. 33
( 115 .)


relatively highly by that person). This individual might be in favor of “sin taxes” (such
as high tobacco and alcohol taxes) to raise money for national health care and also
might be in favor of stronger gun-control laws. A person who considers freedom to
be more desirable than equality probably would be against sin taxes (“Itʼs none of the
governmentʼs business if people want to kill themselves”) and also against govern-
ment regulation of gun ownership.
When asked, do people account for their attitudes by referring to speci¬c values? And
do people on opposing sides of an issue hold opposing values? In one study, research-
ers measured participantsʼ attitudes toward two issues, abortion and nuclear weapons
(Kristiansen & Zanna, 1988). Next, participants were asked to rank the (personal) impor-
tance of 18 values, such as freedom, equality, an exciting life, family security, and so
on, and then relate each value to their attitudes on these two issues.
People with different attitudes consider different values important. People who
oppose the right to abortion, for example, give a higher ranking to certain values
(e.g., mature love, wisdom, true friendship, salvation, and a world of beauty) than do
people who support the right to abortion. Those who support the right to abortion give
Social Psychology

a higher ranking to other values (e.g., happiness, family security, a comfortable life,
pleasure, an exciting life, and a sense of accomplishment) than do those who oppose
the right to abortion.
At the same time, both groups shared many values. Both ranked freedom, inner
harmony, and equality as the values most important to their attitude. Differences in the
rankings of other values were slight. The results also suggest that people on either side
of volatile issues might be much closer in their values than they realize.

Explicit and Implicit Attitudes
In many cases we freely express and are aware of our attitudes and how they in¬‚uence
our behavior. An attitude falling into this category is known as an explicit attitude.
explicit attitude An attitude
that operates on a conscious Explicit attitudes operate on a conscious level, so we are aware of them”aware of the
level via controlled processing. cognitive underpinnings of them”and are conscious of how they relate to behavior.
They operate via controlled processing and take some cognitive effort to activate. For
example, you may know how you feel toward a given political candidate and match
your behavior (e.g., voting for him or her) to that attitude. It is these explicit attitudes
that we often ¬nd having a directive effect on behavior.
Although many of our attitudes operate on this conscious level, there are others
implicit attitude An attitude that operate unconsciously. This form of an attitude is known as an implicit attitude.
that affects behavior Speci¬cally, an implicit attitude is de¬ned as “actions or judgments that are under
automatically, without control of automatically activated evaluation without the performerʼs awareness of that
conscious thought and below
causation” (Greenwald, McGhee, & Schwartz, 1998, p. 1464). In other words, implicit
the level of awareness via
attitudes affect behaviors automatically, without conscious thought, and below the level
automatic processing.
of awareness. For example, an individual may have a quick negative reaction toward a
member of a minority group, even though the individual professes positive and toler-
ant attitudes toward that group. The “gut-level” reaction occurs without thought and is
often distasteful to the individual (Wilson, Lindsey, & Schooler, 2000).
Wilson, Lindsey, and Schooler (2000) proposed a model of dual attitudes to explain
the relationship between explicit and implicit attitudes. They suggested that when one
develops a new attitude, the new attitude does not erase the old attitude. Instead, the
two attitudes coexist. The new attitude serves as the explicit attitude; the old attitude
remains in memory and takes on the role of the implicit attitude. This implicit attitude
can override the explicit attitude when the situation is right. For example, a person who
has changed from a racially prejudiced attitude to a nonprejudiced attitude may still
have an automatic negative reaction to a member of a minority group, despite the newly
formed positive attitude. In this case, the underlying unconscious implicit attitude has
overridden the explicit attitude.
Researchers have usually assumed that when people develop new attitudes, they tend
to override or obliterate the old attitudes. However, Petty, Tormala, Brinol, and Jarvis
(2006) have found that when attitudes change, the old attitude may not only remain in
memory but in fact can affect behavior. Petty and his colleagues did several experiments
in which they created new attitudes in people and then changed those attitudes for some
of the experimental participants and did not change them for others. The researchers
found that when participants were given new attitudes via a “priming” procedure in
which the people were not aware of the in¬‚uence attempt, their response to the person
or object was ambivalent. In other words, if you were conditioned to like Phil but then
were primed with negative words about Phil (presented very quickly, just below the level
of conscious awareness), your attitude should have changed from positive to negative.
We might expect that the new attitude would override the old, as Wilson et al. (2000)
originally suggested. However, that was not quite what happened. The new attitude
Chapter 5 Attitudes 161

