"One downside of Bayesian statistics is that it requires prior information — and often scientists need to start with a guess or estimate. Assigning numbers to subjective judgments is “like fingernails on a chalkboard,” said physicist Kyle Cranmer, who helped develop a frequentist technique to identify the latest new subatomic particle — the Higgs boson.It's actually worse than that. It is a favorite of ideological hacks who place their ideology into the equation up front, thereby getting the results they want rather than objective knowledge. Bayesian calculations are an open invitation to pretend that circular arguments are justified statistically.
Others say that in confronting the so-called replication crisis, the best cure for misleading findings is not Bayesian statistics, but good frequentist ones. It was frequentist statistics that allowed people to uncover all the problems with irreproducible research in the first place, said Deborah Mayo, a philosopher of science at Virginia Tech. The technique was developed to distinguish real effects from chance, and to prevent scientists from fooling themselves.
Uri Simonsohn, a psychologist at the University of Pennsylvania, agrees. Several years ago, he published a paper that exposed common statistical shenanigans in his field — logical leaps, unjustified conclusions, and various forms of unconscious and conscious cheating.
He said he had looked into Bayesian statistics and concluded that if people misused or misunderstood one system, they would do just as badly with the other. Bayesian statistics, in short, can’t save us from bad science."
[emphasis added]
That's not to say that with legitimate use of non-biased input, Bayes calculations can't be valuable, as in the case of the fisherman rescued by the Coast Guard which used Bayes to anticipate the location of the drifting man. But that uses known information, regarding physical data which is not ideological as an input to the calculation. That is far different from trying to calculate, say, the existence of a deity, where any input is prejudiced by definition.
Whenever Bayes is used, the calculations must ALWAYS be scrutinized for bias, because in some venues they always will be. And that is, indeed, like fingernails on a chalkboard.
2 comments:
Metacrock had a debate with Jeff Lowder on Infidels about this:
Metacrock: Answer to Lowder's attack part 1
Lowder's attack part 2
Two things of humorous interest (to me):
(1) It is highly amusing that the Reverend Thomas Bayes, a religious person, is the current "darling" of the atheists, especially over on The Secular Outpost. Did Jeff Lowder not get the Hitchens' "memo" that "religion poisons everything?" (Maybe that should have been written as "the late Hitchen's memo," with the appropriate equivocation about the meaning of "late" but I digress.) If I'm not mistaken, "everything" would include Bayes' Theorem. (At least, that is my assumption without attempting to apply Bayes' Theorem to it.) Perhaps a chorus or two of "Unclean! Unclean!" would be appropriate from the atheist echo chambers.
(2) It is highly amusing to see atheists using physical hammers in a futile attempt to drive metaphysical nails. It is even more amusing (ROTFLMAO funny) to see statistical probabilities being used to "prove" inferences. But then I am easily amused. . .
Post a Comment