Interesting, though I’d be hesitant to read too much into that. To the extent this rationality project is succeeding, I’d expect people here are more likely to be exposed to the full range (or at least a large range) of human variation, and more likely to correctly determine if they’re actually any particular minority group, with people defaulting to not-a-member on priors without significant reflection.
This seems like a really hard thing to survey consistently that’ll be systemically skewed by degree of prior exposure to the topic in question in the survey population. If you ask someone point blank “do you have [minority quirk they’ve probabably never heard of]?”, they’re unlikely to return a meaningful answer in the time surveys have. Folks spend months or years figuring that out. I don’t see how you avoid measuring P(has unusual quirk & has invested the time and developed self awareness to realize A if true) instead of just P(has unusal quirk). Speaking very generally as I expect this holds outside of the realm of sex/gender/etc identity issues too.
misc: I didn’t check out the specific papers linked. I recall Scott commenting on one or more of his yearly surveys the degree to which the LW and SSC communities end up being outliers on just about every measure like this (much higher than base rate) but didn’t find the specific comments back.
Interesting, though I’d be hesitant to read too much into that. To the extent this rationality project is succeeding, I’d expect people here are more likely to be exposed to the full range (or at least a large range) of human variation, and more likely to correctly determine if they’re actually any particular minority group, with people defaulting to not-a-member on priors without significant reflection.
This seems like a really hard thing to survey consistently that’ll be systemically skewed by degree of prior exposure to the topic in question in the survey population. If you ask someone point blank “do you have [minority quirk they’ve probabably never heard of]?”, they’re unlikely to return a meaningful answer in the time surveys have. Folks spend months or years figuring that out. I don’t see how you avoid measuring P(has unusual quirk & has invested the time and developed self awareness to realize A if true) instead of just P(has unusal quirk). Speaking very generally as I expect this holds outside of the realm of sex/gender/etc identity issues too.
misc: I didn’t check out the specific papers linked. I recall Scott commenting on one or more of his yearly surveys the degree to which the LW and SSC communities end up being outliers on just about every measure like this (much higher than base rate) but didn’t find the specific comments back.