Winter 2006Research & Discovery

Snap Judgments

Psychological test reveals hidden bias

TOPICS:

On this test, you can’t really cheat or lie. In fact, it’s impossible to even know how you’re answering the questions. With just the stroke of a computer keypad, scientists are now taking an unusual approach to measure your “implicit” biases.

It’s called the Implicit Association Test, and its aim is to explore the unconscious roots of thinking and feeling. One of its principal investigators is Brian Nosek, a U.Va. assistant professor of psychology, who developed the test with two other colleagues at Harvard and the University of Washington.

The design of this Web-based test is brilliantly simple. You are asked to pair words with images, and the speed at which you sort and categorize is interpreted as an indication of how strongly those concepts are associated in your mind—in other words, your hidden prejudice.

How do you compare?

  • 85% of the test participants are from the U.S.; 15% are international
  • 80% of Web respondents show implicit negativity toward the elderly compared to the young
  • Approximately 75% of whites and Asians show an implicit preference for white people over black people
  • About 80% of heterosexuals show implicit biases for straight people over gay people
  • Minorities internalize the same biases as majority groups. About 48% of blacks show a pro-white implicit bias; 36% of Arab Muslims show an anti-Muslim bias; and 38% of gay people show a bias for straight people over gay people

Source: Project Implicit

So far, more than 3 million individuals have assessed their unconscious preferences in more than 90 subject areas. These run the spectrum, from political issues, ethnic groups, sports teams, age and gender to stereotypes of race and crime.

Subjects view a series of images on a computer screen and are asked to match positive words and negative words with the images. In the race test, for example, words like “wonderful” and “peace” or “evil” and “horrible” must be matched with faces of African Americans or whites. The delay in response time—measured in milliseconds—signifies internal conflict. And, with such a simple categorization, a hesitation measured in milliseconds can reflect a big influence on thought processes, says Nosek.

The focus of Project Implicit is public education and research. Since it was launched online in 1998, the response to the Implicit Association Test has been overwhelming. Nosek and his fellow investigators are sifting through a mountain of data from millions of tests. They now boast the largest database on implicit attitudes ever collected. Educators use the test as a vehicle for discussion; it’s also been used by businesses for diversity training and by law enforcement. The test has been translated into more than a dozen languages, most recently into Polish and Turkish.

Its enormous popularity may be due in part to the fact that it gives the test taker immediate results. The outcomes, however, can be unsettling. Many people believe they are not prejudiced and deny the test results when they indicate otherwise. This introduces important questions about whether evidence of an implicit bias is meaningful or important. Nosek notes that the scientific definition of bias simply means having a preference or inclination for something. Whether that preference is “morally justified,” he says, is another matter. While you might be morally justified to feel a stronger preference for a family member over a stranger, for instance, in American culture, it’s a less tenable position if your preference is, say, for a particular ethnicity or age group.

“The first time I did the race test, I was stunned,” Nosek says. As someone who espouses egalitarianism, Nosek was taken aback to learn from the results that he has a bias against black people. “My explicit beliefs aren’t the sole director of my everyday behavior,” he says. “That’s the humbling part—those associations have an influence without me knowing about it.”

While data shows that implicit biases are pervasive, they don’t necessarily mean that people always behave that way—the tests measure thoughts and feelings, not actions. And the latest research indicates that these biases are not intractable. Nosek says that people may be able to override their bias through sometimes subtle interventions if they know that it’s present, and have an effective means of adjusting it. The screensaver on Nosek’s computer in his Gilmer Hall office depicts photos of prominent blacks; by exposing himself to counterstereotypes, Nosek is attempting to reinforce positive associations to counter the negative ones he accumulated through everyday cultural experience.

People shouldn’t come away feeling pessimistic if they get negative feedback from the test, says Nosek. “I think the message of this work is optimistic, because we now have the potential for awareness of something that we cannot just look inside ourselves to see,” he explains. “With knowledge, I can be more deliberate in directing my behavior to conform to my conscious intentions—despite my automatic reactions.”

“The whole point of consciousness,” he adds, “is to not be slaves to our unconscious.”

For More Info

    SHARE
    • E-mail
    • Digg
    • Del.icio.us
    • LinkedIn
    • Facebook
    • StumbleUpon
    • Reddit
    • PrintFriendly

    Comments

      There are no comments for this article yet. Begin the discussion below!

    Leave a Comment

    U.Va. Magazine welcomes your respectful discussion. Comments are subject to editorial moderation. Review our user guidelines for more information »




    Please enter the word you see in the image below:

    HIGHLIGHTS