Next section: Significance tests in multiple regression

Just as in the case of one-variable regression, the sum of squares in a multiple regression analysis can be partitioned into the sum of squares predicted and the sum of squares error.

Sum of squares total: 55.57 Sum of squares predicted: 22.21 Sum of squares error: 33.36Again, as in one-variable regression, R² is the ratio of sum of squares predicted to sum of squares total. In this example, R² = 22.21/55.57 = 0.40. Sometimes multiple regression analysis is performed on standardized variables. When this is done, the regression coefficients are referred to as beta (ß) weights. The Y intercept (A) is always zero when standardized variables are used. Therefore, the regression equation for standardized variables is:

Y' = ß

Next section: Significance tests in multiple regression