PSY235:Social Psychology

Lecture 4:Group Perception/Concepts in Action

How stereotypes influence the interpretation of information

BehaviorMale employeeFemale employee
The family picture is on her/his desk.Ah, a solid, responsible family man.Hmm, her family will come before her career.
His/her desk is cluttered.He's obviously a hard worker and a busy man.She's obviously a disorganized scatterbrain.
She's/He's talking with coworkers.He must be discussing the latest deal.She must be gossiping.
He/She is not at his/her desk.He must be at a meeting.She must be in the ladies room.
She/He is not in the office todayHe's meeting customers.She must have a family emergency.
He/she is having lunch with the boss.He's on his way up.They must be having an affair.
The boss criticized him/her.He'll improve his performance.She'll be very upset.
She/He got an unfair deal.Did he get angry?Did she cry?
He's/She's getting married.He'll get more settled.She'll get pregnant and leave.
She's/He's going on a business trip.It's good for his career.What does her husband say?
He's/She's leaving for a better job.He recognizes a good opportunity.Women are undependable.

SOURCE: Blau, F. D., & Ferber, M. A. (1986). The economics of women, men, and work. Englewood Cliffs, NJ: Prentice-Hall.

How would you respond?


SOURCE:THE NEW REPUBLIC, 1986, pp. 18-22.

A bitter controversy has broken out in Washington over a column by Richard Cohen in the Washington Post Magazine. Cohen reports that certain jewelry stores in Washington admit customers only through a buzzer system, and that some store owners use this system to exclude young black males on the grounds that these people are most likely to commit a robbery. Cohen defends this practice. He points out that "young black males commit an inordinate amount of urban crime, "that black potential victims as well as white ones often act on this awareness, and that under certain circumstances, "the mere recognition of race as a factor . . . is not in itself racism."

If you were a jewelry store owner, would you use your buzzer system to exclude young black males? You might take other factors into account - not just age and sex, but style of dress and time of day - but would you ever take race into consideration as well? Whatever you yourself would do, is taking race into consideration in these circumstances racist? Understandable? Both? Neither?

Suppose you were a cabdriver cruising for customers in the middle of the night. Or a woman about to get into an elevator with a stranger in a residential apartment building. How does your decision or your analysis change in these circumstances?

We asked some people to reflect briefly on these questions. Here are their responses


Men are not gods. Therefore, men face challenges gods would not have to endure - ignorance and uncertainty. To make decisions, we need to have information about the world around us. The information we gather is not only imperfect, it is costly as well. So we learn to economize by guessing, prejudging, and using stereotypes.

Imagine you are challenged to a basketball game and must select five out of 20 people who appear to be equal in every respect except race and sex. There are 5 black and 5 white females, 5 black and 5 white males. You have no information about their basketball proficiency. There is a million-dollar prize for the contest. How would you choose a team? If you thought basketball skills were randomly distributed by race and sex, you would randomly select. Most people would perceive a strong associative relationship between basketball skills on the one hand, and race and sex on the other. Most would confine their choice to males, and their choice would be dominated by black males.

Can we say such a person is a sexist/racist? An alternative answer is that he is behaving like an intelligent Bayesian (Sir Thomas Bayes, the father of statistics). Inexpensively obtained information about race and sex is a proxy for information that costs more to obtain, namely, basketball proficiency.

There is a large class of human behavior that generally falls into the same testing procedure. Doctors can predict the probability of hypertension by knowing race, and osteoporosis by knowing sex. A white jeweler who does not open his door to young black males cannot be labeled a racist any more than a black taxi driver who refuses to pick up a young black male at night. Black females and white females and white males commit holdups, but in this world of imperfect information cabdrivers and jewelers play the odds. To ask them to behave differently is to disarm them. The other side of the argument is that they are losing money if they err - i.e., turning away an honest customer.

People can be stupid Bayesians. In 1971 I owned a house in Washington's exclusive Chevy Chase suburb. My Saturday chore was to pick up motorists' litter along the side of my house. One time I was approached by a gentleman who stopped his car and said "Sir, when you're finished working at this house, could you come over to my place and work for me?" I politely told the gentleman I would not have time, since the rest of the day would be spent working on my Ph.D. dissertation. The gentleman apologized profusely. (My wife once encountered another incompetent Bayesian. Awaiting a bus during a downpour, she was offered a ride to Chevy Chase by a black woman who turned out to be a domestic servant. "Don't you just hate working for these white people out here?" her benefactor observed.) Mistaken identity is always a bit unpleasant, but it is not the same as racism.

Walter E. Williams is John M. Olin Distinguished Professor of Economics and George Washington University.


In my opinion, crime is the problem. It does not matter if you own a jewelry store, drive a cab, or if a stranger holds an elevator door for you. Crime does not have a race associated with it, yet there are those who would attach race to perpetrators of crime.

