Science still has a fraudulent publication problem
Among European countries retractions of biomedical-related papers have nearly quadrupled in recent years.
I came across this pretty interesting article from Nature which argues that biomedical-related papers have nearly quadrupled in their retractions based on retraction data collected between the years 2000 and 2021:
This report is based on a study published in early May1:
This would seem rather alarming, but research misconduct has been a growing concern within the science community, even though misconduct has likely been around since scientific endeavors first came about.
However, over the past few years greater attention has been made to the fact that paper mills, the use of AI, and other forms of manipulation and tampering of images has increased, meaning that more papers with unreliable information are making their way into publications and being reported on by the media.
For instance, a few years ago it was found that many Chinese papers did not undergo proper peer-review processes and many were found to contain evidence of fraud or misconduct:
In this case it appeared that some researchers were directing emails for reviewer recommendations to either themselves or to 3rd parties who would write glowingly of these papers which helped to get them published:
MOST's 27 July announcement marked the culmination of an investigation into the mass retraction this past April of 107 papers by Chinese authors that appeared in a single journal, Tumor Biology. The papers, published between 2012 and 2016, were pulled after editors found "strong reason to believe that the peer review process was compromised," Editor-in-Chief Torgny Stigbrand, of Umeå University in Sweden, wrote on 20 April on the website of the publisher Springer. (Springer, an arm of Springer Nature, published Tumor Biology until December 2016; the journal is now operated by SAGE Publications.)
Investigators say the authors engaged in an all-too-common scam. Tumor Biology allowed submitting authors to nominate reviewers. The Chinese authors suggested "experts" and provided email addresses that routed messages from the journal back to the researchers themselves, or to accomplices—sometimes third-party firms hired by the authors—who wrote glowing reviews that helped get the papers accepted.
And rather shockingly nearly 10,000 papers were retracted in 2023 alone, with nearly 8,000 of the retracted papers being derived from Hindawi- a subsidiary of the large publication Wiley:
The reason for the retractions related to Hindawi is described as the following:
The bulk of 2023’s retractions were from journals owned by Hindawi, a London-based subsidiary of the publisher Wiley (see ‘A bumper year for retractions’). So far this year, Hindawi journals have pulled more than 8,000 articles, citing factors such as “concerns that the peer review process has been compromised” and “systematic manipulation of the publication and peer-review process”, after investigations prompted by internal editors and by research-integrity sleuths who raised questions about incoherent text and irrelevant references in thousands of papers.
Most of the Hindawi retractions are from special issues: collections of articles that are often overseen by guest editors and that have become notorious for being exploited by scammers to rapidly publish low-quality or sham papers.
Wiley seems to have addressed this concern by shuttering some of Hindawi’s titles and collecting papers and titles under the Wiley umbrella- a process that appears to have cost Wiley tens of millions of dollars.
Altogether, there has been a serious publication crisis going on which only appears to be getting worse (or at least not getting any better). These publication issues can be extremely harmful as it may lead researchers to follow false leads or to direct funding towards hypothesis that may not be accurate, as appears to be the case regarding some of the most pivotal research in Alzheimer’s Disease:
Falsified data can also be compounded when cited by other researchers or placed into review articles. Eventually, this research may be presented by news outlets or published into textbooks further propagating misinformation.
It makes one wonder how much of the science we are taught are actually accurate…
In the case of the more recent study researchers looked at retracted papers starting from January 1, 2000 up to June 30th, 2021. The study focused on “European” papers which included papers in which a corresponding author was affiliated with a European institution. This included countries that are part of the European Union as well as additional countries such as the UK, Turkey, and Russia to name a few.2
When reviewing the 2069 retracted papers the researchers noted that nearly 67% of the papers were cited as being retracted due to misconduct while around 16% were cited as being due to honest research errors.
When broken down researchers noted that the main reason for retraction was duplication:
Duplication is when portions of images or slides are reused across different figures within a study. An example of this can be found within the 2022 Science article from Piller above in which it appears that some of the Western Blot images published may have been duplicated. A portion of that sleuthing section can be seen below, but refer to the article for the entire investigative procedure:
Again, duplication is a rather common method of tampering with results, with some egregious instances being ones in which researchers may take slides and just turn them 180 and pretend as if they have captured something completely different.
Surprisingly, duplication misconduct appears to have dropped off in years post-2015 after increasing for over a decade. Interestingly, falsification/fabrication and use of unreliable data has increased over time.
Note that falsification/fabrication refers to instances in which there was clear evidence that data/figures were manipulated while unreliable data refers to instances in which the source of the data were unknown or not provided for review, meaning that the data could not be verified.
Unreliable data appears to have increased in certain countries such as Italy, and the authors suggest that this may be due to reviewers not being able to investigate the source of questionable data or it could be due to the rise in paper mills which may publish lower quality works.
I generally found this study to be interesting as it concerningly suggests that research misconduct is still an ongoing issue, and in the case of European countries appears to have increased over time.
Now, a caveat to the increase in retractions could be the fact that there are better tools to stop fabrication/duplication in studies, and therefore the increase in retractions could be a consequence of better detection. However, this perspective would call into question how many studies in years past may have slipped through publication even though they may have had egregious errors. Also, bear in mind that this study looked at papers from European countries in particular and doesn’t tell us anything about trends in other countries.
Nonetheless, as skepticism grows towards science and science institutions this should serve as a wake-up call that more due diligence needs to be done in order to ensure that published science has undergone extensive review and is free of misconduct. If institutions continue to pump out bad science and poor data then skepticism will continue to rightfully grow.
We need better science more than ever, and as of now it appears we may be a long way away from that coming to fruition.
If you enjoyed this post and other works please consider supporting me through a paid Substack subscription or through my Ko-fi. Any bit helps, and it encourages independent creators and journalists such as myself to provide work outside of the mainstream narrative.
Freijedo-Farinas, F., Ruano-Ravina, A., Pérez-Ríos, M. et al. Biomedical retractions due to misconduct in Europe: characterization and trends in the last 20 years. Scientometrics 129, 2867–2882 (2024). https://doi.org/10.1007/s11192-024-04992-7
Norway, Iceland, Liechtenstein, Serbia, Switzerland, and Belarus were also included as other countries not part of the EU that were also considered “European” for this study.
Certainly, detection has influenced these trends and there are many unknowns from the age before Photoshop, but... I still sense a growing antiscience culture within science.
I'm interested to know how crooked scientists come to be. Do they enter a research career with the proper motivation, only to be steered awry by external pressures (e.g., publish or perish)? Or do they simply have the wrong idea from the get-go (e.g., PhD as status symbol, research as entertainment)?
--
If even high-quality data is supposed to be replicated -- i.e., treated as untrustworthy -- before being built upon, the least we can do is treat* unverifiable data as falsified data. That would hopefully clear up the lot of papers with vague methods and hidden datasets, and force the fraudsters to work a lot harder.
*Treat, not categorize. Important difference.
The system needs a system to systematically monitor the system.
Or just a new system.