Research Reveals Data on Which Institutions Are Retraction Hotspots (nature.com) 24
Chinese hospitals dominate a first-ever analysis of scientific paper retractions worldwide, with some institutions having retraction rates 50 times higher than the global average, according to data published in Nature.
Jining First People's Hospital in Shandong leads with more than 5% of its research output from 2014-2024 being retracted -- over 100 papers. The hospital had disciplined 35 researchers for publication fraud in late 2021 amid a broader Chinese government crackdown on paper mills selling fake manuscripts.
The analysis, based on data from three research integrity firms, found that about 60% of retracted articles over the past decade had authors affiliated with Chinese institutions. Other retraction hotspots include universities in Saudi Arabia, India, Pakistan and Ethiopia.
While retractions remain rare globally, affecting fewer than 0.1% of published papers, rates have tripled over the past decade. In 2023, over 10,000 papers were retracted, mostly from journals owned by publisher Wiley's now-closed Hindawi subsidiary following widespread peer review fraud. The surge in retractions partly reflects increased scrutiny from research integrity experts who spot problematic papers, rather than definitively indicating which institutions produce fraudulent work, according to Ivan Oransky, co-founder of Retraction Watch.
Jining First People's Hospital in Shandong leads with more than 5% of its research output from 2014-2024 being retracted -- over 100 papers. The hospital had disciplined 35 researchers for publication fraud in late 2021 amid a broader Chinese government crackdown on paper mills selling fake manuscripts.
The analysis, based on data from three research integrity firms, found that about 60% of retracted articles over the past decade had authors affiliated with Chinese institutions. Other retraction hotspots include universities in Saudi Arabia, India, Pakistan and Ethiopia.
While retractions remain rare globally, affecting fewer than 0.1% of published papers, rates have tripled over the past decade. In 2023, over 10,000 papers were retracted, mostly from journals owned by publisher Wiley's now-closed Hindawi subsidiary following widespread peer review fraud. The surge in retractions partly reflects increased scrutiny from research integrity experts who spot problematic papers, rather than definitively indicating which institutions produce fraudulent work, according to Ivan Oransky, co-founder of Retraction Watch.
no surprise (Score:5, Interesting)
Back in the day when I was reviewing papers for publication, I caught some Chinese authors verbatim plagiarizing another paper without attribution or quotes. But the editors of the publication refused to take action, citing sympathy for the Chinese authors.
Re: (Score:1)
What publication was it?
Re: (Score:2)
Arrest rates (Score:4, Insightful)
Re: (Score:2)
Suppose there is a new police chief in Chicago. Two years after he's selected arrest rates are double, and conviction rates are much higher. Can we conclude that he is incompetent?
I'm trying to parse this. What are you talking about? All of the Chinese retractions are likely either fraudulent or incompetence based. My bet is pressure based fraud.
Re:Arrest rates (Score:4, Insightful)
Re: (Score:2)
If they were always a rigorous, perhaps.
But even the summary mentions a recent crackdown by the Chinese government. It would be silly to assume that an increase in retractions shortly after that kind of crackdown cannot possibly be connected.
When changes in leadership occur, quite often, changes in policies and practices occur as well.
Re: (Score:3, Interesting)
In this case I would say China probably does deserve its current reputation, but the retractions are evidence they're trying to improve.
The need to retract and the punishment for having to retract are both increasing, which means the pressure to publish pleasing documents over quality ones (regardless of whom you're trying to please) will be relatively reduced.
This is a good thing.
If the retraction stats stay high for years, that's an indication that the pressures causing low quality papers still overpower
Re: (Score:2)
My point is that the places doing more retracting are likely the ones performing more rigorous checks, while other places not retracting as often may well be publishing flawed research and leaving it out there. The number of retractions is not sufficient to decide the quality of a journal or academic institution.
Retractions in and of themselves are not indications of fraud. There can be accidents found after publication. Now in the case of Jining First People's Hospital, 100 papers over 10 years - That's indicative of something really askew. With that rate, it might be considered to not accept publications for say, 5 years.