toward Phil was ambivalent; you liked him and you didnʼt like him. You werenʼt quite
sure how you felt about Phil. This suggests that the old attitude hasnʼt disappeared and
is still affecting your judgments about Phil. This also suggests that when you take a
test of implicit attitudes, which are discussed later in this chapter, an older prejudicial
attitude may leak and merge with a newer, nonprejudiced one. This may be why lots of
people who take implicit attitude tests are surprised, even astounded, that they are as
prejudiced as the test seems to say they are.

How Are Attitudes Measured?
What happens when investigators want to learn about peopleʼs attitudes on a particular
issue, such as af¬rmative action, illegal aliens, or capital punishment? As pointed out
earlier in this chapter, attitudes are private; we canʼt know what a personʼs attitudes
are just by looking at her or him. For this reason, social psychologists use a variety of
techniques to discover and measure peopleʼs attitudes. Some of these techniques rely
on direct responses, whereas others are more indirect.

The Attitude Survey
The most commonly used techniques for measuring attitudes are attitude surveys. In an
attitude survey A self-
attitude survey, the researcher mails or emails a questionnaire to a potential respondent,
report method of measuring
conducts a face-to-face interview, or asks a series of questions on the telephone. Because
attitudes that involves a
respondents report on their own attitudes, an attitude survey is a self-report measure. A
researcher™s mailing a
respondent indicates his or her attitude by answering a series of questions. questionnaire to a potential
There may be several types of questions on an attitude survey. Open-ended ques- respondent, conducting a
tions allow respondents to provide an answer in their own words (Oskamp, 1991). For face-to-face interview, or
asking a series of questions on
example, respondents might be asked, What quali¬cations do you think are necessary
the telephone.
in a president of the United States? Although this type of question yields rich, in-depth
information, the answers can be dif¬cult to analyze. Consequently, most of the ques-
tions on an attitude survey are close-ended, or restricted, questions such as, Are women
quali¬ed to be president of the United States? Respondents would check a box indi-
cating a response, e.g., yes, no, or donʼt know. Notice that this type of question forces
respondents into making one of a limited number of choices.
Another kind of survey item is the rating scale, in which respondents indicate the
extent to which they agree or disagree with a statement by circling a number on a scale.
One of the most popular of these methods is the Likert scale. Likert items ask the person
to agree or disagree with such attitude statements as the following on a 5-point scale:
“I believe women are quali¬ed to serve in national of¬ce.” Likertʼs technique is a sum-
mated rating scale, so called because individuals are given an attitude score based on
the sum of their responses.
In evaluating election preferences or other attitudes, social psychologists usually
are interested in the attitudes of a large group. Because it is not possible to survey every
member of the group, researchers conducting an attitude survey select a sample or small
subgroup of individuals from the larger group, or population. Donʼt think that you need
a huge sample to have a valid survey. In fact, most nationwide surveys use a sample of
only about 1,500 individuals.
Although a sample need not be large, it must be representative. As you recall from
Chapter 1, a representative sample is one that resembles the population in all impor-
tant respects. Thus, for any category that is relevant to the attitude being measured
(e.g., race and ethnicity, socioeconomic class, gender, age), the sample would contain
Social Psychology

the same proportion of people from each group within the category (e.g., from each
race and ethnic group) as does the population whose attitudes are being measured. A
representative sample contrasts with a biased sample, which is skewed toward one or
more characteristics and does not adequately represent the larger population.

