Scientific misconduct estimated to drain millions each year - SCIENTIFIC AMERICAN

Katherine Harmon

As speculation swirls around the status of possible investigations into research by the prolific Harvard psychologist Marc Hauser, a new study drills down to figure out the true cost of scientific misconduct.

Neither Harvard nor the federal government, which has funded some of Hauser's work that has been retracted or amended, has come forward with statements about the status of the scholar's work. But in the meantime, any investigation is likely costing the university—and possibly the government—a pretty penny, according to the new work, published online August 17 in PLoS Medicine.

Scientific misconduct is defined as "fabrication, falsification or plagiarism in proposing, performing or reviewing research or in reporting research results," according to the U.S. Office of Research Integrity (ORI). A 2009 meta-analysis of misconduct studies found that about 14 percent of responding scientists reported having witnessed falsification by others—and 2 percent confessed (anonymously) to having been involved in fabrication, falsification or modification of data themselves.

An inquiry into scientific misconduct often leads to research disruption, evidence confiscation and lengthy meetings, all of which can add up quickly in terms of expenses such as faculty and staff labor. A typical case might run in the neighborhood of half a million dollars, concluded the authors of the new case study, led by Arthur Michalek of the Roswell Park Cancer Institute in Buffalo, N.Y. Taking as an example a real case from their own institution, they estimated the direct costs of that instance of misconduct to be about $525,000.

Michalek and his colleagues break down the costs into three categories: fraudulent research (grants, investments and equipment), investigation (faculty, personnel and external assistance) and remediation (loss of current or pending grants and others in the affected lab). These calculations don't take into account other potential costs (such as lawsuits and loss of future funding) and intangibles (such as loss of trust, demoralization of associates and any research conducted on the basis of fraudulent results).

In the case of the example at their institution, the researchers estimated that the most expensive aspect of the internal investigation was the demands on faculty, who spent hundreds of hours assessing the case both during and outside of formal meetings.

The new policy forum paper does not aim to be a universal measure of misconduct costs. "Our experience will likely not be wholly representative of other institutions," the researchers noted, acknowledging that their estimates thus far "amount to a 'best guess' scenario."

By their calculations, however, the 217 U.S. cases of misconduct reported to the OIR in 2009 would add up to more than $110 million each year. And the actual rate of misconduct remains uncertain, "owing largely to its clandestine nature as well as to the problem of underreporting," the researchers noted.

Steps to avoid wrongdoing in the first place—such as education, training, mentoring and monitoring—are not free either. But, Michalek and his colleagues estimate that, "the costs of these proactive activities pale in comparison to the costs of a single case of scientific misconduct."