. 36
( 115 .)


In summary, the theory of planned behavior emphasizes that behavior follows from
attitudes in a reasoned way. If a person thinks that a particular behavior associated with
an attitude will lead to positive outcomes, that other people would approve, and that the
behavior can be done readily, then the person will engage in the behavior (Eagly, 1992).
People essentially ask themselves if they can reasonably expect that the behavior will
achieve their individual and social needs.
Letʼs use the theory of planned behavior to analyze voting behavior. Assume you
have a positive attitude about voting (the object). Will you actually vote? Letʼs say you
think that it is the duty of every citizen to vote. Furthermore, your friends are going
to vote, and you believe they will think badly of you if you donʼt (subjective norms).
Finally, you feel that you will be able to easily rearrange your schedule on election day
(perceived behavioral control). If we know all this about you, we can conclude you have
a strong intention to vote and can make a pretty con¬dent prediction that, in keeping
with your attitude, you are likely to vote.
The accuracy of behavioral intentions in predicting behavior is evident in the Gallup
Poll. The Gallup organization has been conducting voting surveys since 1936, the year
Franklin Delano Roosevelt ran against Alf Landon, governor of Kansas. Figure 5.2 shows
the record of the Gallup Poll in national elections from 1968 to 2001. In general, the
polls are quite accurate. Yes, there have been a few exceptions over the past 57 years.
They certainly got it wrong in 1948: The data indicated that Harry Truman did not have
much of a chance to win. But rarely in history books do we hear mention of Dewey, the
governor of New York who ran against Truman and who was projected as the winner.
Chapter 5 Attitudes 175

Outcome Type
Predicted Actual




Figure 5.2 Gallup Poll
data showing predicted
and actual outcomes for
presidential elections from
1968 to 2000. Gallup
10 Polls are remarkably
accurate in predicting not
only the winner but also the
1968 1972 1976 1980 1984 1988 1992 1996 2000 margin of victory. (Note:
Presidential Election Year Average error = “1.93.)

In this case, the pollsters were wrong primarily because they stopped polling a little
too early. They had not yet learned that people have other things on their minds than
elections and may not start to pay serious attention to the campaign until a week or so
before the actual vote. Pollsters will not make that error again.
Although the question, “For whom will you vote, candidate X or candidate Y?”
might appear to be a measure of attitude, it is really a measure of behavioral inten-
tion. Voting is a single act and can be measured by a single direct question. These are
the circumstances in which consistency between attitude and behavior is likely to be
the highest. Pollsters often try to determine the strength of these intentions by asking
such questions as: How strongly do you feel about your preferred candidate? How
intense are your feelings? Although re¬nements like these may add to the accuracy
of voting surveys in the future, what is needed is a concrete way of measuring behav-
ioral intentions.
Recent research has reinforced the notion that emotions are crucially involved in
turning attitude into behavior. For example, Farley and Stasson (2003) examined the
relationship between attitudes and giving blood donations. They found that both donorsʼ
behavioral intentions to give blood and their positive emotions about doing so were
predictive of actually donating blood.

The Importance of Conviction
So what we have seen in the previous section is that the importance of some of our
attitudes is a crucial determinant of how we act. Some of our attitudes are important
to us; others are much less important. One reason researchers underestimated the
attitude“behavior link is because they did not focus on attitudes that are important to
people (Abelson, 1988). Attitudes held with conviction are central to the person holding
them. Examples include attitudes of racial and gender equality, racism and sexism,
patriotism, religious fundamentalism, and occultism. Attitudes held with conviction
are like possessions (Abelson, 1988). Recall that one function of an attitude is that it
Social Psychology

