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

It’s 500k people for over a decade. If any study has the data to do it, it’s this one.



Sample size is useless at removing confounders.


It's an observational study. You can't make conclusions based on it. But it can be a basis for more detailed study.




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

Search: