Political Science 233
Intermediate Statistical Methods
Social Sciences Bldg. 228
406 Perkins Library
Class: M-W-F 10:20-11:10
Office Hours: M & W 1:300-2:30
This course introduces students to the use of regression analysis and its application to political science research. Course topics place an emphasis on the types of data encountered in political science research and the challenges that researchers face in analyzing these kinds of data. The level of mathematical presentation is moderate, and the primary goal is the understanding of conceptual problems and applications rather than an emphasis on derivations. However, students will be expected to understand and follow the basic derivation and properties of the regression estimator in both scalar and matrix notation. The course is structured around the assumptions of the regression estimator, the implications of those assumptions for analyses of political science data, and the methods used to address the violations of assumptions of the regression model. The goal of the course is to provide students with the statistical tools necessary to address research questions that are of interest to them.
Email to PS 233
The following books have been ordered and should be available for purchase at the Bryan Center bookstore.
Student performance will be evaluated on the following basis:
Exercises will consist of problem sets reviewing the statistical properties of the simple regression and general regression estimators. These homework problem sets will be assigned during the first section of the course when we review and derive the regression estimators.
Research notes will consist of brief (3-5 page double-spaced) written reports on analyses of datasets that face some of the estimation problems discussed in this course. Students will be expected to present statistical analyses as if they were writing a research note for submission to a peer referreed journal.
Students will be given an open-book take home final exam. The exam will be distributed on April 27 and will be returned on April 28.
Access to STATA
Homework projects for this course will require access to STATA statistical software package. Datasets for analysis will be available in STATA format. Students can gain access to STATA through the Bunche Institute computer lab in Perkins 214.
You can get help with STATA from the STATA website, including resources for those just learning STATA and answers to frequently asked questions. In addition, you can download a copy of Giacomo Chiozza's STATA Notes or Chris Zorn's "Stata for Dummies" for help with essential STATA commands.
Schedule of Topics, Readings and Assignments
January 17 - Organizational Meeting
January 22 - The Regression Model Approach to Political Science Research
January 24 & 29 - The Simple Linear Regression Model
January 31 & February 5 - The General Linear Regression Model
February 7 & 12 - Assumptions of the Regression Model
February 14 Model Performance, R-squared and Forecasting
February 16 - First Set of Exercises Due
February 19 - Model Performance, R-Squared and Forecasting (Part II)
February 21 - Section Discussion on Stata
February 23 & 26 - Multicolinearity
Febryary 28 - Omitted Variable Bias
March 2 - Second Set of Exercises Due
March 5 - Omitted Variable Bias (Part II)
March 7 - The EE' Matrix
March 12 - 16 - SPRING BREAK!
March 19 & 21 - Autocorrelation Continued - (AND, Omitted Variable Bias Research Note Due March 19)
March 26 - Measurement Errors & Implications
March 28 - Dummy Variables, Interaction Effects & Specifying Functional Form
March 30 - Autocorrelation Research Note Due
April 2 & 4 - Dichotomous Dependent Variables: Logit & Probit
April 9 & 11- Endogeneity Bias & Simultaneous Equations
April 13 - Interactions and Functional Form Research Note Due
April 16 & 18 - Selection Bias and Selection Effects
April 20 - Logit and Probit Research Note Due
April 26 - Final Exam will be made electronically available at NOON.
April 27 - Final Exam Due at NOON (via email)