If the people in the three hypothetical cases are so threatened or intimidated by blacks in a community where blacks make up more than half the population, I would strongly suggest that they sell out, move to an all-white community, or don't drive a cab. If that solution is not feasible, the next best thing is to inquire at their local police department about involvement in community crime prevention programs. This is not a panacea. The police are not perfect or without prejudices either.

What would the jewelry store owner do if his business were located in the middle of America in an all-white community with a crime problem? Would he lock the doors and keep all white males out of his store? No. Nor would the woman stop using elevators. They would not attach race to the crime or to the person who committed it. Yet, as crime has rapidly increased over the past few decades, more and more people seem to believe that only blacks are perpetrators, and that it is an unalterable fact of modern life.

Such perceptions and attitudes prevent us from dealing with the real problem crime poses in America. The answer is not to lock your doors to blacks or stop riding the elevator. It is instead to begin to understand and work within our communities to help reduce the level of crime at work, home, and play.

Ronald Hampton is a D.C. police officer.


A jeweler who bars young black males from his store may or may not be a racist, but he is surely a grand simplifier, a naif in flight from the complexity of urban American.

Anyone who endures the daily alarums of city life will recognize the need to impose order on its scruffy disarray. Confronted every hour with fresh evidence or our own vulnerability, we all erect grids and charts, calculate risks and probabilities, promulgate axioms and formulas to reassure ourselves that we are in control.

One friend in New York refuses to ride the West Side subway but still takes its East Side counterpart. Another habitually crosses the street to avoid the teenagers who loiter outside a certain variety store. A third has switched his morning run from Riverside Park to Central Park. Are these rational responses to perceived danger or totemic rituals designed to ward off evil spirits? A bit of both, I imagine.

But how many of the perils that prowl our streets can really be addressed by such measures? What precautions, for example, could Dan Rather have taken a few weeks ago against two well-dressed white men who accosted him near his home on upper Park Avenue and beat him repeatedly while shouting over and over, "Kenneth, what is the frequency?".

City life cannot be forced into neat little boxes and categories. Thus, though it is undoubtedly true that the banishment of young black males from Washington jewelry stores would reduce the number of robberies in those stores, so would the banishment of young white males; of young Hispanic males; of American Indians off the bus from New Mexico; of Appalachian drifters hitching in from West Virginia; of unemployed Greek dishwashers; of suburban kids out for a thrill; of senior citizens down on their luck.

So if I owned a jewelry store in Washington today, I'm sure I'd be as worried about crime as those whom Richard Cohen interviewed. I'd probably be more than a little nervous when some young black men entered my store, but I'd be just as concerned about some young white men. Finally, I hope I'd have the good sense to eschew a pat nostrum that would only give me a bogus sense of security, and insist instead on confronting my environment in all its disheveled ambiguity.

J. Anthony Lukas

J. Anthony Lukas is the author of Common Ground, a study of three families - black, Irish, and Jewish - during the Boston busing crisis.


I am a retired jeweler in Washington, D.C., and, yes, I had a buzzer system in my store and the same fears as Richard Cohen's jeweler. I experienced several "snatch and grabs" prior to the installation of my buzzer, and they were committed by, yes, young black males age 18 to 25. Like everyone else who contemplates a buzzer, my two fears were: Am I a racist? Or will I be a dead merchant as a result of a robbery that gets out of control? Which fear do you choose to live with? If you're branded a racist, the worst that can happen is that you're out of business in one way or another but you continue to live and breathe. I chose the buzzer.

Rhonda Schoem

Rhonda Schoem wrote this contribution in the voice of her father. Ms. Schoem, former TNR controller, and her husband, a former D.C. policeman, recently took over her father's jewelry store.


Who would object to the refusal of a storekeeper, particularly of a store so obviously a target of crime as a jewelry store, to admit a stranger who looked "suspicious?" What is the purpose of a locked door and a buzzer system unless it is to keep out strangers who might wish to do harm? And since there are by definition strangers, how does anyone expect to distinguish between the innocuous and malevolent strangers except by finding the latter "suspicious?"

I would find anyone suspicious who appeared different in certain ways from most of those who come into my shop. Teenagers, unless they are conventionally dressed, would be suspicious. People whose carriage and movements gave the impression of nervous tension would be suspicious. People whose voice on the intercom sounded tense, artificial, unclear would be suspicious. My decision to admit would be based on an intuitive calculation of the number of categories of suspicion and the intensity of my feeling, balanced by the knowledge that I must take adverse chances or encounter the greater risk of losing my store. And, of course, I add two decisive categories of suspicion, assuming one or more of the categories previously mentioned were positively present: sex and race.

