ABSTRACT

Since regression analysis will be used to analyze repeated measures, Chapter 2 reviewed Bayesian regression techniques for standard models, including logistic for binary responses and simple and multiple linear regression for continuous responses. Also reviewed was the analysis of nonlinear models and last to be presented was an example of repeated measures. As was discussed, the standard regression model assumes a constant variance for the dependent variable and that the observations are independent. For a repeated measures model, correlation is present, and the variance of observations is not the same. In order to execute Bayesian analysis, correlation must be taken into account. Our first example of repeated measures is based on a clinical trial of epilepsy conducted by Lippik1 and analyzed by Thall and Vail.2