RCTs having limitations and/or weak or misleading design does not automatically make observational studies better.

Observational studies having limitations and/or weak or misleading design does not automatically make RCTs better.

Limitations and design isn& #39;t a zero sum game.

I don& #39;t understand the "team RCT" vs "team obs" thing.

RCTs and obs studies are not diametric opposites that must always be battling.

You can critique observational studies without RCTs being the perfect and only answer and vice versa. It& #39;s fine, I promise.

You can also eliminate automatic hierarchy without falsely promoting weaker designs.

ALL causal inference is subject to the same fundamental ruleset.

If you review all studies against those same standards, the weak ones are weak, strong ones strong, period. No dichotomy needed.

Nearly every time someone critiques a weak study, the conversation derails into RCT vs obs camps, where people are arguing about the idea of obs and RCTs, but not the actual limitations of this study itself.

I& #39;ve done it too, but it& #39;s a really unhelpful reaction.

It& #39;s worth taking a moment and asking yourself:

Am I <defending/critiquing> this <RCT/obs study>, or am I <defending/critiquing> <RCTs/obs studies> in general?

If it& #39;s the latter, it might not be relevant or helpful to the discussion at hand.