It’s lesswrong; will you ever find a higher density of Bayesians? We know P-values don’t account for P(H|~E), P(H|E), P(E) or P(H). P(E|H) is not enough information to determine your proper probability assignment to H, given you see E, and P(E|H) is all a P-value tells you.
It’s lesswrong; will you ever find a higher density of Bayesians? We know P-values don’t account for P(H|~E), P(H|E), P(E) or P(H). P(E|H) is not enough information to determine your proper probability assignment to H, given you see E, and P(E|H) is all a P-value tells you.