Understanding Uncertainty with Priors and Posteriors: An Introduction to Bayesian Thinking

Overview

This course focuses on the principles of Bayesian analysis with the aim to enable participants to apply Bayesian methods on their own research and understand other people's results via Bayesian analysis.

Bayesian analysis is an alternative approach to the statistical techniques that are commonly used throughout most of the research world for the analysis of data. It's core principle stems for the idea that experiments are not abstract devices, hence knowledge existent prior to an experiment should be incorporated formally and openly in the analysis.

Learning outcomes

At the end of the course, delegates should be able to think statistically using Bayesian reasoning and have the tools to implement such reasoning formally into a statistical analysis of data. In particular, delegates will be able to:

• Differentiate the principles of both classical and Bayesian statistical frameworks, and the advantages / disadvantages of each.
• Understand the concept of a probability distribution, including the main established distri-butions (e.g. the normal distribution) and their parameters.
• Recite the different contributing probability distributions that play a role in Bayes theorem of conditional probabilities.
• Incorporate their prior beliefs into an analysis alongside evidence presented by the collect-ed data.
• Use the results of Bayesian analysis to answer their research question (through prediction, estimation hypothesis testing and credible intervals).
• Have appreciation for the computational complexity of fitting some Bayesian models.