5. In a study of association between weight (WGT; Y ) and several predictors including height (HGT), age (AGE), and the square of age (AGE2), n children were involved in the study (you don't need to know what n is in order to answer the questions below). The following table is obtained when evaluating the MSE and R2 for different models (i.e., subset of predictors):
Model MSE R2
HGT 29.93 0.6630
AGE 36.18 0.5926
AGE2 36.63 0.5876
HGT, AGE 21.71 0.7800
HGT,AGE2 22.07 0.7764
AGE,AGE2 40.20 0.5927
HGT,AGE,AGE2 24.40 0.7802
a. Which regression model has the smallest SSE? Is this model necessarily the best model? Please explain.
b. Which regression model is the best? Please explain, including indicating the criterion (criteria) used in selecting the optimal model.
