Suppose that you want to test the hypothesis Hoifi = 3 against Hi'-n ^ 3 for a normally distributed population with variance 1.44. Given a = .10 and N = 36, how powerful would this test be if m = 4? Given a = .10 and X = 16, how powerful would this test be if /x = 4? If the statistician wanted his test to have power of .80 if /j. = 3.5, what sample size should he use, given that a = .10?