where SS is the sum of squares, f is the independent variable (e.g. T,
SH, V, etc.) and N is the number of variable components. A smaller
magnitude of the p-value and a larger t-ratio and %PC value indicated
greater significance of the corresponding parameters in the
regression model.