Clinical perspective, offered as context rather than as advice. The honest answer is that the effect is real, the magnitude is contested, and the individual variation is larger than either. Those three things can all be true at once, and most arguments here are two people holding different parts of that.

sean_dublin said:

The honest answer is that the effect is real, the magnitude is contested, and the individual variation is larger than either.

I read this differently from sean_dublin, on substance rather than tone. This treats a plausible mechanism as a demonstrated one. Plausible is where you start looking, not where you stop.

sean_dublin said:

The honest answer is that the effect is real, the magnitude is contested, and the individual variation is larger than either.

Right, and the exception is worth naming rather than left implied, because the exception is where people get into trouble.

sean_dublin said:

The honest answer is that the effect is real, the magnitude is contested, and the individual variation is larger than either.

Thank you for spelling out the reasoning rather than just the conclusion. Sending this to two other people who asked me the same thing last week.

Moderator note: good thread. Keeping it here rather than moving it, because the question is general enough to be useful.