(a) Describe the Durbin–Watson test as it is applied in traditional textbook econometrics, and explain why it might not be an effective M-S test of assumption [4] in the linear regression model (Table 15.6).

(b) Using your answer in (a), discuss the problem with the traditional textbook strategy of adopting the alternative as a way to respecify the original model in order to account for the systematic information in the data.

(c) Imagine you are in a seminar where a researcher is presenting an applied econometrics paper using table after table of estimation and testing results relating to his/her favorite theoretical model, on the basis of which he/she makes a good case for the theory in question; no data plots, no model assumptions, and no M-S tests are presented.

Discuss the merits of the case for the particular theory model made by the presenter.