It appears to be Paper Mill Sweeps Week here at Retraction Watch.
On Tuesday, we reported on an editor who believes one such operation was responsible for the withdrawals of at least two articles in her journal.
Now, the Royal Society of Chemistry is retracting 68 articles, across three of its titles, after an investigation turned up evidence of what it suspects was the “systemic production of falsified research.” The society said it is in the process of beefing up its safeguards against milled papers and plans to train its editors to have “extra vigilance in the face of emerging, sophisticated digital fraud.”
That sounds much like what science sleuth Elisabeth Bik and others have found in hundreds of papers “that all appear to contain Western blots with the exact same background, often accompanied by hairball-like flow cytometry plots,” and other signs of manipulation like templatized text and plagiarism. (Bik tells us she was not behind the 68 new retractions.)
According to a statement released Wednesday by the RSC:
We are retracting 68 articles that have been published in RSC Advances, with a small number of articles also to be retracted from RSC Medicinal Chemistry and Food and Function. These retractions are on the basis of what we believe to be the systemic production of falsified research, and we are one of a number of publishers to have been affected by such activity.
We take a zero tolerance approach to any alleged fraud in our journals and will be informing institutions and funders where evidence and investigation shows an individual has or may have submitted fraudulent research in their name.
This latest incident is the result of an organised and sophisticated operation, summed up as what is known as a “paper mill”. We will be sharing insights and experience of this with colleagues across the publishing community, as part of a concerted, coordinated effort to stamp out falsified research.
The RSC statement — which does not list the retracted papers — said its “extensive investigation” began last year and involved
independent image integrity and scientific experts, and consulting with other publishers who are affected. We have identified common features across these papers, including the subject matter being mainly biomedical, as well as instances of image duplication and manipulation. We have identified that many of these papers are written in very similar structures or templates, despite having no common authors. These papers often appear to be legitimate when viewed on their own, and many of the concerning features only come to light when comparing multiple papers.
This is clearly a serious breach of our ethical policies, and as members of the Committee on Publication Ethics (COPE), we have stringently applied their recommendations during our investigation. Further information on the individual circumstances will be included in the retraction notice for each paper, and we will continue to be extra vigilant, ensuring that no stone remains unturned in papers published in our journals in future.
RSC said it also will be implementing an “action plan” to avoid being stung similarly in the future. Under the plan:
we will continually review our systems to mitigate against fresh attempts to play the system and undermine science.
Enhanced screening has been introduced at the initial stages of assessment, using knowledge we have gained from this investigation, to identify submissions of concern and to reject them. We are also trialling software to detect image manipulation, which will help to identify instances where images have been altered.
We are providing additional training for our Associate Editors, ensuring extra vigilance in the face of emerging, sophisticated digital fraud techniques, giving them the knowledge to be able to identify concerning submissions and take necessary action.
Finally, we would like to apologise to our community that we are having to make these retractions, and to assure our readers that we take the integrity of the scientific record very seriously and will take all necessary steps to ensure our editorial and peer review processes keep pace with the evolving threat and advancements in scientific fraud.
Update, 1330 UTC, 1/20/21: RSC sent us one retraction. It reads:
Retraction of ‘Overexpression of PCDH8 inhibits proliferation and invasion, and induces apoptosis in papillary thyroid cancer cells’ by Liang Chang et al., RSC Adv., 2018, 8, 18030–18037, DOI: 10.1039/C8RA02291G.
The Royal Society of Chemistry hereby wholly retracts this RSC Advances article due to concerns with the reliability of the data. The images in the article were screened by an image integrity expert who identified instances of manipulation of western blots, which undermines the reliability of all the western blots presented in the article.
In Fig. 1D, the PCDH8 band shows signs of image manipulation. The bottom edge of the second band is missing and there are also repeating patterns in the background. In addition, in the GAPDH control panel, the 2nd and 4th bands are duplicates.
The authors contacted the Editor asking to retract the article because they found that PCDH8 has no effect on papillary thyroid cancer cell apoptosis. The authors informed us that they performed this assay again and found that their prior results were wrong.
Given the significance of the concerns about the validity of the data in the article, the findings presented in this paper are not reliable.
The authors did not respond to any correspondence regarding the wording of the retraction notice.
Signed: Laura Fisher, Executive Editor, RSC Advances
Date: 7th January 2021
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
I bet that most of these forged articles come from China.
When can we say “just about any paper published in China is crap?” When people realize there is a culture problem here, which people confound for racism. That ain’t going to happen in PC academia….
Failed scientist, I’ll take your bait. I’m no fan of your calls for caning those who commit misconduct even if I find myself sympathizing with the underlying sentiment and you personally. However, ignorant and offensive statements like the one above will lead me to consider “just about any post from you as crap” from this time forward and I will ignore each of them.
I over-reacted, and I apologize. Thank god I’m leaving academia. Failed scientist.
I appreciate your response and accept your apology.
