GOVT 6059

Panel and Multilevel Data

Course Description

Course information provided by the Courses of Study 2016-2017.

This course continues the path of 6019 and 6029 in offering a hybrid of applied social statistics and econometric modeling for graduate students, with a focus on panel, time-series cross-section, and multilevel data. Such data are now commonplace, presenting additional complications but also offering statistical and quasi-experimental leverage via repeated measurements and change-points. The main topics will be the use and choice among robust standard errors and fixed and random effects estimators, a variety of multi-level specifications and their estimation and interpretation, and a key application called multi-level regression and post-stratification (MRP). At the end of the course, we will examine dynamic models, focusing on their assumptions, interpretation, and related pitfalls. Each student will be required to actively participate in a group presentation of the nuts and bolts of a data analysis using this kind of data, either assigned by the instructor or based off peer research. Students with active quantitative research projects are encouraged to join the course for an opportunity to hone the analysis, and to receive feedback from peers and the instructor.

When Offered Fall (weeks 8-14).

View Enrollment Information