Consider the linear model yi = 0 +1xi1 ++pxip +ei where the errors e1,...,en are i.i.d normal with mean zero and variance 2. a) Let yi denote the ith fitted value and let yi(i) denote the predicted response value for the ith subject without including the ith subject in the regression (in the notation used in class, yi(i) = xTi [i]. Write the difference yi yi(i) in terms of the ith residual and the ith leverage. (0.2 points)