Chapter 9 Model Building

Model building methods are used mainly in exploratory situations where many independent variables have been measured, but a final model explaining the dependent variable has not been reached. You want to build a model that contains enough covariates to explain the model well, but still be parsimonious such that the model is still interpretable.

This chapter introduces different types of covariates that can be used, stratified models, confounding and moderation. We then conclude with measures of model fit and methods to compare between competing models.