3 ms·
Frequentist approaches do not use any prior at all. Indeed some of the issues that frequentists wind up getting genuinely concerned about, such as repeated sig
by btilly 14y ago
Frequentist approaches do not use any prior at all. Indeed some of the issues that frequentists wind up getting genuinely concerned about, such as repeated significance testing errors, are clear logical fallacies if you're doing any sort of Bayesian analysis.
As an example, most of http://www.evanmiller.org/how-not-to-run-an-ab-test.html http://www.evanmiller.org/how-not-to-run-an-ab-test.html will look dead wrong to any informed Bayesian.