As much as social equality is trumpeted in the United States, a new University of Virginia study suggests that besides evaluating their own race and religion most favorably, many people share implicit hierarchies for racial, religious and age groups that may be different from their conscious, explicit attitudes and values.

The study’s findings appear in the journal Psychological Science, a journal of the Association for Psychological Science.

“People from relatively low-status groups can readily report that their group does not have the most power,” explained U.Va. psychology Ph.D. student Jordan R. Axt, who led the study. “At the same time, most groups, even if they have less social power, favor their own group above all others.

“We wanted to investigate how these dual influences – the knowledge that one’s group may not have the most power, but nevertheless favoring that group the most – would reveal themselves on measures of both explicit and implicit attitudes.”

Axt and colleagues analyzed data from hundreds of thousands of American participants who completed online Brief Implicit Association Tests on race, religion and age.

In the first task, participants viewed a male or female face of a particular racial group as well as positive words such as “love,” “pleasant,” “great” and “wonderful,” and negative words such as “hate,” “unpleasant,” “awful” and “terrible.” For each set, participants categorized the positive and negative words with faces belonging to each racial group.

The idea behind the tests is that people are quicker to categorize things with the same response when they are associated more closely in memory, even if they consciously reject that association. If a person has positive associations with a particular racial group, for example, it should take less time to categorize faces from that group together with positive words. A person with negative associations, on the other hand, would need more time to categorize faces from that group together with positive words. Thus, by examining variations in how long it takes to link the words and images, the tests can uncover biases people may not be conscious of and do not endorse.

Axt and his U.Va. colleagues found that participants were most likely to prefer members of their own race. Additionally, members of almost every racial group exhibited an implicit racial hierarchy of positive evaluations: White, then Asian, then black, then Hispanic.

Likewise, people favored their own religion.

After their own group, participants’ implicit hierarchies usually placed Christianity next, followed by Judaism, Hinduism or Buddhism (there were two versions of the test, with either Hinduism or Buddhism as an option), and Islam.

Unlike race and religion, however, people did not show a preference for members of their own age group. Still, every age group demonstrated an implicit age-based hierarchy with children at the top, followed by young adults, middle-aged adults, and, finally, older adults.

Importantly, participants’ implicit associations differed from evaluations they made when asked to report what they consciously thought of various racial, religious and age groups.

“Our explicit, conscious attitudes may be derived more from personal beliefs about others. At the same time, implicit attitudes may arise both from our own identities as well as from widely spread cultural beliefs or values,” Axt said. “While we may disagree with such cultural beliefs, these results illustrate how they can nevertheless shape our minds.”

According to Axt, the findings contribute to the debate over whether people prefer their own groups, or if those on a lower social rung actually esteem high-status groups as a justification for the way things are.

“Like many scientific debates, our results suggest that the answer is both,” he said.

Co-authors on the study include U.Va. psychology graduate student Charles R. Ebersole and psychology professor Brian A. Nosek.