[bcjdjsndon]

> Simply saying a confounding variable was missing isn't helpful because it's impossible to control for all confounding variables

That's the problem with studies though. My point is why defend something so ardently when we all agree it's shit but the best we have?

[estearum]

Uhh... there's a truly gigantic gap between the shit end of "has confounders" and the good end of "has confounders", and locating where a particular study lands on that spectrum is a foundational skill to evaluating virtually any factual claim you'll ever encounter...

I am advocating for people to identify that there's a spectrum instead of just saying "oh well, every study is imperfect" which diverts attention from understanding in what particular ways is it imperfect.