Question 5 [3+3 = 6 points]: Suppose we are interested in the linear model y.- 2 g + lmli + 321:2; + cg. Also suppose the columns x1 and 1:2 of the design matrix for this model have mean I] and length 1. (That is1 xixl = 1 and x'QXQ = 1. This is a very.r particular situation that is unlikel}r to happen in practice; it just makes our arithmetic easier for a moment}. Then if r is the correlation between an and Kg? we [have the following: \".1 [l [l lfn [l 0 XX 2 n 1 r and (rm1 2 n 1; (1 42] r,'{1r2} [l r 1 [l rf{1r2} \"(lr?) {a} In our setup where the predictors have mean [I and length 1, explain wh}r SH 2 1. Use that to show that the 1WF formula on page EDS matches o'2 (X'X)_1 (above). {b} Determine what values of T will make the variance of E1 and '32 large. Explain why, using what you know about the variance of the vector 16