To pretend that a roughly dressed, nervous, slurring, tense young black male has a democratic right to enter a jewelry store is to distort a rational process with an irrelevant abstraction. Does it salve my conscience to say that I exclude him because he's suspicious, not because I ascribe an innate criminal potentiality to males with black skin? Perhaps. My parents used to whisper the phrase "colored people" lest our Negro maid be embarrassed."

I too feel virtuously unprejudiced, and I have convinced myself that only so long as young black males commit an inordinate number of robberies will I continue to find decisive some suspicion-raising attributes I would overlook others.

Roger Starr

Roger Starr is an editorial writer for the New York Times


Neither black nor white store owners are in business to display the virtues of admitting people of all colors, creeds, and fashions to their stores. They are in business to make money. I would want to take precautions to prevent robbery; I would look closely at people entering the store. The race of a potential customer would be one factor among many to be considered as I girded myself against thieves.

But in Washington and almost all other major cities, blacks do patronize jewelry stores. A jeweler in Beverly Hills who closed his door to heavily bejeweled Mr. T would be foolishly closing his cash register. Unless I am a racist, race and age cannot be the sole deciding factors in calculating whom I will and will not let into my store. And I certainly would not close my door to, say, all young black men - not even to those who are casually dressed and behaving nervously. I would act cautiously in dealing with them, as I would with an antic, strangely dressed white man.

As a cabdriver I would apply the same considerations. Discrimination can be used judiciously. I would certainly exclude one class of people: those who struck me as dangerous. Nervous-looking people with bulges under their jackets would not be picked up; nor would those who looked obviously drunk or stoned. It all comes down to a subjective judgment of what dangerous people look like. This does not necessarily entail a racial judgment. Cabdrivers who don't pick up young black men as a rule are making a poorly informed decision. Racism is a lazy man's substitute for using good judgment.

The elevator question is disingenuous. I suspect you are suggesting that i am a white woman getting into an apartment building elevator with a strange black man. Of course, black women have just as much to fear as white women. Nevertheless, black women living in black neighborhoods ride elevators with black men frequently, and do so without being raped. In this situation and all others, common sense in my constant guard. Common sense becomes racism when skin color becomes a formula for figuring out who is a danger to me.

Juan Williams

Juan Williams is a reporter for the Washington Post


What is the point of posing and answering these questions? A direct and simple answer obscures more than it clarifies. If I say that I would refuse admission to young blacks, how do I deal with the justified anger of most of them, who press my buzzer hoping to buy a ring or a watch? If I say that I would routinely buzz back, how do I deal with the real fear that shopkeepers feel (which is hardly self-generated)? I would probably feel it too, if I were a shopkeeper; the stereotypes upon which it feeds are social, not personal creations.

Yes, I would be afraid, and since I would not want to keep a gun under the cash register, or hire an armed guard, I might well install a buzzer system. I thing that I would be governed more by age than by race in deciding whether to buzz back or not - young punks, rather than young black punks, seems the appropriate category for refusal. But I can't say with any assurance that I would withstand the assumptions of the larger culture.

But these aren't the right questions for The New Republic to be asking. The responses that individual make when they are alone and frightened aren't made with sufficient freedom. Moral judgments in such cases pass easily into cheap moralizing. Or we respond in ideological terms, fastening individuals to political platforms. The point is to change the situation of shopkeepers and of the people they fear. In a democratic society, the marketplace, like all other public places, ought to be open to all comers. We should not have to live with buzzer systems; they violate the principle of civility. It's not very important to know what I would or wouldn't do given the violation. We need to think instead about what ought to be done to vindicate the principle.

Michael Walzer

Michael Walzer teaches moral philosophy at the Institute for Advanced Studies in Princeton.


While I am putting myself in the shoes of the jewelry store owner, I will also put myself in the skin of a black person. This makes the dilemma even more interesting.

In downtown Washington, where my jewelry store is located, I know that it is possible, but not likely, that a white man is going to rob me. To the extent that I worry about robbers, I mostly worry about blacks. Thus one effective method of keeping robbers out of my store will be to scrutinize blacks more carefully.

I don't much like to do this because I know how easily such occasional and sensible "discrimination" can turn into widespread and simple bigotry. For example, I read in the Washington Post recently that when I go out to look for a house or apartment in the Washington area, I have a 50 percent chance of being steered away from white neighborhoods or not being shown the nicest apartment or not being told about all the amenities. Now, most of these real estate agents in Washington are white and most of them very courteous. You just never know when one of them may rip you off.

As a businessman, I like to be consistent in my dealings with the public. If I don't let a black guy in the door because he's black and he's wearing a baseball cap, maybe I should refuse to deal with a white real estate agent the next time I am apartment hunting. Or maybe I should refuse to be shown around by white real estate agents with Southern accents who voted for Ronald Reagan. But what would white people think if they knew that successful black men like myself were refusing to do business with them just because of their skin color?

