Robert W. Walker

Methods Courses

Pol Sci / App Stat 450: Panel Data

This serves as the home page for Applied Statistics and Political Science 450, Topics in Applied Statistics: Panel Data, Spring 2007, Washington University in Saint Louis. This page provides the syllabus, links to articles, and relevant course materials.

Administrative Documents

Lectures

After a bit of thought and given the uncertainty of pacing this course, I have decided to simply create one really long .pdf that combines all the slides. Here it is.

Readings

By way of a calendar, the next subject to occupy our attention will be the specification of panel data models. By now, you should be reading Sayrs and the Stimson piece and should give serious consideration to the arguments in Hicks. For more detail and rigour, read Mundlak. In no particular order, we will consider Efficient Estimation of Rarely Changing... and Beck and Katz (1995) at about the same time. On a more casual note, you should read David A. Freedman's piece in the American Statistician in November, the link is provided here. This should take us through the end of February. After treating the standard fixed and random effects models, we will consider the random coefficients models. Once we have dealt with the central issues in big N small T analyses, we will wander over to big T small N and consider dynamic panel data models and revisit our cursory overview of time serial issues. I am still searching for good papers on this that are accessible and thorough.

Midterm Examination

Homework Assignments

Code and Handbooks

Data

Valid HTML 4.01! Valid CSS! Robert W. Walker, 18 January 2007