MTO Colloquium Tuesday, November 10, PZ 45 at 12:45h by Prof. dr. Irene Klugkist (Utrecht University) ‘Updating Bayes factors for informative hypotheses in replication studies’ In the past decade, several researchers have worked on the development of a Bayesian model selection approach for the evaluation of informative hypotheses. An example of such a hypothesis is Hi: mu1<mu2<mu3<mu4, where each mu represents a population mean and the ordering expresses a theory based, a priori expected pattern among these means. In my current research, the goal is to combine evidence from multiple studies investigating the same informative hypothesis, that is, I examine the behavior of the Bayes factor, our tool for model selection, when studies are replicated. Using the binomial model as an example, I will show some preliminary results using different approaches of combining information. I like to discuss these results and possible interpretations as well as implications for users of our methods and software.