I think you get half of the value just by noticing that in general P(A|B) is not P(B|A).
Like when people tell you “this study has p=0.05, therefore there is a 95% chance this is true”, and you know they are wrong, even if you don’t know the actual base probability of “a published study being right”.
I think you get half of the value just by noticing that in general P(A|B) is not P(B|A).
Like when people tell you “this study has p=0.05, therefore there is a 95% chance this is true”, and you know they are wrong, even if you don’t know the actual base probability of “a published study being right”.