Lectures: 2 sessions / week, 1.5 hours / session
This course focuses on the specification and estimation of the linear regression model. The course departs from the standard Gauss-Markov assumptions to include heteroskedasticity, serial correlation, and errors in variables. Advanced topics include generalized least squares, instrumental variables, nonlinear regression, and limited dependent variable models. Economic applications are discussed throughout the course.
Wooldridge, J. M. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press, 2001. ISBN: 0262232197.
Students should be prepared in linear algebra at the level of 14.102 or equivalent. They should also know mathematical statistics at the level of 14.381 or equivalent (Chapter 1-12 in Mood and Graybill). A good short summary is found in Graybill, Linear Models, chapters 1-3.
Course requirements include completion of about six problem sets, a midterm, and a final exam. No term paper is required. Grading of these assignments is weighted equally.