<<

. 28
( 115 .)



>>

We are more likely to give in-group members the bene¬t of the doubt than out-group
but not out-group, members
members (Pettigrew, 1979).
the bene¬t of the doubt for
Once we construct our categories, we tend to hold on to them tenaciously, which
negative behaviors.
may be both innocent and destructive. It is innocent because the process is likely to be
automatic and nonconscious. It is destructive because stereotypes are inaccurate and
often damaging; individuals cannot be adequately described by reference to the groups
to which they belong.
In general, social psychologists have not made a consistent attempt to determine
the accuracy of stereotypes. Much of the early research on stereotypes assumed that
stereotypes were inaccurate by de¬nition. More recently, the issue of stereotype accu-
racy has been addressed by Judd and Park (1993). They suggested several technical
standards against which the accuracy of a stereotype can be measured. For example,
consider the notion that Germans are ef¬cient. One standard that Judd and Park sug-
gested to measure the accuracy of that stereotype is to ¬nd data that answers the ques-
tions: Are Germans in reality more or less ef¬cient than the stereotype? Is the group
attribute (ef¬ciency) exaggerated?
Of course, to apply these standards, we need some objective data about groups.
We need to know how groups truly behave with respect to various characteristics. For
some attributes, say, kindness or sensitivity, it is probably impossible to obtain such
information. For others, there may be readily available data.
In McCauley and Stittʼs 1978 study of the accuracy of stereotypes, white participantsʼ
estimates of certain attributes of the African American population were compared with
public records (as cited in Judd & Park, 1993). The attributes estimated were percent-
age of high school graduates, number of crime victims, and number of people on the
welfare rolls. This study showed that whites underestimated the differences between
Social Psychology
134

African Americans and themselves with respect to these attributes. In other words, whites
thought more African Americans graduated from high school than was true, and they
thought fewer African Americans were victims of crime than the data showed.
Is it important to know if a stereotype is accurate? Technically it is, because many
of the earlier de¬nitions of stereotypes assume that inaccuracy is part of the de¬nition
of the concept (Stangor & Lange, 1994). Most stereotypes are unjusti¬ed generaliza-
tions; that is, they are not accurate. But, even if they are accurate, stereotypes still have
a damaging effect on our perception of others. None of us would wish to be judged as
an individual by the worst examples of the group(s) to which we belong.
In previous chapters, we have seen how automatic and controlled processing enter
into the social cognition process. Some people use controlled processing to readjust
initial impressions of others in instances where new information con¬‚icts with exist-
ing categorization (Fiske & Neuberg, 1990; Trope, 1986). Automatic and controlled
processing again come into play when we consider how stereotypes are maintained and
how prejudiced and nonprejudiced individuals differ.

The Difference between Prejudiced and Nonprejudiced
Individuals
Devine (1989) contends that stereotypes are automatically activated when we encoun-
ter a member of a particular social group. According to Devine, some people are able
to consciously alter their prejudiced responses, whereas others are not. Devine found
that those interested in being nonprejudiced think differently from those who are not.
For example, prejudiced individuals are more willing to indulge in negative thoughts
and behaviors toward members of different racial and ethnic groups than nonprejudiced
individuals. Devine also found that both high- and low-prejudiced whites hold almost
the same stereotypes of African Americans. However, nonprejudiced individuals think
those stereotypes are wrong.
Devine also found that the main difference between prejudiced and nonprejudiced
whites was that nonprejudiced whites are sensitive to and carefully monitor their stereo-
types. The nonprejudiced person wants his or her behavior to be consistent with his or
her true beliefs rather than his or her stereotypes. When given a chance to use controlled
processing, nonprejudiced individuals show behavior that is more consistent with non-
prejudiced true beliefs than stereotyped beliefs. In contrast, the behavior of prejudiced
individuals is more likely to be guided by stereotypes. In another study, nonprejudiced
individuals were more likely than prejudiced individuals to feel bad when they had
thoughts about gay men and lesbians that ran counter to their beliefs (Monteith, Devine,
& Zuwerink, 1993). When nonprejudiced individuals express prejudicial thoughts and
feelings, they feel guilty about doing so (Devine, Montieth, Zuwerink, & Elliot, 1991).
What happens if automatic processing takes over? According to Devine, activat-
ing a stereotype puts a person into automatic mode when confronting a person from
the stereotyped group. The automatically activated stereotype will be acted on by both
prejudiced and nonprejudiced individuals unless there is an opportunity to use controlled
processing (Devine, 1989). Devine found that when participants in an experiment were
prevented from switching to controlled processing, both prejudiced and nonprejudiced
individuals evaluated the behavior of an African American negatively.
We can draw several conclusions from Devineʼs research. First, prejudiced individu-
als are less inhibited about expressing their prejudice than nonprejudiced individuals.
Second, no differences exist between prejudiced and nonprejudiced individuals when
stereotype activation is beyond conscious control. Third, nonprejudiced people work
Chapter 4 Prejudice and Discrimination 135

hard to inhibit the expression of negative stereotypes when they have the opportunity to
monitor behavior and bring stereotypes under conscious control. Fourth, nonprejudiced
individuals realize that there is a gap between their stereotypes and their general beliefs
about equality, and they work continually to change their stereotyped thinking.
How easy is it to identify a prejudiced person? If you see a person in a Ku Klux Klan
out¬t distributing hate propaganda or burning a cross on someoneʼs lawn, thatʼs pretty
easy. However, many people do not express prejudices in such obvious ways. When we
encounter someone who makes racist or sexist comments, we can pretty easily identify
that person as prejudiced (Mae & Carlston, 2005). Further, we will express dislike for that
person, even if he or she is expressing ideas with which we agree (Mae & Carlston, 2005).
So, it seems we are pretty adept at identifying individuals who express negative prejudices.
However, when it comes to detecting positive prejudices, we are less adept. Speakers
who espouse negative prejudices are more likely to be identi¬ed as prejudiced than those
who espouse positive prejudices (Mae & Carlston, 2005).


The Consequences of Being a Target of Prejudice
Imagine being awakened several times each night by a telephone caller who inundates
you with racial or religious slurs. Imagine being a second-generation Japanese American
soldier on December 8, 1941 (the day after the Pearl Harbor attack), and being told you
are no longer trusted to carry a gun in defense of your country. Imagine being an acknowl-
edged war hero who is denied the Medal of Honor because of race-related suspicions of
your loyalty to the country for which you had just fought. In each of these instances, a
person becomes the target of prejudicial attitudes, stereotypes, and discriminatory behav-
ior directed at him or her. What effect does being the target of such prejudice have on an
individual? To be sure, being a target of discrimination generates a great deal of negative
affect and has serious emotional consequences for the target (Dion & Earn, 1975). Next,
we explore some of the effects that prejudice has on those who are its targets.

Ways Prejudice Can Be Expressed
In his monumental work on prejudice called The Nature of Prejudice, Gordon Allport
(1954) suggested that there are ¬ve ways that prejudice can be expressed. These are
antilocution, talking in terms of prejudice or making jokes about an out-group; avoid-
ance, avoiding contact with members of an out-group; discrimination, actively doing
something to deny members of an out-group something they desire; physical attack,
beatings, lynchings, and the like; and extermination, an attempt to eliminate an entire
group. One issue we must address is the reaction shown by members of an out-group
when they are targeted with prejudice. It is fairly obvious that those faced with overt
discrimination, physical attack, and extermination will respond negatively. But what
about reactions to more subtle forms of prejudice? What toll do they take on a member
of a minority group?
Swim, Cohen, and Hyers (1998) characterized some forms of prejudice as every-
day prejudice: “recurrent and familiar events that can be considered commonplace”
(p. 37). These include short-term interaction such as remarks and stares, and incidents
that can be directed at an individual or an entire group. According to Swim and col-
leagues, such incidents can be initiated either by strangers or by those with intimate
relationships with the target and have a cumulative effect and contribute to the targetʼs
experience with and knowledge of prejudice.
Social Psychology
136


