Replication Research Symposium

Reproducibility is a foundational pillar of credible scientific research, yet it remains undervalued in many empirical fields. In this talk, I argue that reproducibility—verifying that original results can be precisely regenerated from the authors' own data and methods— and robustness analysis are not only feasible but cost-effective and scalable across disciplines. I describe how routine reproducibility checks can significantly enhance research quality, help identify inadvertent errors, and strengthen the credibility of scientific findings. Furthermore, I propose that reproducibility should perhaps be a mandatory step before resources are allocated to replication efforts using new data. Drawing on examples from economics, political science, psychology and related disciplines, I make the case that fostering a culture of reproducibility benefits the entire scientific ecosystem.