Introduction to Bayesian Analysis for Medical Decision Science

28 June 2026, 8:35 AM – 12:00 PM (local time)

Faculty

  • Christopher S. Hollenbeak, PhD – Professor and Head, Department of Health Policy and Administration, The Pennsylvania State University, United States

Course Overview

Bayesian methods are increasingly important in medical decision science for integrating prior evidence and quantifying uncertainty in decision-making. This course provides a gentle, practical introduction to Bayesian analysis, contrasting it with classical frequentist methods and demonstrating how to conduct Bayesian analyses with modern software.

Participants will learn to conduct Bayesian analyses using JAGS in R through a hands-on workshop covering common models, including proportions, means, linear regression, and logistic regression. The course is designed for those with no formal Bayesian training and emphasizes interpretation and application in decision-relevant contexts.

Learning Objectives

Participants will learn how to:

  • Explain the philosophical and practical differences between Bayesian and frequentist analysis
  • Understand Bayesian terminology, including prior and posterior distributions and their role in inference
  • Interpret Bayesian analysis results in the context of medical decision science
  • Conduct Bayesian versions of common statistical models using JAGS in R and RStudio
  • Apply Bayesian results to support decision-making in health economic evaluations

Course Format

This course includes a combination of lecture and hands-on workshop components. Participants will be introduced to Bayesian concepts and methods, followed by guided exercises using JAGS in R.

Topics include Bayesian versus frequentist approaches, probability and inference, prior and posterior distributions, simulation methods including Markov chain Monte Carlo (MCMC), and application of Bayesian models to decision-making contexts. The session will conclude with interpretation of results and discussion.

Participant Requirements

Basic knowledge of probability and statistics is required. Familiarity with R (data manipulation and running scripts) is helpful but not essential.

Participants must bring a laptop with R, RStudio, and JAGS installed. Slides, R scripts, datasets, and a reference list will be provided. Installation instructions and optional pre-workshop R tutorials will also be shared in advance.