[hombre_fatal]

Nitpick: all studies only show associations. We draw causal inferences from associations. e.g. Duhem-Quine thesis.

There’s a meme on social media where people think studies break down into association studies vs casual studies or something.

[Natsu]

If you merely observe thermometers and which days feel hot, you'll see a link but won't be able to tell which way causality goes. If you set the thermometer to a high value in a cold room and notice that you don't get any warmer, you'll quickly realize which way causality flows.

So I agree with the GP post, I want to see them modeling & testing causality here.