A paper that claimed to have developed a new method to predict acid drainage from mines was not so novel after all, according to one of its authors.
In a series of emails to Retraction Watch, Dulian Zeqiraj of the Polytechnic University of Tirana, Albania, admitted to lifting figures and tables from other articles and said he might also have left some “text as it is in original.”
His paper, “A Novel Stochastic Approach for Modeling Acid Mine Drainage in Three Dimensions,” was published November 17 in Process Safety and Environmental Protection, an Elsevier title.
That the article managed to clear peer review is astonishing, said Muhammad Muniruzzaman, a senior scientist at the Geological Survey of Finland in Espoo, who discovered last week that Zeqiraj’s team had plagiarized his work.
At first, Muniruzzaman said, he was pleasantly surprised when he got an email from Google Scholar notifying him that some of his studies had been cited. “That did not last beyond 10 seconds when I started browsing through the paper,” he told Retraction Watch.
A dozen figures and tables from two of Muniruzzaman’s papers appeared, without citation, in the new article as examples ostensibly validating the modeling. And several chunks of text were simply paraphrases of sections in Muniruzzaman’s work.
“I tried to read their paper,” said Muniruzzaman, who is also an adjunct professor at the University of Turku. “It doesn’t really make sense. If you try to link from one section to another, there is no link whatsoever. It’s almost like gibberish, you feel like it’s not even human generated.”
He added that one of his coauthors had looked into other publications by Zeqiraj and had found what appeared to be a pair of duplicate papers, both crammed with mathematical formulas, in two Elsevier titles.
Zeqiraj said that although he had failed to cite Muniruzzaman’s work appropriately, “essentially my work is true and I can demonstrate the code in MATLAB. At this point I apologize for not mentioning the authors. I can do this now by asking Elsevier to give me the opportunity to cite the figures.”
In a later email, he wrote:
It is true that figures and tables are not cited. As for the text it is true that I have read his paper. But it works so: You for example read a paper and take from it the resume. Maybe in may [sic] case I have not made the resume but have left the text as it is in original. As for the plagiarism check it cost about 5 euros to control a paper online, practically nothing. I repeat: I have not used ChatGPT AND I AM OPEN FOR ANY DISCUSSION. Please give me the opportunity to speak with the Professor. But one thing is for sure. I will request the retraction from Elsevier if they don’t allow me to make the changes.
We asked Elsevier how the plagiarized paper could have passed peer review and whether it had gone through a plagiarism check. The publisher told us:
The manuscript was run through iThenticate and had a total similarity index of 28%, with a maximum of 5% from any one source. The manuscript was sent for peer-review, and following revision, was accepted for publication in Process Safety and Environmental Protection.
We are now investigating Dr Muniruzzaman’s concerns and have put the paper, which is an Article-in-Press, on hold until our investigation is complete.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, subscribe to our free daily digest or paid weekly update, follow us on Twitter, like us on Facebook, or add us to your RSS reader. If you find a retraction that’s not in The Retraction Watch Database, you can let us know here. For comments or feedback, email us at [email protected].
“That the article managed to clear peer review is astonishing”
After following this blog for some years, I don’t see it as an astonishing thing anymore.
“The manuscript was run through iThenticate and had a total similarity index of 28%, with a maximum of 5% from any one source. The manuscript was sent for peer-review”
How does this sentence even makes sense? What is the point of having maximum similarity limit exceeded, and still sending it for the review?!
Yeah, I’ve been an AE for years and have seldom found iThenticate reports useful. If indeed something was directly plagiarized, so what if the overall percentage of text copied was only 5%? It does tend to catch some degree of paraphrasing, but is dumb about whether it is appropriate. It won’t differentiate, for example, between writing “Evgenii (2023) reported profound, breakthrough stuff …..” (appropriate) versus “Here we have demonstrated exact same same profound, breakthrough stuff….. (plagiarized). iThenticate also matches small sentence fragments such that the total similarity scores for a perfectly legit manuscript will come in with an overall similarity score of 20% or so. Sometimes it figures out to ignore identical text in citations and references, but not always. And of course it can’t help with text translated to English or plagiarized or recycled figures. About the only thing it’s good for is detecting a nearly complete resubmission of previously published article. For that I suppose publishers consider it a useful CYA tool.
I completely disagree. iThenticate is an extremely useful tool if you use it properly (like everything else). It’s not the number that matters, although it can be a red flag. And of course it’s “dumb” so a human eye is essential. Even a match as low as (or even lower than) 1% can be very revealing if you examine the source – there might be a duplicated image to go with that text, or that source might be one of many that form part of a papermill batch for example. This particular case just confirms that most people, including the publishers who provide the resources for us, still don’t understand how to use them properly. At a COPE-organised meeting more than a decade ago I was left completely aghast by a discussion amongst editors and publishers about “what percentage of overlap” is considered acceptable. If they had any idea how to use the tool they would not even be asking that question!
Thanks for the post. I am not sure even if the author cited the figures, it would give him the right to do a carbon copy of the published Figures/Tables. The author needs to carefully read the codes of conduct for scientific publications before making such nonsense comments. The issue is not about the Matlab codes he allegedly has or the references he forgot to cite!! It is about stealing other people’s work and replicating the same data in different papers, which needs to be taken very seriously.