A reference catalogue of documented, formally adjudicated cases — each established by journal retractions, official investigations, or court findings. They illustrate what is at stake when the published record fails.
The cases below are not our investigations. Each has been independently established by the relevant journals, institutions, or courts, and is cited here with its primary sources. We present them because they show, in the public record, why the integrity of science matters — and why sustained misconduct so often goes unseen until great cost has already accrued.
A landmark 2006 Nature paper proposed that a specific amyloid-beta assembly (Aβ*56) impaired memory — a foundational result for a major line of Alzheimer’s research. A 2022 Science investigation, prompted by an independent image analysis, raised extensive concerns; the paper was retracted in 2024 over image manipulation. Per Retraction Watch, it became one of the most-cited papers ever to be retracted.
A rising star at Bell Labs published a rapid series of breakthrough results in molecular electronics and organic superconductivity. In 2002 an independent committee found he had fabricated and manipulated data across many papers — in some cases reusing identical noise traces across supposedly different devices. More than twenty papers in Science, Nature, and Physical Review were retracted; his doctorate was later revoked.
Two celebrated Science papers claimed the creation of patient-specific embryonic stem-cell lines. A university investigation found substantial portions of the data fabricated; the papers were retracted in 2006. The case also raised serious ethical questions over egg sourcing. The lead author was prosecuted and received a suspended sentence.
Two Nature papers claimed that a simple acid bath could reprogram ordinary cells into a pluripotent state. Within weeks, image manipulation and other problems were identified, and independent replication failed entirely. An institutional investigation found falsification and fabrication; both papers were retracted in 2014.
A surgeon performed experimental synthetic-trachea transplants and published encouraging results. The lead patients died. An institutional review found research misconduct, papers were retracted or flagged, and a Swedish court ultimately convicted him, with a custodial sentence. Here the gravest cost was measured in human lives, not money.
A 1998 Lancet paper linking the MMR vaccine to autism was later shown, through investigative reporting, to rest on altered patient histories and undisclosed conflicts of interest. It was fully retracted in 2010 and the lead author lost his medical licence. The BMJ reported he had a commercial plan projected to earn substantial sums — a potential profit, not realised income.
Genomic “signatures” claimed to predict patients’ response to chemotherapy were used to launch clinical trials. The underlying data could not be reproduced; a federal oversight body later found research misconduct. At least eleven papers were retracted, and trials that had enrolled patients were halted. A civil case was settled on undisclosed terms.
An anaesthesiologist who received pharmaceutical funding to run clinical trials was found to have never conducted many of them, fabricating data — in some cases inventing patients. About twenty-five papers were retracted. He was convicted of health-care fraud and served prison time, with court-ordered restitution.
A prominent social psychologist was found to have invented entire datasets, handing fabricated “results” to students and collaborators to analyse. An official inquiry concluded that dozens of papers contained fabricated data; 58 were retracted. The case implicated numerous doctoral theses built on his data.
Two separate, long-running cases of large-scale data fabrication in anaesthesiology hold the modern records for retracted papers. In one, statistical analysis showed the reported data distributions were “too good to be true.” Between them they account for several hundred retracted papers — the largest retraction totals on record.