Glad to see that RSC now has a zero-tolerance policy against fraud. Some time ago, RSC Advances published a review paper whose authors presented as their own ( “To assess the kinetics of the isomerization procedure theoretically, we focused on the determination of the transition state” ) the results of other researchers, and instead of retracting it when made aware of the error, RSC advances allowed the authors to publish a corrigendum, as if that had been a simple typo.
The one example provided is present in a spreadsheet of papermill suspects, linked by stylistic similarities (notably the flow-cytometry apoptosis scatterplots, which juxtapose porcupines and portraits of Thidwick the Big-Hearted Moose). But it was the only case of a RSC journal in my list.
I think what one can reasonably say is that the Chinese government’s policy of requiring MD candidates to publish research papers–without providing them the time or resources to actually do research–needs to stop. It creates an enormous pressure to commit this type of fraud. This is blackening the reputation of Chinese research in general, and is hugely unfair to honest Chinese researchers. It’s a textbook case of “perverse incentive.”
(I am reminded of the Chinese grad student applicant who wrote to me to say, “I know you will think my test scores are fake. But that’s something rich people in Beijing can do, not poor village people like me.” I don’t know if he was telling the truth, but certainly it’s plausible, and if so he was as much a victim of Chinese corruption as we were.)
Mary, your anecdote about the Chinese graduate student applicant brings to mind the US college admissions cheating scandal of some months back: https://www.theguardian.com/us-news/2019/mar/12/college-admissions-fraud-scandal-felicity-huffman-lori-loughlin. For sure, money and power can and does corrupt. And let’s not forget that custom-made academic paper services have been around for a while for those who can afford them, as this article from 1995 illustrates: https://www.thefreelibrary.com/This+pen+for+hire%3A+on+grinding+out+papers+for+college+students.-a016916045. One has to wonder how many of those papers written for graduate students made their way into print.
Mary may have already answered this with her comment but what is the motivation behind paper mills? What do they get out of producing these fraudulent papers?
Coming from our collaborators in PRC, so a hearsay: publications in “Western” scientific journals is tied to promotional and funding considerations, as well as direct monetary awards by Chinese funding agencies and universities. Obviously the connection between publications and funding/career considerations isn’t endemic to PRC, the exacerbating factor there the all-too-symplistic implementation of a novel (regionally) concept. In US the rewards for publication aren’t immediate, there’s considerable stigma associated with retractions and often dire career consequences for fraud. So the temptation to game the system, if such arises, is tempered. PRC science lacks this balance: rewards are immediate, and the repercussions are remote. Personally I believe that the potential for fraud is pretty much the same in all national science systems, but some have better safeguards against it.
Your description of reward schemes is in resonance with the old but still valid definition of what is a conflict of interest [1], given by Dennis F. Thompson in the NEJM in 1993:
“A conflict of interest is a set of conditions in which professional judgment concerning a primary interest (such as a patient’s welfare or the validity of research) tends to be unduly influenced by a secondary interest (such as financial gain)” [2].
This is also why the short statement universally included in the “Conflicts of interest” section at the end of these articles sounds odd. My favorite is: “Conflicts of interest: none”. This is conciseness at its best.
[1] https://en.wikipedia.org/wiki/Conflict_of_interest
[2] DOI: 10.1056/NEJM199308193290812
My understanding that the self-reported conflict of interest is intended to inform about direct interest affecting judgement, such as holding shares in or receiving funding or other considerations from a company whose product you are reviewing. More nebulous interests not directly affecting judgement are always present, but harder to enumerate. One can also argue, that rewards system applied universally doesn’t constitute a personal conflict of interest. I agree tho, conflict of interest none sounds rather meaningless. The statement should say something like “conflict of interest as defined in…”.
Isn’t it a requirement nowadays to provide datasets to publishers when submitting manuscripts for consideration? One would think these would be difficult to falsify as they’d be date/time stamped electronically. Digital audit trails in lab data capture seems to be becoming the norm.
RSC actually provided a partial list of retractions:
//Below is a list of papers retracted in connection with what we believe is a paper mill. Please see http://rsc.li/paper-mill-response for more information.
1. https://doi.org/10.1039/D1RA90011K
2. https://doi.org/10.1039/D1RA90018H
3. https://doi.org/10.1039/D1RA90019F
4. https://doi.org/10.1039/D1RA90023D
5. https://doi.org/10.1039/D1RA90020J
6. https://doi.org/10.1039/D1RA90015C
7. https://doi.org/10.1039/D1RA90013G
8. https://doi.org/10.1039/D1RA90006D
9. https://doi.org/10.1039/D1RA90007B
10. https://doi.org/10.1039/D1RA90010B
11. https://doi.org/10.1039/D1RA90008K
Thanks, this was only posted a day after their announcement, which at the time did not include any such list.
Iy wouldnt take much to get the same think going here. I am a retired scientist….reached a pretty good level. Personal chair in a Russell group university. However, once I achieved this I must admit my star started to wane…..I was pressurised to take early retirement at 60 as I wasnt performing! No real regrets (I’m actually a inal music undergraduate now and enjoying it).
