not gender. "Are you trans" is a different question from "What is your gender". If you need to know, ask a second question.
(Apropos of nothing, the entire dissertation was apparently released under a CC-BY license - https://creativecommons.org/licenses/by/4.0/ - which I wasn't expecting but is kinda cool.)
If I were modifying this question today from this 2016 form, I would probably make two or three changes:
1. The question itself would be "What is your gender?" - people have grown kinda uncomfortable with the "identity" recently.
2. Is "non-binary" hyphenated? I always write it "nonbinary", and so does the Gender Census: https://gendercensus.com/post/183832246805/gender-census-2019-the-full-report-worldwide
3. This may be just me, but a "Genderfluid (specify further if desired)"+free text box option seems like a good idea - genderfluidity might be similarly common to agenderness.
...but I'm gonna be real: we got in touch with Dr. Haupert over email /two years/ after taking this survey, because it was just lightyears ahead of so many gender questions we saw in the time since, and it's got academic research backing it up.
-
@packbat I wonder if it might also be better to allow multiple selections? Or if it'd be better to put that under "other"
For either genderfluid or polygender people
@Felthry *nods*
We went clickbaity in the OP because we were frankly disgusted at some of the horrible messes that we've seen people make of a gender survey question, but yeah, it's definitely worth iterating on this template. This form was one person's best attempt in /March 2016/, and the state of knowledge has progressed since then; there's probably a better balance to be found between ease of scoring for research purposes and philosophical accuracy capturing the truths of the people being surveyed.
-
@packbat yeah, i was just making suggestions because you were talking about how you'd modify it. Everything can be improved.
@packbat
this form is wrong
@efi I mean, valid
@packbat tbqh the only correct form would be an input box that says "type your gender in as few words as possible, thanks"
and then again, it should be removed from anything that's not research
@efi ...it's kind of hilarious that you said that, because that was the exact result Dr. Haupert found in the original studies this survey question was based on: the best form /for the people being surveyed/ was a short free-entry text box. Haupert recommends this one because converting free-entry responses into data that can be collated and correlated is a lot of extra work for researchers.
(And, honestly, probably creates the same problems that we see with the survey in the first place, except without the people filling the survey out having /any/ control over how the researchers understand them.)
...but also yeah - this is a question for researchers asking demographic questions on surveys. Most forms shouldn't have these questions at all.
-