Thread by @nathankw

A quick thread on this new paper which it has been claimed refutes the notion of social contagion in trans adolescents. Even by the poor methodological standards of this topic it's remarkable how weak the evidence is compared to the claims being made. 1/x

https://twitter.com/jack_turban/status/1554696508166967297

The central argument is that according to a single survey source, the rate of trans identification is going down (2017 vs 19) disproving the social contagion (SC) hypothesis. It's pretty obvious that inferring a trend from 2 data points is virtually meaningless. It reminds me 2/x

of the bogus argument used by climate deniers who claimed warming had stopped by looking at a short trend (though to be fair to them even they didn't just use 2 years). Furthermore, even if Jack's right and the rate is now reducing that doesn't disprove SC. It simply 3/x

underlines the key point that much more (& better) data is needed. Another argument is about the sex ratio. Here the paper gets properly bizarre. In this paper Jack is careful to refer to "sex assigned at birth" - to distinguish between the sex that they now identify as. 4/x

But the survey only asked for sex - so we have absolutely no idea whether respondents were replying with their sex before or after transition. Bafflingly Jack just decides they must mean SATB. I've looked at the references he quotes and can't find anything to back this up. 5/x

Certainly all the online discussion I've seen makes it clear that many trans youth do not understand their sex now to be their sex assigned at birth. Isn't that the whole point of the "assigned at birth" terminology? But again, if we give Jack the benefit of the doubt and 6/x

assume his analysis of the survey is correct. What does that tell us? Well, that something interesting is going on. The proportions and rate of change in adolescents identifying as TGD in a survey is very different from the population visiting clinic (see the UK GIDS data) 7/x

So that tells us these two populations are different - which is not at all surprising. There are undoubtedly many different overlapping phenomena here. "Being trans" is a not a single thing and there will not be any single cause. I don't believe social contagion is the only 8/x

explanation for all trans identifying young people. But Jack's paper doesn't come close to ruling it out as a factor. What we desperately need is good impartial research open to finding complex answers, not simplistic activism aiming to shut down legitimate inquiry. 9/9

Quick update - as above, a crucial question is whether trans identifying people replied to the sex question with their sex or gender. @jack_turban says he knows that they answered with their birth sex. But here's a paper by Michelle Johns, the lead LGBT researcher for the CDC ...

who did the original survey. ncbi.nlm.nih.gov/pmc/articles/P… - This is what she says about that issue. We have no idea how they're responding. This is just basic digiligence that even a brief literature review would have uncovered. I'd love to hear from @aap_peds how this was published.

Even more basic errors are emerging. @jack_turban says the data he's using comes from 16 states but in fact only 9 states have available data in 2017 and 14 in 2019. So he's comparing data from different states. The makes any trend (if you can even have a trend with 2 points)...

totally meaningless. See osf.io/preprints/soca… which interestingly is written by researchers who ideologically align with JT. The @jack_turban paper is even worse than I thought (and I thought it was truly awful before). Thanks to @jessesingal for the link.