I'm quite aware of P-values :) I neither mentioned posterior calculation of probabilites nor talked about correlation studies, so I'm not seeing your point in linking to Gelman's article?