Dynamic Models for Social Scientists and Policy Analysts
This is an applied course on dynamic modeling for social scientists and policy analysts. Starting with time series regression analysis, the course considers important dynamic models including ARIMA, ARFIMA, Error Correction, Fractional Error Correction, GARCH, VAR/VECM and State Space models. Key concepts such as (non)stationarity, exogeneity and Granger Causality are used to inform model specification, model comparison/selection and post-estimation diagnostics. The emphasis is on practical applications and participants learn how to use widely available software packages such as STATA, R, Rats and Eviews to analyze dynamic models in their fields of interest. Students are invited to bring their own data sets to daily lab sessions.
Prerequisites: Participants should have taken an introductory course in applied statistics thru OLS regression and be familiar with the standard Windows computing environment. Basic knowledge of a major statistical software package such as R, SPSS or STATA is helpful, but not required.
Fee: Members = $1700; Non-members = $3200