Bayesian statistical methods are an increasingly popular alternative to the “classic” frequentist methods. Proponents of Bayesian methods claim they are superior to frequentist methods because they have better principles for statistical inference. But they go further than this, also claiming that Bayesian methods can deliver what researchers always have wanted but frequentist methods have denied: A truly cumulative science and genuine hypothesis-testing. Sceptics, however, have said that Bayesian methods have been over-sold, claiming that they are more subjective and more difficult to use than traditional methods and don't always deliver what researchers need. This seminar offers an introductory overview and critical appraisal of the prospects for the so-called "Bayesian revolution".
Michael Smithson is a Professor in the Research School of Psychology at The Australian National University in Canberra. He is the author of six books, and more than 160 refereed journal articles and book chapters. His primary research interests are in judgment and decision making under ignorance and uncertainty, and in statistical and quantitative methods for the human sciences.