Shock. Denial. Disbelief. Sadness. Regret. Embarrassment.
Those, according to a commentary published today in Science magazine, are some of the reactions from both scientists and science journals when they are found to be involved in cases of potential plagiarism.
The commentary was offered by researchers at the University of Texas Southwestern Medical Center at Dallas, who used a computer-based text-searching tool to analyze millions of randomly selected research abstracts.
The analysis of Medline, a database of biomedical research articles, found 9,120 entries “with high levels of citation similarity and no overlapping authors,” including 212 pairs of articles “with signs of potential plagiarism,” the researchers wrote.
The lead author, Harold Garner, a professor of biochemistry and internal medicine at the medical center, said he and his colleagues then conducted a survey of the authors and journal editors, promising them anonymity. The survey responses, Mr. Garner wrote, included explanations, denials, embarrassed apologies, and some retractions. Among the original authors, he wrote, 93 percent were not aware of the duplicate article.
Mr. Garner also wrote an article for Nature, published in January 2008, titled “A Tale of Two Citations,” that reported a similar finding: His computerized search of several million scientific-journal articles revealed thousands of cases in which one article had large similarities with another article.
Both of Mr. Garner’s articles were based on research involving the Medline database and UT Southwestern’s computer-based text-searching tool, eTBLAST. And in both articles, Mr. Garner suggested that the size and severity of this problem continued to be ignored by publishers.
The Nature article warned against both plagiarism by another author and “self plagiarism,” in which the same author or authors present duplicate findings to different journals.
Mr. Garner nevertheless said that his Science magazine report represents a significant advance over his earlier article in Nature. The survey published in Science, while anonymous, prompted 83 internal investigations at scientific journals, which in turn led to 43 cases in which an article was retracted, he said.
That compares to only 17 such retractions last year, which is a more typical annual figure, he said.
Such a case of plagiarism or duplication can have serious medical consequences, Mr. Garner said, as it could lead a doctor who is investigating a patient’s condition to believe a scientific finding is more recent, or perhaps more reliable, because of its repeated appearance in medical journals. —Paul Basken