8.2 Multiple Linear Regression Model and General Linear Model

The multiple linear regression model and the general linear model can be written as

yi = β0 + β1x1i + β2x2i + ⋅⋅⋅ + βkxki + εi

where y is still the dependent variable, but now there is more than one independent variable, x1,x2,,xk. In simple linear regression and multiple linear regression, the independent variables are considered continuous, in the general linear model the independent variables can be categorical, thus making GLM, more general. Simple linear regression is a subset of multiple linear regression, which is a subset of general linear models. Although the model has the term “linear” the function for the model does not need to be linear. For example, one of the independent variables denoted by xk could actually represent xk2, or -1
xk. The general linear model where the dependent variable is a quadratic function of a single independent variable could be written as

                    2
yi = β0 + β1xi + β2x i + εi, i = 1,2,...,n.