Estimating Optimal Transformations for Multiple Regression and Correlation: Comment