> Yet if you want to set your threshold for evidence so high that nothing short of a large scale RCT will answer the question you can ignore all of this, because such an RCT will never be produced for obvious ethical reasons.

Yes, but this means existing studies can't be relied on to justify a conclusion that would have required the impossible study. Saying that a study isn't practical can't be used to justify making a policy decision based on circumstantial evidence.

My use of marijuana studies is meant only to show how bad science steers public policy, not to advocate for its use. There are other, better reasons to avoid that drug.

> This pattern is common in topics where people simple don’t want to believe a casual relationship might exist.

This is why science exists. But if the science cannot be carried out, then we have no right to draw conclusions that would require actual science to be legitimate.

This is moving away from the original topic, which is the kind of sloppy science common in modern psychology, where people wave their hands instead of their scientific results.

> I knew a guy who ...

A red herring, but I think you knew that.

> Yes, but this means existing studies can't be relied on…

Exactly my point with this line of thinking: It’s used to argue that you can’t rely on any studies, therefore it becomes a free license to reject everything and inject your own desired conclusions.