W08: Updating undergraduate courses to include Bayesian inference

Traditional undergraduate statistics curricula tell only part of the story of statistics: frequentism. The Bayesian approach is often left for seasoned graduate students. This leaves most students unprepared for the increasing wealth of Bayesian work used in industry! Moreover, Bayesian methods have more convenient interpretations. We show that Bayesian methods are accessible early on in undergraduate education.

Specifically, we will summarize Bayesian inference, taking care to differentiate it from frequentist inference. We will then follow up with how to incorporate prior information into a model and combine it with data. We will construct a Bayesian simple linear regression model and spend time conducting inferences on both parameters and on predicted values.