Associate Professor Sama Low-Choy
Dr Clair Alston-Knox
This workshop helps newcomers to Bayesian analysis transition from their classical understanding of statistical modelling to a Bayesian one. Regression provides a simple and versatile context for enabling this transition. We explain the subtle differences in interpretation of model outputs, which are most evident in small samples. Examples are used to illustrate the impact of prior choices, from non-informative to informative, conjugate and otherwise. You will gain experience in how the prior can be harnessed to support a continually adaptive approach to updating your model as new information arrives. Code to reproduce examples will be provided in R.
These sessions will be interactive and will involve discussion in small groups.
Recommended prerequisite: Expanding your Inference through Modelling.