Tag: ai

A round-up from last month, with news articles about science integrity, retractions of several highly cited papers in the fields of Alzheimer’s disease and stem cell research, generative AI and misconduct, and another retraction for Didier Raoult.

General Science Integrity News

• An epidemic of scientific fakery threatens to overwhelm publishers. More than 10,000 scientific papers were retracted last year as “paper mills” exploit the system. Adam Marcus and Ivan Oransky – Washington Post
• How restorative justice could help to heal science communities torn apart by harassment misconduct– Could a programme used by police to tackle repeat offending be used in academia? Some institutions are keen to find out. – Sarah Wild – Nature Career Feature
• The case for criminalizing scientific misconduct – Since universities and journals have failed to regulate themselves, we should strengthen our external layers of oversight. Below are two complementary ideas – Chris Said
• Pressure From the Top – Chinese institutions and authorities are indeed becoming worried about the increasing flood of low- (or zero-) quality papers that come from all sorts of Chinese sources.  – Derek Lowe – In the Pipeline – Science
• 62% of Indians demand mandatory research integrity training for early-career researchers – A survey indicated the need to have broad support for mandatory in-person research integrity training across researcher demographics – Education Times

Highly cited Alzheimer’s paper retracted by Nature

• Daily briefing: Landmark Alzheimer’s paper will be retracted. Senior author plans to retract her team’s highly cited study on the cause of Alzheimer’s disease after acknowledging that the paper contains manipulated images. Plus, why cicadas shriek so loudly and how CO2 helps viruses stay alive in the air. – Flora Graham – Nature Briefing
• Alzheimer’s study pulled over data tampering. American university researchers insist their main findings still stand in dementia paper that had been cited 2,500 times. It is the most-cited paper ever to be retracted, according to Retraction Watch, a group that flags dubious research and regularly finds signs of evidence tampering. – Rhys Blakely – The Times UK
• After Amyloid. Authors of a landmark Alzheimer’s disease research paper published in Nature in 2006 have agreed to retract the study in response to allegations of image manipulation. University of Minnesota (UMN) Twin Cities neuroscientist Karen Ashe, the paper’s senior author, acknowledged in a post on the journal discussion site PubPeer that the paper contains doctored images. – Mike Barr – POZ
• Do we have Alzheimer’s disease all wrong?
  Retracted studies and new treatments reveal the confusing state of Alzheimer’s research. The authors of the retracted paper in question, which was published in Nature in 2006 and claimed to identify a specific target for future drug development, agreed to withdraw their research in full, two years after a stunning investigation by Science found that key images had been doctored.
  Dylan Scott – Vox
• Alzheimer : les dessous de la fraude scientifique derrière l’une des études les plus influentes. Il y a deux ans, un chercheur jusqu’ici inconnu formulait des accusations de fraude contre l’un des articles scientifiques les plus importants au monde. Il vient d’obtenir gain de cause. Antoine Beau – L’Express [In French]
• Retraction, 26 June 2024: https://www.nature.com/articles/s41586-024-07691-8
  

Even higher cited stem cell paper retracted by Nature

About 17 years after New Scientist investigators Eugenie Reich and Peter Aldhous revealed problems in a Nature 2002 paper about making adult stem cells pluripotent (see 2007 articles in New Scientist and New York Times and this recent X thread here), and more than four years after I found additional problems in that paper (PubPeer), Nature decided to retract it. Meanwhile, it has been cited nearly 4,500 times.

• Nature retracts highly cited 2002 paper that claimed adult stem cells could become any type of cell. Nature has retracted a 2002 paper from the lab of Catherine Verfaillie purporting to show a type of adult stem cell could, under certain circumstances, “contribute to most, if not all, somatic cell types.”
  The retracted article, “Pluripotency of mesenchymal stem cells derived from adult marrow,” has been controversial since its publication. Still, it has been cited nearly 4,500 times, according to Clarivate’s Web of Science – making it by far the most-cited retracted paper ever. – Ellie Kincaid – Retraction Watch
• ‘Dit is heel pijnlijk’: veelbesproken stamcelstudie van Catherine Verfaillie herroepen door vakblad ‘Nature’. Het is de finale klap voor het artikel, waar sommigen al langer vragen bij stelden. Als reden verwijst Nature naar onregelmatigheden in twee afbeeldingen die de Nederlandse Elisabeth Bik, een microbiologe gespecialiseerd in de materie, 4,5 jaar geleden had gesignaleerd.
  – Barbara Debusschere – De Morgen [In Dutch]
• Intrekken van paper over “baanbrekend ” stamcelonderzoek komt niet helemaal onverwacht voor Catherine Verfaillie –  Maxie Eckert – Nieuwsblad  [In Dutch]
• Waarom wordt baanbrekend onderzoek van professor Catherine Verfaillie na 22 jaar van tafel geveegd? “Het is een smet op mijn blazoen” – Koen Baumers – Nieuwsblad
• Copy-pasten met Photoshop: ook vakblad Nature trekt paper met “baanbrekend” stamcelonderzoek van Catherine Verfaillie terug – Maxie Eckert – De Standaard
• Paper by researcher Catherine Verfaillie (67) withdrawn by the journal Nature – Leuven Actueel

