Sultans of swap: Turkish researchers plagiarized electromagnetic fields-cancer paper, apparently others

The Bosnian Journal of Basic Medical Sciences has retracted a paper it published in August by Turkish researchers on the potential cancer risks associated with exposure to electromagnetic fields, or EMFs.

The reason: Other people wrote nearly all of it.

According to an editor’s notice:

After the editorial office was alerted of possible plagiarism in the article, it conducted thorough investigation and concluded that the article apparently represents plagiarized material from two World Health Organization reports, one European Commission report and other sources. Since this is considered scientific plagiarism and scientific misconduct, Editor-in-chief has decided to withdraw the article. The authors have agreed with the editorial office decision.

As we’ve written in recent posts, publishers and editors increasingly are turning to software such as CrossCheck to detect plagiarists. But in this case, good old Google would have been sufficient. We did a little experiment, cutting a few sentences at random from the manuscript and pasting them into Google. We immediately came up with matches of verbatim theft from the European Commission report mentioned in the retraction notice, a 2006 document from the EC’s Scientific Committee on Emerging and Newly Identified Health Risks.

The Turkish authors didn’t just lift snippets. They stole entire columns worth of text.

For example, from the plagiarists:

Acoustic neuromas, benign tumours that develop very slowly, arise from the Schwann cells, which enfold the vestibulocochlear nerve. They are of particular interest because of their location. The Hardell-group from Sweden has in several studies reported raised relative risk estimates for acoustic neuroma and also with very short induction periods (19). Two of the Interphone components, Denmark and Sweden, have reported their country specific acoustic neuroma results (20, 21). Lönn et al. reported a doubling of the relative risk estimate after ten years of regular mobile phone use compared to subjects who never used a mobile phone regularly.

From the source:

Acoustic neuromas, benign tumours that develop very slowly, arise from the Schwann cells, which enfold the vestibulocochlear nerve (VIII. cranial nerve). They are of particular interest because of their location. The Hardell-group from Sweden has in several studies reported raised relative risk estimates for acoustic neuroma, also with very short induction periods (Hardell et al. 2005b). Three of the Interphone components, Denmark, Sweden, and Japan, have reported their country specific acoustic neuroma results (Christensen et al. 2004, Lönn et al. 2004, Takebayashi et al. 2006). Lönn et al. (2004) reported a doubling of the relative risk estimate after ten years of regular mobile phone use compared to subjects who never used a mobile phone regularly.

Although the Turkish authors try to reference the passage, only the 2004 Christensen article is consistent with the EC document.

The only surprise here is not that they were caught, but that the theft wasn’t caught during peer review. Five minutes of Googling led us not only to these passages but to what appear to be other instances of word theft by the same researchers. (We recently polled Retraction Watch readers on whether they checked for plagiarism during peer review; almost 70% of 85 respondents said they didn’t.)

Consider:

One of the authors, Vahdettin Bayazit, a biologist at Alparslan University in the town of Mus, and a co-author, Vahit Konar, of Firat University in Elazig, published an article earlier this year in the Journal of Animal and Veterinary Advances (Medwell Journals) with the dramatic title: “Biochemical and Physiological Evaluations of Limonoids as Potential Cancer Destroyers.”

Here’s a passage we pulled at random from the paper:

Recent research suggests that citrus fruit consumers may be getting another health benefit from orange juice and other citrus products called limonoids, which appear to possess substantial anticancer activity.

We threw it into the Google machine and got back this:

Recent research suggests that U.S. consumers may be getting another health benefit from orange juice and other citrus products – phytochemicals called limonoids which appear to possess substantial anticancer activity.

That line comes from a May 2000 report in a University of California, Davis publication, Perishables Handling Quarterly, issue 102, by USDA researchers.

Another random search from Bayazit and Konar turned up this:

This observation has added significance considering the predominance of limonin glucoside among the large amount of limonoid glucosides present in fruit tissues and juice.

And the USDA paper:

This observation has added significance considering the predominance of limonin glucoside among the large amount of limonoid glucosides present in fruit tissues and juice.

Although the Turkish authors cite the USDA researchers in their references, they don’t do so for their use of material above — which, while not a kosher way of dealing with verbatim material, might have mitigated the sin a bit.

