Scientific misconduct harms prior collaborators

Scientists should choose their associates carefully, according to researchers at the University of Luxembourg and the Centre for European Economic Research (ZEW), Mannheim, as future misconduct by colleagues could seriously impact the reputations of their former collaborators.

While there has always been anecdotal evidence that this is the case, a study by Prof. Katrin Hussinger (University of Luxembourg) and Dr Maikel Pellens (ZEW, Mannheim and KU Leuven, Belgium) now provides empirical evidence. "Guilt by Association: How Scientific Misconduct Harms Prior Collaborators" was based on the misconduct filings of the U.S. Office of Research Integrity from 1993 to 2008. A group of 856 prior research collaborators of fraudulent scientists was identified by using publication records dating back five years before the case of misconduct. The study considered only the cases in which a retraction or correction of the research occurred owing to scientific misconduct.

Compared to a control group, the results showed an average drop in citations of 8 to 9 percent for previous colleagues. Citations play an important role in science as they show the impact of research in the scientific community. Researchers with a high citation count are usually also more successful in attracting funding and receive more lucrative job offers. The reduced citation count could therefore have significant career implications.

"The results of the study are worrisome," explained Prof. Hussinger. "Our research shows that guilt by association stretches back to projects prior to the fraud case and thereby to unsuspecting and uninvolved co-workers."

While stigmatization by association has been observed in different settings and contexts, the results from the field of academia are problematic in their own ways, according to Prof. Hussinger: "Trust is a crucial aspect of communicating science and conveying research results to the public. The ripple effects of one misconduct case can put at risk the reputation of a much larger group of scientists and even institutions."

Even though the researchers cannot provide a simple solution to the issue, guilt by association should be treated seriously, Prof. Hussinger and Dr. Pellens argue. An unwanted implication, Prof. Hussinger concluded, could be the underreporting of actual fraud causes: "Knowing that they might be penalised for mere association might make researchers think twice before speaking out."