There is a certain level of trust that has been broken, and at least, they need to be more careful in preparing their papers. Because the peer review is one thing - but after publication, the
Measure of Quality (Score:2)
The number of retractions is not sufficient to decide the quality of a journal or academic institution.
It may not be sufficient to decide the quality of the journal but it absolutely is enough to say something about the quality, or lack thereof, of the academic institute. Researchers have access to the same journals worldwide to publish in so if one institute is getting a 5% retraction rate for its papers averaged over a 10 year period when the global average rate of retraction for journal papers is about 0.02% [researchgate.net] something is very, very wrong at that institute.
The most generous interpretation possible is t
Re: (Score:2)
My point is that the places doing more retracting are likely the ones performing more rigorous checks,
Or those places are not doing rigorous checks at all. They are just responding after international teams of volunteers using digital tools to spot plagiarism, photoshopped data, and other signs of fraudulent research make the problems public.
Re: (Score:2)
Re: (Score:2)
Congrats on winning the most retarded post of they week award on /.
Re: (Score:2)
Suppose there is a new police chief in Chicago. Two years after he's selected arrest rates are double, and conviction rates are much higher. Can we conclude that he is incompetent?
This is an inapplicable comparison, and here's why: The journals are themselves responsible for deciding what to print, or not, and for assigning papers to reviewers.
For your comparison scenario to be apt, the police chief would have to be the step-parent of the criminals being arrested.
And the reasons? (Score:3)
Psychology, for example, might have very few retractions simply because nobody attempts to replicate obviously ridiculous research, even when that research makes headlines and is used to set public policy.
Re:And the reasons? (Score:4, Interesting)
To some degree. When Wiley (old, big publisher) bought Hindawi (young, fast-growing upstart Open Access publisher), they quickly discovered that the entire publishing house was infiltrated by paper mills. They retracted thousands of papers, and closed many journals. However, some of their own journals are also heavily infiltrated by paper mills, and those had far fewer retractions.
Conversely, another young upstart, MDPI, has very few retractions even though they also have a high number of paper mill productions, including some that they know about very well and have "investigated".
Wiley is obviously a much more serious publisher than MDPI, albeit more hesitant to clean their old house than the newer that they bought.
Computer science, by the way, has a far higher rate of retractions for academic misconduct [doi.org] than other disciplines, and it's not because it's so easily replicated, it's because it's rampant with fraud. I'll give you an example of ridiculous verbiage [doi.org] that somehow stays in the academic literature thanks to the non-efforts of IEEE and an academic community that will publish anything but read nothing. You don't need a replication study to see that this isn't a serious academic work. It's most likely a patchwork of plagiarised text that's been fed through some paraphrasing filters to avoid automatic detection.
But yeah, psychology is surely not serious and computer science is very smart.
Re: (Score:2)
Article: Evidence Suggests Misconduct (Score:2)
Surely retractions can be a sign of a healthy science?
Yes, in the same way that a fever is a sign that your body is healthy because it is fighting off a nasty infection. It means that science is working to protect itself from being affected by bad results but, like an infection, you ultimately need to find and potentially treat the source of the problem because if left untreated it can get bad enough to overwhelm your defences and cause serious harm.
The fact that certain institutes have had 5% of their papers retracted over the 10 years period examined is
Re: (Score:2)
I think it may not be obvious to people outside of science what leads to a retraction.
Unless the author retracts, generally good faith papers are left alone. They might be wrong (bollocksed p experiments, errors in a proof), or just shit (like a cack handed rediscovery of an existing technique). They might also attract comments from other authors.
Generally something egregious had to have happened and be noticed, something that often doesn't happen if the paper is a bit crap and boring and no one ever looks
Misleading (Score:2)
"While retractions remain rare globally, affecting fewer than 0.1% of published papers, rates have tripled over the past decade."
The article points out that some institutions have far higher rates of retraction than others. However, this nonuniformity also extends to fields and conferences/journals. In the computer architecture field and especially for the more prestigious conferences, retractions are extremely rare. One reason is that acceptance rates are low and reviews tend to be rigorously scrutinized.