The last thing black people need today is a lot of white people getting pissed off at them. So when facing a situation in which I might get ripped off, I think it's prudent to take a chance. The humiliation of getting ripped off by some speedy black kid who snatches one of my gold necklaces is comparable to the humiliation of getting discriminated against by some smiling white real estate agent. But, of course, the odds that black kid outside is going to rob me are nowhere near 50 percent. Someday I hope I won't have to wonder about white real estate agents. So I figure: unless he's really a dangerous-looking kid, let him in. A lot of white people won't understand that it's them, more than the kid, I'm giving a break. But maybe someday.

Jefferson Morely


Among our contributors, Ronald Hampton, the police officer, takes the hardest line against buzzer bias. Even he, though, fails to note that it's illegal. A jewelry store, as a public accommodation, is forbidden to discriminate on the basis of race under the Civil Rights Act. If race is ever a factor in deciding whom you admit to your store, you are breaking the law. Proof and prosecution would be virtually impossible, of course. But it's interesting that the law imposes duty that all of our contributors except Hampton, Anthony Lukas, and some of the cabdrivers (who may have been disingenuous) concede they would sometimes feel justified in disregarding. (The elevator hypothetical is different in this respect.)

So is the law wrong to impose this duty? Walter Williams argues that generalization from race can be rational behavior in a world of imperfect information, and therefore not racism. But this argument "proves too much," as lawyers like to say. Only petty apartheid of the "whites only" drinking fountain variety is completely irrational in this narrow sense. Most of the racial discrimination American society is struggling to overcome - in schools, housing, employment - could be justified as the efficient use of statistically sound generalizations. Even the biased realtors of Jefferson Morley's example might just be making an economically rational accommodation to the irrational prejudice of their white customers. Walter Williams may nor find this objectionable, but most TNR readers will.

Anthony Lukas essentially shares Professor William's analysis, but reaches the opposite conclusion: the store owner who considers race is an inefficient Bayesian, irrationally generalizing from a factor that is of no predictive value. Most of us would love to believe this, but can't.

Michael Walzer insists that we must "change the situation of the shopkeepers and of the people they fear." No one would argue with that. But moral questions are not pointless simply because they will disappear when we eliminate racial prejudice and urban crime. It's true that the questions we pose are likely to arise when people are "alone and frightened." In those circumstances, a racial response may be understandable without moral being morally correct. That is precisely why we posed the questions to people at the time when they were comfortable and safe, free to accept or reject "the assumptions of the larger culture." And the questions remain.

Our other symposiasts stake out various areas of the middle ground, admitting that they are uncomfortable about taking race into account, but admitting also that they might use race as a major factor (Roger Starr, Rhonda Schoem) or a minor one (Juan Williams, Morley) in deciding whom to admit to their store. Maybe there's no way to do better than this. We would add just three thoughts.

First, returning to Walter William's framework of analysis, there is surely in all of us - especially white people - a tendency to overgeneralize from race. The intelligent Bayesian will compensate for this. A jewelry store owner who thinks he's got the factors toted up correctly might well want to discount the race factor by a large fraction before deciding not to hit the buzzer - just to make sure he's actually making the rational decision.

Second, citizenship imposes duties, and citizenship in a multiracial society requires us to act sometimes in ways that are not in our rational self-interest, narrowly defined. That duty may not require the jewelry store owner to ignore the race of young males wishing to enter his store, but it does call for another discount on the race factor in making the calculation.

Third, the burden of the race problem is one we share as a society. The solutions that will be least successful - and will stir the most resentments - are the ones that attempt to impose a disproportionate part of that burden on random individuals. Expecting a jewelry store owner to risk his life in the service of color-blind justice is expecting too much. Acknowledging this reality, however, also requires acknowledging that that society as a whole, principally through the government, must do all the more.

Be anything but fat: Prejudice against fat people

In this era of "political correctness," one group remains an "acceptable" target of prejudice - the obese. Are you prejudiced against obese people? Find out be answering the following questions:

  1. strongly agree
  2. agree
  3. neither agree nor disagree
  4. disagree
  5. strongly disagree

SOURCE: Crandall, C. S. (1994). Prejudice against fat people: Ideology and self-interest. Journal of Personality and Social Psychology, 66, 882-894.

Illusory Correlations

We sometimes perceive a relationship that does not exist, a phenomenon known as the illusory correlation. Illusory correlations may occur because the two variables being related are both distinctive. When two distinctive variables occur at the same time, we tend to see them as related, even when they are not. Thus, paired distinctiveness may cause an illusory correlation.

In 1976, Hamilton and Gifford argued that paired distinctiveness may explain, at least in part, some of the negative stereotypes of minority groups. Their reasoning was as follows:

  1. Majority-group members have relatively little contact with minority-group members. Thus, minority group members are distinctive for members of the majority.

  2. Negative behavior is relatively infrequent compared to positive or neutral behavior in everyday life. Thus, negative behavior is distinctive.

