Question 3. 3(a) (2.5 points. Suppose that in a weighted least-squares problem you have weighted the observa- tions by wj = 001, ... ,w, =co, instead of the weights w =,..., , = 0, 2, where c is any positive number; (see the notes/text for the notation). Show that the resulting weighted L-S estima- tor, bwis, of and its covariance matrix are unaffected with the choice of c. Clearly show your work. 3(b) (2.5 points. Consider the model Y=BX+€, where e's are independent N(0, 02X?) random variables; here o? is unknown. Find the weighted least squares estimator of the parameter B and its variance. Clearly show your work.