. Consider the linear regression model y = X+E, where y (Y,..., Y) is an n x 1 response vector, = 1 11 212 21,p-1 1 221 T22 12,p-1 X= [1,x1, x2, , xp-1] = In In2 is an n xp design matrix with rank(X) = p, regression coefficient vector, and the error term (a) It can be shown that Inp-1, == (Bo, B1, B2, B-1) is a px 1 = (1,..., En) ~ N(0,0In). (B-B)TXTX (B-3) ~ 02 Use (1) (and other known facts) to find the distribution of (B-B)TXTX(B-3) po where 2 MSE. = (1) 1.204: