Public Release: 

Infants show racial bias toward members of own race and against those of other races

Racial bias begins earlier than previously thought, new insights into cause

University of Toronto

TORONTO, ON - Two studies by researchers at the Ontario Institute for Studies in Education (OISE) at the University of Toronto and their collaborators from the US, UK, France and China, show that six- to nine-month-old infants demonstrate racial bias in favour of members of their own race and racial bias against those of other races.

In the first study, "Older but not younger infants associate own-race faces with happy music and other-race faces with sad music", published in Developmental Science, results showed that after six months of age, infants begin to associate own-race faces with happy music and other-race faces with sad music.

In the second study, "Infants rely more on gaze cues from own-race than other-race adults for learning under uncertainty", published in Child Development, researchers found that six- to eight-month-old infants were more inclined to learn information from an adult of his or her own race than from an adult of a different race.

(In both studies, infants less than six months of age were not found to show such biases).

Racial bias begins at younger age, without experience with other-race individuals

"The findings of these studies are significant for many reasons," said Dr. Kang Lee, professor at OISE's Jackman Institute of Child Study, a Tier 1 Canada Research Chair and lead author of the studies. "The results show that race-based bias already exists around the second half of a child's first year. This challenges the popular view that race-based bias first emerges only during the preschool years." Hear Dr. Lee discuss the research results.

Researchers say these findings are also important because they offer a new perspective on the cause of race-based bias.

"When we consider why someone has a racial bias, we often think of negative experiences he or she may have had with other-race individuals. But, these findings suggest that a race-based bias emerges without experience with other-race individuals," said Dr. Naiqi (Gabriel) Xiao, first author of the two papers and postdoctoral fellow at Princeton University.

This can be inferred because prior studies from other labs have indicated that many infants typically experience over 90 per cent own-race faces. Following this pattern, the current studies involved babies who had little to no prior experience with other-race individuals.

"These findings thus point to the possibility that aspects of racial bias later in life may arise from our lack of exposure to other-race individuals in infancy," Dr. Lee said.

Study results could be significant in prevention of racial bias

He continued to explain that overall, the results of these studies are critically important given the issues of wide-spread racial bias and racism around the world.

"If we can pinpoint the starting point of racial bias, which we may have done here, we can start to find ways to prevent racial biases from happening," he said.

"An important finding is that infants will learn from people they are most exposed to," added Dr. Xiao, indicating that parents can help prevent racial bias by, for example, introducing their children to people from a variety of races.

First study: Face-race and music

In the first study, infants from 3 to 10 months of age watched a sequence of videos depicting female adults with a neutral facial expression. Before viewing each face, infants heard a music clip. Babies participated in one of the four music-face combinations: happy music followed by own-race faces, sad music followed by own-race faces, happy music followed by other-race faces, and sad music followed by other-race faces. The study found that infants at six to nine months of age looked longer at own-race faces when paired with happy music as opposed to with sad music. By contrast, six- to nine-month-olds looked longer at other-race faces when paired with sad music compared to with happy music.

Second study: Face-race and learning

The second study examined whether infants were biased to learn from own-race adults versus other-race adults. Six to eight-month-old infants saw a series of videos. In each video, a female adult looked at any one of the four corners of the screen. Following the look, in some videos, an animal image appeared in the looked-at location (a reliable gaze). In other videos, an animal image appeared at a non-looked-at location (an unreliable gaze). The results showed that six to eight-month-old infants followed the gaze of members of their own race more than they followed the gaze of other-race individuals. This occurred when the faces were slightly unreliable, as they are in the natural environment. This result suggests that, under uncertainty, infants are biased to learn information from own-race adults as opposed to other-race adults.

Racial bias can 'permeate almost all of our social interactions' Dr. Lee said it's important to be mindful of the impact that racial bias has on our everyday lives, stressing that not only is explicit bias a concern, but so too are implicit forms.

"Implicit racial biases tend to be subconscious, pernicious, and insidious. It permeates almost all of our social interactions, from health care to commerce, employment, politics, and dating. Because of that, it's very important to study where these kinds of biases come from and use that information to try and prevent racial biases from developing," he said.

###

MEDIA CONTACTS:

Dr. Kang Lee
Professor, Dr. Eric Jackman Institute for Child Study, OISE/University of Toronto
Email: kang.lee@utoronto.ca (Email best way to reach Dr. Lee to set up interview)

Naiqi (Gabriel) Xiao
Postdoctoral Researcher, Princeton University
Email: naiqi.xiao@princeton.edu
Phone: +1-609-608-6248

Media Relations Coordinator

Lindsey Craig
Communications & Media Relations Coordinator, OISE/University of Toronto
Email: lindsey.craig@utoronto.ca
Phone: 416-978-1127

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.