Financial advising
Starting from the model summarized in Table 4, what happens if you add the explanatory variable State to the model? (Limit your estimation to the n = 2,000 cases in the training sample.)
(a) Does the R2 increase by a statistically significant amount? Briefly interpret your result.
(b) Is any individual t-statistic for a dummy variable representing a state statistically significant at the 0.05 level? How does your answer depend on the choice of the baseline state?
(c) How might you refine the model that includes State?