HPSS 2008
 PRINTER FRIENDLY VERSION
P1 - Introducing Bayesian Data Analysis to Non-Statisticians in a Frequentist Course: Illustration with NDNQI Data
View Presentation View Presentation
Nancy Dunton, University of Kansas School of Nursing  *Byron J Gajewski, University of Kansas Schools of Nursing and Allied Health, Center for Biostatistics   Stephen D. Simon, Department of Medical Research 

Keywords: Shrinkage, Prior, Posterior, ANOVA, Teaching, Hierarchical Model

This poster is targeted towards health policy statisticians who want to teach Bayesian ideas to non-statisticians. Analysis of Variance (ANOVA) is arguably the most widely used statistical model in practice. In this poster we discuss one-way ANOVA with the goal of introducing Bayesian statistics to non-statisticians in a one-hour seminar. The Bayesian seminar is built upon interesting ties to the classic ANOVA creating a smooth transition for students in a frequentist-based course to be introduced to Bayesian ideas. We discuss the limitations of this seminar in the context of our assumptions of what the students can/can't do. Job enjoyment survey data from registered nurses (RN) who work in critical care units in hospitals that are members of the National Database of Nursing Quality Indicators (NDNQI) illustrate the approach. We invite other instructors to try our proposed method.