So we should use the repeated measures model for this data, i.e., to consider the split plot diets as three repeated measures. A. The book has a program on page 153 for this, but the program was written is an unusual way. The subject should be mouse. So how should we dene mouse in the data? B. The book used compound symmetry (cs) for the diet correlation. Why? C. Is cs a reasonable model? Can you nd a better one?