Personality and Social Psychology Bulletin

http://psp.sagepub.com

The Role of Self-Control in Resistance to Persuasion

Edward Burkley
Pers Soc Psychol Bull 2008; 34; 419
DOI: 10.1177/0146167207310458

The online version of this article can be found at:

http://psp.sagepub.com/cgi/content/abstract/34/3/419

Published by:

http://www.sagepublications.com

On behalf of:

Society for Personality and Social Psychology, Inc.

Additional services and information for Personality and Social Psychology Bulletin can be found at:

Email Alerts: http://psp.sagepub.com/cgi/alerts

Subscriptions: http://psp.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations (this article cites 29 articles hosted on the

SAGE Journals Online and HighWire Press platforms):
http://psp.sagepub.com/cgi/content/refs/34/3/419

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 The Role of Self-Control in Resistance
to Persuasion
Edward Burkley
University of North Carolina at Chapel Hill

Four studies investigated a self-control theory of resis- Self-Control as a Limited and General Resource
tance to persuasion. This theory asserts that resistance to
persuasion requires and consumes self-control resources. Often people do not want to be persuaded. This fact
Study 1 showed that resistance to a persuasive message has prompted researchers to explore the means by which
reduced the ability to engage in a subsequent self- people resist (e.g., Petty & Cacioppo, 1986a) and the
control task. Studies 2 and 3 showed that self-control strategies that can increase resistance (e.g., McGuire,
depletion leads to increased persuasion. Study 4 showed 1964). Valuable insights have been gained from these
that self-control depletion increased persuasion, partic- lines of research, but a very basic fact at times has failed
ularly under effortful resistance (i.e., strong arguments). to shine through: Resistance to persuasion can be effort-
Together, these findings suggest that self-control plays a ful. Resistance is not simply a cognitive process that
vital role in the process of resistance to persuasion. occurs under specified conditions. Resistance to persua-
People must have self-control resources to fend off per- sion is also a response that occurs when a person is suf-
suasive appeals; without them, they become susceptible ficiently motivated to work against an agent of influence.
to influence. If resistance is an effortful, goal-directed action, it likely
requires self-control.
Keywords: self-control; ego depletion; self-regulation; resistance; Self-control is commonly viewed as the active inhibi-
persuasion; attitude change tion of unwanted responses that might interfere with the
achievement of desired goals (Baumeister & Vohs,
2005). Research clearly indicates that the ability to exert
self-control acts as a limited resource (Baumeister et al.,
It is easier to resist at the beginning than at the end. 1998; Muraven et al., 1998). Thus, a person who exerts
self-control in the pursuit of a goal may be less able to
Leonardo da Vinci
exert self-control a short time later. Consider a man who
is dieting. Suppose he finds chocolate cake tempting, and
so it would require some restraint to resist eating a piece

R esisting a persuasive influence is often hard work.

One must have not only a desire to resist but also
the willpower to hold fast in the face of a potentially Author’s Note: Edward Burkley is now at Oklahoma State University.
This research was based on the author’s doctoral dissertation and he
tempting alternative—giving in. Thus, in many circum-
would like to thank the members of his committe, Melanie Green,
stances, it may be easier to allow oneself to be per- Chester Insko, Lawrence Sanna, Abigail Panter, and James Shah, for
suaded than it would be to resist. This suggests that their guidance. The author would also like to thank Hart Blanton, Keith
resistance to persuasion is a goal-directed action and, Payne, and Melissa Burkley for comments and suggestions on previous ver-
more specifically, that it requires self-control resources sions of this article; Richard Petty, Chester Insko, and Mark Muraven for
providing research materials; and Trey Williams, John Tsahakis, and Trevor
(Baumeister, Bratslavsky, Muraven, & Tice, 1998;
King for their assistance in conducting this research. Correspondence con-
Muraven, Tice, & Baumeister, 1998). The following cerning this article should be addressed to Edward Burkley, Department
studies test the hypothesis that self-control resources of Psychology, 116 North Murray, Oklahoma State University, Stillwater,
serve a vital role in the resistance process. When people OK, 74078; e-mail: ed.burkley@okstate.edu.
lack self-control resources, their ability to resist persua- PSPB, Vol. 34 No. 3, March 2008 419-431
sion will be diminished and their attitudes will be open DOI: 10.1177/0146167207310458
to change. © 2008 by the Society for Personality and Social Psychology, Inc.

419

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
420 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

of cake. Research on self-control predicts that this indi- More striking links to self-control failure can be found
vidual would have a more difficult time refusing this in studies of intense indoctrination. Brainwashing, police
cake if he recently exerted self-control by reluctantly interrogations, and fraternity hazings all involve an initial
eating an undesired salad (thereby depleting self-control “wearing down” phase that consists of food and sleep
resources; Vohs & Heatherton, 2000). deprivation. The purpose of this technique is to fatigue
More germane to the current research, an individual targets so that they will become open to various forms of
who has exerted self-control in the pursuit of one goal “re-education” (Hunter, 1960; Taylor, 2004). As one
might be less able to inhibit unwanted responses in the thought-reform victim described his experience, “You are
pursuit of another, seemingly unrelated goal. For annihilated, exhausted, you can’t control yourself. . . .
example, Muraven et al. (1998) demonstrated that par- You accept anything he says” (Lifton, 1961, p. 23).
ticipants who were asked to refrain from expressing Empirical evidence suggests that this process does in fact
emotion as they watched a humorous video showed a lead to a heightened vulnerability to influence (Baron,
performance decrement on a subsequent self-control 2000). Interestingly, fraternities also employ the use of
task (holding a handgrip as long as possible). Those unsavory foods in their hazing as a “wearing down” tech-
participants who had been “depleted” of their self- nique (Cialdini, 2001), a procedure that has been used to
control resources while watching the film were less able deplete self-control resources (Baumeister et al., 1998).
to hold the handgrip for an extended period. This type Collectively, these lines of work suggest that it takes
of response led Muraven and Baumeister (2000) to posit energy to resist and that people become open to influ-
that self-control resembles a muscle. If a person wears it ence once they have been depleted. The current studies
out by “exercising” it on one task, this might lead to a investigate forms of self-control depletion that are far
loss of “self-control strength” on a subsequent task. less dramatic than those used to indoctrinate, interro-
Consistent with the notion of muscle fatigue, people gate, or initiate. Nonetheless, these large differences in
who are depleted seem to become more passive and no the procedures employed belie a similar underlying
longer work as hard to achieve their goals (Baumeister theory about the nature of persuasion—that people
et al., 1998; Twenge & Baumeister, 2002). become less resistant if they are first depleted.
When presented with a counterattitudinal message,
people often prefer to resist the persuasive influence. In Present Theory
this context, the individual’s goal is either to retain a
prior opinion or to resist the adoption of a new opinion. It has been suggested that for attitude researchers, “the
However, to achieve this goal, the individual must exert main analytical task becomes understanding the processes
self-control to resist the temptation to give in. Thus, if that enable resistance and the persuasive techniques by
resistance to persuasion requires self-control resources, which resistance might be overcome” (Eagly & Chaiken,
and the ability to exert self-control is a limited resource 1998, p. 310). As one possible answer to this mandate, I
that is general in nature, then it should be possible to propose that resistance to persuasion (a) requires self-
lower a person’s resistance to persuasion through the control resources and (b) consumes self-control resources.
exertion of self-control on an unrelated task. Consider To be clear, this is not to say that all forms of resis-
again the dieter who just ate a bland salad. If the self- tance require self-control. Undoubtedly, some forms of
control analysis is correct, he would not only be less resistance are more effortful than others. However, with
able to resist cakes or hold handgrips, he should also be some persuasive appeals (particularly appeals with
more vulnerable to persuasion. strong, cogent arguments), an individual must exert
effort and thus self-control to successfully resist.
Depletion and Resistance to Persuasion Furthermore, I assert that regulatory depletion affects
resistance in a manner independent of cognitive responses.
Research in the persuasion literature provides anec- This is not to say that depletion will never influence
dotal support for the link between self-control and resis- attitude-relevant thoughts (e.g., Wheeler, Briñol, &
tance to persuasion. For example, it has been shown that Hermann, 2007); however, the point of the present
participants are more favorable toward an advertisement work is to demonstrate that depletion can lead to per-
when it is presented late in a series of ads rather than if suasion independent of thoughts and thus directly influ-
presented early (Knowles & Linn, 2004). This effect sug- ence one’s ability to resist persuasive messages. For
gests that people may lose their ability to resist after example, when people fail at their diets, it is not because
repeated exposure to persuasive appeals. Interestingly, they are unable to generate the thought “the dessert is
this effect bears a noticeable resemblance to self-control bad for me and will ruin my diet.” People generate these
research, where earlier self-control tasks lead to poorer thoughts and eat the dessert anyway. This is because it
performance on subsequent self-control tasks. is not enough to hold these thoughts in their mind; they

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 421

must also exert self-control (willpower) to actually study for course credit. Participants were randomly
implement these thoughts into behavioral change. I pro- assigned to one of two conditions: a 2-year or 10-year
pose that this is also the case for persuasive situations. university policy implementation (Petty, Cacioppo, &
When people agree to help their friend move into an Goldman, 1981). The primary dependent variable was
apartment, it is not because they are unable to generate persistence on an unsolvable task.
unfavorable thoughts about moving. They generate these
thoughts and help anyway. Even when unfavorable Procedure
thoughts are generated, it takes some resource-dependent
Attitude phase. Participants were presented with a
work to implement those thoughts and thereby actively
cover sheet stating that the university wanted to assess
resist. To summarize, self-control is asserted to be an
students’ responses to a policy change to shorten the sum-
underlying component of the resistance process, both
mer vacation to 1 month (Insko, Turnbull, & Yandell,
required and consumed, and as such it plays a funda-
1974; Zimbardo, Snyder, Thomas, Gold, & Gurwitz,
mental role in persuasion.
1970). For half of the participants, the cover sheet indi-
Four studies tested predictions derived from this theory.
cated that the policy would be implemented in 2 years
Study 1 examined the effect resistance has on self-control
and therefore would take effect while they were students
resources. I predicted that those who resisted a persuasive
(personally relevant). For the rest of the participants, the
appeal would exhibit fewer self-control resources on a sub-
cover sheet stated the policy would be implemented in 10
sequent task that was not related to persuasion. Studies 2
years (not personally relevant). Next, participants were
and 3 explored the effect that self-control exertion has on
presented with an essay containing arguments for short-
resistance to persuasion. I predicted that those who had
ening the summer vacation to 1 month (e.g., “earlier
their self-control resources depleted would be less success-
graduation for students,” “reduction in student fees”).
ful at resisting a persuasive appeal. Study 4 extended these
Pretest participants (N = 12) rated this essay as moderately
findings and examined effortful versus noneffortful forms
strong in terms of argument strength (M = 1.64, SD =
of resistance via argument quality. Because strong argu-
1.29) on an 11-point scale ranging from –5 (extremely
ments are more difficult to disregard and should require
weak) to +5 (extremely strong).
more resources to resist, I predicted that depletion would
After reading the essay, participants completed a five-
increase vulnerability to strong as opposed to weak argu-
item semantic differential scale that assessed attitudes
ments. Furthermore, this study relied on mediational
regarding the policy (bad/good, unfavorable/favorable,
analyses to assess whether the impact of self-control on
negative/positive, against/in favor, and harmful/beneficial;
resistance to persuasion occurs independent of message-
Tormala & Petty, 2002). Responses were made on an 11-
relevant thoughts.
point scale (–5 = extremely bad, –3 = bad, 0 = neutral, 3 =
good, 5 = extremely good). Overall, the five questions had
STUDY 1 high internal consistency (α = .95). They were combined to
obtain an overall mean composite score that served as an
The purpose of Study 1 was to examine how resistance index of a participant’s attitude toward the policy.
to a persuasive message affects participants’ self-control
Self-control phase. To assess self-control resources,
resources. According to the present theory, resistance to
participants were given a list of unsolvable anagrams
persuasion consumes self-control resources; therefore,
and their time spent persisting on the task served as a
greater resistance should result in less self-control. To test
measure of self-control. Persistence on unsolvable ana-
this assertion, this study relied on the idea that people are
grams is a common way to assess self-control resources
more resistant toward counterattitudinal messages when
(see Muraven et al., 1998). Participants were told that
the topic is likely to influence them personally. It therefore
this task was a “filler” task and that when they had fin-
was predicted that participants who were given a persua-
ished, were tired, or had worked on the anagrams long
sive message on a topic that would affect them personally
enough, they should open their door and alert the
would be more resistant and subsequently have fewer self-
experimenter in the main room. The experimenter
control resources than participants who were given a per-
timed from the start of the anagrams task to the point
suasive message that would not affect them personally.
when the participant opened the door. A maximum
time limit of 10 min was given and those who did not
Method finish within this time were stopped and given a score of
10 min.1 During the debriefing, none of the participants
Participants and Design
reported being aware that the self-control task was
Seventy-two students (mean age = 19) from the University related to the policy task and no one reported aware-
of North Carolina at Chapel Hill (UNC) participated in the ness that the anagrams were unsolvable.

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
422 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

Lastly, a standard task perception assessment, taken not differ across conditions, alternative explanations,
directly from Muraven, Shmueli, and Burkley (2006), such as participants giving up sooner due to frustration
was administered. Participants evaluated their mood or anger, are unlikely.
and rated the anagrams task with regard to how diffi-
cult, boring, demanding, and frustrating it was per-
ceived to be. These questions were rated on a 25-point STUDY 2
scale ranging from not at all to extremely (e.g., “How
much did your mood change as you were working on Although Study 1 suggests that resistance to persua-
the task?” “How difficult was it to work on the task?” sion consumes self-control resources, the role that these
“How frustrating was that task?”). resources play in one’s ability to successfully resist per-
suasion was not addressed directly. The purpose of
Results Study 2 was to investigate the impact of self-control on
resistance to a persuasive message. If such resistance
Task Perceptions requires self-control, then individuals who are depleted
of their self-control resources should be less able to
Responses across the two conditions were compared
resist a subsequent persuasive message.
to ensure that the groups did not differ in terms of their
mood and perceptions regarding the anagrams task. As
predicted, there were no significant differences (ps > .12). Method
Thus, it appears there were no mood differences between
Participants and Design
the two conditions and that all participants viewed the
anagrams as equally difficult, boring, demanding, and Twenty-two male students (mean age = 19 years)
frustrating.2 from UNC participated in the study.3 Participants were
randomly assigned to a depletion or control condition.
Attitude Ratings The primary dependent variable was level of agreement
with a persuasive essay.
Those who were told that the policy would be imple-
mented in 10 years were significantly in more agreement
Procedure
with the policy (M = –.15, SD = 2.80) than were those
who were told the policy would be implemented in 2 Self-control phase. Participants were told that the
years (M = –1.46, SD = 2.27), t(70) = –2.18, p < .05. As study was investigating various cognitive processes
expected, individuals in the 2-year condition were more assessed through multiple tasks. Half of the participants
resistant to the policy than those in the 10-year condition. first completed a self-control task (depletion condition)
and half (control condition) skipped this task and con-
Persistence tinued to the attitude phase of the study.
Participants in the depletion condition were required
The results indicated that there was a significant differ-
to exert self-control by squeezing a handgrip for as long
ence in persistence between the two groups, t(71) = –1.96,
as possible. This handgrip task is commonly used in
p < .05. Those in the 10-year condition (M = 8.38 min,
studies of self-control (Muraven et al., 1998). The hand-
SD = 2.10) persisted more than 1 min longer on the 10-
grip consists of two handles connected by a wound
min anagrams task than did those in the 2-year condi-
spring and is designed to build muscles in the forearm.
tion (M = 7.12 min, SD = 3.25). Cohen’s d = .47, which
Individuals grasp the handgrip, which brings the han-
is a medium effect size based on Cohen’s (1988)
dles together and compresses the spring, thereby creat-
conventions.
ing resistance. Maintaining a grip is tiring for the
forearm muscles and once these muscles become
Discussion fatigued, the grip relaxes. Thus, continuously squeezing
the handgrip requires the individual to overcome the
Study 1 revealed that those who were more resistant urge to relax the forearm and as such, self-control is
toward a persuasive message had less self-control required to inhibit this desire to let go. To measure how
resources available to use on a subsequent unrelated long participants gripped the handgrip as accurately as
persistence task. Participants who believed that the pol- possible, a folded piece of paper was placed between the
icy would take effect in 2 years were more against the handles. Once a participant gripped the handles
policy and thus more resistant. Subsequently, these par- together with the paper in between, the experimenter
ticipants had fewer resources in reserve to apply to the began timing. When the participant’s grip loosened
persistence task. Because mood and task perceptions did and the handles opened enough so that the paper

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 423

dropped from the handgrip, the timing was stopped and One limitation of Study 2 was that the handgrip was
recorded. not held for an equal duration across all participants.
Therefore, the purpose of Study 3 was to use a self-
Attitude phase. During the attitude phase, all partic- control manipulation that would more uniformly influ-
ipants received the same persuasive essay used in Study ence participants. A second limitation of this study was
1. The key difference was that all participants were told that participants in the control group did not complete
that the policy to shorten the summer would be imple- a task before reading the essay, whereas those in the
mented in 5 to 6 years. This implementation was done depletion condition did. It is possible that completing
to create a moderate degree of involvement and motiva- any task before the essay may have decreased resistance,
tion. The concern here was that low involvement would regardless of whether the task involved self-control.
not invoke resistance and that high involvement would Study 3 sought to address this possible alternative
motivate too much resistance and override the small account by having half the participants complete a self-
self-control manipulation (see Muraven & Slessareva, control task and the other half complete a task that does
2003). Next, participants completed the same attitude not require self-control. Finally, because self-control
measure used in Study 1. It was assumed that partici- can be exerted in a number of ways (e.g., emotional,
pants would generally disagree with this policy and so it behavioral, cognitive), its effect on persuasion was
would require persuasion to move their attitudes in the explored using a different operationalization. Rather
direction of endorsement. A pilot study (N = 22) con- than using a behavioral manipulation of self-control,
firmed this intuition, demonstrating that on average, Study 3 used a cognitive manipulation.
participants rated the policy change negatively (M =
–2.08, SD = 2.02) on a scale from –5 (disagreement) to
+5 (agreement). Thus, the persuasive message was per- STUDY 3
ceived as counterattitudinal.
After reading the essay, participants completed the The design of this study was similar to Study 2
five attitude questions used in Study 1 (α = .95). Finally, except that a new self-control manipulation was used. A
participants completed the same task perception ques- common technique for distinguishing the effects of self-
tionnaire used in Study 1 to assess differences in mood control from nonregulatory effects (e.g., cognitive fatigue)
and perceptions of the attitude task. is to employ two tasks, both requiring exertion but only
one involving self-control exertion. For example, previous
Results research indicates that solving math problems is rated
just as difficult and effortful as suppressing one’s
thoughts (Muraven et al., 1998); however, only thought
Task Perceptions suppression requires self-control. Thus, to assess the
effect of self-control on resistance to persuasion in this
There were no significant differences in mood or per-
study, a thought suppression task was used for the
ceptions of the task, ps > .12. This pattern again ensures
depletion condition and a math problem task was used
that differences in attitude ratings are not attributable
for the control condition.
to differences in mood or task perceptions.

Attitude Ratings Method

As predicted, those who squeezed the handgrip for as Participants and Design
long as possible and thereby exerted self-control were in less
Seventy-eight students (29 men, 49 women; mean
disagreement with the policy (M = 0.16, SD = 2.41) com-
age = 19) from UNC participated in this study.
pared to those who did not exert self-control (M = –1.90,
Participants were randomly assigned to the depletion or
SD = 2.10), t(20) = –2.12, p < .05 (Cohen’s d = .91).4
control conditions. The primary dependent variable was
level of agreement with a persuasive essay.
Discussion
Procedure
Study 2 revealed that participants who were more
depleted of their self-control resources were more per- Self-control phase. This experiment used a bogus
suaded by the counterattitudinal message. Because resis- two-study procedure. The experimenter explained that
tance to persuasion requires resources, those who the first study was investigating various cognitive
exerted self-control through the handgrip task were processes. During this phase, participants completed a
more in agreement with the policy. self-control depletion task or control task for 5 min.

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
424 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

Participants in the depletion condition completed a (M = 9.80, SD = 4.39). These findings offered reassurance
thought suppression task. They were instructed to list that the manipulation targeted self-control and not more
any thoughts that came to mind on a blank paper but general cognitive or affective states.
were told to actively avoid thinking of a white bear dur-
ing this task. Whenever they had unwanted thoughts of Attitude Ratings
a white bear, they were told to place a check mark on
As predicted, those who were told to inhibit their
the page for each occurrence (see also Wegner, 1992;
thoughts were in less disagreement with the policy
Wegner, Schneider, Carter, & White, 1987). Those
(M = –0.43, SD = 2.02) than those who worked on solving
in the control condition were instructed to work on a
math problems (M = –1.40, SD = 2.25), t(76) = –2.01,
list of moderately difficult math problems (Muraven
p < .05 (Cohen’s d = .45). Indeed, those in the math
et al., 1998).
problems condition tended to show an attitude level
Following this initial task, participants completed the
(M = –1.40) much like that shown in the pilot study
task perception questionnaire used in the previous stud-
(M = –2.08), t(54) = 1.15, p = .26.
ies. However, unlike the previous studies where this mea-
sure was used to assess mood and perceptions regarding
the final task, in this study it was used to assess mood and Discussion
perceptions regarding the self-control task. The task per-
ception measure therefore assessed the effectiveness of the Study 3 provided a conceptual replication of Study 2
self-control manipulation. This is consistent with how by showing that participants who exerted self-control
this measure is typically used in the self-control literature through thought suppression were more persuaded by
(see Muraven et al., 2006). In addition, two questions the counterattitudinal message. This study also ruled out
regarding self-control exertion were added (i.e., “How a possible confound in Study 2 by showing that only the
hard did you work at controlling your thoughts?” and task requiring self-control impaired resistance to persua-
“How much did you have to inhibit your thoughts?”). sion. Importantly, the findings cannot be explained by
These two questions were significantly correlated with cognitive fatigue because both tasks required cognitive
one another (r = .59, p < .001) and were averaged to form exertion but only the thought suppression task required
a composite measure of self-control exertion. Unlike the regulatory exertion. Finally, this study extended the pre-
other assessment items, differences between the two con- vious findings by demonstrating this effect with a dif-
ditions were expected because they were thought to rep- ferent manipulation of self-control.
resent self-control resources used. As previously stated, I predicted that this effect of
depletion on resistance to persuasion, as a regulatory
Attitude phase. Next, participants completed the atti- process, should occur independent of cognitions. The
tude phase of the study. A different experimenter was assertion is that self-control exertion impairs one’s
introduced and presented the cover story from Study 1. willpower to stand against an agent of influence and
A packet was given that contained a new consent form, that it does so in a direct manner. However, an alterna-
a cover sheet, the persuasive essay (same one used in tive explanation is that self-control leads to changes in
Studies 1 and 2), and questions assessing the student’s atti- cognitions and that these cognitive changes lead to
tude toward the policy change (α = .93). As in Study 2, all changes in attitudes. For example, it could be that the
participants were told that the policy to shorten the thought suppression in Study 3 led to a “rebound”
summer would be implemented in 5 to 6 years. During effect. That is, when individuals inhibited thoughts on
the debriefing sessions, no participant reported aware- the first task, this led to more thorough thinking about
ness that the two studies were related. the message on the subsequent task. If this were the
case, the increased persuasion found in Study 3 would
be due to increased message scrutiny and not impaired
Results self-control. Although this suggestion may seem plausi-
Task Perceptions ble, the research on rebound effects, as well as the pre-
sent theory, suggests it is an unlikely explanation.
As in the previous studies, and consistent with Thought rebound is defined as the propensity for
Muraven and colleagues (Muraven et al., 1998; Muraven thought suppression to lead to increased intrusions of the
et al., 2006), there were no significant differences in previously forbidden thought (Wegner et al., 1987). This
mood or perceptions of the task (ps > .31). Nevertheless, effect does not necessarily imply that inhibition of
the two conditions did differ in their reported exertion of thoughts in one domain will lead to greater thinking in
self-control, t(76) = –2.45, p < .05. Those in the depletion another domain. Rebound effects have only been found to
condition reported exerting more self-control (M = be domain specific. For example, suppression of stereo-
12.76, SD = 5.88) than those in the control condition types about a skinhead led to greater stereotypic thoughts

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 425

about skinheads on a subsequent task (Macrae, Boden- have little or no effect. I also predicted that there would
hausen, Milne, & Jetten, 1994), but it has not been found be no effect of depletion on participants’ message-
to result in greater thinking overall. If this were the case, relevant thoughts and that depletion would affect
stereotype suppression would likely result in more indi- attitudes directly (i.e., independent of thoughts).
viduated and less stereotypic responses on a subsequent
task. Thus, the suppression of a white bear may lead to a
rebound of greater thoughts of a white bear, but there is METHOD
no evidence to suggest that it would lead to greater over-
all thinking about the persuasive message. Nevertheless, Participants and Design
this alternative account was explored in Study 4.
A second purpose of Study 4 was to assess the impact One hundred and twenty students (23 men, 95
of depletion on effortful versus noneffortful forms of women, 2 failed to report gender; mean age = 19) from
resistance. Although the concept of resistance denotes UNC participated for course credit. Participants were
an effortful quality, clearly not all forms of resistance randomly assigned to one of four conditions in a 2 (self-
are extremely effortful. For example, it should take a control: depletion vs. control) × 2 (argument strength:
great deal of effort to resist a message that contains strong vs. weak) factorial design. The primary dependent
strong, cogent arguments, but it should not take much variable was level of agreement with a persuasive essay.
effort to combat weak, spurious arguments. Thus, by
examining argument strength from a regulatory per- Procedure
spective, Study 4 explored the use of effortful and non-
effortful resistance in relation to self-control resources. The procedure of Study 4 was similar to Study 3,
with one addition. After rating their attitude, partici-
pants completed a thought listing task that followed the
STUDY 4 format used by Petty and Cacioppo (1986b).

The purpose of Study 4 was twofold. First, this study Self-control phase. For greater precision in the manip-
examined the impact that self-control resources have on ulation of self-control, it is common to hold the task con-
resistance to weak and strong persuasive messages. It stant across conditions, where one group is instructed to
was expected that participants would be motivated to exert self-control and the other group does not receive
resist both weak and strong arguments. However, weak these instructions (Muraven et al., 1998). For example, all
arguments can be easily disregarded (i.e., less effortful) participants may be asked to freely list their thoughts on
and therefore should not tax self-control resources. paper, with only half of them instructed to inhibit the
Strong arguments cannot be easily dismissed and should thought of a white bear. In this instance, suppressing
require active resistance. As such, I predicted that for thoughts requires greater regulatory effort than not sup-
strong arguments, depletion would impair resistance pressing thoughts. Unlike two tasks that are perceived to
compared to the control condition. For weak argu- be similar on effort and difficulty and seemingly differ only
ments, depletion should have little or no effect on in the exertion of self-control (e.g., math vs. thought sup-
resistance. pression), the same task performed with or without the
Another purpose of this study was to assess whether exertion of self-control is expected to differ in the amount
depletion, as a regulatory process, can affect attitudes of perceived effort and difficulty (Muraven et al., 1998).
independent of cognitions. Recall that it is not enough to For this reason, Study 4 had participants in the con-
hold message-relevant thoughts in our mind; we must trol condition freely list their thoughts instead of work-
also exert self-control to actually implement these ing on math problems. Thus, half of the participants
thoughts into resistance. If depletion affects attitudes suppressed the thought of a white bear (depletion con-
independent of cognitions, it should not influence dition) when listing their thoughts on a blank piece of
message-relevant thoughts. Furthermore, the impact of paper and the other half freely wrote down their
depletion on attitudes should occur directly and therefore thoughts (control condition; Muraven et al., 1998). Five
should not be mediated by changes in message-relevant minutes were given to work on this task. Next, partici-
cognitions. To test these issues, participants’ message- pants completed the same task perception questionnaire
relevant thoughts were directly assessed and compared used in Study 3 that included the two self-control exer-
across conditions. Also, mediational analyses were used tion questions (r = .68, p < .001).
to see whether the effect of depletion on persuasion
occurred independent of cognitive responses. Thus, I Attitude phase. For the attitude phase of the study,
predicted that with strong arguments, self-control depletion a new persuasive topic was adopted and the procedure
would impair resistance, but for weak arguments it would followed that of Petty and Cacioppo (1986b). After the

