How do I deal with lack of fit data if they actually looking good in the surdface profiler and actual by predicted plot
Dear JMP Pros, I have a dataset (See attachment) with two discrete factors (X_1 and X_2) and after model running, the fit looks good in the actual by predicted as well as in the surface profiler. However, I have a significant lack of fit here. I am wondering what the reason is here and how do I deal with this dataset? Can I still use and believe the fit? Please note, that I had to use the discrete...