  3. According to the paired distinctiveness explanation of the illusory correlation, pairing minority group members with negative behavior - both distinctive - will lead to perceiving a relationship between the two - an illusory correlation.

Consider two fictitious groups - the Wallytogs and the Pennyfords. You are given information about 26 Wallytogs and 13 Pennyfords. Thus, the Wallytogs are the majority group in this example and the Pennyfords are the minority grouper.

Here are one-word descriptions of the 26 Wallytogs:


What is your impression of Wallytogs as a group? Take a minute to record your overall impression of this group.

Here are one-word descriptions of the 13 Pennyfords.


What is your impression of Pennyfords as a group? Take a minute to record your overall impression of this group.

Are your impression of the two groups different? HOW? Is your impression of the Pennyfords more negative/less favorable than your impression of the Wallytogs? If it is, then you succumbed to the illusory correlation. Here's why!

Consider the nature of the information you have about the two groups, which is summarized in the following table:


Notice that the ratio of undesirable to desirable characteristics is 8/18 (4/9) for the Wallytogs, and 4/9 for the Pennyfords - exactly the same. Although you do have more information about the Wallytogs, because you know 26 Wallytogs but only 13 Pennyfords, you should rate the two groups similarly because the ratio of undesirable/desirable characteristics is the same for the two groups. However, the distinctiveness of the minority group (the Pennyfords) AND the distinctiveness of undesirable characteristics - both infrequent events - leads to an illusory correlation between group and characteristics, when there is actually no relationship between the two. Thus, the Pennyfords are rated more unfavorably than the Wallytogs.

SOURCE: Hamilton, D. L., & Gifford, R. K. (1976). Illusory correlation in interpersonal perception: A cognitive basis of stereotypic judgments. Journal of Experimental Social Psychology, 12, 392-407.

Moods and stereotyping

Some nonobvious findings in social psychology

Commonsense suggests that people in a good mood should be less likely to use negative stereotypes than people in a bad mood. Just the opposite has been found in research by Galen Bodenhausen and his colleagues. They "induced" research participants to feel happy or sad by having them recall events from their lives that made them happy or sad. Participants then rated the guilt or innocence of an Hispanic accused of a crime consistent with Hispanic stereotypes. Happy participants were more likely to find the Hispanic guilty than were sad participants. Other findings suggested that happy people prefer to process information heuristically, and use stereotypes as heuristics, or shortcuts, rather than processing systematically, which might distract them from their happy thoughts. Sad people, in contrast, prefer the distraction of systematic processing.

SOURCE: Bodenhausen, G. V., Sheppard, L. A., & Kramer, G. P. (1994). Happiness and stereotypic thinking in social judgment. Journal of Personality and Social Psychology, 66, 621-632.

Are stereotypes accurate? The Case of Gender Stereotypes

The question of whether stereotypes accurately reflect group differences is an important one, but often a difficult one to answer. Part of the difficulty lies in the fact that so few studies have addressed the accuracy question. Another part of the difficulty is that accuracy itself is difficult to assess. However, the study of gender stereotypes is an exception. There is ample research on actual sex differences that may be compared to gender stereotypes, or consensual beliefs about how the sexes differ.

Which sex do you think scores higher on the following characteristics and behaviors?

If you answered "men" to all but "smiling", your beliefs about gender differences match the reality of gender differences perfectly. But you probably got at least one wrong, most likely the one about verbal SAT scores. Gender stereotypes hold that women are more verbal then men - they are believed to be more talkative, to enjoy reading more, and so on. And women seem to do better in school subjects that require verbal skills. So why don't they score higher than men on verbal portion of the SAT? Since 1970, the actual difference between men and women has been 10 to 12 points, in favor of men.

One explanation for the surprising gender difference in verbal SAT performance is that the questions used on the SAT are male-oriented, despite efforts by test-makers to eliminate gender bias. Another explanation is that the questions tap analytic ability more than verbal ability. Since men are, on average, higher in analytic ability than women, this would account for their superior performance on the verbal portion of the SAT. Yet another explanation is that more older women than men are returning to school. Older SAT-takers do not perform as well, on average, than younger ones, probably because some verbal skills are lost over time through disuse.

Were you surprised to find that men are happier than women? And if women are less happy, why are they smiling so much more than men (about 10 times more)? Answers to these question are complex. We do know that women are twice as likely to suffer from depression as men. But it may be that women are more willing to admit to depression and to seek treatment for it. And we do know that smiling is not always a sign of happiness. Evolutionary psychologists have observed that in some primate species the smile is a sign of appeasment to ward off attacks by more aggressive members of the species. Whether women's smiles function in this way is uncertain. But it is certain that the smile serves multiple functions for both sexes.


The Self-fulfilling Prophecy: How Stereotypes Become Realities

SOURCE: Snyder, M. L. (1982, July). Self-fulfilling stereotypes. Psychology Today, 60-68.

