The description of the classes or ranks of equations is as follows:

     For the standard deviations parameters (coefficient, intercept, and estimate) the equation 
     having the smallest value was considered better. 

     The PRESS parameter (Predicted Residual Error Sum of Squares) indicates the sum of the 
     differences between the actual point and the predicted points. The regression is run 
     iteratively by taking out the ith data (for all data points), calculating the difference 
     between the predicted point of the new regression and the actual point, and then adding up 
     all these residuals. The equation having the smallest PRESS value was considered best for 
     this parameter. 

     The power (with alpha=.05) is the probability that the model correctly describes the 
     relationship of the variables, if there is a relationship. The equation showing the 
     largest value for this parameter was considered the best.

     The R and Rsqr together are measures of how well the regression model describes the data. 
     The equation showing the largest value for this parameter was considered the best.