Predictor Variable Transformations

Abstract

In this chapter, we discuss transformations of predictor variables. One popular transformation consists of dummy variables, which are variables that allow for the effect of categorical variables to be considered in regression modeling. Dummy variables can be used in regression analysis as both predictor and response variables, but we will limit our discussion to predictor variables. Using dummy variables as the response variable is often referred to as classification, which will remain outside of the scope of this book. In previous chapters, we assumed linear models with untransformed predictor variables. Here, we introduce nonlinear transformations of predictor variables, thereby, making the model linear.