A recent study published in the Proceedings of the National Academy of Sciences finds that the majority of retractions are caused not by error, but by fraud.
According to the study, which looked at all of the 2,047 biomedical and life-science research articles indexed by PubMed as retracted, only 21.3% of the retractions were attributable to error. And 67.4% were due to misconduct, including 43.4% for fraud or suspected fraud, and 25% duplication, which includes both plagiarized and duplicative publication.
The study contradicts earlier research that showed most of the retractions were due to errors.
The study also found that retractions rose sharply after 2005, which was the year the Federal government drastically reduced funding for biomedical research, indicating the possibility that increased competition for limited resources helped to spur unethical behavior.
Also noteworthy was that 43% of the retractions came from just 38 labs, despite there being thousands of labs worldwide, indicating that a small number of “bad actors” may be creating lion’s share of the problem.
One of the challenges the study had was determining the actual cause of the retraction. According to the study’s senior author Dr. Arturo Casadevall, many of the journals would either hide or mislead as to why the article was retracted. This prompted the researchers to turn to the National Institutes of Health Office of Research and Integrity and Retraction Watch to learn the true cause.
Though the study paints a generally bleak picture for retractions, including one of an increased pace of retractions and a growing problem with unethical behavior as government budgets continue to shrink, it offers at least some hope for progress.
With so many of the retractions being caused by duplication, it means a large number can likely be avoided through proper use of duplication detection software, which can help filter out both plagiarism and, in many case, duplicative publishing before they become they complete review and require a retraction.
It also indicates that the ethical issues come, disproportionately, from a small number of sources. This makes communication between publications and awareness of retractions important. Through closer scrutiny of known problem labs, other issues may also be prevented.
So while there are no simple solutions to the problem and only limited things that can be done to address the external forces that are helping to compound the issue, the research does show clear steps that can be taken to help reduce and mitigate the problem, potentially eliminating a large number of retractions.
Still, even with those steps, research fraud is likely to continue to a growing problem and one that will require constant vigilance by everyone involved in the publication process, including researchers, journals and readers, to help address the issue and keep unethical research to a minimum.
Related