Reproducibility in the Social Sciences: Lessons learned for students, researchers, and institutions from 1,000 articles

Originally posted:

Talk at Stockholm School of Economics, Stockholm, Sweden

Abstract

The American Economic Association’s Data Editor has reviewed more than 1,000 empirical articles since July 2019, and worked with authors to improve the reproducibility of their research compendia (replication packages). Some lessons emerge from this work. In this presentation, I will focus on lessons for young scholars (students and young researchers), on possible lessons even for more seasoned researchers. Students and researchers are embedded within institutions, and I will discuss the kind of support that institutions (universities, data providers, compute services) should be providing to students, faculty, and researchers, for a robust, reproducible, and transparent science enterprise.