I came from a generation where you published solid papers in good journals (in my area J Gen Physiol, J Physiol, Am J Phsiol, Brit J Pharm). The emphasis was on getting it right…. 2 – 3 good papers a year was a good benchmark.
About 15 years ago we had a very bright and ambitious medic doing a PhD with us. I remember him quite calmly telling me that he had to publish at least 10 papers from his PhD, or his career would flounder. This is *before* he’d done any real experimental work.
I was gobsmacked but I now see that he’d seen a reality that i had completely missed.
Despite my success….I do see myself as a failed scientist. Ultimately Icouldnt adapt to the new realuty. In the last 10 years of my career I saw a huge amount of blatant academic fraud. Even though two of these did develop into major investigations…..the university senior admin basically passed the buck between each other. Nobody was prepared to take responsibility and openly admit that this had been going on. In the end, basically nothing was done.
The list of retracted papers now includes 70 articles. The work done by the RSoC staff is impressive, since in each instance, a detailed notice gives the ground for retraction.
1. https://doi.org/10.1039/D1RA90011K
2. https://doi.org/10.1039/D1RA90018H
3. https://doi.org/10.1039/D1RA90019F
4. https://doi.org/10.1039/D1RA90023D
5. https://doi.org/10.1039/D1RA90020J
6. https://doi.org/10.1039/D1RA90015C
7. https://doi.org/10.1039/D1RA90013G
8. https://doi.org/10.1039/D1RA90006D
9. https://doi.org/10.1039/D1RA90007B
10. https://doi.org/10.1039/D1RA90010B
11. https://doi.org/10.1039/D1RA90008K
12. https://doi.org/10.1039/D1RA90012A
13. https://doi.org/10.1039/D1RA90014E
14. https://doi.org/10.1039/D1RA90016A
15. https://doi.org/10.1039/D1RA90017J
16. https://doi.org/10.1039/D1RA90021H
17. https://doi.org/10.1039/D1RA90022F
18. https://doi.org/10.1039/D1RA90024B
19. https://doi.org/10.1039/D1RA90025K
20. https://doi.org/10.1039/D1RA90026A
21. https://doi.org/10.1039/D1RA90027G
22. https://doi.org/10.1039/D1RA90028E
23. https://doi.org/10.1039/D1RA90030G
24. https://doi.org/10.1039/D1RA90031E
25. https://doi.org/10.1039/D1RA90046C
26. https://doi.org/10.1039/D1RA90033A
27. https://doi.org/10.1039/D1RA90034J
28. https://doi.org/10.1039/D1RA90035H
29. https://doi.org/10.1039/D1RA90032C
30. https://doi.org/10.1039/D1RA90036F
31. https://doi.org/10.1039/D1RA90037D
32. https://doi.org/10.1039/D1RA90038B
33. https://doi.org/10.1039/D1RA90039K
34. https://doi.org/10.1039/D1RA90040D
35. https://doi.org/10.1039/D1RA90047A
36. https://doi.org/10.1039/D1RA90048J
37. https://doi.org/10.1039/D1RA90049H
38. https://doi.org/10.1039/D1RA90050A
39. https://doi.org/10.1039/D1RA90051J
40. https://doi.org/10.1039/D1RA90052H
41. https://doi.org/10.1039/D1RA90041B
42. https://doi.org/10.1039/D1RA90042K
43. https://doi.org/10.1039/D1RA90043A
44. https://doi.org/10.1039/D1RA90044G
45. https://doi.org/10.1039/D1RA90045E
46. https://doi.org/10.1039/D1RA90054D
47. https://doi.org/10.1039/D1RA90055B
48. https://doi.org/10.1039/D1RA90056K
49. https://doi.org/10.1039/D1RA90057A
50. https://doi.org/10.1039/D1RA90058G
51. https://doi.org/10.1039/D1RA90059E
52. https://doi.org/10.1039/D1RA90060A
53. https://doi.org/10.1039/D1RA90061G
54. https://doi.org/10.1039/D1RA90062E
55. https://doi.org/10.1039/D1RA90063C
56. https://doi.org/10.1039/D1RA90064A
57. https://doi.org/10.1039/D1RA90065J
58. https://doi.org/10.1039/D1RA90071D
59. https://doi.org/10.1039/D1RA90072B
60. https://doi.org/10.1039/D1RA90073K
61. https://doi.org/10.1039/D1RA90074A
62. https://doi.org/10.1039/D1RA90075G
63. https://doi.org/10.1039/D1RA90076E
64. https://doi.org/10.1039/D1RA90066H
65. https://doi.org/10.1039/D1RA90067F
66. https://doi.org/10.1039/D1RA90068D
67. https://doi.org/10.1039/D1RA90069B
68. https://doi.org/10.1039/D1RA90070F
69. https://doi.org/10.1039/D1FO90004H
70. https://doi.org/10.1039/D1MD90001C