Ranking Fields by p-Value Suspiciousness

The article discusses the issue of p-hacking in scientific research, where researchers manipulate data to achieve statistically significant results. The focus is on the distribution of p-values in various fields, with an emphasis on the "uncanny mountain" of p-values between 0.10 and 0.01, suggesting potential data manipulation. The analysis reveals that many fields exhibit suspicious clustering of p-values, indicating possible bias or selective reporting. Economics stands out as having relatively more credible results compared to other disciplines, attributed to rigorous reporting practices and a focus on statistical robustness. The replication crisis is highlighted as a universal challenge across all fields due to p-hacking and questionable research practices. The study examines large datasets from diverse fields to assess the prevalence of suspicious p-values and emphasizes the need for improved research integrity and transparency to address these issues.