The Risks of Bias Testing
Many tests may make unconscious biases visible—which is why using them can be risky for employers.
When hiring and promoting people, HR professionals and managers know that certain factors—such as gender, race, disability status and age—cannot be considered in the decision-making process under the law. What they may not understand, however, is that, without any conscious awareness, they actually may be considering these precise factors. For example, a white male manager may know that gender and race lawfully cannot be considered in a hiring situation. But he, as a white man, may favor a white male candidate over a woman of color based on how “comfortable” he is with each of them, even though he has no idea that his comfort level may relate to race and/or gender. Unconscious bias, often referred to as implicit bias, is bias that we are unaware of. It happens automatically and without any conscious thought process and is triggered by our brain making snap judgments formed, at least in part, as a result of the messages that we received growing up, as well as our own experiences, culture, mass media and other influences. But how can we address a form of bias that we are unaware we have? The first step is to acknowledge that bias exists and that no one is immune from it. Good people can—and do—make biased decisions. That’s one reason a training approach with a punitive tone won’t work well and, in fact, is likely to be counterproductive. Implicit bias is both personal, in that the various stereotypes that employees have internalized can vary based on their experiences, and ubiquitous, in that we all harbor unconscious assumptions; it’s part of being human.That doesn’t mean we should sugarcoat the issue—but it’s important to understand the dynamics driving unconscious bias and to think about how to best raise the topic so that individuals will be motivated to learn more about it and how it influences their decision-making. Raising awareness through bias testing, however, is a risky approach.Understand the Evidence You can start by familiarizing yourself with the scientific evidence demonstrating the existence of unconscious bias. In a 2002 study by the University of Chicago Booth School of Business, “Are Emily and Brendan More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination,” researchers sent recruiters virtually identical resumes that differed in only one way: the applicants’ names.They found that individuals who submitted resumes with “white-sounding” names received 50 percent more callbacks than those whose monikers were more likely to be associated with black applicants. In other words, all else being equal, Karen will fare far better in her job search than Keisha will. Another classic study, 1997’s “Orchestrating Impartiality: The Impact of ‘Blind’ Auditions on Female Musicians” by the National Bureau of Economic Research, involves an orchestra that wanted to increase gender diversity but was having trouble finding women with the requisite talent.After the orchestra had no luck with the traditional audition process, individuals were asked to perform behind a curtain. Magic! As a result of this step alone, women were 50 percent more likely to proceed to the final round of auditions.Stereotypes that are perpetuated and reinforced by messages we receive through the media, societal cues or our individual experiences can become embedded in the amygdala region of the brain.In both cases, conscious and unconscious bias were likely at work. But as the latter example—in which the orchestra stated a genuine desire to increase gender diversity—shows, it is often people’s implicit prejudices that steer them toward partial outcomes. Know the Basis for the BiasWhat is the basis for implicit bias? Part of the answer may lie in neurology and social psychology.Stereotypes that are perpetuated and reinforced by messages we receive through the media, societal cues or our individual experiences can become embedded in the amygdala region of the brain, which is activated when we make snap judgments based on pictures, images or names that can correlate with a characteristic such as race, gender or age. As a result, we make unconscious associations, which can be ither positive or negative, based on these factors. Understand the Risks Of TestingThere are a number of tests that individuals can take to become more conscious of their unconscious biases. One is the Harvard Implicit Awareness Test (IAT), which measures implicit bias in 14 areas, including gender, race, religion, age and sexual orientation.The test shows images and words associated with a color, gender, religion, etc. Generally, the slower the test-taker is in pairing certain words or images with a specific group, the more likely he or she has an implicit bias, according to the assessment.A report published in the European Review of Social Psychology in 2007, “Pervasiveness and Correlates of Implicit Attitudes and Stereotypes,” found that:68 percent of respondents had a more favorable automatic association with people who are white than they did with people of color.76 percent made a greater connection between the words “men” and “careers” and “women” and “families” than they did with the converse concepts (“men” and “families,” and “women” and “careers”).But while the IAT and other such tests may be instructive for individuals, these assessments are far from conclusive. For example, there is considerable debate among psychologists about their reliability. (A discussion on this topic appears in “IAT: Fad or Fabulous?” in the July/August 2008 issue of the American Psychological Association’s Monitor on Psychology.) Moreover, the tests may create legal risk because the results are discoverable in litigation.Consider the following example: Greg chose Jim over Jane for a promotion. Soon afterward, Greg is deposed in a claim by Jane alleging sex discrimination. The plaintiff’s attorney initiates the following line of questioning: “You took an implicit awareness test?” the attorney asks.“Yes,” Greg replies. “Did you have any implicit bias in terms of gender?” the lawyer continues.“I was surprised to learn that I favor men over women,” Greg says. “And you picked Jim over Jane. Is that correct?” At this point, the plaintiff’s attorney is thinking about using the money her firm wins in this case to purchase a summer home. “Yes, that is correct, too,” Greg says.“But you admit both were qualified?” she asks.“Yes, but …” Greg trails off. Now Greg’s company would need to bring in an expert to discredit the test that was used, and even then the jury could find the test to be more valid than some psychologists may find it to be. So think carefully before reflexively using implicit awareness tests. At a very minimum, consider their potential usefulness as well as the legal perils that they may create. Fortunately, there are less-risky steps employers can take that are helpful in identifying and addressing unconscious bias. I will cover them in next month’s issue. Jonathan A. Segal is a partner at Duane Morris in Philadelphia and New York City. Follow him on Twitter @Jonathan_HR_Law.Was this article useful? SHRM offers thousands of tools, templates and other exclusive member benefits, including compliance updates, sample policies, HR expert advice, education discounts, a growing online member community and much more. Join/Renew Now and let SHRM help you work smarter.
An organization run by AI is not a futuristic concept. Such technology is already a part of many workplaces and will continue to shape the labor market and HR. Here's how employers and employees can successfully manage generative AI and other AI-powered systems.