PSC 535SEM - Experimental Design and Analysis

Causal claims regarding the relationship between different variables of interest are manifest in political science research, and yet it is only under a very particular set of conditions that we are able to move beyond correlation and get at causality. Experiments are one popular way of moving from cause to effect, and random assignment is a powerful tool that can be utilized to isolate the impact of one variable on another, net of other possible confounds. Yet in a technical sense, nearly every study that we conduct in the discipline could be called an experiment. This course therefore focuses on the conditions under which an experiment can be used to forge a path between cause and effect. We will focus on elements of experimental design in the first part of the course. The second part of the course will march students through a host of different ¿quasi-experimental¿ techniques: matching, instrumental variables, regression discontinuity, and differences-in-differences. Taken together, these techniques comprise much of the toolkit that can be used by researchers to make causal claims.

Credits: 3

Requisites: Pre-Requisite: PSC 508 or GEO 505 or PSY 607 or SOC 504.