Thread by @jessesingal

1/ Jack Turban ran a correlational study on a known-to-be-broken self-report dataset. He found no statistically significant correlation between youths' access to hormones and the most important suicidality indicators -- ideation with plan or completed attempt -- w/one exception:

2/16-17 year olds in his study who DID go on hormones were more than twice AS LIKELY to have a suicide attempt requiring hospitalization. By his own (completely flawed) reasoning, that causation can be extracted from this mess, access to hormones caused serious suicide attempts.

3/ How does he describe this study in his new WaPo article? As one of "at least 14 studies [that] have examined the impact of gender-affirming care on the mental health of youths with gender dysphoria and have shown improvements in anxiety, depression and [link:]suicidality."

4/ That's only true if you cherrypick -- if you accept as true the statistically significant correlations with suicidal ideation (which is there), but ignore all the null results, and that one scary positive correlation, when it comes to the more serious suicidality measures.

5/ In a sense this is all moot, though. Michael Biggs reanalyzed the Turban team's data and as soon as he separated out testosterone from estrogen (why wouldn't you?), he found that their claims fell apart disastrously.

journals.plos.org/plosone/articl…

6/ I wrote a whole book about half-baked science. I'm familiar w/cherrypicking and covering up deficiencies or inconvenient results in the data. Doing this w/power-posing or grit is unfortunate and causes some harm. Doing this with data on *adolescent suicide* is unconscionable.

7/As far as I can tell @jack_turban does not ever publicly respond to the many critiques of his work. I think folks should pin him down on one question: Is he concerned the 16-17 yos in his study who went on hormones were twice as likely to be hospitalized for a suicide attempt?