Content uploaded by Julia Kneer
Author content
All content in this area was uploaded by Julia Kneer on Nov 05, 2014
Content may be subject to copyright.
J. Kneer & S. Glock: Reduction of Cogn itive Dissonance in SmokersSocialP sychology 2012; Vol. 43(2):81–91© 2012 Hogrefe Publishing
Original Article
Fast and Not Furious?
Reduction of Cognitive Dissonance in Smokers
Julia Kneer
1
, Sabine Glock
2
, and Diana Rieger
1
1
University of Cologne, Germany,
2
University of Luxembourg, Luxembourg
Abstract. Three studies explored whether cognitive dissonance in smokers is reduced immediately or remains constant due to the per-
ceived health risk. Because dissonance-reducing strategies might occur very quickly and previous research has focused only on ratings
concerning health risk, we additionally analyzed response latencies and psychophysiological arousal as more implicit measurements. In
Study 1, 2, and 3, participants rated their smoking-related health risks twice for different diseases. Ratings, response latencies (Study 1,
2), and psychophysiological arousal (Study 3) differed during the first testing. Differences in response latencies and psychophysiological
arousal diminished during the second testing, whereas ratings did not change. The results are discussed in terms of implicit methods as
measurements for cognitive dissonance and in terms of prevention and intervention programs.
Keywords: smoking, cognitive dissonance, response latencies, arousal, ratings
In general, healthy human beings have the need to perceive
themselves as intelligent, reasonable, and decent people.
Therefore, we strive to maintain a positive concept of our-
selves even when we exhibit behavior or an attitude that
contradicts this rosy self-image (Aronson, 1969). Think of
a person who smokes cigarettes. Smoking is known to con-
tribute to many different diseases, for instance, lung cancer
or circulatory disorders. Smokers are confronted with their
health-damaging behavior every day. Although most peo-
ple are nonsmokers, nearly 25% of the German population
still smoke (Statistisches Bundesamt, 2010). Smoking re-
sults in tremendously negative health consequences; in
2008, more than 5% of all deaths resulted from a smoking-
related disease (Statistisches Bundesamt, 2010). But smok-
ers, too, do not want to harm their health on purpose or
assume a negative self-concept due to unreasonable behav-
ior. Therefore, smoking behavior and a health-related self-
concept are inconsistent with each other. Because the self
is threatened when confronted with own unhealthy behav-
ior, this conflict between attitude and behavior results in a
negative drive state (Stone & Cooper, 2001).
Festinger (1957) defines such negative drive states as
cognitive dissonance. The theory of cognitive dissonance
is one of the most important and influential theories in so-
cial psychology (e.g., Aronson, 1992; Cooper, 1992; Jones,
1976; Visser & Cooper, 2003). It describes how counterat-
titudinal behavior results in cognitive dissonance. Such a
state of mind is disturbing and motivates people to reduce
their cognitive dissonance. Some theories of cognitive dis-
sonance assume that the involvement of the self is crucial
in dissonance processes (Aronson, 1969; Stone & Cooper,
2001). The smoker’s dilemma is ideal for investigating
cognitive dissonance processes (Steele, 1988) because
smokers are highly involved with their health-damaging
behavior and the resulting negative consequences. Nega-
tive smoking-related information constantly reminds
smokers of a discrepancy between self-standards and cur-
rent behavior.
Another approach assumes that dissonance arises be-
cause of the knowledge about one’s personal responsibility
for the possible aversive consequences (Cooper & Fazio,
1984). Regardless of the theoretical dissonance approach,
smoking is a dissonance-inducing behavior because of the
involvement of the self in health-damaging behaviors and
the personal responsibility for the health consequences in-
volved. Smokers could use many strategies to reduce cog-
nitive dissonance: They could quit smoking, they could de-
ny the relationship between diseases and smoking (Gib-
bons, Gerrard, Lando, & McGovern, 1991), or they could
emphasize other (more positive) cognitions (Kneer &
Glock, 2008). For instance, Van Harreveld, Van der Pligt,
and De Vries (1999) found that smokers tend to emphasize
other cognitions by stressing short-term benefits (like re-
laxation), whereas only nonsmokers take health-damaging
consequences into account.
Another strategy to reduce cognitive dissonance is to
think less about conflicting information. Smokers seem to
suppress thoughts about negative consequences because of
their personal involvement and the resulting cognitive dis-
sonance. Since actively avoiding a thought or a stressful
topic needs cognitive capacity, it could be shown that peo-
ple react more slowly to the suppressed thought (Macrae,
DOI: 10.1027/1864-9335/a000086
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
Bodenhausen, & Milne, 1994). Recent research on the suc-
cess of thought suppression is ambiguous. Many studies
showed that attempting to suppress thoughts (e.g., stereo-
types) results in a higher activation known as rebound ef-
fect (Macrae et al., 1994; Monteith, Spicer, & Tooman,
1998): Trying not to think about something seems to make
such information more accessible. Of course, practice
makes perfect. Especially, internally motivated people
seem to be able to suppress activated thoughts (Monteith,
Spicer et al., 1998; Plant & Devine, 1998). As Macrae et
al. (1994) claim, thought suppression first needs cognitive
capacity as it is a controlled process (Bargh, 1989; Logan,
1988; Shallice, 1972) but becomes automatized, and thus
unconstrained, by attentional capacity (Wegner & Erber,
1992). People might suppress activated concepts and might
avoid their application, but they are not able to circumvent
the initial activation (Monteith, Sherman, & Devine, 1998).
What is known so far is that thought suppression and re-
bound effects are also found when people are not directly
instructed to suppress certain thoughts. Therefore, thought
suppression seems to be a robust concept (Wenzlaff &
Wegner, 2000). Personal behavior affected by a specific
issue and related negative association seems to be a very
good reason motivating one to learn how to suppress re-
sulting negative thoughts (Newby-Clark, McGregor, &
Zanna, 2002). Smokers should be motivated to suppress
thoughts about smoking-related hazards, which might be
an effective way to reduce cognitive dissonance. This
should result in decreased processing time when someone
is confronted with negative health information. Neverthe-
less, unpleasant thoughts must be activated for building a
motivation to suppress. Processing time for negative infor-
mation should be high at the beginning but should change
over time due to thought suppression and reduced process-
ing time. For example, Salkovskis and Reynolds (1994)
found that smokers trying to suppress thoughts about smok-
ing showed higher cravings than smokers who did not try
to suppress their thoughts. What is not known so far is how
quickly these dissonance-reducing strategies of smokers
occur when one is confronted with the health risks.
Another line of research could help in addressing the
topic of more implicit dissonance-reduction strategies.
Cognitive dissonance is said to lead the individual into a
state of arousal and discomfort. Elliott and Devine (1994)
showed that people who are put into a dissonance-arousing
state report more discomfort and agitation than people who
do not experience dissonance. Other studies underline this
research by applying more implicit measures, such as psy-
chophysiological measurements, to the research of the
arousing qualities of cognitive dissonance. For example, it
has been shown that dissonance increases general activity
(Cook, Pallak, Storms, & McCaul, 1977; Pallak, Brock, &
Kiesler, 1967; Pallak & Kiesler, 1968), as indicated by the
fact that it affects several indicators such as heartbeat
(Mann, Janis, & Chaplin, 1969), galvanic skin responses
(Elkin & Leippe, 1986), and the frequency of nonspecific
skin conductance responses (Croyle & Cooper, 1983).
Hence, psychophysiological measurements could help in
analyzing dissonance arousal and its reduction over a given
time period.
We conducted three explorative studies to analyze
whether smokers immediately reduce cognitive dissonance
by activating defense mechanisms or whether thought sup-
pression does not work as a dissonance-reduction strategy
when one is confronted with their health risks. In Study 1
and Study 2 we investigated response latencies; in Study 3
we studied the psychophysiological arousal as a measure-
ment for dissonance induction and reduction.
Background
Previous research on the smoker’s dilemma (e.g., Lee,
1988; Simmons, Webb, & Brandon, 2004) used rating
scales to examine risk perception among smokers. High
risk perception was regarded as cognitive dissonance,
whereas low risk perception was interpreted as a denial of
smoking-related risks and perceived immunity, indicating
a strategy to reduce cognitive dissonance. Gibbons, Eggles-
ton, and Benthin (1997) conducted a study in a smoking-
cessation program, using different questionnaires to assess
risk perception, commitment to quitting, and self-concept.
High risk perception was the primary reason for quitting.
This was attributed to increased cognitive dissonance (Gib-
bons, McGovern, & Lando, 1991; Gibbons et al., 1997;
Klesges et al., 1988; Lando, 1989).
Contrary to nonsmokers, smokers are highly involved
when considering smoking-related hazards that should
result in cognitive dissonance. Smoking risks as disso-
nance-enhancing information contradict health-related
self-concepts and can thus be regarded as inconsistent. It
is well known that the processing of inconsistent infor-
mation results in additional processing efforts and is
time-consuming (e.g., Hemsley & Marmurek, 1982;
Sherman, Lee, Bessenoff, & Frost, 1998). Most signifi-
cantly, Festinger (1964) concluded that “. . . dissonance
reduction does, indeed, require that time be spent in
thinking about the characteristics of the alternative”
(p. 59). In addition, higher involvement increases infor-
mation processing (Petty & Cacioppo, 1986). If people
are highly involved and relevant information is disso-
nance-enhancing, processing time should slow down.
Since arousal energizes all concurrent response tenden-
cies, while a task elicits few competing responses of low
strength, the effect of high arousal is expected to facili-
tate performance or information processing. Conversely,
when the dissonance-arousing topic concerns highly self-
relevant ones, such as risk perception of diseases, perfor-
mance is impaired (Pallak & Pittman, 1972; Waterman,
1969). Indeed, when a task elicits many competing re-
sponse tendencies, arousal energizes them all, so that per-
formance, such as fast response latencies, should be hin-
dered. Waterman (1969) was able to confirm that disso-
82 J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers
Social Psychology 2012; Vol. 43(2):81–91 © 2012 Hogrefe Publishing
nance arousal may alter performance (such as informa-
tion processing of inconsistent information) in ways un-
related to dissonance reduction. Thus, it can be expected
that with high-demanding or high-aversive (inconsistent)
information, where no dominant response (in terms of
dissonance reduction) is present, the effects of arousal
should disappear. In that case, no response facilitation
with dissonance should be observed (Martinie, Olive, &
Milland, 2010).
As mentioned above, dissonance needs to be reduced. If
smokers use thought suppression as a strategy to reduce
cognitive dissonance, processing time of related negative
information should decrease over time. This change of
arousal and reduction of dissonance can be measured via
response latencies (Fazio, 1990).
Compared to conventional methods of assessing cog-
nitive dissonance, response latencies provide some ad-
vantages:
1) Ratings might not be able to reflect changes in cognitive
dissonance because of people’s tendency to be seen as
stable and coherent in their attitudes and beliefs (Back-
man, 1988; Fazio, 1990; Tedeschi, Schlenker, & Bono-
ma, 1971).
2) The use of questionnaires and rating scales to examine
cognitive dissonance remains questionable because it is
not clear whether higher risk perception really indicates
a long-term state of cognitive dissonance. Response la-
tencies might reflect internal processes, indicating
whether changes in cognitive dissonance occur or
whether the aversive state is consistent.
3) In addition, response latencies are more implicit in com-
parison with rating data (Arcuri, Castelli, Boca, Loren-
zi-Cioldi, & Dafflon, 2001) and are less influenced by
social desirability and response bias (Fazio, 1990; Sher-
man, Rose, Koch, Presson, & Chassin, 2003).
However, response latencies have rarely been used to in-
vestigate cognitive dissonance, especially when investigat-
ing changes in cognitive dissonance or even the reduction
of this dissonance. This lack of research is striking because
it is already known that dissonance creates intrusive
thoughts that are irrelevant to the task in hand yet are re-
lated to the management or reduction of this discomfort.
These intrusive thoughts siphon off resources from the on-
going task (Engle, Tuholski, Laughlin, & Conway, 1999;
Rosen & Engle, 1998; Wegner, 1994). One could argue that
response latency is a necessary measure to account for im-
paired information processing due to inconsistent informa-
tion at the moment of dissonance arousal. Therefore, our
first two explorative studies are also meant to test response
latencies as a measure of cognitive dissonance. The main
purpose is to investigate whether response latencies are
more sensitive than questionnaires and rating scales and
whether smokers reduce cognitive dissonance directly as it
occurs. They are also used to test if they could contribute
to research concerning arousal patterns of dissonance and
its reduction.
Study 1
The first study assessed response latencies as well as rat-
ings to analyze arousal and changes in cognitive disso-
nance. We assessed smokers’ and nonsmokers’ risk percep-
tion by evaluating each of 12 smoking-related diseases on
a 10-point Likert-scale. Simultaneously, we recorded re-
sponse latencies for these ratings. To investigate whether
the dissonance aroused by the confrontation with personal
risks of developing a smoking-related disease is immedi-
ately reduced, we assessed the risk perception twice.
Based on previous results and our theoretical framework
outlined above we formulated 3 hypotheses and one addi-
tional research question:
Hypothesis 1: Because of the personal involvement of
smokers, the health risk ratings of typical smoking dis-
eases should be higher compared to nonsmokers
Hypothesis 2: Response latencies should increase for
smokers compared to nonsmokers.
People want to be seen as stable regarding their beliefs
(Backman, 1988; Fazio, 1990; Tedeschi et al., 1971). Be-
cause smokers’ conscious risk perception should not vary
over time, we further expected the following:
Hypothesis 3: There is no difference between the ratings
of the first and second test.
What is not known is how quickly smokers are able to sup-
press negative thoughts. Hence, we wanted to answer the
following research question:
Research Question 1: Do smokers suppress negative
thoughts to immediately reduce cognitive dissonance?
And is cognitive dissonance in smokers in fact immedi-
ately reduced by suppressing negative thoughts?
Differences in response latencies reflect the processing of
dissonant information (Newby-Clark et al., 2002). As men-
tioned above, thought suppression first requires the initial
activation of the concerned concepts. Smokers should re-
spond more slowly when first confronted with smoking-re-
lated negative health information. If cognitive dissonance-
reducing strategies are applied immediately after health-
damaging confrontation, we would expect differences
between the first and second assessment. If smokers are not
trained to suppress negative thoughts as a dissonance-re-
duction strategy, no differences in response latencies be-
tween the first and second testing phase should be found.
Method
Participants and Design
A group of 40 students (20 smokers and 20 nonsmokers)
from a German university participated in Study 1. We used
a 2 × 2 mixed design with the smoking status (smoker vs.
J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers 83
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
nonsmoker) as the between-subject factor and testing (T1
vs. T2) as the within-subject factor. Smokers had smoked
on average for 6.05 years, SD = 2.72, with a mean of 10.25,
SD = 5.68, cigarettes per day. Nonsmokers did not smoke
and had never smoked before.
Procedure
To activate the smoking-related self-concepts, participants
had to fill out a questionnaire assessing their smoking be-
havior, namely, the number of years they had been smoking
and the number of cigarettes they smoked on average per
day. The names of typical smoking-related diseases were
presented, in series, on a computer screen (random order)
for a maximum of 3000 ms. Participants responded by
pressing a number key on a range between 0 (no risk of
developing this disease) and 9 (highest risk of developing
this disease). To familiarize participants with the task, we
introduced two practice items at the beginning of each list.
Practice items were excluded from further analyses. Re-
sponses were assessed by using the number pad of the key-
board. Response latencies and rating data were recorded.
In an interpolated task, participants had to fill out a 9 × 9
Sudoku grid of medium complexity. They spent 10 minutes
on this task. After the interpolated task, participants again
rated their risk of developing the same diseases.
Materials
To construct valid stimulus materials, 20 students from a
German university rated 40 diseases regarding their depen-
dency on smoking on a Likert-Scale ranging from 1 to 7 (1
= not caused by smoking,7=caused by smoking). Six dis-
eases with means of 5.75–6.95, average SD = 0.71, were
included in the study. The diseases were rated as typically
stimulated by smoking and included, for instance, lung can-
cer and circulatory disorders.
Results
Rating Data
The rating data were submitted to a 2 (Smoking Status:
smoker vs. nonsmoker) × 2 (Testing: T1 vs. T2) mixed-
model analysis of variance with the last factor varying
within participants.
Analysis of variance revealed a significant main effect
of Smoking Status, F(1, 38) = 87.43, p < .001, η
2
p
= .69.
Smokers rated their risk of developing a smoking-related
disease higher (M = 6.60) than did nonsmokers (M = 2.90).
Neither the main effect of Testing, F(1, 38) = .83, ns, nor
the interaction between Smoking Status and Testing,
F(1, 38) = 1.89, ns, reached a significant level.
Response Latencies
To reduce skewness, response latencies were log-trans-
formed (Fazio, 1990); they were then submitted to a 2
(Smoking Status: smoker vs. nonsmoker) × 2 (Testing: T1
vs. T2) mixed-model ANOVA with repeated measure-
ments on the last factor.
Analysis of variance revealed a significant main effect
for Testing, F(1, 38) = 25.81, p < .001, η
2
p
= .40. Smokers
and nonsmokers responded faster during T2, as compared
to T1. ANOVA yielded a significant interaction between
Smoking Status and Testing, F(1, 38) = 4.24, p < .05, η
2
p
= .10. Planned simple effect tests, comparing response la-
tencies of smokers and nonsmokers, indicated that smokers
responded more slowly in T1 (M = 2597.18) than nonsmok-
ers (M = 2232.66), t (38) = 3.12, p < .01. This difference
diminished in T2. Smokers (M = 2075.93) responded as
fast as nonsmokers (M = 2040.31) during T2, t (38) = .70,
ns. Response latencies for smokers decreased in T2 com-
pared to T1, t (19) = 35.23, p < .001, while nonsmokers
response latencies did not change from T1 to T2, t (19) =
2.07, ns. The main effect of Smoking Status did not reach
significance, F(1, 38) = 3.23, ns.
Discussion
As assumed, smokers rated their risks higher than non-
smokers (Hypothesis 1). This suggests that cognitive dis-
sonance had been aroused by the confrontation with their
smoking-related health risks due to differences in involve-
ment between smokers and nonsmokers. According to the
rating data (Hypothesis 3), risk perception of both smokers
and nonsmokers did not vary over the time. Corresponding
to the rating results, it seems that smokers cannot suppress
their health risks. However, data regarding the response la-
tencies paints a slightly different picture: Smokers respond-
ed more slowly than nonsmokers during the first test (Hy-
pothesis 2), but these differences diminished during the
second test. Our findings support the assumption that cog-
nitive dissonance in smokers, induced by the confrontation
with smoking-related health risks, is only a short-term ef-
fect. Because the dissonance seems to be reduced immedi-
ately, one can conclude that smokers are able to suppress
the processing of negative information and to apply this
dissonance-reducing strategy whenever they are confront-
ed with health risks. Slower response latencies during the
first testing point to additional mental capacity needed for
intrusive thoughts (Engle et al., 1999; Rosen & Engle,
1998; Wegner, 1994) and competing response tendencies
(Pallak & Pittman, 1972; Waterman, 1969). Nonsmokers’
ratings and response latencies did not differ because they
were neither involved in nor exposed to cognitive disso-
nance in the first place. Response latencies seem to be able
to more precisely reflect subtle processes in cognitive dis-
sonance.
Our findings suggest that response latencies measure-
84 J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers
Social Psychology 2012; Vol. 43(2):81–91 © 2012 Hogrefe Publishing
ment is a useful method for examining changes in cognitive
dissonance. However, our first study is limited in some re-
spects:
1) Participants were confronted with the same 12 diseases
twice. One could speculate that reduced response laten-
cies of smokers during the second test were due to fa-
miliarity with the presented diseases (practice effects).
Although nonsmokers’ response latencies did not de-
crease, which argues against this assumption, one could
change the stimulus materials from T1 to T2 to rule out
any practice effects.
2) Furthermore, one could argue that cognitive dissonance
in the smokers was caused by general health risk percep-
tions, activated by ratings, and are not due to the specific
smoking risk because of the activation of smoking-relat-
ed self-concept. We conducted Study 2 in order to rule
out the assumption that Study 1 only reflects habituation
to the items, and that cognitive dissonance was due to
general health risk perception.
Study 2
Unlike in Study 1, in Study 2 the 12 diseases were separated
into two different lists. Additionally, we presented the same
number of other diseases not known to be related to smok-
ing. Hence, participants were confronted with different
smoking-related and other diseases during the first and the
second testing. In the case of successful replication of
Study 1, differences in response latencies between the first
and second testing cannot occur because of familiarity with
the target diseases, but rather must be attributed to a height-
ened arousal and subsequent reduction of cognitive disso-
nance. Furthermore, cognitive dissonance in smokers
should be restricted to smoking-related diseases only. Other
diseases should not be inconsistent with the activated
smoking-related self-concept; accordingly, the confronta-
tion with these diseases should not result in cognitive dis-
sonance due to the lack of personal involvement.
If the results of Study 1 are not due to familiarity with
the target diseases, Study 2 should reveal the same differ-
ences in ratings and response latencies. In addition, if cog-
nitive dissonance is due to the activation of the smoking-
related self-concept, differences between smoking-related
and other diseases should emerge. The comparison of dif-
ferent types of diseases allows us to investigate the sensi-
tivity of response latencies as a measure of cognitive dis-
sonance. Ratings and response latencies of smokers should
differ with respect to the different types of diseases. Find-
ings from Study 1 led to two additional hypotheses for
Study 2:
Hypothesis 4: Ratings should be higher for smoking-re-
lated diseases than for other diseases, but they should
not vary over time.
Hypothesis 5: There should be an interaction effect be-
tween type of disease and testing regarding response la-
tencies. Response latencies for smoking-related diseases
should increase at T1, while there should be no differ-
ence between different types of diseases at T2.
Method
Participants and Design
Twenty smokers were recruited from the students of a
German university. The study had a 2 (Type of Disease:
smoking-related versus other) × 2 (Testing: T1 versus T2)
within-subject design. Participants had smoked on aver-
age for 6.2 years, SD = 3.12, with a mean of 10.40, SD
= 5.48, cigarettes per day. Regarding smoking habits, the
sample of Study 2 was very similar to the smoker group
in Study 1.
Procedure
The procedure in Study 1 and Study 2 differed in three
aspects:
1) Two different lists (A and B) were needed for Study 2
in order to rule out effects caused by familiarity with the
target diseases. The 12 smoking-related and other dis-
eases were assigned randomly to each list. Participants
were confronted with one list before and with the other
after an interpolated task. Each list consisted of three
smoking-related and three other diseases. Whether list
A or list B appeared as first or second test was counter-
balanced.
2) As outlined above, nonsmokers do not experience cog-
nitive dissonance when confronted with smoking-relat-
ed health risks. Therefore, only smokers participated in
Study 2.
3) The names of smoking-related and non-smoking-related
diseases were presented to investigate whether aroused
cognitive dissonance was due to the confrontation with
smoking-related diseases and the activation of a smok-
ing-related self-concept or due to general health risk and
the activation of a health-related self-concept.
Participants first filled out the questionnaire assessing their
smoking behavior to activate the smoking-related self-con-
cept. The names of the smoking and other diseases were
presented in random order on a computer screen one after
another. Participants used the number pad and responded
by pressing a number key between 0 and 9. No risk was
represented by 0, and 9 indicated the highest risk of devel-
oping the disease. Response latencies and rating data were
recorded. After filling out a 9 × 9 Sudoku grid for 10 min-
utes, participants were confronted with the second list of
diseases with the same instructions. To familiarize partici-
pants with the task, we introduced two practice items at the
beginning of each testing. Practice items were excluded
from further analyses.
J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers 85
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
Materials
In the pretest described above in Study 1, 20 students rated
40 diseases with respect to their stimulation by smoking.
Six diseases with means from 1.15 up to 2.20, average SD
= 1.42, were rated as not related to smoking and were in-
cluded in Study 2 as other diseases (e.g., glaucoma or mul-
tiple sclerosis).
Results
Rating Data
Rating data were submitted to a 2 (Type of Disease: smok-
ing-related vs. other) × 2 (Testing: T1 vs. T2) analysis of
variance with both factors varying within participants.
ANOVA revealed a significant main effect for Type of dis-
ease, F(1, 19) = 85.94, p < .001, η
2
p
= .82. Smokers rated
their risk of developing a smoking-related disease higher
(M = 5.11) than their risk of developing another disease (M
= 1.77). Neither the main effect of testing, F(1, 19) = 1.72,
ns, nor the interaction between type of disease and testing,
F(1, 19) = 0.38, ns, reached significance.
Response Latencies
Log-transformed response latencies were submitted to a 2
(Type of Disease: smoking-related vs. other) × 2 (Testing:
T1 vs. T2) within-model analysis of variance. The ANOVA
revealed a significant interaction between type of disease
and testing, F(1, 19) = 5.82, p < .05, η
2
p
= .23. Planned
simple effect tests indicated that smokers responded faster
to other diseases (M = 2166.50) compared to smoking-re-
lated diseases (M = 2561.5) at T1, t (19) = 2.37, p < .05. In
T2 these differences diminished. Responses to smoking-re-
lated diseases (M = 2128.40) did not differ from those of
other diseases (M = 2060.20), t (19) = .16, ns. Participants
responded to smoking-related diseases inT2 faster than at
T1, t (19) = 3.59, p < .01. Neither the main effect of type
of disease (F(1, 19) = 3.73, ns), nor the main effect for
testing (F(1, 19) = 3.87, ns.) did reach significance.
Discussion
Differences in response latencies between types of diseases
were found only in T1 but not in T2; there were no differ-
ences in ratings at T1 and T2. Thus, the findings in Study
2 replicated the results of Study 1, allowing us to to rule
out the familiarity assumption. Response latencies changed
during testing because of the immediate reduction of cog-
nitive dissonance and not because of familiarity with the
items. This suggests that changes in response latencies be-
tween the first and second assessment were not due to prac-
tice or familiarity. Results also show that induction of cog-
nitive dissonance resulted from the confrontation with
smoking-related health risks and the activation of the
smoking-related self-concept. Smokers seem to be effec-
tive in suppressing negative thoughts about the health risk
and use this strategy to reduce cognitive dissonance imme-
diately when confronted with the health risks.
However, there are some limitations with regard to the
arousal properties and dissonance-reduction strategies in
Study 1 and Study 2. Differences in response latencies be-
tween smokers and nonsmokers are explained so as to re-
veal implicit dissonance-reduction strategies such as
thought suppression. This new technique to assess disso-
nance arousal and its reduction should be underlined by
comparing it with a more established implicit measurement
of cognitive dissonance. It could be argued from Study 1
and Study 2 that an increase in response latencies during
the second testing was due to a habituated processing of
inconsistent information, and that altogether no dissonance
occurred during the first testing. We regard dissonance to
occur during the first testing, and its reduction, namely,
thought suppression, to take place during second testing.
As explained above, recent research used psychophysio-
logical arousal to assess dissonance reduction. We conduct-
ed a third study to clarify the open question of whether
dissonance arousal and its immediate reduction actually
cause the differences in response latencies. As for issues of
the operation mode, to measure psychophysiological
arousal, one hand needs to remain motionless (namely, the
hand to which the sensor is attached), and participants can
only use the other hand for responses. Because of technical
requirements, assessing response latencies at the same time
as psychophysiological arousal would bias effects on re-
sponse latencies. Therefore, we assessed ratings and psy-
chophysiological arousal, but not response latencies, in
Study 3.
Study 3
Six smoking-related diseases were separated into two dif-
ferent lists, though we did not show diseases unrelated to
smoking. This procedure was chosen because arousal data
are known to produce errors if stimulus materials are inco-
herent. The use of a between-subject design to apply psy-
chophysiological measures was important in order to en-
sure that dissonance arousal was induced by the experi-
mental manipulation (Losch & Cacioppo, 1990). We
recorded galvanic skin response (GSR) to measure psycho-
physiological arousal during perception and evaluation of
smoking-related diseases for smokers and nonsmokers.
Electrodermal activity was chosen as a coherent psycho-
physiological indicator because it has an especially sensi-
tive information-processing capacity for stimulus and thus
best reflects data obtained with response latencies (Hop-
kins & Fletcher, 1994; Ravaja, 2004). Given the suggested
association of electrodermal activity with several psycho-
86 J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers
Social Psychology 2012; Vol. 43(2):81–91 © 2012 Hogrefe Publishing
logical processes, it has sometimes been criticized as not
being a clearly interpretable measure of any particular psy-
chological process. However, the psychological meaning
of an electrodermal response becomes interpretable by tak-
ing into account the stimulus condition or experimental
paradigm in which the response occurred (Dawson, Schell,
& Filion, 2000). In order to infer the psychophysiological
process mediating the resultant electrodermal response,
one has to monitor experimental conditions carefully (Ra-
vaja, 2004). For example, arousing pictures result in in-
creased electrodermal activity compared to a decrease with
low-arousal pictures. Mixing smoking-related diseases and
nonsmoking-related diseases is therefore likely to even out
the effects on the arousing properties of the experience of
dissonance in smokers. Hence, nonsmokers also participat-
ed as a control for dissonance effects. Participants received
three different smoking-related diseases during the first and
the second testing in random order. If smokers experience
dissonance during first testing but not during second testing
(which is assumed on the basis of the first two experi-
ments), we would expect the following hypothesis:
Hypothesis 6: The rise in psychophysiological arousal
should decrease in T2 compared to T1 for smokers. Con-
frontation with smoking-related diseases for nonsmok-
ers should not result in cognitive dissonance and the rise
in psychophysiological arousal should be the same at T1
and T2 due to lack of personal involvement.
Method
Participants and Design
Ten smokers and ten nonsmokers were recruited from stu-
dents of a German university. The study had a 2 (Smoking
Behavior: smoker versus nonsmoker) × 2 (Testing: T1 ver-
sus T2) mixed design. Participants had smoked on average
for 10.90 years, SD = 3.03, with a mean of 12.50, SD =
6.54, cigarettes per day.
Procedure
The procedure of Study 3 was similar to that of Study 1 and
Study 2. Participants first filled out the questionnaire as-
sessing their smoking behavior to activate the smoking-re-
lated self-concept. The names of three smoking diseases
were presented in random order on a computer screen one
after another during first and second testing runs. GSR was
measured with the Wild Divine IOM Lightstone Biometrics
USB Widget. This device provides two hard plastic clips
to measure GSR. The device is connected to the computer
via USB. Participants used the number pad and responded
by pressing a number key between 0 and 9. After filling out
a 9 × 9 Sudoku grid for 10 minutes, participants were con-
fronted with the second list of diseases under the same con-
ditions.
Preprocessing of Data
The GSR mean for each disease was aggregated for pre-
sentation duration. GSR measures for the first disease of
each participant were excluded for T1 and T2 to account
for orientation reaction. Orientation reactions can lead to a
peak in psychophysiological arousal independent of con-
tent of stimulus material (Ravaja, 2004). GSR during the
first listed diseases may not be interpreted in terms of cog-
nitive dissonance for smokers and nonsmokers. Supporting
the evidence that dissonance has energetic properties that
increase individuals’ arousal, one might also expect that
arousal decreases because dissonance is a psychological
aversive experience that prompts individuals to seek ways
of reducing it. Beauvois and Joule (1996) reported that
arousal, due to dissonance induction, already started to de-
cline after 5 minutes. Therefore, differences in psycho-
physiological arousal between the third and second diseas-
es were computed as a measurement for an increase, or
decrease, respectively, in psychophysiological arousal.
Results
Ratings
Rating data were analyzed via a 2 (Smoking Status: smoker
vs. nonsmoker) × 2 (Testing: T1 vs. T2) mixed-model anal-
ysis of variance with the last factor as repeated measure-
ment.
We found a significant main effect of Smoking Status,
F(1, 19) = 25.69, p < .001, η
2
p
= .57. As in Study 1 and
Study 2, smokers (M = 6.05) rated their risk of developing
a smoking-related disease higher than nonsmokers did (M
= 2.78). Neither the main effect of Testing, F(1, 19) = .45,
ns, nor the interaction between Smoking Status and Test-
ing, F(1, 19) = 1.89, ns, reached significance.
Psychophysiological Arousal
Data were submitted to a 2 (Smoking Status: smokers vs.
nonsmokers) × 2 (Testing: T1 vs. T2) analysis of variance
with repeated measurement of the second factor. ANOVA
revealed a significant main effect for Smoking Status,
F(1, 18) = 7.11, p < .05, η
2
p
= .28. Smokers showed signif-
icantly more increase in arousal (M = 1961.50) than did
nonsmokers (M = 334.09). The main effect for Testing
reached significant levels, F(1, 18) = 5.10, p < .05, η
2
p
=
.22. Increase in arousal was higher during T1 (M =
1564.69) than during T2 (M = 730.90). The interaction be-
tween Smoking Status and Testing did not reach a signifi-
cant level, F(1, 18) = 1.41, p > .29, η
2
p
= .06, but planned
simple effect tests confirmed H6. Smokers showed higher
increase in arousal during T1 (M = 2575.63) than during
T2 (M = 1347.37), t (10) = 2.75, p < .05. Nonsmokers did
not differ during T1 (M = 553.75) and T2 (M = 114.43), t
J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers 87
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
(10) = 0.75, ns. During T1 we found a higher increase in
arousal for smokers compared to nonsmokers, t (18) = 3.22,
p < .01. This effect diminished during T2, t (10) = 1.56, ns.
Discussion
Rating data supported our findings from Study 1 and Study
2. Smokers are conscious about their health-damaging be-
havior. Cognitive dissonance-reduction processes were not
observed regarding explicit measurements. However, as in
Study 1 and Study 2, we did find effects for implicit mea-
surements. Differences in psychophysiological arousal be-
tween smokers and nonsmokers were found only for T1.
This effect diminished in T2, thus supporting the findings
on response latencies in Study 1 and Study 2. Results agree
with the findings of Martinie et al. (2010), who showed
that, when arousal is highly increased, information process-
ing is impaired due to a lack of available resources. Results
also showed that the arousal properties found for disso-
nance arousal, diminished after a first cognitive involve-
ment with the dissonant arousing topic (Martinie et al.,
2010). Our findings support the notion that during disso-
nance arousal, galvanic skin response is higher and infor-
mation processing of inconsistent information is impaired.
This higher galvanic skin response is not found for non-
smokers since they lack personal involvement with the dis-
played diseases. Since effects already decrease during sec-
ond testing, it can be assumed that an implicit dissonance
reduction, namely, thought suppression, has taken place.
General Discussion
We conducted three studies to investigate whether smokers
are able to reduce cognitive dissonance immediately, or
whether the state of dissonance persists. Because smoking
is a health-damaging behavior, smokers must use disso-
nance-reducing strategies other than, for example, attitude
change in order to maintain their behavior.
We used response latencies as well as ratings for the
investigation of changes regarding an increase or reduction
of cognitive dissonance. Cognitive dissonance arises when
two cognitions are inconsistent with each other. In the first
two studies, we confronted smokers with their risk of de-
veloping a smoking-related disease. The ratings and re-
sponse latencies of risk perception were measured twice.
Previous research indicated that smokers experience cog-
nitive dissonance when confronted with the harmful con-
sequences of their behavior (e.g., Gibbons et al., 1997). Our
results support these findings. Both the ratings and re-
sponse latencies indicated the arousal of cognitive disso-
nance. During the second testing, differences between re-
sponse latencies and ratings emerged: Ratings remained
constant, whereas response latencies changed. Response
latencies implied that smokers reduce cognitive dissonance
immediately after being confronted with their health risks
via thought suppression. First, response latencies increase
for negative smoking-related information due to personal
involvement and reduced processing time of dissonance
enhancing information; then thought suppression as disso-
nance-reduction strategy leads to decreased information
processing and changes in response latencies. These differ-
ences in response latencies suggest applying implicit mea-
sures to assess changes in cognitive dissonance. Our find-
ings agree with Lang’s (2006) proposition that aversive ac-
tivation reaches a point at which there is some defensive
withdrawal of cognitive resources from encoding a mes-
sage. At T1, the point when smoking-related diseases were
first encoded, response latencies for smoking-related dis-
eases increased due to additional resources constrained by
thought suppression. In general, negative messages capture
more attentional resources compared with neutral or even
positive messages (Bolls, Lang, & Potter, 2001). To fully
connect this novel technique to assessing implicit disso-
nance-reduction strategies, physiological arousal was mea-
sured using the same paradigm as for the response laten-
cies. This research agrees with the findings of Lieberman,
Ochsner, Gilbert, and Schacter (2001), who also concluded
that the processing components of cognitive dissonance
need clearer specification insofar as the dependence of cog-
nitive dissonance theory on both explicit memory measures
and conscious processing should yield clearer specifica-
tion. Also, it should be figured out what effect exactly ac-
counts for an increase in response latencies from disso-
nance arousal. Further studies should address whether
mental capacity is occupied because of intrusive thoughts
(Engle et al., 1999; Rosen & Engle, 1998; Wegner, 1994)
or because of competing response tendencies (Pallak &
Pittman, 1972; Waterman, 1969) – or whether a confound-
ing of several strategies explains the effect.
There are four main results regarding response latencies:
1) As an implicit method, response latencies can illustrate
changes in cognitive dissonance processes of smokers.
2) Response latencies help to distinguish between smokers
who experience cognitive dissonance in terms of smok-
ing-related health risks and nonsmokers who do not.
3) This measure allows discriminating between the pro-
cessing of smoking-related health risks, which induces
cognitive dissonance, and the processing of other infor-
mation which does not.
4) GSR as a measure for psychophysiological properties of
cognitive dissonance underlines the effects found in re-
sponse latencies measures, thus accounting for the value
of a new methodology to assess implicit dissonance re-
duction.
The simultaneous application of explicit and implicit meth-
ods can provide a deeper insight into cognitive dissonance
processes. Understanding the changes in cognitive disso-
nance processes can inform strategies to positively influ-
ence attitudes and behavior. However, further research
should include both implicit and explicit measures of dis-
88 J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers
Social Psychology 2012; Vol. 43(2):81–91 © 2012 Hogrefe Publishing
sonance reduction, such as combining attitude change (Fes-
tinger & Carlsmith, 1959), act rationalization (Beauvois,
Joule, & Brunetti, 1993), trivialization (Simon, Greenberg,
& Brehm, 1995), and denial of responsibility (Gosling, De-
nizeau, & Oberlé, 2006) with implicit measures such as
arousal (Martinie et al., 2010) or response latencies (Glock,
2010).
Our explorative studies are limited in several other ways.
First, we discovered cognitive dissonance changes, but it
remains unclear what specific reduction strategies smokers
actually apply. In addition, we analyzed only health-related
risk perceptions, but people do not smoke in order to be-
come ill: They emphasize positive outcomes and learn to
ignore their health risks (Van Harreveld et al., 1999). Other
studies revealed a decreased risk perception when people
are confronted with a dissonance-arousing topic (e.g., Sher-
man, Nelson, & Steele, 2000). Nevertheless, one could ar-
gue that these results are not comparable with those found
in our studies since we did not include self-affirmation ma-
nipulations in the study. Instead of concluding that Sher-
man et al. (2000) found a decreased risk perception in non-
affirmed participants who experienced dissonance, it is
rather more the case that affirmed participants overevalu-
ated their risk percentage (Sherman et al., 2000). Further-
more, the authors themselves claim that the chosen para-
digm, for instance, coffee-drinkers and the likelihood of
getting fibrocystic disease, is neither as salient to a college-
age population as are other health concerns, nor is it a topic
they are commonly confronted with – as opposed to the
smoking-related issues smokers are up against every day
(Glock, 2010).
Regarding smoking prevention and intervention, health-
related information could be useful for nonsmokers to re-
main so because it confirms the health-related benefits of
nonsmoking. For smokers, health-related information
about smoking seems only to result in cognitive disso-
nance-reduction processes, leading to the acceptance of the
smoking-related health risk because smoking has positive
aspects (Hendricks & Brandon, 2005; Mullenix, Kilbey,
Fisicaro, Farnsworth, & Torrento, 2003) and because
smokers use various dissonance-reducing strategies in or-
der not to change their behavior. Further research should
focus on the variety of strategies smokers use to reduce
dissonance in order to determine which information could
be useful for the elicitation of a persistent state of cognitive
dissonance that cannot be easily reduced. In such cases,
quitting would be the only way to stop this unpleasant state
of mind.
Our results have implications for intervention and pre-
vention programs focusing on the health-damaging aspects
of smoking. Such programs try to prevent people from
smoking by inducing cognitive dissonance and fear by em-
phasizing the health consequences of smoking, for in-
stance, through warning labels on cigarette packages. Our
data imply that these programs may fail because of the
short-lived cognitive dissonance induced through confron-
tation with the health risks. Smokers feel cognitive disso-
nance, think about their health-risk, and reduce cognitive
dissonance immediately. Thus, inducing cognitive disso-
nance through health-related warning labels may not influ-
ence actual smoking behavior because dissonance is re-
duced immediately. Therefore, warning labels and other
programs focusing on the health-related aspects of smoking
may not be sufficient to motivate smokers to quit. Future
investigation is necessary to test whether prevention and
intervention programs should include different negative as-
pects. As mentioned above, smokers do not smoke to dam-
age their health but rather to achieve positive outcomes, for
example, stress reduction or social advantages. Threaten-
ing these outcomes might induce longer lasting cognitive
dissonance than health-related information and therefore
could help to change smoking behavior.
References
Arcuri, L., Castelli, L., Boca, S., Lorenzi-Cioldi, F., & Dafflon,
A. (2001). Fuzzy gender categories: How emotional expres-
sion influences typicality. Swiss Journal of Psychology, 60,
179–191. doi: 10.1024//1421-0185.60.3.179
Aronson, E. (1969). The theory of cognitive dissonance: A cur-
rent perspective. In L. Berkowitz (Ed.), Advances in experi-
mental social psychology (Vol. 4, pp. 1–34). San Diego, CA:
Academic Press.
Aronson, E. (1992). The return of the repressed: Dissonance the-
ory makes a comeback. Psychological Inquiry, 3, 303–311.
doi: 10.1207/s15327965pli0304_1
Backman, C. W. (1988). The self: A dialectical approach. In L.
Berkowitz (Ed.), Advances in experimental social psychology
(Vol. 21, pp. 229–260). New York: Academic Press.
Bargh, J. A. (1989). Conditional automaticity: Varieties of auto-
matic influence in social perception and cognition. In J. S. Ule-
man & J. A. Bargh (Eds.), Unintended thought (pp. 3–51).
New York: Guilford.
Beauvois, J.-L., & Joule, R. V. (1996). A radical dissonance the-
ory. London: Taylor & Francis.
Beauvois, J.-L., Joule, R.V., & Brunetti, F. (1993). Cognitive ra-
tionalization and act rationalization in escalation of commit-
ment. Basic and Applied Social Psychology, 14, 1–17. doi:
10.1207/s15324834basp1401_1
Bolls, P.D., Lang, A., & Potter, R.F. (2001). The effects of mes-
sage valence and listener arousal on attention, memory and
facial muscular responses to radio. Communication Research,
28, 627–651.
Cook, D., Pallak, M., Storms, M., & McCaul, K. (1977). The
effect of forced compliance on attitude change and behavior
change. Personality and Social Psychology Bulletin, 3, 71–74.
doi: 10.1177/014616727600300110
Cooper, J. (1992). Dissonance and the return of the self-concept.
