Syllabus

This syllabus is also available in PDF format.

Overview

This course will prepare you to conduct empirical research in political science, with a focus on linear regression models. You should come away from this course an informed consumer and user of the most prevalent statistical techniques in political science. You will also learn to appreciate the connections between statistical practices, research ethics, and the ongoing crisis of confidence in the sciences.

Grading

Your grade will be based on:

I will not accept late assignments except in case of a documented family or medical emergency.

Software

All analysis will be conducted in R. You must write and submit your problem sets in R Markdown format, which allows you to embed R code and its output, including graphs, directly in a document. You will submit assignments by pushing to a GitHub repository. I will not accept assignments by email. Don’t even think about printing them out. We will discuss homework submission policies—and, along the way, the basics of Git and GitHub—in the first class or recitation. There will be a separate handout laying out the details.

If you are not yet comfortable with the basics of R, here are some hands-on tutorials I recommend completing before the start of the semester:

Some other useful resources on R include:

Collaboration Policy

Your work in this course must be the product of your own intellectual labor. Although it is important to learn how to collaborate and co-author, at this stage of your academic careers it is even more crucial that you personally comprehend the basic principles of statistics and data analysis. I expect you to follow these rules:

I consider any violation of these guidelines a violation of the university’s Honor Code, and I will deal with such a violation accordingly.

Books

Two textbooks are required:

I recommend, but do not require, supplementing the selections from Wooldridge with their corresponding treatments in any or all of these more advanced books:

The last three of these are particularly useful supplements to Wooldridge, since they’re by statisticians rather than econometricians.

Schedule

This schedule is tentative and is subject to change.

Basics

January 14: Working with Data

January 21: (Re-)Introduction to Regression

January 28: Making Inferences

Beyond the Standard Assumptions

February 4: Specification and Misspecification

February 11: Non-Constant Variance

February 18: Panel Data

February 25: Panel Data, continued

March 3: Nonlinear Models: A Brief Overview

Take-home midterm sometime this week—exact timing TBD.

Causal Inference

March 17: Introduction to Causal Inference

Turn in final paper proposals.

March 24: Instrumental Variables

March 31: Instrumental Variables, continued

Advanced Topics

April 7: Computationally Intensive Methods

Turn in initial drafts of final papers.

April 14: Model Selection

Turn in peer reviews.

April 21: Missing Data

  1. Chapter numbers in the syllabus correspond to the 6th edition, but any edition is fine.