Module Aims: This module introduces students to Bayesian statistical methods in biomedical settings and their advantages and challenges, and provides skills for designing, assessing and interpreting Bayesian analyses using standard Bayesian statistical software.
Module Learning Outcomes:
By the end of the module, students should be able to:
- Describe the pros and cons of the Bayesian statistical paradigm, and discuss its suitability in different health science contexts
- Appreciate the influence of the prior distributions, and apply standard approaches for selecting and critically appraising prior distributions
- Design appropriate Bayesian models for health science problems, and implement them in standard software
- Apply and assess the performance of computational algorithms used for Bayesian analysis using a standard software package
- Summarise and accurately interpret the output of Bayesian analyses
- Effectively appraise and criticise Bayesian models both qualitatively and quantitatively using standard measures.
Pre-requisites: Statistics for HDS (core module) and a good understanding of classical (i.e. non-Bayesian) statistical modelling: likelihood (as in maximum likelihood estimation) and sampling distributions
Teaching Strategy: Lectures and computer practicals. Group work, e.g. on prior elicitation.
Assessment: Assessed take-home practical involving analysis of dataset(s) in R and Bayesian software
Module Length: 4 days