Potential Biases in Attitude Surveys
Although attitude surveys, containing various types of questions, are very popular, they
do have several problems that may make any responses made by research participants
invalid. Schwarz (1999) suggested that the way a person responds to a survey question
depends on a variety of factors, including question wording, the format of the question,
and the context within which the question is placed.
For example, presidential candidate Ross Perot commissioned a survey in March
1993 that included the following question: Should laws be passed to eliminate all pos-
sibilities of special interests giving huge sums of money to candidates? Ninety-nine
percent of the people who responded to the survey said yes. A second survey done by
an independent polling ¬rm asked the same question in a different way: Do groups have
the right to contribute to the candidate they support? In response to this question, only
40% favored limits on spending. This is a textbook example of how the wording of the
question can in¬‚uence polling data (Goleman, 1993).
Phrasing is important, but so are the speci¬c words used in a question. For example,
in one survey commissioned some years ago by the American Stock Exchange, respon-
dents were asked how much stock they owned. Much to everyoneʼs surprise, the highest
stock ownership was found in the Southwest. It seems that the respondents were think-
ing of stock of the four-legged kind, not the Wall Street type. The moral is that you
must consider the meaning of the words from the point of view of the people answer-
ing the questions.
Finally, respondents may lie, or to put it somewhat differently, they may not remem-
ber what they actually did or thought. Williams (1994) and his students asked voters
whether they had voted in a very recent election; almost all said they had. Williams was
able to check the actual rolls of those who had voted (not how they voted) and found
that only about 65% of his respondents had voted. Now, some may have forgotten, but
many simply did not want to admit they had failed to do a socially desirable thing”
to vote in an election (Paulhus & Reid, 1991).

Behavioral Measures
Because of the problems associated with self-report techniques, social psychologists
have developed behavioral techniques of measuring attitudes. These techniques, in one
way or another, avoid relying on responses to questions.
unobtrusive measure Unobtrusive measures assess attitudes by indirect means; the individual whose
A method of assessing attitudes are being measured simply is never aware of it. For example, in one early study,
attitudes such that the investigators measured voting preferences by tallying the number of bumper stickers for
individuals whose attitudes
a particular candidate on cars in a parking lot (Wrightsman, 1969). Other researchers
you are measuring are not
measured attitudes toward competing brands of cola by searching through garbage cans.
aware of your interest in them.
Still others attempted to determine the most popular exhibit at a museum by measuring
the amount of wear and tear on various parts of the carpet (Webb, Campbell, Schwartz,
Sechrist, & Grove, 1981).
Another example of unobtrusive measurement of attitudes is the lost-letter technique
(Milgram, Mann, & Hartner, 1965). If a researcher wants to measure a communityʼs
attitudes toward, say, its foreign residents, she might not get honest answers on a Likert-
type questionnaire. But, if she has some stamps and envelopes, she can try the lost-letter
Chapter 5 Attitudes 163

technique. This is what the researcher does: She addresses an envelope to someone with
a foreign-sounding name at a local address. She puts a stamp on the envelope and then
drops it on a crowded street near the post of¬ce so that it can easily be found and mailed.
As her baseline control, she drops a stamped envelope addressed to someone whose name
doesnʼt sound foreign. She repeats the procedure as many times as necessary to get a large
enough sample. Then all she has to do is count the envelopes that turn up in the mail and
compare the number with the names that sound foreign to the number with names that
doesnʼt. This is her measure of that communityʼs attitude toward foreigners.

