Methodology

The department is committed to providing rigorous, cutting edge methodological training to our graduate students. We provide a full sequence of courses in quantitative methods and formal theory, and offer a Political Methodology concentration in the Ph.D. program.

Our faculty includes leading methodologists who have created methods, models, and software packages widely used in political science and other fields. The methods group's research and teaching interests cover all major methodological fronts, with substantive interests spanning political science and beyond. Our methods faculty include several Fellows of the American Academy of Arts and Sciences and the National Academy of Sciences, serve on the editorial boards of leading political science and methodology journals, and have won numerous major grants, awards and prizes. Faculty members regularly involved in the teaching of graduate methods courses required by the methods concentration include:

Gary Cox Scott Desposato James Fowler Clark Gibson

Thad Kousser David Lake Mat McCubbins Keith Poole

Sebastian Saiegh Branislav Slantchev Langche Zeng

In addition, many other faculty members have serious interests in applied methods work and routinely employ sophisticated quantitative, formal, or field-research methods in their research, including:

Marisa Abrajano Jeeyang Baum Lawrence Broz Wayne Cornelius Karen Ferree

Erik Gartzke Zoltan HajnalGary Jacobson Samuel Kernell David Law

Edmund Malesky Megumi Naoi Samuel Popkin Philip Roeder Darren Schreiber

Matthew Shugart Peter Smith Kaare Strom Barbara Walter

For detailed individual research and teaching profiles follow the name links.

Requirements of the Political Methodology Concentration:

(1) Required Courses

Students must take all of the following courses:

• 204A. Research Design. Principles of research design and social scientific research, focusing on issues common to research in political science and the choice of alternative research designs and methods. Experimental, quasi-experimental, quantitative and qualitative designs will be discussed.

• 204B. Quantitative Methods I. The use of basic quantitative methods (particularly multiple regression and its extensions) in political science research. Introduction to statistical computing. Emphasis on applications.

• 204C Game Theory 1. This course introduces students to the fundamentals of decision theory and game theory. Emphasis will be placed on modeling and solving games.

• PS270. Mathematical and Statistical Foundations. Introduction to probability theory (probability rules, random variables, univariate and multivariate distributions) and mathematical statistics (sampling distributions, estimation and inference frameworks). Also review of essential calculus and linear algebra.

• PS271B. Advanced Statistical Applications. Generalized linear models for discrete choice, ordinal, count, duration/survival data, truncated/censored/sample selected data, and times series cross section/panel data. Inference via maximum likelihood estimation, Bayesian posterior sampling, or bootstrapping. Introduction to selected topics such as missing data treatment and nonparametric methods and models for causal inference and prediction.

(2) Additional Course Requirements

Students must take at least one additional course in the 270-279 range, such as:

• PS276. Mathematical Modeling. Demonstrates how to construct mathematical models of phenomena of interest to political science. Methods are drawn from the spatial theory of politics; game theory; social choice theory; and algorithmic game theory. Specific applications examined may include models for the distributions of state size, war magnitude, and democracy over time and space. Spatial models of party choice, models of organization, the theory of agency, and the theory of collective action may also be covered.

• PS277. Measurement Theory. Methods for estimating latent dimensions of preference and similarity from observed choices and judgments. Factor Analysis, Multidimensional Scaling, and related techniques are studied with both classical maximum likelihood and Bayesian methods.

• PS279. Special Topics in Methodology. Some topics being offered or planned for the near future include:

  • What's New in Econometrics. Covers the following topics: Generalized Method of Moments and Empirical Likelihood; Estimation of Average Treatment Effects Under Unconfoundedness; Linear Panel Data Models; Nonlinear Panel Data Models; Regression Discontinuity Designs; Instrumental Variables with Treatment Effect Heterogeneity; Weak Instruments and Many Instruments; Local Average Treatment Effects; Control Function and Related Methods; Bayesian Inference; Cluster and Stratified  Sampling; Partial Identification; Difference-in-Differences Estimation;  Discrete Choice Models; Missing Data; and Quantile Methods.
  • Social Network Analysis. Introduction to mathematical social network theory from sociology and physics with a special emphasis on applications to large data sets.
  • Graphical Models and Statistical Learning. Introduction to graphical models, a general framework for representing and applying diverse probabilistic models including Bayesian networks, causal graphs, neural networks, and social networks; Aspects of statistical learning theory and methods, focusing on nonlinear models for supervised learning and model ensembling and selection methods.

Students must earn a grade of at least a "B" in all courses and pass a comprehensive exam in methods. The methods comprehensive examination will be offered in the fall quarter (to avoid conflict with the spring general exams). Students planning to take the exam shall notify the methods field coordinator in writing before the end of the preceding spring quarter of their desire to take the exam, providing information on the methods courses completed (course numbers/names, instructors, syllabi). The field coordinator nominates, and the department Chair appoints, four members to form the exam committee which, in consultation with other members of the methods faculty, prepares and grades the written examination. The written exam lasts four hours and may be either open-book, closed-book or a mixture. The coverage may be adjusted for each individual student to accommodate the particular set of methods courses the student has completed. In addition to the written exam, the student may be required to submit an empirical research paper demonstrating methods skills in application. The exam committee can assign a grade of fail, pass, or distinction. Committee decision is based on majority vote and the student will receive written notification of the outcome. A student who fails the exam will be given a second chance (but no more than that) when the exam is next offered.