2. We write a linear regression with intercept a and slope coefficient (vector) 7 (that is, we write B = (a, 7')'): yi = a +17+ u; with E(uz) = 0 and E(1:1;) = 0. (a) When x; is a scalar, show that a = E(yi) E(T;) and B = Cov(Ii, Yi)/Var(T). (6) When Ii is a q x 1 vector with q> 1, find out the expressions for a =? and a =? Note that this problem asks you to find out expressions for true values of a and , not the least squares estimators of them. 2. We write a linear regression with intercept a and slope coefficient (vector) 7 (that is, we write B = (a, 7')'): yi = a +17+ u; with E(uz) = 0 and E(1:1;) = 0. (a) When x; is a scalar, show that a = E(yi) E(T;) and B = Cov(Ii, Yi)/Var(T). (6) When Ii is a q x 1 vector with q> 1, find out the expressions for a =? and a =? Note that this problem asks you to find out expressions for true values of a and , not the least squares estimators of them