Gordon Allport, the Harvard psychologist who wrote a classic work on the nature of prejudice, told a story of a child who had come to believe that people who live in Minneapolis were called monopolists. From his father, moreover, he had learned that monopolists were evil folk. It wasn't until many years later, when he discovered his confusion, that his dislike of residents of Minneapolis vanished.

Allport knew, of course, that it was not so easy to wipe out prejudice and erroneous stereotypes. Real prejudice, psychologists like Allport argued, was buried deep in human character, and only a restructuring of education could begin to root it out. Yet many people whom I meet while lecturing seem to believe that stereotypes are simply beliefs or attitudes that change easily with experience. Why do some people express the view that Italians are passionate, blacks are lazy, Jews materialistic, or lesbians mannish in their demeanor? In the popular view, it is because they have not learned enough about diversity among these groups and have not had enough contact with members of the groups for their stereotypes to be challenged by reality. With more experience, it is presumed, most people of good will are likely to revise their stereotypes.

My research over the past decade convinces me that there is little justification for such optimism - and not only for the reasons given by Allport. While it is true that deep prejudice is often based on the needs of pathological character structure, stereotypes are obviously quite common even among fairly normal individuals. When people first meet others, they cannot help noticing certain highly visible and distinctive characteristics: sex, race, physical appearance, and the like. Despite people's best intentions, their initial impressions of others are shaped by their assumptions about such characteristics.

What is critical, however, is that these assumptions are not merely beliefs or attitudes that exist in a vacuum; they are reinforced by the behavior of both prejudiced people and the targets of their prejudice. In recent years, psychologists have collected considerable laboratory evidence about the processes that strengthen stereotypes and put them beyond the reach of reason and good will.

My own studies initially focused on first encounters between strangers. It did not take long to discover, for example, that people have very different ways of treating those whom they regard as physically attractive and those whom they consider physically unattractive, and that these differences tend to bring out precisely those kinds of behavior that fit with stereotypes about attractiveness.

In an experiment that I conducted with my colleagues Elizabeth Decker Tanke and Ellen Berscheid, pairs of college-age men and women met and became acquainted in telephone conversations. Before the conversations began, each man received a Polaroid snapshot, presumably taken just moments before, of the woman he would meet soon. The photograph, which had actually been prepared before the experiment began, showed either a physically attractive woman or a physically unattractive one. By randomly choosing which picture to use for each conversation, we insured that there would be no consistent relationship between the attractiveness of the woman in the picture and the attractiveness of the woman in the conversation.

By questioning the men, we learned that even before the conversation began, stereotypes about physical attractiveness came into play. Men who looked forward to talking with physically attractive women said that they expected to meet decidedly sociable, poised, humorous, and socially adept people, while men who thought that they were about to get acquainted with unattractive women fashioned images of rather unsociable, awkward, serious, and socially inept creatures. Moreover, the men proved to have very different styles of getting acquainted with women whom they thought to be attractive and those whom they believed to be unattractive. Shown a photograph of an attractive woman, they behaved with warmth, friendliness, humor, and animation. However, when the woman in the picture was unattractive, the men were cold, uninteresting, and reserved.

These differences in the men's behavior elicited behavior in the women that was consistent with the men's stereotyped assumptions. Women who were believed (unbeknown to them) to be physically attractive behaved in a friendly, likeable, and sociable manner. In sharp contrast, women who were perceived as physically unattractive adopted a cool, aloof, and distant manner. So striking were the differences in the women's behavior that they could be discerned simply by listening to tape recordings of the women's side of the conversations. Clearly, by acting upon their stereotyped beliefs about the women whom they would be meeting, the men had initiated a chain of events that produced behavioral confirmation of their beliefs.

Similarly, Susan Anderson and Sandra Bem have shown in an experiment at Stanford University that when the tables are turned - when it is women who have the pictures of men they are to meet on the telephone - many women treat the men according to their presumed physical attractiveness, and by so doing encourage the men to confirm their stereotypes. Little wonder, then, that so many people remain convinced that good looks and appealing personalities go hand and hand.


It is experiments such as these that point to a frequently unnoticed power of stereotypes: the power to influence social relationships in ways that create the illusion of reality. In one study, Berna Skrypnek and I arranged for pairs of previously unaquainted students to interact in a situation that permitted us to control the information that each received about the apparent sex of the other. The two people were seated in separate rooms so that they could neither see nor hear each other. Using a system of signal lights that they operated with switches, they negotiated a division of labor, deciding which member of the pair would perform each of several tasks that differed in sex-role connotations. The tasks varied along the dimensions of masculinity and femininity: sharpen a hunting knife (masculine), polish a pair of shoes (neutral), iron a shirt (feminine).