Psychological Inquiry, 3, 320–323. doi: 10.1207/s15327965
pli0304_5
Cooper, J., & Fazio, R. H. (1984). A new look at dissonance the-
ory. In L. Berkowitz (Ed.), Advances in experimental social
psychology (Vol. 17, pp. 229–260). San Diego, CA: Academic
Press.
Croyle, R. T., & Cooper, J. (1983). Dissonance arousal: Physio-
J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers 89
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
logical evidence. Journal of Personality and Social Psycholo-
gy, 45, 782–791. doi: 10.1037//0022-3514.45.4.782
Dawson, M. E., Schell, A. M., & Filion, D. L. (2000). The elec-
trodermal system. In J. T. Cacioppo, L. G. Tassinary, & G. G.
Berntson (Eds.), Handbook of psychophysiology (2nd ed.,
pp. 200–223). New York: Cambridge University Press.
Elkin, R., & Leippe, M. (1986). Physiological arousal, dissonance,
and attitude change: Evidence for a dissonance-arousal link and
a “don’t remind me” effect. Journal of Personality and Social
Psychology, 51, 55–65. doi: 10.1037//0022–3514.51.1.55
Elliot, A.J., & Devine, P. G. (1994). On the motivational nature
of cognitive dissonance: Dissonance as psychological discom-
fort. Journal of Personality and Social Psychology, 67,
382–394. doi: 10.1037//0022-3514.67.3.382
Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. A.
(1999). Working memory, short-term memory and general flu-
id intelligence. A latent variable approach. Journal of Experi-
mental Social Psychology: General, 128, 309–331. doi:
10.1037//0096-3445.128.3.309
Fazio, R. H. (1990). A practical guide to the use of response la-
tencies in social psychological research. In C. Hendricks &
M. S. Clark (Eds.), Research methods in personality and social
psychology (pp. 74–97). Newbury Park, CA: Sage.
Festinger, L. (1957). A theory of cognitive dissonance. Evanston,
IL: Row, Peterson.
Festinger, L. (1964). Conflict, decision, and dissonance. Stanford,
CA: Stanford University Press.
Festinger, L., & Carlsmith, J.M. (1959). Cognitive consequences
of forced compliance. Journal of Abnormal Social Psychology,
58, 203–210. doi: 10.1037/h0041593
Gibbons, F. X., Eggleston, T. J., & Benthin, A. C. (1997). Cogni-
tive reactions to smoking relapse: The reciprocal relation be-
tween dissonance and self-esteem. Journal of Personality and
Social Psychology, 72, 184–195. doi: 10.1037//0022-3514.
72.1.184
Gibbons, F. X., Gerrard, M., Lando, H. A., & McGovern, P. G.
(1991). Social comparison and smoking cessation: The role of
the typical smoker. Journal of Experimental Social Psycholo-
gy, 27, 239–258. doi: 10.1016/0022-1031(91)90014-W
Gibbons, F. X., McGovern, P. G., & Lando, H. A. (1991). Relapse
and risk perception among members of a smoking cessation
clinic. Health Psychology, 10, 42–45. doi: 10.1037//0278-
6133.10.1.42
Glock, S. (2010). Rauchen ist uncool! Die Repräsentation der ver-
schiedenen Aspekte des Rauchens im Selbstkonzept und ihre Be-
deutung für die Veränderung des Verhaltens [Smoking is uncool!
The smoking self-concept and the impact of different smoking
aspects on behavior change]. Hamburg: Dr. Kovac.
Gosling, P., Denizeau, M., & Oberlé, D. (2006). Denial of respon-
sibility: A new mode of dissonance reduction. Journal of Per-
sonality and Social Psychology, 90, 722–733. doi: 10.1037/
0022-3514.90.5.722
Hemsley, G. D., & Marmurek, H. H. C. (1982). Person memory:
The role of consistent and inconsistent person information.
Personality and Social Psychology Bulletin, 8, 433–438. doi:
10.1177/0146167282083007
Hendricks, P. S., & Brandon, T. H. (2005). Smoking expectancy
associates among college smokers. Addictive Behaviors, 30,
235–245. doi: 10.1016/j.addbeh.2004.05.006
Hopkins, R., & Fletcher, J. E. (1994). Electrodermal measure-
ment: Particularly effective for forecasting message influence
on sales appeal. In A. Lang (Ed.), Measuring psychological
responses to media messages (pp. 113–132). Hillsdale, NJ:
Erlbaum.
Jones, E. E. (1976). Forward. In R. A. Wicklund & J. W. Brehm
(Eds.), Explorations in cognitive dissonance (pp. x–xii). Hills-
dale, NJ: Erlbaum.
Klesges, R. C., Somes, G., Pascale, R. W., Klesges, L. M., Murphy,
M., Brown, K., & Williams, E. (1988). Knowledge and beliefs
regarding the consequences of cigarette smoking and their rela-
tionships to smoking status in a biracial sample. Health Psychol-
ogy, 7, 384–401. doi: 10.1037/0278-6133.7.5.387
Kneer, J., & Glock, S. (2008). Werden Raucher von Warnhinwei-
sen auf Zigarettenschachteln beeinflusst? Eine sozialpsycho-
logische Perspektive. [Is there any impact of cigarette warning
labels on smokers? A perspective from social psychology].
Zeitschrift für Sozialmanagement, 6, 99–108.
Lando, H. A. (1989). Treatment outcome evaluation methodology
in smoking cessation: Strengths and key issues. Advances in
Behavior Research Therapy, 11, 201–214. doi: 10.1016/0146-
6402(89)90023-4
Lang, A. (2006). Using the limited capacity model of motivated
mediated message processing to design effective cancer com-
munication messages. Journal of Communication, 56, 57–80.
doi: 10.1111/j.1460-2466.2006.00283.x
Lee, C. (1988). Perceptions of immunity to disease in adult smok-
ers. Journal of Behavioral Medicine, 12, 267–277. doi:
10.1007/BF00844871
Lieberman, M.D., Ochsner, K.N., Gilbert, D.T., & Schacter, D.L.
(2001). Do amnesics exhibit cognitive dissonance reduction?
The role of explicit memory and attention in attitude change.
Psychological Science, 12, 135–140. doi: 10.1111/1467-9280.
00323
Logan, G. D. (1988). Toward an instance theory of automatiza-
tion. Psychological Review, 95, 492–527. doi: 10.1037//0033-
295X.95.4.492
Losch, M. E., & Cacioppo, J. T. (1990). Cognitive dissonance may
enhance sympathetic tonus, but attitudes are changed to reduce
negative affect rather than arousal. Journal of Experimental
Social Psychology, 26, 289–304. doi: 10.1016/0022-1031 (90)
90040-S
Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1994). Out
of mind, back in sight: Stereotypes on the rebound. Journal of
Personality and Social Psychology, 67, 808–817. doi:
10.1037/0022-3514.67.5.808
Mann, L., Janis, I., & Chaplin, R. (1969). Effects of an anticipa-
tion of forthcoming information on predecisional processes.
