Table 5.
Variables | Sum of Squares | Degrees of Freedom | Mean Square 1 | Partial Correlation | F-Value 2 | p-Value |
---|---|---|---|---|---|---|
X1 | 0.8406 | 1 | 0.8406 | 0.4857 | 4.0143 | 0.0664 |
X2 | 0.8740 | 1 | 0.8740 | 0.4930 | 4.1741 | 0.0619 |
X3 | 0.9753 | 1 | 0.9753 | 0.5136 | 4.6578 | 0.0502 |
X1 2 | 2.6006 | 1 | 2.6006 | 0.6990 | 12.4197 | 0.0037 |
X2 2 | 6.1470 | 1 | 6.1470 | 0.8325 | 29.3565 | 0.0001 |
X3 2 | 14.1505 | 1 | 14.1505 | 0.9158 | 67.5797 | 0.0001 |
X1X2 | 1.3448 | 1 | 1.3448 | 0.5750 | 6.4225 | 0.0249 |
X1X3 | 0.1012 | 1 | 0.1012 | 0.1894 | 0.4835 | 0.4991 |
X2X3 | 0.3872 | 1 | 0.3872 | 0.3529 | 1.8492 | 0.1970 |
Regression | 27.1532 | 9 | 3.0170 | F2 = 14.40859 | 0.0002 | |
Remainder | 2.7221 | 13 | 0.2094 | |||
Misfit | 1.1025 | 5 | 0.2205 | F1 = 1.08921 | 0.4114 | |
Error | 1.6196 | 8 | 0.2024 | |||
Sum | 29.8752 | 22 |
1 MS, mean square values. 2 A preliminary determination of the suitability of the selected regression model was made according to the misfit test of equation (F1 = MSmisfit/MSerror). The significance test of the regression equation was measured by F2, which equated to the ratio of MSregression to MSremainder. The significant difference between the regression model and the coefficient was evaluated by the F test, and p < 0.05 was considered significant.