We regard confidence distributions as the ‘gold standard’ for empirically based distributions for a parameter

“The two main schools of statistics, the Bayesian and the frequentist, are not quite at peace with each other. Statisticians are to a considerable degree split between these schools, and so are the areas of application. Most statistical work in the field of genetics is, for example, of a Bayesian character, while analysis of clinical trials in medicine is almost entirely frequentist. Both methodologies have their specific advantages. The Bayesian provides distributional inference (posterior distributions), and the frequentist avoids colouring the results by subjective input in the form of prior distributions. But is this split necessary?”

“The chasm between frequentist and Bayesian methodology is in our view unfortunate. We regard confidence distributions as the ‘gold standard’ for empirically based distributions for a parameter, whether found by fidicual, Bayesian or other methods. We hope the present book is a contribution to a synthesis of the diverse methodologies for statistical inference in science. ”

Leave a comment

Your email address will not be published. Required fields are marked *


+ two = 8

Leave a Reply