Applied Time Series Analysis for Social Science Contexts

Catalog description

This course is designed to introduce the modeling challenges presented by time series data, and panel data with a large time dimension. Emphasis is placed on developing an understanding of the distinction between stationary and nonstationary (constant mean, variance vs. nonconstant mean and variance data) for estimation purposes in time series contexts. Since nonstationarity is very prevalent in social science applications, understanding estimation approaches to nonstationary data is crucial for nonspurious estimation results. The course will place considerable emphasis on the development of maximum likelihood vector error correction mechanism estimation as a solution to nonstationarity. Throughout, problem sets and exercises place emphasis on real world applications and estimation challenges.