Widespread failure to reproduce the results of published scientific studies
The replication crisis refers to the discovery that a significant proportion of published scientific findings cannot be reproduced by independent researchers — calling into question the reliability of the scientific literature across multiple disciplines. The crisis was brought into sharp focus by several landmark attempts to replicate published studies.
The Open Science Collaboration's 2015 effort to reproduce 100 psychology studies found that only 36% yielded statistically significant results on replication. A 2012 Amgen study found that only 6 of 53 "landmark" cancer biology papers could be replicated. Bayer HealthCare reported that approximately 75% of published findings in their areas of interest could not be validated internally.
The replication crisis has profound implications for public trust in science and for policy decisions based on scientific evidence. If a significant percentage of published findings are wrong — due to p-hacking, publication bias, small sample sizes, or outright fraud — then regulations, medical treatments, and public health policies built on those findings may be causing harm rather than preventing it. The crisis also provides legitimate ammunition for those who question scientific consensus, complicating efforts to distinguish between healthy skepticism and bad-faith denialism.