One member of the team was led to believe that the other was, in one condition of the experiment, male; in the other, female. As we had predicted, the first member's belief about the sex of the partner influenced the outcome of the pair's negotiations. Women whose partners believed them to be men generally chose stereotypically masculine tasks; in contrast, women whose partners believed they were women ususally chose stereotypically feminine tasks. The experiment thus suggests that much sex-role behavior may be the product of other people's stereotyped and often erroneous beliefs.

In a related study at the University of Waterloo, Carl von Baeyer, Debbie Sherk, and Mark Zanna have shown how stereotypes about sex roles operate in job interviews. The researchers arranged to have men conduct simulated job interview with women supposedly seeking positions as research assistants. The investigators informed half of the women that the men who would interview them held traditional views about the ideal woman, believing her to be very emotional, deferential to her husband, home-oriented, and passive. The rest of the women were told that their interviewer saw the ideal woman as independent, competitive, ambitious, and dominant. When the women arrived for their interviews, the researchers noticed that most of them had dressed to meet the stereotyped expectations of their prospective interviewers. Women who expected to see a traditional interviewer had chosen very feminine-looking makeup, clothes, and accessories. During the interviews (videotaped through a one-way mirror) these women behaved in traditionally feminine ways and gave traditionally feminine answers to questions such as "Do you have plans to include children and marriage in your career plans?"

Once more, then, we see the self-fulfilling nature of stereotypes. Many sex differences, it appears, may result from the images that people create in their attempts to act out accepted sex roles. The implication is that if stereotyped expectations about sex roles shift, behavior may change too. In fact, statements by people who have undergone sex-change operations have highlighted the power of such expectations in easing adjustment to a new life. As the writer Jan Morris said in recounting the story of her transition from James to Jan: "The more I was treated as a woman, the more woman I became."

The power of stereotypes to cause people to confirm stereotyped expectations can also be seen in interracial relationships. In the first of two investigations done at Princeton University by Carl Word, Mark Zanna, and Joel Cooper, white undergraduates interviewed both white and black job applicants. The applicants were actually confederates of the experimenters, trained to behave consistently from interview to interview, no matter how the interviewers acted toward them.

To find out whether or not the white interviewers would behave differently toward white and black job applicants, the researchers secretly videotaped each interview and then studied the tapes. From these, it was apparent that there were substantial differences in the treatment accorded to blacks and whites. For one thing, the interviewers' speech deteriorated when they talked to blacks, displaying more errors in grammar and pronunciation. For another, the interviewers spent less time with blacks than with whites and showed less "immediacy." a the researchers called it, in their manner. That is, they were less friendly, less outgoing, and more reserved with blacks.

In the second investigation, white confederates were trained to approximate either the immediate or the nonimmediate interview styles that had been observed in the first investigation as they interviewed white job applicants. A panel of judges who evaluated the tapes agreed that applicants subjected to the nonimmediate styles performed less adequately and were more nervous than job applicants treated in the immediate style. Apparently, then, the blacks in the first study did not have a chance to display their qualifications to the best advantage. Considered together, the two investigations suggest that in interracial encounters, racial stereotypes may constrain behavior in ways that cause both blacks and whites to behave in accordance with those stereotypes.


Having adopted stereotyped ways of thinking about another person, people tend to notice and remember the ways in which that person seems to fit the stereotype, while resisting evidence that contradicts the stereotype. In one investigation that I conducted with Seymour Uranowitz, student subjects read a biography of a fictitious woman named Betty K. We constructed the story of her life so that it would fit the stereotyped images of both lesbians and heterosexuals. Betty, we wrote, never had a steady boyfriend in high school, but did go out on dates. And although we gave her a steady boyfriend in college, we specified that he was more of a close friend than anything else. A week after we had distributed this biography, we gave our subjects some new information about Betty. We told some students that she was now living with another woman in a lesbian relationship; we told others that she was living with her husband.

To see what impact stereotypes about sexuality would have on how people remembered the facts of Betty's life, we asked each student to answer a series of questions about her life history. When we examined their answers, we found that the students had reconstructed the events of Betty's past in ways that supported their own stereotyped beliefs about her sexual orientation. Those who believed that Betty was a lesbian remembered that Betty had never had a steady boyfriend in high school, but tended to neglect the fact that she had gone out on many dates in college. Those who believed that Betty was now a heterosexual tended to remember that she had formed a steady relationship with a man in college, but tended to ignore the fact that this relationship was more of a friendship than a romance.

The students showed not only selective memories but also a striking facility for interpreting what they remembered in ways that added fresh support for their stereotypes. One student who accurately remembered that a supposedly lesbian Betty never had a steady boyfriend in high school confidently pointed to that fact as an early sign of her lack of romantic or sexual interest in men. A student who correctly remembered that a purportedly lesbian Betty often went out on dates in college was sure that these dates were signs of Betty's early attempts to mask her lesbian interests.

