Reference priors and Bayes factors provide links between frequentist and classical statisticians. Reference priors `let the data speak for itself' in accord with the frequentist paradigm, and Bayes factors provide an analogue to frequentist hypothesis testing. Because of these links, pressure to adopt their use is strong when one opts for a Bayesian analysis. In this talk I will discuss the use of reference priors and Bayes' factors in the analysis of clinical trial data, explore what we gain and lose in doing so, and suggest alternative directions.

Back to Workshop Schedule