In recent webinars titled “Blindspot: Hidden Biases of Good People,” more than 1,300 Harvard faculty and others tuned in to hear from Mahzarin Banaji, PhD, professor of social ethics in Harvard’s Psychology Department. Banaji described her work in understanding implicit, or unconscious, bias—why everyone has it, and what participants can do to recognize it in themselves. The seminars, sponsored by Harvard’s Office of the Senior Vice Provost for Faculty Development and Diversity, were open to all Harvard faculty, including MGHers with Harvard Medical School appointments. 

Mahzarin Banaji, PhD
Mahzarin Banaji, PhD

Banaji is a co-developer of the internationally renowned implicit association test, in which test takers must rapidly associate certain words or images with “good” or “bad” words. Whereas “Red Sox good” and “Yankees bad” pairings might be innocuous, said Banaji, many others throughout her decades of research have been sobering and surprising. When shown faces of white people and of a minority group, viewers routinely show a preference for the white faces—sometimes even viewers in that minority group. In a study of “American-ness,” subjects were shown pictures of well-known white Americans, Asian Americans and white Europeans—and found that white Europeans were perceived as more American than Asian-Americans, revealing an “American=white” bias. 

Implicit bias is no less a problem in health care, said Banaji. Black patients routinely receive less pain medication than do whites, but not because their physicians are overtly racist, she says. Her proposed solution? Intervene in the moment of decision-making. As a prescription for pain medication is being entered in the patient’s record on a computer, a graph highlighting the actual data on race disparities in that health care system is made available, with no specific instruction other than awareness. “Make data available that will assist good people to act in accord with their values,” she said.

Our unconscious associations are a result of how our brains work, said Banaji; as we absorb social norms around us, those associations become ingrained: “Neurons that fire together wire together.”

But there’s good news. Banaji cited 13th century poet Rumi: “Every single one of us is a jackass but with wings of angels tacked onto us.” Humans have the capacity to grow and change, she noted; “conscious awareness allows us to analyze data...and change our minds.”

Can frequently taking implicit association tests create a feedback loop? To an extent, said Banaji, “but I need everyday reminders to move me in a direction that’s healthy and good even if my brain snaps back to the cultural default. Social and individual change are intertwined.”

And Banaji has been heartened to see the results of social change revealed in her research. Between 2006 and 2015, there has a demonstrable decrease in implicit biases toward gay people. “We’re changing the world and it’s changing us back,” said Banaji.

To take an implicit association test, visit To learn more about implicit bias, visit