Cognitive Measures: The Implicit Association Test (IAT)
In recent years a new test has been developed to tap our implicit attitudes, self-concepts,
and other important aspects of our cognitive system. The term implicit in this context
refers to relatively automatic mental associations (Hofman, Gawronski, Gschwendner,
Le, & Schmitt, 2005). The most well-known implicit measures test is the Implicit Implicit Association Test
(IAT) The most widely known
Association Test (IAT) (https://implicit.harvard.edu/implicit/) developed by Greenwald,
measure of implicit attitudes.
McGhee, and Schwartz (1998). Implicit attitudes, as we suggested earlier, are attitudes
that we hold but are not aware of, so that you are not able to directly report that atti-
tude. These implicit attitudes can only be measured by indirect means. The IAT aims
at determining the strength of connection between two concepts. For example, the IAT
asked test-takers to assign a stimulus, which can be word or pictures, as quickly as they
possibly can, to a pair of targets. Consider the following example.

Barry Bonds vs. Babe Ruth
As I write this chapter, the controversial San Francisco Giant left ¬elder, Barry Bonds,
has passed Babe Ruth for second place on the all-time homerun list. Bonds is an African
American and Ruth was white, playing in an era when African Americans were barred
from playing in the major leagues. On the IAT Web site, you are asked to respond as
quickly as you can to different photos of Barry or the Babe. In addition, you are asked
to respond to the pairing of the words good or bad when used with photos of the two
stars. The strength of connection (associative strength) between two concepts is therefore
assessed by combining a pair of categories”in this case, race (African American vs.
Caucasian) and a pair of attributes (good-bad). These are combined in both association
compatible (Babe”good [presumably]) and incompatible (Babe”bad). The scoring
of these associations may take a number of different forms, but basically, the differ-
ences in the time it takes to respond to these pairings (mean response latencies) is the
measure of the relative strength between the two pairs of concepts (Greenwald, Nosek,
& Banaji, 2003). The fundamental assumption behind the IAT is that we “donʼt always
»speak our minds,ʼ ” and as is noted on the IAT Web site, we may not even know our
own minds. The IAT is an attempt to tap into our unconscious associations. It has been
used to explore the unconscious bases of prejudicial attitudes of all kinds.

What Has the IAT Taught Us about Our Racial and Ethnic Attitudes?
The results of the millions of tests on IAT Web sites showed that 88% of white people
had a pro-white or antiblack implicit bias; nearly 83% of heterosexuals showed implicit
biases for straight people over gays and lesbians; and more than two-thirds of non-Arab,
non-Muslim volunteers displayed implicit biases against Arab Muslims.
In addition, similar results were obtained for religious, gender, and socioeconomic
attitudes. The most interesting ¬nding is that these results contrast not only with what
people say about their own attitudes but also with what they actually believe about their
true attitudes. Marajin Banaji, who helped develop the IAT, has said that “The Implicit
Social Psychology

Association Test measures the thumbprint of the culture on our minds. If Europeans
had been carted to Africa as slaves, blacks would have the same beliefs about whites
that whites now have about blacks” (Vedantam, 2005).

How Are Attitudes Formed?
We can see now that attitudes affect how we think, feel, and behave toward a wide
range of people, objects, and ideas that we encounter. Where do our attitudes come
from? Are they developed, as Allport suggested, through experience? If so, just how
do our attitudes develop through experience? And are there other ways in which we
acquire our attitudes?
The term attitude formation refers to the movement we make from having no atti-
tude toward an object to having some positive or negative attitude toward that object
(Oskamp, 1991). How you acquire an attitude plays a very important role in how you
use it. In this section, we explore a range of mechanisms for attitude formation. Most
of these mechanisms”mere exposure, direct personal experience, operant and classi-
cal conditioning, and observational learning”are based on experience and learning.
However, the last mechanism we will look at is based on genetics.

Mere Exposure
mere exposure The Some attitudes may be formed and shaped by what Zajonc (1968) called mere exposure,
phenomenon that being which means that simply being exposed to an object increases our feelings, usually
exposed to a stimulus positive, toward that object. The mere-exposure effect has been demonstrated with a
increases one™s feelings,
wide range of stimuli, including foods, photographs, words, and advertising slogans


. 33
( 115 .)