Welcome (?), Infowars Readers
Scott Alexander addresses Infowars readers about a misattributed survey analysis, providing caveats and emphasizing proper attribution and vaccine safety.
Hello to all the new readers I’ve gotten from, uh, Paul Watson of Infowars. Before anything else, consider reading this statement by the CDC about vaccines.
Still here? Fine.
Infowars linked here with the headline Survey Finds People Who Identify As Left Wing More Likely To Have Been Diagnosed With A Mental Illness. This is accurate only insofar as the result uses the publicly available data I provide. The claim about mental illness was made by Twitter user Philippe Lemoine and not by me. In general, if a third party analyzes SSC survey data, I would prefer that media sources reporting on their analysis attribute it to them, and not to SSC.
As far as I can tell, Lemoine’s analysis is accurate enough, but needs some clarifications:
1. Both extreme rightists and extreme leftists are more likely than moderates to have been diagnosed with most conditions.
2. Leftists might be more likely to trust the psychiatric system and get diagnosed. My survey shows some signs of that. Liberals are 60% more likely than conservatives to have formal diagnoses of depression, but only 30% more likely to have a self-diagnosis of depression.
3. Leftists might be more likely to think of their issues through a psychiatric lens than rightists, meaning that even the self-diagnosis numbers might be inflated.
4. The SSC survey is a bad sample to use for this, not just because it’s unrepresentative, but because it might be unrepresentative of different political affiliations in different ways. For example, SSC Marxists really are surprisingly depressed, but maybe the only Marxists who would read an anti-Marxist blog are depressed Marxists looking for things to be miserable and angry about (though see below for some counterevidence).
5. A commenter on Lemoine’s tweet links to this blog post by someone who found the same thing in the General Social Survey. The General Social Survey is much larger and more rigorous than my survey, and there’s no reason to care what my survey has to say when there are GSS results available.
In general, if a survey analysis is posted on this blog, it’s mine. If not, then it isn’t mine and you should link to whoever performed it and let them clean up their own mess. Thanks – and seriously, vaccines are fine.