In this episode, Thomas Plümper and Eric Neumayer explore the hidden challenges in modern science, from outright fraud to the subtler practice of “tweaking” data that distorts results. They examine why the self-correcting nature of science often falls short, how incentives and academic pressure drive misconduct, and the double-edged role of AI in both enabling and detecting fraud. The conversation also tackles debates around p-values and statistical reasoning, shares cautionary case studies, and proposes solutions like greater data transparency and stronger verification standards.

Chapters

00:00 Introduction to Fraud in Research

06:21 The Nature of Fraud Detection

08:56 Incentives and Motivations for Fraud

10:43 Self-Correction in Science

12:13 Understanding Statistical Significance

13:04 The Role of Replication in Research

14:32 Bayesian vs Frequentist Approaches

23:09 Understanding Bayesian Statistics and Its Implications

26:24 The Humility of Empirical Science

27:16 Concrete Examples of Scientific Fraud

32:52 Proposed Solutions to Scientific Fraud

Credibility Crisis in Science with Thomas Plümper and Eric Neumayer