The study, by authors at Harvard University, Yale University, and the University of Toronto, was conducted at Yale using functional magnetic resonance imaging (fMRI), and is reported in the December issue of the journal Psychological Science.

"Physical properties that make up a person cannot be disregarded in face-to-face interactions, and the imprint of culture is what's reflected in the response to a 30-millisecond subliminal exposure," says co-author Mahzarin R. Banaji, Richard Clarke Cabot Professor of Social Ethics in the Department of Psychology in Harvard's Faculty of Arts and Sciences and Carol K. Pforzheimer Professor at Harvard's Radcliffe Institute for Advanced Study. "However, seeing the face consciously, for as little as a half second, allows a more reasoned response to the face in view."

The researchers performed fMRIs on 13 white participants. During the scans, participants viewed a series of faces -– some of which could be consciously seen and some of which were presented so quickly that participants did not report seeing them. The researchers found that for the ultra-brief subliminal images, amygdala activity was greater in response to black faces than to white faces, suggesting that at least initially, black faces provoked a stronger emotional reaction than white faces.

"These results suggest that automatic and controlled evaluative processes may be more connected than previously thought. These processes may not be isolated or fully separate -– they may interact in more dynamic ways during evaluations of our social environment," says lead author William A. Cunningham, assistant professor of psychology at Toronto. "Even for non-prejudiced individuals, early perceptual processing may result in an automatic emotional response that may direct attention toward people of stigmatized social groups. Yet, with the opportunity to change or modify this initial impulse, they have the ability to do so."

The researchers found that the difference in amygdala response to black and white faces was greater among individuals displaying higher degrees of racial bias on a test that assesses the extent to which individuals are faster in a task associating black with bad and white with good compared to black with good and white with bad. However, all of the participants expressed disagreement with prejudiced statements and a personal interest in egalitarian behavior. Consistent with these conscious beliefs, in the fMRI phase of the study, when faces were viewed for longer periods of time, areas of the brain's frontal cortex that are involved in inhibition and control took over and amygdala activity displayed less bias.

Co-author Marcia K. Johnson, Charles C. and Dorathea S. Dilley Professor of Psychology at Yale, noted that the researchers also found that greater amygdala response to black than white faces was associated with less activity to black than white faces in the fusiform gyrus, a brain area associated with face processing. Thus, lack of "expertise" about other-race faces may trigger an early emotional response that can be modulated by more conscious processing.

The current work builds upon a tool for assessing bias developed in 1998 by Banaji and Anthony Greenwald of the University of Washington, which is now available online at http://implicit.harvard.edu. This self-administered test, which has been taken by close to 3 million individuals, provides an index of unconscious prejudice against various groups such as elderly, gays, dark-skinned people, the obese, and a wide range of other groups.

###

This work was supported by the National Institutes of Health, the National Institute of Mental Health, and the National Science Foundation.

Note: If no author is given, the source is cited instead.

• Psychology
• Social Psychology
• Racial Issues
• Intelligence
• Neuroscience
• Perception

• Familiarity increases liking
• Functional neuroimaging
• Social cognition
• Confirmation bias
• Milgram experiment
• Double blind