Economics 270D: Research Transparency Methods in the Social Sciences

This interdisciplinary methodology course introduces students to tools and resources in the conduct of transparent research within all social science fields, with specific applications to Economics, Political Science, Psychology, Sociology, and Demography. The class addresses principles, issues, and practices around research openness and integrity including: meta-analysis, multiple testing corrections, differential privacy data concerns and frontier topics and next steps in research transparency. This course is taught at the Ph.D. level and was developed by the BITSS Faculty Director, Edward Miguel.

The 14 lectures in this course were delivered at the University of California, Berkeley in Spring 2015. The course covered a range of approaches that aim to enhance the transparency and reproducibility of social science research. These lectures are appropriate for students, faculty, and researchers in social science disciplines and related fields. We hope that these materials can provide a resource to other instructors interested in teaching about research transparency.

Course Materials

View the full syllabus here. Course outline:

I. Introduction to Issues of Research Transparency and Reproducibility

• Lecture 1: Understanding the problem
• Lectures 2-3: Publication bias and the file-drawer problem

II. Approaches to Pre-registration

• Lectures 4-5: Using pre-analysis plans
• Lecture 6: Transparency in non-experimental research
• Lecture 7: Data Adaptive Pre-specification Approaches

III. Building Scientific Knowledge

• Lecture 8: Approaches to the replication of research
• Lectures 9-10: Meta-analysis techniques

IV. Open Data and Materials

• Lecture 11: What does open data do?
• Lecture 12: Differential privacy and the cost of openness

V. Looking Forward

• Lecture 13: Presenting and visualizing data
• Lecture 14: Next steps in changing scientific research practices