Clearly, the students had allowed their preconceptions about lesbians and heterosexuals to dictate the way in which they interpreted and reinterpreted the facts of Betty's life. As long as stereotypes make it easy to bring to mind evidence that supports them and difficult to bring to mind evidence that undermines them, people will cling to erroneous beliefs.


The power of one person's beliefs to make other people conform to them has been well demonstrated in real life. Back in the 1960s, as most people well remember, Harvard psychologist Robert Rosenthal and his colleague Lenore Jacobson entered elementary-school classrooms and identified one out of every five pupils in each room as a child who could be expected to show dramatic improvement in intellectual achievement during the school year. What the teachers did not know was that the children had been chosen on a random basis. Nevertheless, something happened in the relationships between teachers and their supposedly gifted pupils that led children to make clear gains in test performance.

In can also do so on the job. Albert King, now a professor of management at Northern Illinois University, told a welding instructor in a vocational training center that five men in his training program had unusually high aptitude. Although these five had been chosen at random and knew nothing about their designation as high-aptitude workers, they showed substantial changes in performance. They were absent less often than were other workers, learned the basic of the welder's trade in about half the usual time, ans scored a full 10 points higher than other trainees on a welding test. Their gains were noticed not only by the researcher and by the welding instructor, but also by other trainees, who singled out the five as their preferred co-workers.

Might not other expectations influence the relationships between supervisors and workers? For example, supervisors who believe that men are better suited to some jobs and women to others may treat their workers (wittingly or unwittingly) in ways that encourage them to perform their jobs in accordance with stereotypes about differences between men and women. These same stereotypes may determine who gets which job in the first place. Perhaps some personnel managers allow stereotypes to influence, subtly or not so subtly, the way in which they interview job candidates, making if likely that candidates who fit the stereotype shows up better than job-seekers who do not fit them.

Unfortunately, problems of this kind are compounded by the fact that members of stigmatized groups often subscribe to stereotypes about themselves. That is what Amerigo Farina and his colleagues at the University of Connecticut found when they measured the impact upon mental patients of believing that others knew their psychiatric history. In Farina's study, each mental patient cooperated with another person in a game requiring teamwork. Half of the patients believed that their partners knew they were patients; the other half believed that their partners thought they were nonpatients. In reality, the non patients never knew a thing about anyone's psychiatric history. Nevertheless, simply believing that others were aware of their history led the patients to feel less appreciated, to find the task more difficult, and to perform more poorly. In addition, objective observers saw them as more tense, more anxious, and more poorly adjusted than patients who believed that their status was not known. Seemingly, the belief that others perceived them as stigmatized caused them to play the role of stigmatized patients.


Apparently, good will and education are not sufficient to subvert the power of stereotypes. If people treat others in such a way as to bring out behavior that supports stereotypes, they may never have an opportunity to discover which of their stereotypes are wrong.

I suspect that even if people were to develop doubts about the accuracy of their stereotypes, chances are they would proceed to test them by gathering precisely the evidence that would appear to confirm them.

The experiments I have described help to explain the persistence of stereotypes. But, as is so often the case, solving one puzzle only creates another. If by acting as if false stereotypes were true, people lead others, too, to act as if they were true, why do the stereotypes not come to be true? Why, for example, have researchers found so little evidence that attractive people are generally friendly, sociable, and outgoing and that unattractiven people are generally shy and aloof?

I think that the explanation goes something like this: Very few among us have the kind of looks that virtually everyone considers either very attractive or very unattractive. Our looks make us rather attractive to some people but somewhat less attractive to other people. When we spend time with those who find us attractive, they will tend to bring out our more sociable sides, but when we are with those who find us less attractive, they will bring out our less sociable sides. Although our actual physical appearance does not change, we present ourselves quite differently to our admirers and our detractors. For our admirers we become attractive people, and for our detractors we become unattractive. This mixed pattern of behavior will prevent the development of any consistent relationship between physical attractiveness and personality.

Now that I understand some of the powerful forces that work to perpetuate social stereotypes, I can see a new mission in my research. I hope, on the one hand, to find out how to help people see the flaws in their stereotypes. On the other hand, I would like to help the victims of false stereotypes find ways of liberating themselves from the constraints imposed on them by other members of society.

PSY235 Home PageClick on the Home to go to the PSY235 Home Page

Lecture 4/OutlineClick on the Notebook to go to Lecture 4/Outline

Lecture 4/Thought QuestionsClick on the Brain to go to Lecture 4/Thought Questions

Lecture 4/PowerPoint SlidesClick on the Sphere to go to Lecture 4/PowerPoint Slides.

conference roomClick on Smiley to enter the Conference Room

Lecture 4/Quiz 4Click on the Pencil to take Quiz 4 for Lecture 4

emailClick on the Envelope to send EMAIL to Prof J