Consider the following model with exchangeable correlation: prot ij =0 1 barleyij 2 lupins ij 3 week ij 4 barley ij *week ij 5 lupinsij *week ij ij where barley and lupins are the indicator variables for the two diet, ie. Barley=1 if the diet is barley, 0 otherwise. Similar for variable lupins. a) Test for interaction of diet and week, i.e., H0 : 4 = 5 =0 vs. H1 : 40 or 50 b) Test the protein difference between barley and mixed at week 10 H0: 1 104 =0 vs. H1: 1 1040 c) Estimate the protein difference between lupins and mixed at week 12, that is to estimate 2 125