Been in another peeing contest with a Bayesian. They're as bad as the tea baggers in believing so many things that just ain't so.
The fundamental flaw in Bayes is the assumption that investigators are both objective and honest. Humans always have an agenda. In the pharma world, sponsors will do most anything to get a drug approved, even glossing deaths. FDA, on the other hand, will renege on agreements (SPA, generally) if they see new data that is worrisome.
Since there are an infinite number of priors available to any study, there's every reason to believe a sponsor will choose the one which best fits the agenda. Good luck with that.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment