Relating multiple regression to categorical variables

Categorical variables and multiple regression

This image contains the data for the first example for Class 6. Going up in the vertical direction is the Y-variable RATING. In the X1 direction we have IN-HOUSE? and in the X2 direction is SALARY. There are two X-variables so the objective is to place the best (least squares) plane through the points. As IN-HOUSE? can only take on one of two values, NO, which is coded as -1 (red balls) and YES which is coded as +1 (green balls) all the points must lie in one of two vertical planes (you can see this when all the red balls and all the green balls line up in two columns). The least squares plane, shown here in blue, has to cross each of these two vertical planes and these crossings are drawn with the yellow lines. It is these two yellow "parallel lines" that you will see when I draw them on the board in class.

-->