Guillaume Cabanac, a computer scientist from the University of Toulouse in France, created an online tool capable of examining a scientific paper and identifying any references to articles that have been retracted. Named the Feet of Clay Detector, in reference to a Biblical idiom about statues built on fragile bases, the software has already found at least 1,700 papers with a high number of retracted references. Depending on the importance of the citations, such cases can compromise the conclusions of the articles.
Cabanac shared a list of extreme examples of these articles with the journal Nature. One was published in 2023 by two computer engineers at the Vels Institute of Science, Technology, and Advanced Studies in Pallavaram, India, in which 18 of the 30 references—60% of the total—had been retracted. The article was published in the proceedings of an international conference on intelligent structures and systems hosted by the Institute of Electrical and Electronics Engineers (IEEE). “We are not accusing anybody of doing something wrong. We are just observing that in some bibliographies, the references have been retracted or withdrawn, meaning that the paper may be unreliable,” Cabanac told Nature.
Every author has experience with articles that they have cited being retracted—the need to publish a correction to remove a retracted reference is not uncommon. But when a large number of citations in one paper have been retracted, it can be a sign that the author was not rigorous about the papers they chose to reference and did not thoroughly check their validity. It may also be a reflection of more serious misconduct.
At the top of Cabanac’s list is an article that has had 65% of its 51 references retracted. The paper was published in the journal Materials Research in 2012 by Ali Nazari, who was accused of scientific misconduct and fired from Swinburne University of Technology in Melbourne, Australia, in 2019. Of the 33 retracted citations, 32 were articles authored by Nazari himself, who is fifth on a list compiled by Retraction Watch of the authors who have had the most scientific articles retracted. He has had 104 articles retracted for image manipulation, duplication of results, and interference with the peer-review process.
The Feet of Clay Detector is Cabanac’s second contribution to tackling problematic papers. In 2021, he and his colleagues developed a program that detects what they called “tortured phrases”—strangely or poorly translated expressions that reveal an attempt to circumvent plagiarism detection systems or the use of artificial intelligence software to generate text (see Pesquisa FAPESP issue n° 317).
The story above was published with the title “Online tool identifies articles with retracted references” in issue 344 of October/2024.
Republish