New Analysis Challenges Study Suggesting Racial Bias at NIH

A new analysis challenges a troubling 2011 study that suggested that black researchers encounter racial bias when they seek funding from the National Institutes of Health (NIH). The report finds that African Americans do just as well as whites at similar institutions who have an equivalent research record.

In the 2011 study published in Science, economist Donna Ginther of the University of Kansas, Lawrence, and co-workers probed years of confidential NIH grants data. They found that black scientists' chances of winning a grant for their research idea were a startling 10 percentage points lower than for white scientists even after controlling for their institution, research training, and publication record. NIH Director Francis Collins was shaken by the results; one possible explanation was that peer reviewers might have an unconscious bias against African Americans, he said.

Collins and his advisers came up with an action plan. Last December, NIH announced a $500 million, 10-year program aimed at boosting the number of young minority scientists and improving mentoring for minority researchers. To reduce the potential for racial bias, NIH also plans a pilot project that will test peer review of research proposals that have been made anonymous.

Questions have persisted, however, about the Ginther study (see letters here). Now, another research group has examined the possible role of racial bias at NIH by comparing head-to-head the productivity and funding of black and white medical researchers at the same institutions.

Biomedical engineer Ge Wang of the Virginia Polytechnic Institute and State University (Virginia Tech) in Blacksburg and co-authors at several institutes randomly selected 40 black faculty members and 80 white faculty members at the top-ranked 92 U.S. medical schools. They developed an algorithm for assessing a scientist's productivity that took into account such factors as number of publications, the scientists' role on each paper, the journals' impact factor, and citations.

Compared with white faculty members in similar positions at the same set of institutions, the black scientists were less productive, Wang's team found. They then looked a subset of 11 black and 11 white researchers with NIH funding. Compared to white scientists with the same productivity index, the black researchers had just as many, if not more, grants and research dollars. "Individual performance determines NIH award success, while the evidence for racial bias is scant," says Wang, who this week moved from Virginia Tech to Rensselaer Polytechnic Institute in Troy, New York.

Ginther has reservations about the study, published yesterday in the Journal of Informetrics. She notes that the study looked at a small number of researchers and didn't examine their chances of winning funding, just how much they ultimately got. She also questions Wang's team's decision to weight first and last authors equally, because the principal investigator is usually the last author.

Even if the findings are valid, they "complement" rather than contradict her work, Ginther says. As her team looks more closely at productivity, they, too, are finding that it explains some of the gap in blacks' and whites' success rates. But it's not the whole answer, she says, because "there's still a significant gap."

Ginther also points out that in their Science paper, her team said another explanation for the racial gap could be that black scientists submit weaker proposals. That might reflect the fact that blacks have less access to mentors than their white colleagues—a possible problem that NIH's plan will address. "I think they're [Wang's team] making a mountain out of a molehill when it comes to the bias question," Ginther says.

Follow News from Science

Latest News

A 3D plot from a model of the Ebola risk faced at different West African regions over time.
Dancing sneakers on pavement