Robust Bayesian Analysis of the Linear Model with an Application to a Problem in Psychology Open Access
Downloadable ContentDownload PDF
This thesis discusses the Bayesian analysis method for multi-variable normal linear regression. We use the normal distribution with known variance as the prior for the parameters in the linear model. This is the base prior for the robust Bayesian analysis. The epsilon-contamination class and the density ratio class of normal and Cauchy distribution are considered for a robust Bayesian analysis. We also discuss Bayesian methods for hypothesis testing.