BUILDING, TESTING, AND USING MULTIPLE LINEAR REGRESSION MODELS

The least-squares approach to estimating parameters of a multiple regression model is a fairly straightforward extension of simple linear regression. What is not so easy is the extension of the statistical testing procedures, which present more variants when multiple variables are involved. Nevertheless, the necessary intuition for understanding what commercially available statistical software tools offer is not difficult to grasp, if one is armed with a solid understanding of simple regression. However, when multiple variables are involved, new issues arise, which we outline in the following subsections.