Interview with Claude Steele

By Coert Visser (2010)

Professor Claude Steele is a social psychologist and the Provost of Columbia University. He has written the book Whistling Vivaldi: And Other Clues to How Stereotypes Affect Us about the work he and his colleagues have done on a phenomenon called 'stereotype threat'. Stereotype threat is the tendency to expect, perceive, and be influenced by negative stereotypes about one’s social category, such as one’s age, sex, sexual orientation, ethnicity, profession, nationality, political affiliation, mental health status, and so on. Stereotype threat can be harmful by creating racial, gender, and social class achievements gaps in schools and in the workplace and tensions across group lines. In this interview Claude Steele explains, among other things, what stereotype threat is and what can be done about it.

How would you explain in simple terms to people like teachers, managers, and policymakers what stereotype threat is and why it is important for them to be informed about it?

First, an illustration; the story of New York Times editorialist Brent Staples. Some years ago, when he began graduate school at the University of Chicago, he noticed that as African American dressed in student informality, he was making white pedestrians a little nervous as he walked the streets of Chicago’s Hyde Park, especially after dark. People seemed to steer away from him; avoid eye contact; even cross the street sometimes to avoid him. He realized he was being seen stereotypically as a possibly threatening African American male. It worried and scared him. Will this happen every time he walks the streets of his neighborhood? One day he whistles Beatle tunes and Vivaldi. Suddenly his passersby relax; say “hello.” Some want to stop and talk. He realizes that whistling Vivaldi has preempted his being seen stereotypically as threatening figure. The relief he felt told him how pressured he had been, how upset he had been all along by the possibility of being seen through the lens of a negative stereotype about his racial identity.

That pressure is what we call “stereotype threat”: being in a situation or doing something for which a negative stereotype about one of your identities (your age, religion, sex, race, political orientation, etc.) could apply. When this happens you know you could be seen or treated in terms of that stereotype and if you care about what you are doing, the prospect of being seen and treated that way can be upsetting, upsetting enough to make you uncomfortable and interfere with your functioning right then and there.

Now imagine you are a member of a group whose intellectual abilities are negatively stereotyped in an area you care about, as for example, being a woman trying to succeed in a quantitative field. The stereotype threat you feel—the possibility of confirming or being seen to confirm the negative stereotype about your group’s abilities that is an ordinary accompaniment of your taking on challenges in this field—could be an all too constant pressure in your life. Sometimes the pressure might directly interfere with your performance. You might, as people often do, want to disprove the stereotype. But this extra effort makes you multi-task. Now you are trying to perform the task and trying to disprove the stereotype—all at the same time with all of the self-monitoring that goes with that. Such extra effort is tiring and, as our research shows, capable of depressing your performance. That’s how stereotype threat—something “in the air” and seemingly quite ephemeral—can have real effects in a person’s life, effecting important performances, and eventually even the walks of life they chose.

Together with various colleagues you have been doing research on how stereotypes may affect performance since the late 1980’s. How did you start to find out about the workings and strong impact of stereotype threat? Did you more or less expect to find what we know now or did you kind of stumble on it?

The word “stumble” is a little strong. But when we started we certainly had no clear idea of the phenomenon we now call stereotype threat. We were trying to understand the real-world problem of underperformance: that at every skill level, as measured by standardized tests, groups whose intellectual abilities were negatively stereotyped in the larger society (e.g., African Americans, women in math) got worse subsequent grades than non-stereotyped students with the same skills. Something was depressing their performance other than a lack of skill or knowledge. We didn’t know what it was. We guessed that it had something to do with the pressure of being stereotyped. But the process of getting clear about stereotype threat was a long one—several years anyway.

What helped us I think was being precise about what the problem was; the mysterious fact that ability-stereotyped students were getting lower subsequent grades than other students with the same test scores. This specification of the problem pointed to some kind of stereotype pressure. We began designing experiments to test this pressure. Pretty quickly we came up with stereotype threat experiments, only, not having the concept of stereotype threat at the time, we didn’t call them that. It took a while to develop a clear, testable conception or theory about what we had. But that was the process: first a real-world observation, then the effort to replicate that real-world effect in the laboratory (so we could study it closely); then developing an interpretation or theory of the effect; then explicitly testing that theory through conditions that would remove this effect to see if that would improve the performance of these groups. Finally, we thought we had a phenomenon. Then, over the years, we and many other researchers have tested its generality, the internal processes that mediate it and so on.

It seems like a lot has become clear about the mechanisms that underlie stereotype threat. Could you try to summarize what research has revealed about how and to what extent the impact of stereotype threat can be reduced? I am curious about what individuals, organizations and society might do to help reduce it.

I think of remedy as a matter of strategy and tactics; general principles of understanding derived from the research and specific tactics of implementation that, to date, have also been supported in research.

The general principle begins with a recognition: that in “diverse” settings, our identities—given their history in our society and the stereotypes about them that bring that history into the present—have the power to threaten each other, to make us feel uncomfortable and worry about how we will be treated and judged based on an identity we have. This is a recognition that rather than diverse settings being innocent of tension if people have good will, they can always have tension unless something is done to prevent or reduce it. It’s a sobering recognition, but I think a correct one. Thus the general principle of remedy is that something has to be done to allow people to trust the setting, to trust that they won’t be judged or treated badly in the situation based on an identity they have. The setting and the people in it have to convey a sense of “identity safety.” That’s the principle, the goal that derives from our research. Now the tactics.

A group of ingenious and now distinguished investigators have isolated a number of specific tactics for achieving identity safety. The story of this work is described in the book. Here I will mention a few examples.

