I have a survey with multiple demographics, including gender, age group, and state (USA).
Because I want my sample to be more representative of the United States as a whole, I am using raking to weight different responses. Normally I use a 50.8%:49.2% female:male ratio, but I include an "other" category in my surveys for inclusiveness.
At a glance, it appears that about 0.6% of the US population is transgender (source), so I would be quick to use that. I'm not sure that's what many/most people are trying to indicate when they select "other", though: i.e., would a trans man would select "male" for gender, or would a trans woman would select "female" for gender?
According to this study from the Williams Institute of Law at UCLA (pdf), about 0.6% of individuals in the US identify as transgender, according to this question:
Do you consider yourself to be transgender?
Yes
No
[If Yes] Do you consider yourself to be male-to-female, female-to-male, or gender non-conforming?
If the interviewer is asked for a definition of transgender, they respond: Some people describe themselves as transgender when they experience a different gender identity from their sex at birth. For example, a person born into a male body, but who feels female or lives as a woman would be transgender. Some transgender people change their physical appearance so that it matches their internal gender identity. Some transgender people take hormones and some have surgery. A transgender person may be of any sexual orientation – straight, gay, lesbian, or bisexual.
So this figure appears to include transgender, as well as anyone who considers themselves gender non-conforming. I don't know how the former group (if they don't consider themselves part of the latter group) would identify themselves on a survey, although I assume "gender non-conforming" would respond "other".
At the end of the day, I'd ideally like to avoid under/over-weighting this group by a significant margin, especially since I've noticed certain US states have a much higher incidence of the "other" gender category, which can fudge weights created from raking.
What is the best approach for weighting the "other" category in this sort of circumstance? Does anyone have any experience with weighting responses like this?