Prejudice-Based Jokes
How do encounters with everyday prejudice affect the target? Letʼs start by looking at
one form of antilocution discussed by Allport that most people see as harmless: prejudice-
based jokes. Most of us have heard (and laughed at) jokes that make members of a group
the butt of the joke. Many of us may have even told such jokes, assuming that they do
no harm. But how do those on the receiving end feel? Women, for example, ¬nd sexist
jokes less funny and less amusing than nonsexist jokes (LaFrance & Woodzicka, 1998).
They also tend to report feeling more disgusted, angry, hostile, and surprised by sexist
versus nonsexist jokes. They also tend to roll their eyes (indicating disgust) and touch
their faces (indicating embarassment) more in response to sexist than to nonsexist jokes
(LaFrance & Woodzicka, 1998).
Ryan and Kanjorski (1998) directly compared the reactions of men and women
to sexist jokes. They found that compared to men, women enjoyed sexist humor less
and found it less acceptable and more offensive. Interestingly, men and women did not
differ in terms of telling sexist jokes. A more ominous ¬nding was that for men, there
were signi¬cant positive correlations between enjoyment of sexist humor and rape myth
acceptance, adversarial sexual beliefs, acceptance of interpersonal violence, likelihood
of engaging in forced sex, and sexual aggression. In another study, the exposure of
men with sexist attitudes to sexist jokes was related to tolerance for sexism and fewer
negative feelings about behaving in a sexist manner (Ford, Wentzel, & Lorion, 2001).
These ¬ndings may lend some credence to Allportʼs (1954) idea that antilocution, once
accepted, sets the stage for more serious expressions of prejudice.
A study reported by Thomas and Esses (2004) con¬rms the relationship between
sexist attitudes and enjoyment of sexist humor. Male participants completed measures
of sexism and authoritarianism. They then evaluated two types of sexist jokes. Half
of the jokes were degrading to women and half degrading to men. The results showed
that male participants who scored highly on the sexism scale found the jokes degrad-
ing females funnier and were more likely to repeat them than male participants who
were low on the sexism measure. Sexism did not relate to the evaluation of the jokes
that degraded men.

Stereotype Threat
As noted earlier, af¬liation with groups often contributes to a positive social identity.
What about membership in a group that does not confer a positive social identity? Not
all social groups have the same social status and perceived value. What happens when
an individual is faced with doing a task for which a negative stereotype exists for that
personʼs group? For example, it is well-established that blacks tend to do more poorly
academically than whites. What happens when a black individual is faced with a task
that will indicate academic aptitude?
One intriguing hypothesis about why blacks might not score well on standard tests
of IQ comes from an experiment conducted by Steele and Aronson (1995). According
to Steele and Aronson, when a person is asked to perform a task for which there is a
negative stereotype attached to their group, that person will perform poorly because
the task is threatening. They called this idea a stereotype threat. To test the hypothesis
stereotype threat
The condition that exists when that members of a group perform more poorly on tasks that relate to prevailing negative
a person is asked to perform stereotypes, Steele and Aronson conducted the following experiment. Black and white
a task for which there is a participants took a test comprising items from the verbal section of the Graduate Record
negative stereotype attached
Exam. One-third of the participants were told that the test was diagnostic of their intel-
to their group and performs
lectual ability (diagnostic condition). One-third were told that the test was a laboratory
poorly because the task is
threatening.
Chapter 4 Prejudice and Discrimination 137

