Methods Workshop: Daniel Stegmueller (University of Mannheim)

Methods Workshop: Daniel Stegmueller (University of Mannheim)| 270 Gross Hall | 1PM-4PM | "Multilevel Models in Political Science: Basics and Extensions." Hierarchical data are ubiquitous in all subfields of political science, for example in form of panel data (of, say, countries over time) or nested structures, such as individuals within regions. This workshop introduces the basics of hierarchical models taking both theoretical and applied perspectives. We will discuss the basic idea of hierarchical or ¿multilevel¿ models for both linear and limited dependent variables. Special attention will be given to estimating multiple sources of variability and "cross-level" interactions between macro and micro units. We will also discuss practical issues arising in estimating and interpreting these models. Finally, hierarchical models are closely linked to the Bayesian framework. Understanding this relationship will allow us to explore powerful extensions, such as ideal point and measurement error models.