Workshop: Regression Models for Time Series Analysis

Session Submission Type: Meeting

Abstract

This workshop will consider several of the many methods available for analyzing data collected on a single unit over time. Time series analysis is useful, for example, in studying the variability in crime trends within a city, state, or nation. The workshop will focus on regression models and will devote particular attention to the concepts of serial correlation and stationarity. The presentation will assume that attendees possess a solid knowledge of ordinary least squares regression analysis but have no prior experience with time series data. Specific topics will include models for correlated errors, nonstationarity and its consequences, cointegrated series, error correction models, and vector autoregressions. Although this is not a hands-on workshop, it will present multiple empirical examples using Stata and other time series software.

Sub Unit

• Meetings

Speaker

• David McDowall, University at Albany, SUNY