Scientific Reports, a Springer Nature title, has retracted an article a group of sleuths described as “a kind of case study of all the red flags for fraud that we look for” in an open letter to the publisher’s head of research integrity.
The article, “Hybrid CNN-LSTM model with efficient hyperparameter tuning for prediction of Parkinson’s disease,” appeared in September 2023. It has been cited 11 times, according to Clarivate’s Web of Science.
In December 2023, a PubPeer user commented on 13 tortured phrases the Problematic Paper Screener had flagged in the article, such as the use of “Parkinson’s illness,” “Parkinson’s infection,” and “Parkinson’s sickness” rather than Parkinson’s disease.
“These typically result from an attempt to avoid plagiarism detection using a paraphrasing software,” the commenter wrote about the phases. “How come these incorrect wordings survived proofreading by the coauthors, editors, referees, copy editors, and typesetters?”
An “author correction” to the article appeared the following April, stating “several non-standard terms” had been “replaced with established scientific terminology,” including all of the incorrect references to “Parkinson’s disease.” The correction listed eight sections with updated wording for other phrases.
The original article also incorrectly identified the software the authors used in their analysis as python rather than SPSS, according to the correction, and five sentences were updated.
In an August comment on the correction, the original PubPeer commenter wrote:
Astonished that these tortured phrases and other issues got through peer review. Astonished that a correction was accepted. Astonished that some tortured phrases are still there, e.g., 10-overlay instead of 10-fold.
The correction was “outrageous,” Guillaume Cabanac, a computer scientist and sleuth who developed the Problematic Paper Screener, posted on X October 9:
That same day, the journal added an editor’s note to the article, which stated it was under investigation, stating: “Appropriate editorial action will be taken once the investigation has concluded.”
Meanwhile, multiple sleuths added comments to the PubPeer pages for the article and its correction, pointing out more issues.
On October 16, sleuth Dorothy Bishop posted “An open letter regarding Scientific Reports” on her blog, addressed to Chris Graf, Springer Nature’s director of research integrity. In it, she and 22 other “sleuths and forensic meta-scientists” wrote they were “concerned that Springer Nature is failing in its duty to protect the scientific literature from fraudulent and low quality work” due to problems with Scientific Reports.
The corrected article was “a striking example” of the sometimes “inadequate” response of the journal editor or integrity team to concerns the sleuths have raised, they wrote:
As is evident from comments on PubPeer, it turned out to be a kind of case study of all the red flags for fraud that we look for. As well as (still uncorrected) tortured phrases, it contained irrelevant content, irrelevant citations, meaningless gibberish, a nonsensical figure, and material recycled from other publications.
This is perhaps the most flagrant example, but we argue that it indicates problems with your editorial processes that are not going to be fixed by AI.
The sleuths concluded with their recommendations for addressing the issues they raised, including “an urgent audit of all editors” for Scientific Reports.
On November 7, the editors retracted the offending article “due to concerns regarding the validity and veracity of the work presented,” with a list of six issues sleuths had posted on PubPeer. The authors “have not responded to correspondence from the Editors about this retraction,” according to the notice. Corresponding author Pugazhenthan Thangaraju of the All India Institute of Medical Sciences in Raipur, did not respond to our request for comment.
Rafal Marszalek, chief editor of Scientific Reports, told Retraction Watch in a statement:
The paper was already under investigation prior to the publication of the letter. After carefully considering the facts of the matter, we concluded that retracting the paper was the appropriate editorial action to take.
Three Springer Nature journals, Soft Computing, Optical and Quantum Electronics, and Environmental Science and Pollution Research, have been retracting articles en masse beginning last year, approaching 900 combined.
As we reported previously, the publisher listed paper mills as a significant threat to its business in paperwork for its October initial public offering. The company identified “a significant number” of articles published in its journals that came from paper mills, and stated:
These investigations continue and could result in many retractions, which could negatively impact our reputation.
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].
Springer will charge you $59 to read a book from the 1930s, then they publish this drek – what are they doing with all that money?
If the authors didn’t respond to Springer’s enquiries, who made the correction?!
papers like this, of such low quality that they are transparent frauds, serve the purpose of attracting payments from academics who need publication credits. how many co-authors did this paper have?
“how many co-authors did this paper have?”
Ten.
This shows how troubling the quality of peer review is. And of course it does not always happen, and it is not always bad. But peer review does not seem to have a consistent quality which you can rely on. It’s sad that there are so many cases that are so obvious and still are missed by peer reviews.