For so many decades this was binary: male and female. For an increasingly well recognised, and, one hopes, an increasingly better respected, set of the population this binary categorisation is no longer enough. However, this brings challenges, disupted space.
This has become pretty complicated! I’ll try to keep this to the implications for statistical analyses of psychological intervention data.
Details #
“Genetic, Genomic and biological” sex #
I think the first thing to get out of the way is the distinction between genetic sex and the whole complexity of “gender” which is so much more than genetic or biological. Even genetic sex is not binary though the variants other than of having XX sex chromosomes (genetically female) and XY (genetically male) probably represent over 0.03% of us, i.e. over 3 in 1,000. Chromosomal sex is not just about long recognised disorders of the number of sex chromosomes (“aneuploidy” is the technical term for having some number of chromosomes other than the typical number for the species, i.e. some number other than 46 for humans and includes having other than two sex chromosomes). It can be complicated by “mosaicism” when not all the cells in the body have the same chromosomes. Genetic sex is actually more than chromosomal sex as it’s increasingly understood that genetic issues beyond the simple count of X and Y sex chromosomes, genomics, determines “biological sex”, i.e. anatomical and physiological aspects of “sex/gender”.
Gender #
OK so what’s gender? I used to be fairly comfortable with a phrase something like “it’s the psychological expression of, experience of, well, gender influenced by biology but going way beyond that”. I don’t think that’s good enough now and I’d like to have “within the person’s cultural and subcultural construction(s) of gender.
We have an expanding recognition of “disorders of sexual development” in the more biological realm but also of how different cultures and different individuals have gender options way beyond “male/female” or even a “male/female” dimension and we have increasing recognition of the misery of individuals who experience their gender as outside the options their [sub]culture offers and the viciousness of much cultural reaction to this. So I know of and respect “non-binary”, “other” and I recognise that people transition, i.e. have their anatomical and physiological sexual features changed surgically and pharmacologically. I also recognise that this intersects with sexual interests, and people saying they don’t experience any sexual interests: that seems to me an intersection but of two moderately distinct issues of gendered personal identity and sexual preferences.
How does this have implications for quantitative therapy data? #
Given that gender in all its emerging complexities impinges radically on social, economic and emotional states I think we should be trying to find variables that capture it usefully and I feel increasingly that the old binary of “M/F” no longer suffices. However, there are two issues here:
- the hurt that only offering a binary category causes for some (and, though I sympathise far, far less with this, the wrath that going beyond binary elicits from some)
- how to use whatever categories we have
The challenges are the classic problems for quantitative analysis of categorical data:
- having your data corrupted by people using categories that aren’t really accurate for them
- having people omit the question
- having people simply not participate in any data collection when they hit a category system that they feel doesn’t allow them to answer honestly
- having categories that are imposed by a country that clearly aren’t OK for particular subpopulations (clearly a huge issue for services for gender dysphoria)
- having categories that are so complicated that a high proportion of participants won’t use them
- ending up not knowing what analyses of our data, using whatever categories we have, is impacted by those three
- a danger of people looking for categories that are universal, independent of culture/country
- a related danger of assuming that categories that work for one age group are optimal for a different age group even withhin a single culture/country
At the moment I see no simple resolutions of these issues and think we simply have to become much more careful with our analyses whatever categories we have in our data. If the dataset and data collection method allows it I think we should use categories that allow comparison with official national statistics (which may or may not mean using the legal gender options for the country though sometimes official national surveys may use far more complex categories than the legal ones but the national statistical ones may be too complication for use in therapy services and the legal categories often binary and unacceptable for many services and their clientele) I think we should allow not answering gender questions as it at least allows us to assess the rate of that decision. Generally I think we should look at opting out against age. Sometimes it may be possible to look at total refusal to participate against age though often the constructon of the data collection may not give an ethical option for that. Where opting out of gender categories is allowed in the data collection that should be looked at against age.
I think it will be decades, perhaps generations before we perhaps reach categorisations that are accepted by most people within a culture/country so that we can also have good referential data using those categories that won’t be plagued by the biases of misclassification and/or of missing data. Meanwhile, let’s be particularly careful to respect the complexities and ethics of all of this even if that may mean that bland reassurances that all our offers of interventions for gender dysphoria will be “evidence based”, if that means based in unarguable and generalisable data from quantitative datasets.
Try also #
Nothing yet.
Chapters #
Gender comes up in a number of chapters, particularly chapter 5 but I’m now ashamed to say that all mention of gender in OMbook is binary.
Online resources #
None currently.
Dates #
First created 19.vii.24.
