In scientific method we don’t just trust: or why replication has more value than discovery

Robert S. Gordon Jr. Lecture | Wednesday, March 13, 2019 | 3:00 to 4:00 p.m. ET

John P.A. Ioannidis, M.D., D.Sc.

Professor of Medicine, Professor of Health Research and Policy, Professor of Biomedical Data Science, and Professor of Statistics

Co-Director, Meta-Research Innovation Center (METRICS)

Stanford University

Website

Summary

Science is the best thing that has happened to Homo sapiens. It is important to apply the scientific method in ways that are the most efficient in leading to—and translating—important discoveries. However, this goal is not easy. There are many situations where research practices are applied in suboptimal ways, resulting in a reproducibility crisis in which trust in scientific findings is diminished. In his lecture, Dr. Ioannidis will discuss how we can improve the robustness, efficiency, and transparency of research practices. In particular, emphasis on replication may improve the credibility track record of current investigations across many different disciplines. Several disciplines—such as genetic epidemiology—that have already incorporated replication as a sine qua non have seen a dramatic improvement of their credibility. In most scientific disciplines, replication has far more scientific and practical value than discovery. In the current environment, in most fields, discovery without replication is mostly a nuisance.