Other notable misconduct findings

• Faked results lead to retraction of high-profile cancer neuroscience study. An investigation found that the experiments required more animals than the scientists had purchased. Dalmeet Singh Chawla – The Transmitter
• Star botanist likely made up data about nutritional supplements, new probe finds – After 2022 exoneration, fresh inquiry concludes Steven Newmaster fabricated data that questioned quality of echinacea, ginkgo, and other substances – Charles Piller – Science
• Findings of research misconduct have been made against Shaker Mousa, Ph.D., M.B.A., FACC, FACB (Respondent), who was a Professor, Chairman, and Executive Vice President of the Pharmaceutical Research Institute, Albany College of Pharmacy and Health Sciences (ACPHS) – Federal Register

Another retraction for Didier Raoult

The International Journal of Obesity retracted an article on which Didier Raoult was the corresponding and last author, “because the authors could not confirm approval from an appropriate ethics committee to recruit healthy donors to provide samples for use in this research.”According to the Retraction Watch Database, this is his 13th retraction. The paper had been flagged on PubPeer because it was one of five papers with the same local approval number, while it appeared to not have CPP approval (regional IRB-like approval needed in France for human subjects research). Read more about the problems in Raoult’s papers here, here, and here.

Artificial intelligence and science misconduct

A (now-retracted) paper featuring an AI-generated rat with a giant penis has caused a lot of laughs on social media. But AI is capable of making much more believable images. Is the scientific publishing world ready for a tsunami of fake photos and dataset? And how can we use AI tools to actually detect fraud? Also see below in the new publication list.

• Smelling a rat with AI – Judith Hochstrasser – Horizons Switzerland
• Dieses Bild ist gefälscht – Moritz Marthaler – Tagesanzeiger (in German)
• AI vs. AI: How to detect image manipulation and avoid academic misconduct – Deepesh Bodekar – Enago Academy
• Academics despair as ChatGPT-written essays swamp marking season
  ‘It’s not a machine for cheating; it’s a machine for producing crap,’ says one professor infuriated by rise of bland scripts – Jack Grove – Times Higher Education
• Online course provider launches AI plagiarism detection tool – Breccan F. Thies – Gazette
• Springer Nature unveils two new AI tools to protect research integrity – Geppetto and SnappShot playing important role in stopping fake research from being published – Springer Nature press release
• Universities regulate use of AI in writing – More than 80 percent of college students have used AI tools to assist their assignments, according to a survey conducted by China Youth Daily in November. – Zhao Yimeng – China Daily

Science Integrity around the world

• Elite researchers in China say they had ‘no choice’ but to commit misconduct – Smriti Mallapathy – Nature News
• Barriers to Investigating and Reporting Research Misconduct. Between January 2023 and March 2024, the UK Research Integrity Office (UKRIO) convened a working group to explore what barriers there may be to investigating and reporting research misconduct. 
• China ramps up fight against online lies and misconduct – Cao Yin – China Daily

Interviews with science sleuths

• Ils traquent les études bancales ou frauduleuses : la croisade des chevaliers blancs de la science. Chiffres tronqués, copier-coller grossiers, textes absurdes écrits par l’intelligence artificielle : des chercheurs traquent dans les études scientifiques les pratiques douteuses de leurs confrères. L’infectiologue Didier Raoult fait régulièrement les frais de leur intransigeance. – Gaël Lombart – Le Parisien
• Frode Scientifica: Una intervista ad Elisabeth Bik – Open Science @UNIMI

New publications and editorials

• Proposed Increases in Government Authority Over Research Misconduct Proceedings – Caron et al., JAMA
• The Impact of Generative Artificial Intelligence in Scientific Content Synthesis for Authors – Chhavi Chauhan – American Journal of Pathology
• The Impact of AI on Academic Research and Publishing – Brady Lund et al., arXiv
• Codes of conduct should help scientists navigate societal expectations – Ambrosi et al. – Humanities & Social Sciences Communications
• Editorial: Adapting to AI – Artificial intelligence tools have the potential to revolutionize how scientists work and publish. We share our ground rules for managing the inherent risks. – Nature Geoscience

Retraction Watch’s Weekend Reads:

• June 1: Weekend reads: Another vaccine-autism retraction; ‘Why We Commit Academic Fraud’; on a psilocybin-depression correction
• June 8: Weekend reads: Major Alzheimer’s paper slated for retraction; research dog breeder pleads guilty; biomedical retractions quadruple
• June 16: Weekend reads: ‘An epidemic of scientific fakery’; death threats for critics; Cleveland Clinic settles mismanagement allegations for $7.6 million
• June 22: Weekend reads: A researcher explains how he publishes every three days; scientific bounty hunters; criminalizing scientific misconduct
• June 29: Weekend reads: Huge cash bonuses for publishing in Nature; Alzheimer’s researcher charged with fraud; fines for buying authorship