Comment by vkou

No study should be good enough for HN. If a single non-obviously-flawed study is enough to convince you to something, then you can be convinced of anything and everything under the sun.

One study can find any effect it's looking for.

A study shouldn't move consensus. A study finding an effect is a signal that more studies should be done.

Once they are done, and people who know their stuff pour through them and reach some consensus is the sort of bar that needs to be crossed for a reasonable non-expert to 'follow the science'.

And sometimes those experts get it wrong, and accepting that degree of uncertainty is part of it.