Management of a company divided its sales force into two groups. The situation is precisely that of the previous exercise, only now a manager decided the assignment of the groups. The 95% confidence interval for μ A – μ B was found to be [$500, $2,200].
(a) Does the lack of randomized assignment of the two groups suggest that we should be more careful in checking the condition of similar variances?
(b) The manager responsible for Group B complained that her group had been assigned the poor performers and that the results were not her fault. How would you respond?