tool for studying problem solving (nondiagnostic condition). The ¬nal third were told
that the test was of problem solving and would present a challenge to the participants
(nondiagnostic”challenge condition). Steele and Aronson then determined the average
number of items answered correctly within each group.
The results of this experiment showed that when the test was said to be diagnostic of
oneʼs intellectual abilities, black and white participants differed signi¬cantly, with black
participants performing more poorly than white participants. However, when the same
test was presented as nondiagnostic, black and white participants did equally well. There
was no signi¬cant difference between blacks and whites in the nondiagnostic-challenge
condition. Overall across the three conditions, blacks performed most poorly in the
diagnostic condition. In a second experiment, Steele and Aronson (1995) produced
results that were even more pronounced than in their ¬rst. They also found that black
participants in the diagnostic condition ¬nished fewer items and worked more slowly
than black participants in the nondiagnostic condition. Steele and Aronson pointed out
that this is a pattern consistent with impairments caused by test anxiety, evaluation
apprehension, and competitive pressure.
In a ¬nal experiment, Steele and Aronson (1995) had participants perform word-
completion tasks (e.g., ” ” ce; la ” ”; or ” ” ack that could be completed in a
racially stereotyped way (e.g., race; lazy; black) or a nonstereotyped way (e.g., pace;
lace; track). This was done to test if stereotypes are activated when participants were told
that a test was either diagnostic or nondiagnostic. Steele and Aronson found that there
was greater stereotype activation among blacks in the diagnostic condition compared
to the nondiagnostic condition. They also found that in the diagnostic condition, blacks
were more likely than whites to engage in self-handicapping strategies (i.e., develop-
ing behavior patterns that actually interfere with performance, such as losing sleep the
night before a test). Blacks and whites did not differ on self-handicapping behaviors in
the nondiagnostic condition.
These ¬ndings help us understand why blacks consistently perform more poorly
than whites on intelligence tests. Intelligence tests by their very nature and purpose
are diagnostic of oneʼs intellectual abilities. According to Steele and Aronsonʼs (1995)
analysis, when a black person is faced with the prospect of taking a test that is diagnostic
of intellectual ability, it activates the common stereotype threat that blacks are not sup-
posed to perform well on tests of intellectual ability. According to Steele and Aronson,
the stereotype threat impairs performance by generating evaluative pressures. Recall
that participants who were under stereotype threat in the diagnostic condition spent more
time doing fewer items. As they became more frustrated, performance was impaired.
It may also impair future performance, because more self-handicapping strategies are
used by blacks facing diagnostic tests. In short, the stereotype threat creates an impair-
ment in the ability to cognitively process information adequately, which in turn inhibits
performance. So, lower scores on IQ tests by blacks may relate more to the activation
of the stereotype threat than to any genetic differences between blacks and whites.
Steele and his colleagues extended the notion of the stereotype threat to other
groups. For example, Spencer, Steele, and Quinn (cited in Aronson, Quinn, & Spencer,
1998) found that men and women equated for math ability performed differently on a
math test, depending on whether they were told that there were past results showing
no gender differences in performance on the test (alleviating the stereotype threat) or
given no information about gender differences (allowing the stereotype threat to be
activated). Speci¬cally, when the “no gender differences” information was given, men
and women performed equally well on the test. However, when the stereotype threat
was allowed to be activated (i.e., that women perform more poorly on math tests than
Social Psychology
138

do men), men scored signi¬cantly higher than women. Aronson and Alainas reported
similar effects for Latino versus white participants and white males versus Asian males
(cited in Aronson et al., 1998).
In a more direct test of the relationship between gender, stereotype threat, and math
performance, Brown and Josephs (1999) told male and female students that they would
be taking a math test. One-half of the participants of each gender were told that the test
would identify exceptionally strong math abilities, whereas the other half were told that
the test would uncover especially weak math skills. Brown and Josephs reasoned that
for males the test for strong math skills would be more threatening, because it plays into
the stereotype that males are strong in math. On the other hand, the test for weakness
would be more threatening to females, because females stereotypically are viewed as
being weak in math. Their results were consistent with Steele and Aronsonʼs stereotype
threat notion. Males performed poorly on the test that supposedly measured exceptional
math skills. Conversely, females performed poorly on the test that was said to identify
weak math skills. In both cases, a stereotype was activated that was relevant to gender,
which inhibited performance. According to Brown and Josephs, the stereotype threat
for math performance is experienced differently for males and females. Males feel more
threatened when faced with having to prove themselves worthy of the label of being
strong in math skills, whereas females feel more threatened when they face a situation
that may prove a stereotype to be true.
Stereotype threat also operates by reducing positive expectations a person has going
into a situation. For example, based on a personʼs previous experience, he or she may
feel con¬dent about doing well on the SATs, having a positive expectation about his or
her performance on the exam. Now, letʼs say that a stereotype of this personʼs group is
activated prior to taking the exam. The resulting stereotype threat may lower that personʼs
expectations about the test, and as a consequence, the person does not do well.
The fact that this scenario can happen was veri¬ed in an experiment by Stangor,

<<

. 28
( 115 .)



>>