View comments(4) | Comment on this blog   

If you're found guilty of plagiarism or scientific fraud, is your academic career as a researcher over? Not according to a study published in Science tomorrow (August 8), which contradicts a long-standing assumption by suggesting that rebuilding a career after a misconduct finding is difficult, but not impossible.

"While the punishments [for misconduct] are severe, there are hopes for redemption," said Jon Merz of the University of Pennsylvania, a coauthor on the study.

Merz and Barbara Redman, who has a joint appointment at Wayne State University and the University of Pennsylvania, examined the fates of researchers who, as independent investigators (as opposed to graduate students or postdocs), were found guilty of misconduct by the Office of Research Integrity, the NIH branch that polices misconduct, between 1994 and 2001.

They report that 19 of the 37 scientists for whom they were able to find publication data continued to publish at least once per year. "People who were found guilty of plagiarism [as opposed to expressly fabricating or falsifying data] get less severe of a punishment, so they were more likely to continue to publish," Redman noted. Ten of the 28 scientists whose employment information they were able to trace continued to hold academic appointments after the ORI ruling. Originally, 23 out of those 28 had worked in academia.

However, Merz and Redman's data, as well as interviews they conducted with the seven researchers who agreed to speak with them, indicate that recovering from the misconduct ruling was extremely difficult. Unsurprisingly, the group's average publication rate was significantly lower after the ruling, dropping from 2.1 to 1.0 publications per year. Twelve of the scientists ceased to publish completely. In interviews with Merz and Redman, researchers described extensive personal and financial hardships due to the ruling.

"Some felt that they didn't know what they were doing was wrong," or otherwise disagreed with the ORI's conclusions, said Redman. Also, she noted, several would have liked to appeal the ORI's decision, but couldn't afford to do so. "We don't really know if the decision would have been different" if they had appealed, she said. "They weren't able to use the full system."

The seven scientists that were willing to discuss their experiences may not be representative of scientists in the group, said Redman. "We think the seven are probably survivors," said Redman. Indeed, those researchers did say that they were able to rebuild their careers, in some cases with the help of their institutions. "We don't have data on this, but it would be my sense that some institutions are better at helping people to rehabilitate," she added.

Redman and Merz acknowledge that the data are limited, but suggest they raise questions about the system's fairness and whether or not researchers are being punished too harshly. "It's not clear that the [researchers] who did the worst things necessarily got the worst punishments, or were unable to regroup," said Redman. "What to do about [misconduct] is a value question. Should people be allowed to redeem themselves?"

She said that the study is too limited to warrant a change in how penalties for misconduct are meted out, but should be expanded. "It's pretty important to the scientific community and to individual scientists to find out empirically what is happening" to researchers after they are found guilty of misconduct, she said.

Following up on the findings is going to be difficult, however, in part because of researchers' reluctance to discuss their alleged misconduct. Also, Redman and Merz limited the study to cases decided by 2001, in order to give publication rates time to rebound. Now, they'll need to wait a few years before conducting a follow-up study, said Redman.

The results of an anonymous survey which asked researchers about incidents of misconduct they have observed, published in Nature in June, suggested that such incidents often go unreported. Merz said he wishes that survey had probed a little deeper, and had asked survey respondents to report what the outcome of those instances had been. The authors, he said, "should have asked, 'You knew about this case. So, what happened? Did the person get fired?'"