Kinds of Bayesians (comment on articles by Berger and by Goldstein)
Robert E. Kass · Bayesian Analysis · 2006
Philosophy is consequential because it guides action.In statistics, our beliefs about the nature of statistical inference guide us both in selecting statistical methods to apply and in choosing problems to research.Jim Berger and Michael Goldstein have articulated clearly two opposing viewpoints.Yet, both statisticians are highly-experienced, widely-respected Bayesians and, faced with a particular set of data, it is very likely that the two of them would reach similar conclusions.So where, specifically, do they disagree?And which points of disagreement are most consequential?To help in identifying key issues I have constructed a list of questions.Answers to these may be used to classify various kinds of Bayesians.