Type I error, testing power, and predicting precision based on the GLM and LM models for CATA data--Further discussion with M. Meyners and A. Hasted