300292 Method Class: Introduction to Causal Inferences in Experimental and Observational Data (S) (WiSe 2022/2023)

In the past years, rigorous identification of causal effects received increasing attention in applied research in the social sciences as well as neighboring disciplines, such as psychology and economics. This trend is accompanied by a popularity of research designs that are more sensitive to causal interferences than correlational studies. On the one hand, randomized controlled trials (RCT) –implemented either in a lab-context (e.g., behavioral social sciences) or in the field (e.g., intervention studies)– became popular even in disciplines such as sociology and political sciences that relied heavily on observational data. On the other hand, the identification of natural experiments and exogenous variation in the treatment (e.g., instrumental variables regression (IV), regression-discontinuity-designs (RDD)) improved causal inferences in observational data.

This Methods Class provides an applied introduction to the concept(s) of causality in the social sciences as well as frequently applied research designs sensitive to causal inference. The seminar covers both the designs, implementation, and analysis of RCTs as well as causally sensitive regression methods in observational data, in particular IV and RDD. In hands-on sessions, we will use example data to apply these methods and discuss their implementation using Stata. Also, we discuss ways how to present causal interpretations in own research, such as doctoral dissertations, submissions to journals, and conference talks.

This class is open to doctoral students of the faculty and advanced MA students as well as researchers interested in the topic. Dependent upon the preferences of participants, we may organize the seminar in weekly meetings or in (half)day workshops.