. 43
( 115 .)


Social Psychology

When this classic experiment was done in 1959, almost all the participants agreed
to misrepresent how much fun the experiment was (Festinger & Carlsmith, 1959).
Several weeks later, the participants were contacted by a third party and asked whether
they had enjoyed the study. Their responses turned out to depend on how much money
they had been paid. You might predict that the participants who got $20 said that they
enjoyed their experience more than those who got only $1. Well, thatʼs not what hap-
pened. Participants paid $20 said the tasks were boring, and those paid $1 said they
had enjoyed the tasks. A third group, the control participants, were given no reward
and were not told that anyone else had received one. Like the $20 group, they said the
tasks were boring.
Cognitive dissonance theory argues that change occurs when people experience
dissonance. Where is the dissonance in this experiment? Being paid $1, a tri¬‚ing sum
even in 1959, was surely insuf¬cient justi¬cation for lying. If a $1 participant analyzed
the situation logically, it would look like this: “I lied to someone because the experi-
menter asked me to, and I got paid only a buck.” Conclusion: “Either I am a liar or I
am stupid.” Neither conclusion ¬ts with what we generally think of ourselves. The dis-
sonance is between what we want to think of ourselves and how we have behaved. So,
how does the participant resolve the dissonance? The behavior canʼt be undone, so the
participant engages in self-persuasion: “Iʼm not a liar or stupid, so I must have meant
what I said. I enjoyed the experiment.” The $20-participant has an easily available, if
not very ¬‚attering, justi¬cation for the lie: “I needed the money.”

The Reverse Incentive Effect
The implications of this study and many more that have replicated the effect over the years
are intriguing. One concept that came from the original study is the reverse-incentive
effect: When people are given a large payment for doing something, they infer that the
activity must be dif¬cult, tedious, or risky (Freedman, Cunningham, & Krismer, 1992).
Thus, professional athletes who once played the game just for fun may now moan about
playing the game for $5 million a year. People seem to get suspicious when they are paid
large sums for doing something they enjoyed doing in the ¬rst place. They feel a little
apprehensive and develop a less positive view of the activity (Crano & Sivacek, 1984).
Dissonance theory argues, then, that the less the reward or the less the threatened
punishment used to make people behave counter to their attitudes, the more people have
to provide their own justi¬cations for their behavior. The more they have to persuade
themselves of the rightness of the behavior, the more their attitude is likely to change.

The Importance of Free Choice
An important condition in the arousal of dissonance is whether behavior is freely chosen
or coerced. In another study of cognitive dissonance, participants were asked to write
an essay arguing a position that ran counter to their real beliefs (Elkin & Leippe, 1986).
Furthermore, they did this attitude-inconsistent act when they felt they had freely chosen
it. Dissonance theorists call this situation induced compliance. The researchers found
that when participants wrote an essay counter to their beliefs, they showed greater
physiological arousal than if they had written an essay consistent with their beliefs. This
¬nding is compatible with predictions from cognitive dissonance theory, speci¬cally
that dissonance increases feelings of tension (physiological arousal).
This study reinforced the ¬nding that people do not experience dissonance if they
do not choose the inconsistent behavior (Brehm & Cohen, 1962). If they are forced
to do something, the coercion is a suf¬cient external justi¬cation for the attitude-
Chapter 6 Persuasion and Attitude Change 211

discrepant actions. If they donʼt have to justify their behavior to themselves, there is no
self-persuasion. This suggests that attribution processes may play a role in mediating
dissonance arousal and reduction. We explore this possibility later in this chapter.

Postdecision Dissonance
Free choice relates to dissonance in another way when you have to choose between two
mutually exclusive, equally attractive, but different alternatives (e.g., between two cars
or two jobs). After a choice is made, dissonance is experienced. It is important to note
that postdecision dissonance is not the same as predecision con¬‚ict, where you vacillate
between the two alternatives. Postdecision dissonance comes after your decision.
Here is how it works: Letʼs say you have enough money to buy a car. There are two
cars you are considering that are equally attractive to you. For each car, there is a set of
positive cognitions. Once you have made your choice (letʼs say you picked car 1), all
the positive cognitions associated with your chosen alternative are consistent with your
choice. However, all the positive cognitions associated with the unchosen alternative
are now inconsistent with your choice. Dissonance theory predicts that you will take
steps to reduce the dissonance associated with the unchosen alternative. One way to
reduce dissonance would be to change your decision (that is, choose car 2). Of course,
this wonʼt work, because now all of the cognitions associated with car 1 are inconsis-
tent with your new decision, and the dissonance remains. More likely, you will begin to
think of negative things about the unchosen car to reduce dissonance. For example, you
may reason that the insurance costs would be higher, the color isnʼt exactly what you
wanted, and the warranty is not as good. At the same time, you may also think of more
positive things about the chosen car. For example, you may point out how comfortable
the seats are, how good the stereo sounds, and how the color ¬ts you perfectly.
The arousal of postdecision dissonance and its subsequent reduction was demon-
strated in a classic experiment by Brehm (1956). In this experiment, female participants
¬rst rated the desirability of several household products (e.g., a toaster). Brehm then
offered the women one of the two products they had rated very closely or they had rated
very differently. After the women made their choices, they again rated the products.
Brehm found that when the two choice alternatives were close in desirability (a dif¬-
cult decision), ratings of the chosen alternative became more positive, compared to the
original ratings. At the same time, the ratings of the unchosen product became more
negative. This effect was less pronounced when the choice was between two products
that varied more widely in desirability (easy decision).
Generally, the greater the separation between alternatives, the less dissonance will
be produced after a decision. After all, a choice between a highly desirable product and
an undesirable product is an easy one. On the other hand, the closer the alternatives are
to one another (assuming they are not identical), the more dif¬cult the decision and the
more postdecision dissonance will be aroused. Thus, the greatest postdecision disso-
nance will be realized when you have to choose between two mutually exclusive (you
can only have one), equally attractive, but different alternatives.
How do we explain these free-choice dissonance situations? Shultz and Lepper
(1999) suggested that an analogy can be made between dissonance phenomena and the
operation of arti¬cial intelligence neural networks. Networks of cognitions underlie
states of consonance and dissonance and are activated by a set of constraints imposed
by a problem. For example, in a choice between two cars, you may be constrained by
¬nances, model preference, and color desirability. According to Shultz and Lepper, the
decision we make attempts to satisfy as many of the constraints as possible. In short,
Social Psychology

