Causal Inference

Introduksjon

Please note that this is a preliminary course description. The final version will be published in June 2026.

In the 21st century, information is created, digitized, and stored at unprecedented rates. The availability of high-dimensional, large-scale datasets has transformed the landscape of business analytics and data science, providing new avenues for insightful decision-making. However, vast datasets alone are insufficient for addressing fundamental analytical questions central to contemporary research and practice.

This course introduces the potential outcome framework as a rigorous foundation for causal inference. We explore microeconometric techniques designed to uncover causal relationships using experimental and quasi-experimental data. We discuss the promise and pitfalls of large-scale experimentation and consider empirical applications relevant for business and policy analysis.

Kursets innhold

The course covers the following topics:

• The potential outcome framework
• Large-scale experimentation
• Noncompliance
• Treatment effect heterogeneity
• Regression discontinuity designs
• Difference-in-differences and event studies
• False positives, p-hacking and publication bias
• Supplementary analysis and replication