de¬nes us; it tells people who we are. The person owns his or her attitudes, proudly
displaying them to those who would appreciate them and defending them against
those who would try to take them away. For example, someone deeply committed
to one side or the other of the abortion issue will likely defend his view against the
other side and show his solidarity with those on the same side. Such attitudes will
be hard to change, as a change would mean a major alteration in the way the person
sees the world.
Because attitudes to which people are strongly committed are hard to manipulate in
a laboratory experiment, researchers tended to stay away from them. As a result, social
psychologists overestimated the ease with which attitudes might be changed and under-
estimated the relationship between attitudes and behavior. If an attitude is important
to people, they expect that behavior in agreement with that attitude will help them get
what they want. Thus, important attitudes and behavior tend to be closely linked.
An attitude held with conviction is easily accessible. This means that if you discuss
with someone a subject about which they feel strongly, they respond quickly and have a
lot of ideas about it. Moreover, attitude accessibility”the ease with which one can bring
a particular attitude to mind”is increased by constant use and application of that atti-
tude (Doll & Ajzen, 1992). In a study several years ago, researchers measured latencies
(speed of response) with respect to questions about womenʼs rights, abortion, and racial
integration (Krosnick, 1989). Whatever the issue, people who considered an attitude
important responded more quickly than those who considered it unimportant. Important
attitudes are more available in memory and are more likely to correspond to behavior.
If your stand on abortion, womenʼs rights, gun ownership, or the Dallas Cowboys is
important, you are more likely to act in a manner consistent with that attitude.
You can get a sense of how accessible an attitude is by noting how long it takes
you to recall it. For example, notice how long it takes you to recall your attitude toward
the following: living wills, parent-teacher associations, the death penalty, aisle seats,
snakes, water ¬lters, political action committees, the clergy, daylight-savings time,
baseball. Some of these notions brought feelings and thoughts to mind quickly; others
may not have.
If attitude accessibility indicates strength of conviction, we might expect attitudes
high in accessibility to be better predictors of behavior than attitudes lower in accessi-
bility. Fazio, who has extensively studied attitude accessibility, investigated this issue in
connection with the 1984 presidential election (Fazio & Williams, 1986). The summer
before the election, potential voters were asked whether they agreed with each of the
following two statements: “A good president for the next 4 years would be Walter
Mondale (the then Democratic nominee),” and “A good president for the next 4 years
would be Ronald Reagan (the elected Republican).” The respondents had to indicate
how strongly they agreed or disagreed by pressing one of ¬ve buttons: strongly agree,
agree, donʼt care, disagree, strongly disagree.
The researchers measured the time that passed before respondents pressed the
button. The delay interval between the moment you are confronted with an object and
the moment you realize your attitude is called the latency (Rajecki, 1990). The longer
respondents took to hit the button, the less accessible the attitude. Not only were the
researchers able to get a reading of the attitude toward the candidates, but also they
were able to get a measure of accessibility.
On the day after the election, respondents were asked whether they had voted and,
if so, for whom they had voted. Was there a relationship between latency times and
voting behavior? That is, did attitude accessibility predict behavior? The answer is, yes,
it did. Attitude accessibility measured in June and July 1984 accurately predicted voting
Chapter 5 Attitudes 177

behavior in November. Those who had responded quickly for Reagan were more likely
to vote for him than those who had taken longer to respond. The same relationship held,
although not quite as strongly, for Mondale supporters.

The Nonrational Actor
The theories and ideas about attitudes and behavior so far tend to assume a rational,
almost calculated approach to behavior. In the theory of planned behavior, if you can
get measures of peopleʼs attitude toward a behavior, their perception of how important
others might approve or disapprove of what they do, and their sense of control over
that behavior, then you can predict their intentions and, therefore, their likely behavior.
If there is a signi¬cant criticism of the theory of planned behavior, it is that when you
ask people to tell you about the components of their intentions, they know that their
answers should be logical. If you reported that you voted but you had no interest in
the candidates and you thought all candidates were crooks, this hardly makes you look
like a logical individual.
Some theories have taken the opposite approach: They assume that human beings
nonrational actor A view
are nonrational actors (Ronis & Kaiser, 1989), and our attitudes may often be totally
that humans are not always
irrelevant to our behavior. Cigarette smoking, for example, is so habitual as to be auto-
rational in their behavior
matic, totally divorced from any attitude or behavioral intention the smoker may have.
and their behavior can be
Most of our behaviors are like that (Ronis & Kaiser, 1989). We do them over and over
inconsistent with their attitudes.
without thought (Gilbert, 1991). You ¬‚oss your teeth, but your attitude and intentions
about dental hygiene are activated only when you run out of ¬‚oss. Even though you
believe ¬‚ossing is important, and even though you remember that sign in your dentistʼs
of¬ce that reads, “No, you donʼt have to ¬‚oss all you teeth”only the ones you want to
keep,” you now have to act on your attitude. Are you willing to get in the car at 11 P.M.
and drive to the store to buy more dental ¬‚oss? Similarly, if your regular aerobics class
becomes inconvenient, is your attitude about the importance of exercise strong enough
that you will rearrange your whole schedule?
In sum, people usually behave habitually, unthinkingly, even mindlessly. They make
active decisions only when they face new situations. Thus, there is a good chance of
inconsistencies between our attitudes and our behavior.

Mindless Behavior in Everyday Life
Have you ever arrived home after work or school and not been able to recall a single thing
about how you got there? In everyday life, we often run on a kind of automatic pilot.
Our behavior becomes so routine and automatic that we are hardly aware of what we are
doing. We are in a state of mind that Ellen Laner (1989) termed mindlessness, one that
involves reduced attention and loss of active control in everyday activities. Mindlessness
occurs when weʼre engaging in behaviors that have been overlearned and routinized. In
this state, we carry out the behaviors rigidly, according to a preconceived pattern and
without thought or appraisal. Mindlessness is fairly common in our everyday interac-
tions. The cashier at a restaurant asks you, “How was everything?” You say that your
steak was overcooked, your potato was cold, and the service was terrible. The cashier
replies, “Hereʼs your change, have a nice day.” In this example, the cashierʼs question
and response were automatic; she really didnʼt care how you enjoyed your meal.
Langer was interested in studying this state of mind (Langer, Blank, & Chanowitz, 1978).
She had a researcher approach people waiting to use a copy machine in the library and ask
to use it ¬rst. The request was phrased in one of several ways: “Excuse me, I have ¬ve
pages to copy. May I use the machine because I am in a rush?” “Excuse me, I have ¬ve
Social Psychology