“the motivation to increase cognitive consonance, and thus to reduce dissonance, results
from the various constraints on the beliefs and attitudes that a person holds at a given
point in time” (p. 238). Consonance results when similar cognitions are activated and
inconsistent cognitions are inhibited. Thus, in the free-choice situation, linkages among
positive cognitions associated with an alternative produce consonance. However, for the
unchosen alternative, the linkages between inconsistent elements (the unchosen alterna-
tive and the positive cognitions associated with it) produce dissonance.
Shultz and Lepper (1996) performed computer simulations of Brehmʼs (1956)
original experiment and produced results that matched quite well with Brehmʼs results.
That is, ratings of the unchosen alternative became more negative, and ratings of the
chosen alternative became only slightly more positive. However, Shultz and Lepper
pointed out that in Brehmʼs experiment, participants always made a decision that was
both dif¬cult (two products that were rated very similarly) and between two highly
desirable products. Schultz and Lepper found that when participants had to choose
between two similarly rated but undesirable products, the ratings of the chosen product
became much more positive, but the ratings of the unchosen product became only
slightly more negative.
An experiment by Shultz, Leveille, and Lepper (1999) sought to test the results from
computer simulations of free-choice experiments against actual behavior of individuals.
Participants in this experiment were given the choice between two posters after indicating
on a rating scale how much they liked each poster. The choice parameters varied in dif-
¬culty. An easy choice was one between two posters”one with a high initial rating and
one with a low initial rating. In the “high-dif¬cult” condition, a choice was to be made
between two posters that had been rated very positively by participants. Finally, in the
“low-dif¬cult” condition, participants had to choose between two posters that had been
poorly rated. Following the choice procedure, participants again rated the posters. The
results paralleled the computer simulations. In the high-dif¬cult condition, ratings of the
unchosen alternative became substantially more negative, whereas ratings of the chosen
alternative became only slightly more positive. In the low-dif¬cult condition, the oppo-
site was true; ratings of the chosen alternative became much more positive. However,
ratings of the unchosen alternative became only slightly more negative. These results are
consistent with Shultz and Lepperʼs (1996) consonance constraint satisfaction model.
Finally, the way that postdecision dissonance operates may depend partly on oneʼs
culture (Hoshino-Browne et al., 2005; Kitayama, Snibbe, Markus, & Suzuki, 2004).
In Western culture personal dissonance reduction dominates. This means that when
we are selecting between two alternatives for ourselves, we are likely to experience
dissonance and resolve it in the manner predicted by dissonance theory. However, in
Eastern cultures (e.g., Japan) personal choices do not arouse as much dissonance as
they do in Western cultures. Instead, interpersonal dissonance tends to be more impor-
tant. Interpersonal dissonance arises when an individual is required to make a choice
for someone else. In the Hoshino-Browne et al. (2005) study, for example, European
Canadians (i.e., Canadians born in Canada) and Asian Canadians (Canadians born in
an Asian country) were asked to rank 10 Chinese cuisine entrees that would be served
at an on-campus restaurant. The rankings were done under two conditions. In one con-
dition, participants were instructed to rank the entrees based on their own personal
preferences (self-preferences). In the other condition, participants were instructed to
rank the entrees according to the preferences of their best friend (other preferences).
After completing some other measures, participants were offered two gift certi¬cates
for entrees they had ranked (their ¬fth and sixth choices were offered). Participants had
to choose one of the gift certi¬cates for themselves (in the self-preference condition)
Chapter 6 Persuasion and Attitude Change 213

European Asian



Difference From 0


Figure 6.9 Dissonance
when making a choice for
oneself or a friend among
Asian and European
Self Friend
Based on data from Hoshino-Browne et al.
Person Gift Chosen For (2005).

or their friend (in the other preference condition). The results, as shown in Figure 6.9
showed that when European Canadians were making a choice for themselves, more
dissonance reduction was shown than when making a choice for the friend. The oppo-
site was true for the Asian Canadians. They showed more dissonance reduction when
making a choice for their friend than for themselves.

