1. Specification Bias
a. Define "specification bias". Describe why it is considered to be a "fatal" error when estimating an econometric model.
b. Describe two ways the Granger Causality Test can help in the construction of a correctly specified econometric model.
c. When can a Durbin-Watson statistic indicate a "spurious" regression or possible specification error? What is a limitation of the DW statistic to detect such an error?
d. If you reject the Ho for the Omitted Variable Test, does that necessarily imply that you have a correctly specified model? Why or Why not?