1. What problems are caused by omitting a relevant variable? What problems are caused by including an irrelevant variable? Under what
circumstances does omitting a relevant variable not cause problems with the other estimates?
2. Why should one not omit the intercept term in a regression. What happens when you do omit the intercept term? What components
potentially go into the constant term in a regression?
3. In what way is a linear regression estimated by OLS linear? Can one still estimate a linear regression if as in the case of cost curves, theory says the relationship is non-linear? What is a dummy variable and how are they used
4. Explain the simple regression model, Yi = E EXi + HI, what is Yi? What are E and E" What is Xi? What is HI? Which of these items do you collect data for and which of them do you estimate? If this changes to a multiple regression model, Yi = E EXi + EXi + .... + ENXi + HI, what is different?
5. What are the assumptions of the classical regression model? What does it mean to say an estimator is BLUE? What does the central limit theorem say about the distribution of the estimated E and why is this important?