|
|
|
|
|
by laichzeit0
188 days ago
|
|
Unfortunately, this is an observational study and when you get to the confounding part, they kind of shrug their shoulders and say “well, we included a bunch of covariates that should reduce make the bias go away”, but there’s no causal diagram so we have no idea how they reasoned about this. If you’ve read even something layman friendly like Pearl’s Book of Why you should be feeling nervous about this. |
|