POLS 603
Quantitative
Political Analysis II
Fall 2004
Dave Peterson
Office: 2037 Bush
Phone: 845-6783
Office Hours: by appointment.
Email: dave@polisci.tamu.edu
The main purpose of this class is to provide the student with a mastery of the basic regression model—understanding both the statistical theory behind its use in political science and a deep understanding for application to your own work. My assumption is that what happens to be in vogue in political methodology will change rapidly and, more importantly, that the methods you will need in your own research will lead you in unpredictable directions that cannot be anticipated in this class. Mastery of the underlying statistical theory behind regression will provide the background necessary to quickly master new applications as you need them.
To that end, the class starts with an introduction to statistical, probability, and distributional theory. In many ways, this may be the most important portion of the class. I do not believe you need to know this material at the top of your head (I certainly do not), but it is important that you understand how to read works that use this theory and know where to look to understand what is occurring “behind the scenes” of our statistical models. Then the course moves to a mathematical introduction of the general linear regression model and some of its numerous extensions. We then move to Generalized Method of Moments (GMM) and Maximum Likelihood Estimation (MLE) approaches to model estimation. Both of these types of models are flexible approaches to many different types of statistical models.
The textbook for this course is Fumio Hayashi’s Econometrics. This can be a difficult read and is not written explicitly for political scientists. I will supplement the text with applications that should make the points clearer. Additionally, there will be a list of recommended readings that are more basic if you have trouble with Hayashi. Finally, I find Peter Kennedy’s A Guide to Econometrics a nice reference which attempts to provide some intuition about what is going on.
Grades will be assigned based on a combination of weekly homeworks for the early portion of the class, a final exam, and your preparation for and involvement in class.
Matrix algebra, distribution theory and probability, inference. Why do we use statistical methods and what do we want from the methods we use?
Kennedy Chapter 1-2
Greene Econometrics Chapters 1-4
Namboodiri Matrix Algebra: An Introduction Sage Monograph
Fox Applied Regression Analysis, Linear Models and Related Methods, App D
Hanushek & Jackson Statistical Methods for Social Scientists Chapter 1, App II
2. The classical linear regression model (finite sample
properties)
Assumptions, algebra, finite samples, hypothesis testing, relation to MLE, GLS.
What is regression, why is the default method of choice?
Hayashi Chapters 1&2.
Other introductions:
Kennedy Ch 3
Fox Ch 5, 6, 9, 10
Hanushek & Jackson Ch 5
3. Large Sample Theory
Limit theorems, time series, large sample properties of OLS, Hypothesis test, consistency of s2, heteroskedasticity, serial correlation. What happens if I have lots of data, how bad are some violations of error assumptions then?
Hayashi Chapter 2
Other Treatments:
Kennedy Ch 5, 6, 9 & 11
Fox Ch 6.3 13
4. Introduction to GMM
A few examples, general formulation, definition, large sample properties, testing restrictions, hypothesis testing, heteroskedasticity, 2SLS,
Hayashi Chapter 3
5. Multiple-equation
GMM
Definition, What is the difference
between single and multiple equation, 3SLS, SUR,
Hayashi Chapter 4
6. Panel Data
What is it, error components model, fixed effects, random effects, unbalanced panels
Hayashi Chapter 5
7. Serial Correlation
Modeling serial correlation, estimating
Hayashi Chapter 6
8. Extremum Estimators
MLE, Consistency, Hypothesis testing, Binary choice, truncation, censoring, multivariate regression, FIML, LIML
Hayashi Chapter 7 & 8.
9. Unit Roots and
Error Corrections
More Times series stuff that I doubt we will get to.
Hayashi Chapter 9 & 10.
Other possible topics we may think about:
Models of timing until an event occurs. Becoming increasingly common in the discipline.
Box-Steffensmeier and Jones “Time is of the Essence: Event History Models in Political Science” AJPS 1997 pp. 1414-1461
Beck Katz and Tucker, “Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable” AJPS 1998 p. 1260-1288
Katz and Sala “Careerism,
Committee Assignments, and the Electoral Connection” APSR 1996 pp. 21-33
Interpretation.
An new way to interpret the results graphically.
King, Tomz, and
Using iterative approaches to solving problems and making inferences. Also bootstrapping and other solutions to distributional problems.
Greene Ch 4
Mooney “Bootstrap Statistical Inference: Examples and Evaluations for Political Science” AJPS 1996 p. 570-602
Erikson, MacKuen and Stimson “A Model of American Macro Politics” 2000 available at http://www.unc.edu/~jstimson/papers.htm
Matlzman “Meeting Competing Demands: Committee Performance in the Postreform House” AJPS 1995 653-682
Bruce Western “Concepts and Suggestions for Robust Regression Analysis” AJPS 1995 786-817 p. 786-817.
I expect that all students will
conduct themselves in a manner that is consistent with the Aggie Code. Any lying or cheating in this class will be
handled in accordance with Texas A&M policy.
The Americans with Disabilities Act
(ADA) is a federal anti-discrimination statute that provides comprehensive
civil rights protection for persons with disabilities. Among other things, this legislation requires
that all students with disabilities be guaranteed a learning environment that
provides for reasonable accommodation of their disabilities. If you believe you have a disability
requiring an accommodation, please contact the Department of Student Life,
Services for Students with Disabilities in Room 126 of the