Journal of Personality and Social Psychology, 11, 10–16. doi:
10.1037/h0026967
Martinie, M.-A., Olive, T., & Milland, L. (2010). Cognitive dis-
sonance induced by writing a counterattitudinal essay facili-
tates performance on simple tasks but not on complex tasks
that involve working memory. Journal of Experimental Social
Psychology, 46, 587–594. doi: 10.1016/j.jesp.2009.10.018
Monteith, M. J., Sherman, J. W., & Devine, P. G. (1998). Suppres-
sion as a stereotype control strategy. Personality and Social Psy-
chology Review, 2, 63–82. doi: 10.1207/s15327957pspr0201_4
Monteith, M. J., Spicer, C. V., & Tooman, G. D. (1998). Conse-
quences of stereotype suppression: Stereotypes on and not on
the rebound. Journal of Experimental Social Psychology, 34,
355–377. doi: 10.1006/jesp.1998.1355
Mullenix, J. W., Kilbey, M.M., Fisicaro, S. A., Farnsworth, L. M.,
90 J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers
Social Psychology 2012; Vol. 43(2):81–91 © 2012 Hogrefe Publishing
& Torrento, R. (2003). The representation of smoking expec-
tancies by single-word adjectives and the relationship to smok-
ing characteristics. Nicotine and Tobacco Research, 5,
682–694. doi: 10.1080/1462220031000158654
Newby-Clark, I.R., McGregor, I., & Zanna, M. P. (2002). Think-
ing and caring about cognitive inconsistency: When and for
whom does attitudinal ambivalence feel uncomfortable? Jour-
nal of Personality and Social Psychology, 82, 157–166. doi:
10.1037//0022-3514.82.2.157-166
Pallak, M., Brock, T., & Kiesler, C. (1967). Dissonance arousal
and task performance in an incidental verbal learning para-
digm. Journal of Personality and Social Psychology, 7, 11–20.
doi: 10.1037/h0024894
Pallak,M., & Kiesler, C. (1968). Dissonance arousal, task evaluation
and task performance. Psychonomic Science, 11, 197–198.
Pallak, M., & Pittman, T. (1972). General motivational effects of
dissonance arousal. Journal of Personality and Social Psy-
chology, 21, 349–358. doi: 10.1037/h0032266
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood
model of persuasion. In L. Berkowitz (Ed.), Advances in ex-
perimental social psychology (Vol. 19, pp. 123–205). New
York: Academic Press.
Plant,E. A., & Devine, P. G. (1998). Internal andexternal motivation
to respond without prejudice. Journal of Personality and Social
Psychology, 75, 811–832. doi: 10.1037//0022-3514.75.3.811
Ravaja, N. (2004). Contributions of psychophysiology to media
research: Review and recommendations. Media Psychology, 6,
193–235. doi: 10.1207/s1532785xmep0602_4
Rosen, V., & Engle, R.W. (1998). Working memory capacity and
suppression. Journal of Memory and Language, 39, 418–436.
doi: 10.1006/jmla.1998.2590
Salkovskis, P. M., & Reynolds, M. (1994). Thought suppression
and smoking cessation. Behavioral Research and Therapy, 32,
193–201. doi: 10.1016/0005-7967(94)90112-0
Shallice, T. (1972). Dual functions of consciousness. Psycholog-
ical Review, 79, 383–393. doi: 10.1037/h0033135
Sherman D. A. K., Nelson L. D., & Steele, C. M. (2000). Do mes-
sages about health risks threaten the self? Increasing the ac-
ceptance of threatening health messages via self-affirmation.
Personality and Social Psychology Bulletin, 26, 1046–1058.
doi: 10.1177/01461672002611003
Sherman, J.W., Lee, A.Y., Bessenoff, G.R., & Frost, L.A.
(1998). Stereotype efficiency reconsidered: Encoding flexibil-
ity under cognitive load. Journal of Personality and Social
Psychology, 75, 589–606. doi: 10.1037//0022-3514.75.3.589
Sherman, S. J., Rose, J. S., Koch, K., Presson, C. C., & Chassin,
L. (2003). Implicit and explicit attitudes toward cigarette
smoking: The effects of context and motivation. Journal of
Social and Clinical Psychology, 22, 13–39. doi: 10.1521/
jscp.22.1.13.22766
Simmons, V. N., Webb, M.S., & Brandon, T. H. (2004). College-
student smoking: An initial test of an experimental dissonance-
enhancing intervention. Addictive Behaviors, 29, 1129–1136.
doi: 10.1016/j.addbeh.2004.03.005
Simon, L., Greenberg, G., & Brehm, J. (1995). Trivialization: The
forgotten mode of dissonance reduction. Journal of Personal-
ity and Social Psychology, 68, 247–260. doi: 10.1037//0022-
3514.68.2.247
Statistisches Bundesamt. (2010). Pressemitteilung Nr. 190 vom
28.05.2010: Drei Viertel der Bevölkerung sind Nichtraucher
[Press release No. 190, May, 28, 2010: Three quarters of the
German Population are nonsmokers]. Retrieved from
http://www.destatis.de/jetspeed/portal/cms/Sites/destatis/Inte
rnet/DE/Presse/pm/2010/05/PD10__190__239,templateId=
renderPrint.psml
Steele, C. M. (1988). The psychology of self-affirmation: Sustain-
ing the integrity of the self. In L. Berkowitz (Ed.), Advances
in experimental social psychology (pp. 261–302). New-York:
Academic Press.
Stone, J., & Cooper, J. (2001). A self-standards model of cogni-
tive dissonance. Journal of Experimental and Social Psychol-
ogy, 37, 228–243. doi: 10.1006/jesp.2000.1446
Tedeschi, J. T., Schlenker, B. R., & Bonoma, T. V. (1971). Cognitive
dissonance: Private ratiocination or public spectacle? American
Psychologist, 26, 685–695. doi: 10.1037/h0032110
Van Harreveld, F., Van der Pligt, J., & De Vries, N. K. (1999). Atti-
tudes toward smoking and the subjective importance of attri-
butes: Implications for changing risk-benefit ratios. Swiss Jour-
nal of Psychology, 58, 65–72. doi: 10.1024//1421-0185.58.2.65
Visser, P. S., & Cooper, J. (2003). Attitude change. In M. A. Hogg
& J. Cooper (Eds.), The Sage handbook of social psychology
(pp. 211–231). London: Sage.
Waterman, C. K. (1969). The facilitating and interfering effects
of cognitive dissonance on simple and complex paired-asso-
ciate learning tasks. Journal of Experimental Social Psychol-
ogy, 5, 31–42. doi: 10.1016/0022-1031(69)90004-3
Wegner, D. M. (1994). Ironic processes of mental control. Psy-
chological Review, 101, 34–52. doi: 10.1037//0033-295X.101.
1.34
Wegner, D. M., & Erber, R. (1992). The hyperaccessibility of sup-
pressed thoughts. Journal of Personality and Social Psychol-
ogy, 63, 903–912. doi: 10.1037//0022-3514.63.6.903
Wenzlaff, R. M., & Wegner, D. M. (2000). Thought suppression.
Annual Review of Psychology, 51, 59–91. doi: 10.1146/annurev.
psych.51.1.59
Received August 18, 2010
Final revision received February 23, 2011
Accepted March 17, 2011
Julia Kneer
Department Psychology
University of Cologne
Richard-Strauss-Str. 2
50931 Cologne
Germany
Tel. +49 221 470-6512
Fax +49 221 470-5299
E-mail jkneer@uni-koeln.de
J. Kneer & S. Glock: Reduction of Cognitive Dissonance in Smokers 91
© 2012 Hogrefe Publishing Social Psychology 2012; Vol. 43(2):81–91
A preview of this full-text is provided by Hogrefe Publishing.
Content available from Social Psychology
This content is subject to copyright. Terms and conditions apply.
