Statistical Foundations and the Philosophy of I. J. Good

• Activity Number: 149
• Type: Invited
• Date/Time: Monday, August 2, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Memorial
• Author(s): James Berger*+
• Companies: Duke University
• Keywords: Good ; Bayesian ; Foundations

Abstract

"An honest person is something of a scientist; an honest scientist is something of a statistician; an honest statistician is something of a Bayesian; an honest Bayesian is something of a subjectivist; and an honest subjectivist knows that judgments are seldom, if ever, precise." This quote from Jack Good - like all of his work - covers a lot of ground. Indeed, Good worked over such a vast canvas that any type of overview of his work and impact would be woefully inadequate. The talk will thus instead focus on discussion of Good's statistical philosophy, as evidenced by the above quote. While Good was an unequivocal Bayesian, he was the first to clearly articulate the limitations of the purist Bayesian position; to seek incorporation of non-Bayesian techniques to help overcome these limitations; and to try to translate crucial Bayesian understandings into non-Bayesian language.