Hacker News new | past | comments | ask | show | jobs | submit login

I don't know what she could possibly mean by 'proper controls'. They're twins. They are better controls for each other (full siblings raised simultaneously of exactly the same age) than you see in 99.9% of sociology or education studies; in some respects they are the best controls that are possible, since you can't run a person through school multiple times to see if they get into college or not. And this study also came with polygenic scores so you can check the genetics directly if you've fallen for handwavy criticisms like of the EEA.

(Heck, she's not even right when she mouths 'correlation != causation', because if you used the PGS to compare siblings, their PGS differences are randomized at conception and the correlation = causation! As far as I can tell, OP didn't do this, just scoring unrelated individuals, but they could have, many other studies have, and she doesn't seem to've checked herself.)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: