Regression Basics, 3rd Edition

6Nonlinear and Logarithmic Models and Dummy and Interaction Variables

DOI: 10.4324/9781003349174-6

In the previous chapter, we built multiple regression models for our four examples. In each case, new independent (X) variables were introduced that we hypothesized were important in explaining the dependent variable (Y). In all these examples, the new X variables were continuous measures, and they entered our model linearly (i.e., with the power of 1). There are, however, other types of variables we can consider that may be desirable under certain circumstances. For example, consider our simple two-variable regression model that was introduced in Chapter 2,

$$Y_i=\alpha +\beta X_i+u_i$$(6.1)

Thus far we have only considered continuous values for Y_i and X_i without ...