POL2578H1S L0101

Topics in Methods

Models for Categorical Dependent Variables and Panel Data

Themes

This course provides graduate students with an advanced training in quantitative methods focusing on the two families of models most commonly used in political science: 1) models for categorical dependent variables, and 2) models for panel data and time-series. The course also includes an introduction to maximum likelihood estimation. The course starts where POL 2504 ends, and represents a natural continuation for PhD students looking to develop their methodological skills and use statistical methods in their own research. POL 2578 is also suitable for MA students who received training in quantitative methods at the undergraduate level and interested in an advanced course going beyond linear regression. The course comprises lectures presenting the theory behind each statistical model, discussions of concrete examples based on actual data, as well as interactive sessions with statistical packages, in particular the R programming language.

Texts

Two textbooks particularly relevant for this course are:

J. Scott Long, 1997. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks: Sage Publications.

Jeffrey M. Wooldridge. 2002. Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press, Chapters 10-20.

Required and suggested readings will be detailed in the syllabus.

Format and Requirements

The course comprises lectures and labs. The requirement for this course is POL 2504 or some equivalent background (for MA students, this would typically be a third or fourth year undergraduate course in quantitative methods). Interested students may contact the professor about requirements.