We performed the same exercise with another paper by Vahdettin Bayazit (along with Murat Bayazit and Elif Bayazit, whom we take to be related but aren’t sure of that), titled “Evaluation of Bioceramic Materials in Biology and Medicine,” which appeared in the March 2010 issue of the Digest Journal of Nanomaterials and Biostructures.

All of these materials are biocompatible and osteoconductive. However, they differ considerably in the rate of resorption. HA resorbs very slowly compared with β-TCP, and bioactive glass.

The virtually identical passage appears in a 2004 article by Christine Knabe and colleagues in the Journal of Biomedical Materials Research:

All of these materials are biocompatible^1,3,4and osteoconductive.^1,3–13 However, they differ considerably in the rate of resorption. HA resorbs slowly compared to β-TCP1,4,7 and bioactiveglass.^5,6,12,13

Not the lack of citations in the former, by the way; the Knabe paper doesn’t even appear in the reference section.

We could go on with this, but we won’t.

Now, some authors might argue that government documents are fair game for reuse. That’s true only to a limited degree, but it certainly doesn’t apply to the verbatim repetition of text nor the lack of proper referencing. Yet the Turkish authors here not only plagiarized from government documents, they did so from “private” scholarship, too. So even the flaccid ignorance-of-the-rules-regarding-public-intellectual-property excuse doesn’t fly.

We’ve reached out to Bayazit and Knabe, as well as the editors of the journals involved here and will update this post if we hear from them.

Update, 5 p.m. Eastern, 11/30/10: We came across another paper in which Bayazit and his co-authors had apparently lifted material from an earlier work. The paper, which appeared in the Digest Journal of Nanomaterials and Biostructures, included language similar to that in a paper in Biomaterials. It was co-authored by Knabe, the source of other apparently lifted content.

We contacted the editor-in-chief of Biomaterials, David Williams of Wake Forest, for comment. His response:

I have now checked this.  It is true that Bayazit et al appear to have used similar language to that of Knabe et al in  a few sentences.

Knabe etal (2008) published a paper dealing with a clinical study of the use of beta tricalcium phosphate in maxillofacial surgery.  This was a very good, extensive, paper.

Bayazit et al published a very poor, rather elementary review-style paper, in a journal that I have never heard of before. Their work mostly concentrated on the standard technology of some bioceramics. They summarized some clinical results but failed to quote their source(s)-however the information they included was consistent with some of the conclusions of the Knabe paper.

I view this as an inappropriate use of someone else’s language in a  minor way. Bearing in mind that the use of these words did not attempt to wrongly ascribe someone else’s work as their own, but was more a clumsy lack of a citation, I do not see this as a major issue. Also, taking into account the relative obscurity of the ‘offending’ authors, their institution, the journal and the paper, I do not intend taking any action here.

Obviously you have a more serious case to investigate, in which you refer to ‘lifting substantial amounts of text’, which you are clearly in your rights to do, but I do not see any advantage in enjoining the very minor case in my journal with your more significant enquiry.

Williams’ response suggests that there is a threshold for plagiarism, and that an author’s or journal’s obscurity has something to do with that threshold. We’ll continue to monitor the Bayazit et al papers, but in the meantime would welcome Retraction Watch readers’ comments and thoughts.

Williams sent us further comments in response to an email about these issues:

It may appear arbitrary – and indeed as all editors know there are rarely absolutely clear cut issues in which the line is unequivocally drawn in the sand – but my position is based on a combination of many years experience and an element of pragmatism.

Just so you are aware, I stand very firmly on the established principles of scientific publishing, I suspect far more so than most editors. For many years I ran a school for referees and give workshops all around the world on publishing procedures and ethics. Since my journal is one of the leading scientific journals by any metric, I have to deal with many ethical and legal issues associated with authors who desperately want to publish with us.  I do get distressed about cases of authorship, which often arise with some of the world’s leading institutes, and take these issues very seriously. I do the same with clear cases of intentional plagiarism and fraud, and other matters of scientific misconduct – it takes up a great deal of editorial time, but this is a necessary part of the job.

I could not get seriously distressed with this case. As I mentioned to you, if you have evidence of significant lifting of work from other papers, you are correct to pursue this.  But I also repeat that this was a clumsy, and possibly inexperienced, lack of citation to some work that was not a mainstream issue with their paper, just discussed in passing. It makes no difference to the validity of either paper. If I were to deal with such cases on a very formal basis, there would be little time for anything else.