Geoff Cohen, Lee Ross and I asked the question “how should a white professor give critical feedback to a black student in such a way that the feedback will be trusted and will motivate efforts to improve the work?” In an intriguing experiment that Geoff designed and implemented, he got a clear answer: say that you are using high standards in evaluating their work (a clear signal that you are not seeing them through the lens of an ability-demeaning stereotype) and that you have looked at their work and believe that they can meet those standards (another signal that you are not seeing them stereotypically). The high standards/affirmation of potential tactic has proven to be a powerful way of motivating and engaging students who would otherwise be repressed by stereotype and identity threat.

Critical mass is another effective tactic. A setting that is self-evidently inclusive is a powerful signal that one has identity safety in the setting. Since all of us “count” when we’re in settings where an identity we have puts us in the minority, seeing others there with that identity is a powerful identity safety cue.

The presence of identity role models in the setting is another cue that signals identity safety.

Reminding people in school and workplace settings that ability is an expandable resource reduces the impact of stereotype threats focused on ability and achievement. Derived from the work of Carol Dweck and Josh Aronson, this tactic has been shown to greatly diminish the performance-impairing effects of stereotype threat.

Geoff Cohen, Julio Garcia and Valerie Purdie-Vaughns have accomplished an amazing set of field studies showing that the performance-impairing effects of identity threat in K through 12 classrooms can be dramatically reduced by allowing students to briefly affirm their most important values. It seems to allow students a more trusting understanding of the classroom culture—something their identities would otherwise disrupt—and this greatly improves their engagement and grades—achieving a 40% reduction of the racial achievement gap.

Greg Walton and Geoff Cohen developed a similar strategy in a college-level intervention to lead students toward a more trusting interpretation of their college environment. It too dramatically improved minority student grades; reducing the racial achievement gap by 46% for the whole last 3 years of college.

The book goes into each of these in more detail and describes how to get the original reports of this work, several of which have now been published in Science—I am proud to say. But this should give the reader a sense that remedying these pressures is possible as well as some idea about how to go about designing their own remedies.

Imagine, in the next ten years or so, we would indeed become better at doing these things and at reducing the negative effects of stereotype threat. Could you try to describe what would be some of the important benefits you’d expect to individuals, organizations and society?

If this happened we would likely see a great reduction in racial, gender, and social class achievements gaps, especially in advanced areas of performance, both in schools and in the workplace.

I also think we would have a lot easier time relating to each other across group lines. A good deal of the tensions that can beset our intergroup relations come from the sense of identity threat we can feel in each other's presence. With this threat gone, these relations would be immensely easier and the true, rewarding nature of our diversity would come to light.

Every individual, organization, and the entire society would be able to function more effectively if these threats were diminished. And this would allow us all to better draw the benefits that this diverse society has to offer.

That sounds like an attractive perspective. Now, I have a question about psychology. In the book, you say psychologists tend to focus on the internal, the psychological, when looking for plausible explanations for behavior and deemphasize, as causes of behavior, the things we can't see very well, namely, the circumstances to which we are adapting. Could you share some thoughts about what psychology might look like if – to some extent - we´d get rid of this psychologist´s bias to underestimate the importance situational behavior determinants?

When I use that term I am referring to the idea that we psychologists, like we laymen, tend to make the “fundamental attributional error” in explaining behavior, we tend to see human behavior as stemming from internal, psychological processes, rather than from external, situational conditions and pressures. It’s an honest bias. After all, our mission and focus is to understand internal functioning. So it’s only natural to privilege this perspective in our thinking. But some of the time this “bias” causes us to miss the real causes of behavior—especially when the real causes are indeed situational, or social, or cultural, or normative or structural pressures rather than internal processes; when the internal is only mediator of the external.

This lesson was made clear to us in stereotype threat research when we tried to answer the question of what determined the strength of stereotype threat, that is, the strength of its impact on performance for example. Was it something internal to the person, something like their level of self-esteem, the strength of their skills, their performance expectations etc.? Or was it something external to the person, something like the number and strength of the cues in the immediate situation suggesting the likelihood of being negatively stereotyped there. It turned out to be the latter, and we were initially surprised—thus revealing to ourselves how the perspective we took as psychologists biased us to look for internal causes of behavior. The experience stands as a cautionary tale for us: beware the psychologist’s error, one can fall victim to it without even realizing it.

Here is a last question: how do you see the research into stereotype threat develop in the coming years? And could you say something about your own plans?

Hopefully researchers will continue to explore its role in a broad variety of phenomena. For example, Priyanka Carr and I just published several experiments showing its role as a cause of risk aversion among women making economic decisions. I am heartened to see the utility of the concept in explaining such a variety of phenomena.

For my own part these days, I am most fascinated by the growing body of research showing how small interventions designed to reduce the experience of stereotype threat in classrooms and schools can dramatically improve the academic performance of groups whose abilities are negatively stereotyped—minority students in K through 12 schooling and college, and women in advanced math and science courses. To see these interventions-- by people such as Geoff Cohen, Josh Aronson, Steve Spencer, Greg Walton, Valerie Purdie-Vaughns, Julio Garcia, Carol Dweck, and others--achieve improvements against such tenacious patterns of underperformance is inspiring. I am thrilled to see it, even though these days, from my current position as a university Provost, I am only able to participate primarily as a cheerleader, commenter and occasionally as a publicist—see chapter 9 of Whistling Vivaldi.

From the standpoint of research, my great blessing in the stereotype threat area (and the other areas of research that I have been involved in) has been the quality of my students and colleagues. One of the joys of writing this book was the opportunity to tell the story behind the research, the story of how it actually unfolded and developed. From that vantage point, what becomes vividly clear is the role of collaboration and the contribution of so many people to the development and application of this idea. When I stick my head up and take an overview look at the enterprise of stereotype threat research, that’s what I am proudest of.


Popular publications
Academic publications