US scientists are more likely than their counterparts in other countries to publish research studies containing fabricated or falsified data, according to a study by R. Grant Steen. And scientists who perpetrate fraudulent research like this are more likely than most to repeat the egregious behavior.
To reach these conclusions, Steen reviewed all 788 English language research papers that had been retracted from the PubMed database during the past decade. Compiling relevant information from the associated retraction notices, Steen classified the causes for retraction as being due to either fraud or an honest procedural mistake.
Steen found that about 75% of the retractions resulted from procedural error, and 25% were caused by fraud. Remarkably, papers retracted because of fraud were published in journals with a higher “impact factor,” that is, journals containing articles that are cited more frequently by other scientists.
Steen also found that articles retracted because of fraud tended to have more co-authors and to take longer to retract than the ones retracted because of procedural mistakes.
Of the 788 retracted papers, fully one-third featured a US-based scientist as the first author, and one-third of these were retracted because of fraud. Asian nations accounted for an additional 30% of the retracted papers, although only one-fourth of them were retracted because of fraud. In fact, the US was the only country in which significantly more papers were retracted because of fraud than procedural errors.
A stunning 53% of all fraudulent papers had been penned by a “repeat offender,” a first author that had been associated with a retracted paper in the past. In contrast, only 18% of the papers retracted because of mistakes featured first authors that had been previously involved with a retracted paper.
“The duplicity of some authors is cause for concern,” Steen wrote. The results of this study, “suggest that papers retracted because of data fabrication or falsification represent a calculated effort to deceive. It is inferred that such behavior is neither naïve, feckless nor inadvertent.”