pages to copy. May I use the machine?” and “Excuse me, I have ¬ve pages to copy.
May I use the machine because I have to make copies?” The researcher also asked to
make 20 copies in these three different ways. Request 2 offers no reason for using the
copier ¬rst, and request 3 offers a mindless reason (“because I have to make copies”);
only request 1 provides a minimally acceptable reason (“because I am in a rush”). If the
participants in this situation were dealing with the request in a mindless fashion, they
would fail to distinguish between legitimate and illegitimate (or ridiculous) reasons.
As it turns out, any kind of excuse works as long as the request is small. When the
request was to make ¬ve copies, people apparently did not appraise the quality of the
excuse as long as one was offered: Having to make copies was just as good as being in
a rush. People snapped out of their mindless state, however, when the request was to
make 20 copies. It is clear that when the behavior (the request) had a signi¬cant impact,
people paid more attention to the difference between bad and good excuses. Although
we usually pay close attention to good and bad reasons for peopleʼs behavior, it may
be that the request to copy ¬ve pages isnʼt worth the effort. When the ante is raised to
20 pages, then we are more mindful.
The fact that we hold a number of attitudes without really thinking about them
means there can be some interesting consequences once we are forced to think about
them. Thinking about our attitudes and the reasons we hold them can sometimes be
disruptive and confusing (Wilson, Dunn, Kraft, & Lisle, 1989). More generally, the
process of introspecting”of looking into our own mind, rather than just behaving”can
have this effect.
Timothy Wilsonʼs work showed that thinking about the reasons for our attitudes
can often lead us to behave in ways that seem inconsistent with those attitudes (Wilson
et al., 1989). For example, if you are forced to think about why you like your romantic
partner, you might wind up ending the relationship in the near future. Much depends on
the strength of the relationship. If the relationship is not strong, thinking about reasons
might weaken it. If it is pretty strong, then reasoning might further strengthen it. The
stronger our attitude or belief, the more likely that thinking about it will increase the
consistency between it and our behavior (Fazio, 1986).
Why should thinking about reasons for our attitudes sometimes lead to inconsis-
tency between our attitudes and behavior? The basic answer is that if we have never
really thought about an attitude before, then thinking about it may cause us to change it
(Wilson et al., 1989). If you are forced to count the ways you love your current partner,
and it takes you a lot of time to use all the ¬ngers on one hand, you have gotten some
insight into how you really think about the relationship.
This explanation was supported by a study in which people were asked their attitudes
about social issues, such as the death penalty, abortion, and national health insurance, in
two separate telephone surveys conducted a month apart (Wilson & Kraft, 1988). In the
¬rst survey, some people were asked to give their reasons for their opinions, whereas others
were just asked their opinions. A month later, those people who had been asked to give
reasons proved more likely to have changed their opinion. So thinking about reasons seems
to lead to change. Why? The full explanation might lie in the biased sample hypothesis,
proposed by Wilson and colleagues (1989). It goes like this: If you ask people why they
believe something, they are not likely to say, “I donʼt know.” Instead, they will conjure
up reasons that seem plausible but may be wrong or incomplete. That is, because people
often do not know their true reasons, they sample only part of those reasons. Thus, they
present a biased sample of their reasons. People then assume the reasons in the biased
sample are their true reasons for holding the belief. If these reasons donʼt seem compel-
ling, thinking about them may persuade people to change their belief.
Chapter 5 Attitudes 179

The Rational and Nonrational Actors: A Resolution
Sometimes we are rational actors; sometimes we are nonrational actors. Sometimes our
behavior is “coupled” to our attitudes; sometimes it is “uncoupled” from them. Isnʼt this
where we began? Letʼs see if we can now resolve the apparent con¬‚ict. It makes sense
to see attitudes and behavior as ordinarily linked, with uncoupling occurring primarily
under two kinds of circumstances.
The ¬rst circumstance is when an attitude is not particularly important to you. You
may not have thought about the attitude object much or have expressed the attitude very
often. So in this case, you donʼt really know what you think. True, capital punishment
and national health care are important issues. But many of us may not have thought
them through. When you are forced to consider these issues, you may be surprised by
what you say. This may make you reconsider your attitude.
The second circumstance is slightly more complicated. Essentially, it is when you
donʼt have a clear sense of your goals and needs. Letʼs go back to the theory of planned
action for a moment. The theory says if you expect that a behavior can help you achieve
your goals and social needs, you will do it. But people are often not clear about their
goals and needs (Hixon & Swann, 1993). When you are not clear about what you want
to accomplish, then your behavior will be relatively unpredictable and might well be
uncoupled from your attitudes.
For example, we exercise, but only sporadically, because we are mainly concerned
about looking good in front of our health-obsessed friends. Our reasons are weak, not
clear to us, and therefore our exercising behavior is infrequent and unpredictable. But
if we or a friend the same age has a heart attack, we develop a much stronger attitude


. 36
( 115 .)