Responsibility: Another View of Cognitive Dissonance
Another view suggests that cognitive dissonance occurs only when our actions produce
negative consequences (Cooper & Scher, 1992). According to this view, it is not the
inconsistency that causes dissonance so much as our feelings of personal responsibility
when bad things happen (Cooper & Fazio, 1984).
Letʼs say, for example, that you wrote a very good essay in favor of something you
believed in, such as not raising tuition at your school. You knew that the essay could be
presented to the schoolʼs board of trustees, the body that determines tuition rates. You
then learned that your essay was actually used to convince the board to raise tuition.
Or perhaps you were asked to write an essay taking a position you did not believe
in”raising tuition. You then learned that the essay convinced the board to raise tuition.
How would you feel?
According to this responsibility view, simply doing something counter to your
beliefs will not produce dissonance unless there are negative results. If you are opposed
to tuition hikes and write an essay in favor of them, but there are no hikes as a result,
you do not experience dissonance. In several similar studies, people were asked to write
essays advocating a position”raising tuition”that con¬‚icted with their beliefs. When
rates were increased and essayists felt responsible for the outcome, they resolved the
dissonance by changing their attitude in the direction of the outcome. That is, they began
to say they were now more in favor of a fee increase than before they wrote the essay.
When students wrote essays in favor of a fee increase, and fees were not increased,
they did not experience dissonance and did not change their attitudes. When there is no
tension, there is no attitude change.
Social Psychology

So, what creates dissonance, inconsistency, or a sense of responsibility? There have
been hundreds, perhaps thousands, of experiments that support the basic ideas of cogni-
tive dissonance theory”namely, that inconsistency leads to attitude change. That there
are valid alternatives simply means the theory may have to incorporate those ideas and
continue to be revised.

Attribution Processes and Dissonance
We noted earlier that dissonance is unlikely to be aroused when a person has a suf¬-
cient external justi¬cation (attribution) for his or her attitude-discrepant behavior. An
experiment by Cooper (1998) highlighted the role of attribution processes in mediating
dissonance reactions. Cooper had participants write a counter attitudinal essay advocat-
ing the institution of 7:00 A.M. classes on campus (something students opposed). They
wrote the essays under either a high-choice (participants were asked to write the essay
“if you are willing”) or a low-choice condition (the “if you are willing” phrase was left
out). Participants were also randomly assigned to a misattribution condition (an instruc-
tion that inconsistent lighting makes many feel tense and aroused) or a no-misattribution
condition (the instruction about the lighting effects was deleted). The main measure was
the participantsʼ ratings (positive or negative) about instituting 7:00 A.M. classes.
Cooper found that greater attitude change occurred under the high-choice condition.
This con¬rms our earlier statement that under conditions of free choice, dissonance is
more likely to be aroused and attitude change more likely to occur. Additionally, there
was less attitude change in the direction of the essay under the misattribution condition
than the no-misattribution condition. Participants in the misattribution condition had
an external explanation for their arousal (dissonance), and were consequently less
likely to change their attitude. The greatest amount of attitude change in the direction
of the essay was realized in the high-choice (participants chose to write the essay)/
no-misattribution condition. In a follow-up experiment using a different task, Cooper
found that participants who had previously misattributed their arousal to the lighting
did not show dissonance-consistent attitude change.
Attribution style also relates to the arousal of dissonance. Stalder and Baron
(1998) investigated the relationship between attributional complexity (AC) and
dissonance-produced attitude change in a series of experiments. Speci¬cally, attributional
complexity refers to how complex a personʼs attributions are for explaining behavior and
events. High-AC individuals are those who normally engage in thorough attributional
searches for information. Thus, a high-AC person will search long and hard for the
source of arousal in a given situation (e.g., a situation that arouses dissonance). A low-AC
person is less likely to engage in such a search.
The results from their ¬rst experiment con¬rmed the idea that high-AC individuals
show little dissonance-related attitude change, most likely because they are able to generate
a wide variety of possible causes for the arousal associated with dissonance (Stalder &
Baron, 1998). Having attributed the arousal to something other than the dissonance-arousing
situation, the high-AC individual would not be expected to show much attitude change.
In their second experiment, Stalder and Baron found that low-AC individuals showed the
typical dissonance-related attitude change after dissonance arousal.
The two experiments just discussed suggest strongly that dissonance-related attitude
change is mediated by the attributions made about the dissonance situation. If an alterna-
tive to dissonance is provided for an explanation for dissonance-related arousal, the typical
dissonance result does not occur. Stalder and Baronʼs study shows us that there are indi-
vidual differences in attributional style, which correlates with dissonance-related attitude
Chapter 6 Persuasion and Attitude Change 215


. 43
( 115 .)