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
426 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

self-control phase, the experimenter informed participants

that the next task was about impression formation and 3
that they were to read an essay and form an impression
about the author. All were then presented with an essay 2
that consisted of weak or strong arguments for a manda-
tory senior exam policy (Petty & Cacioppo, 1986b). After

Attitude
1
reading the essay, participants rated their attitude toward
the senior exam policy. These questions were taken from
Petty and Cacioppo (1986b) and were similar to those used 0
in the previous studies. Specifically, they consisted of four
questions (bad/good, unfavorable/favorable, harmful/ben- −1
eficial, and foolish/wise) on an 11-point semantic differen-
tial scale ranging from –5 to +5. The four questions had −2
high internal consistency (α = .94) and they were combined
weak strong
to obtain an overall index of a participant’s attitude.
Control Depletion
Thought listing task. After rating their attitudes, par-
ticipants completed a thought listing task. This task is Figure 1 Attitude regarding senior exam policy by self-control and
commonly used to assess how much participants are argument quality conditions.
thinking about the persuasive message (e.g., Cacioppo
& Petty, 1981; Insko, Lind, & LaTour, 1976; Tormala,
pattern of results is shown in Figure 1. The main effect
Brinol, & Petty, 2006). Participants were given a sheet
of argument strength was significant, with strong argu-
of paper with blank boxes and were asked to list their
ments being more persuasive (M = 1.74, SD = 1.82) than
thoughts regarding the policy they just read. After list-
weak arguments (M = –0.78, SD = 2.21), F(1, 116) =
ing their thoughts, participants then classified each
47.49, p < .001. The main effect of self-control condi-
thought as positive (in favor of the policy), negative
tion was not significant, F < 1. Most important, there
(opposed to the policy), or neutral (irrelevant).
was a significant interaction, F(1, 116) = 15.77, p < .05.
As predicted, those who read the strong argument essay
Results and who were in the depletion condition showed greater
agreement with the policy (M = 2.25, SD = 1.36) com-
Task Perceptions pared to those in the control condition (M = 1.23, SD =
2.09), F(1, 116) = 3.85, p < .05 (Cohen’s d = .59).
As in the previous studies, there was no significant dif-
However, for those who read the weak argument essay
ference by condition in mood or perceptions of the task
there was not a significant difference between those in the
in general (ps > .33). However, consistent with prior self-
depletion (M = –1.00, SD = 2.23) and control (M = –0.57,
control findings using this manipulation (see Muraven
SD = 2.22) conditions, F(1, 116) = .70, p > .40.5
et al., 1998), participants in the depletion condition
rated the task as more effortful (M = 13.20, SD = 6.61)
Thought assessment. To determine whether depletion
than did participants in the control condition (M = 9.63,
influenced participants’ cognitions, the number of positive
SD = 6.49), t(118) = –2.98, p < .01, and they rated the
(favorable) and negative (unfavorable) thoughts listed by
task as more difficult (M = 9.07, SD = 6.31) than did
participants were separately analyzed in 2 (condition:
participants in the control condition (M = 5.62, SD =
depletion vs. control) × 2 (argument: strong vs. weak)
5.41), t(118) = –3.22, p < .01. As in Study 3, the two
ANOVAs. For positive thoughts, only a main effect of
tasks differed in their ratings of self-control required,
argument strength was significant, such that there were
t(118) = –3.59, p < .001, with those in the depletion con-
more positive thoughts for strong (M = 1.12, SD = .94)
dition reporting that they exerted more self-control (M =
than for weak (M = 0.33, SD = 0.66) arguments, F(1, 116)
12.86, SD = 5.82) than those in the control condition
= 27.94, p < .001. For negative thoughts, once again there
(M = 9.16, SD = 5.45). These results supported the con-
was only a main effect of argument strength, such that
tention that the two conditions differed primarily in the
there were more negative thoughts for weak (M = 2.87,
amount of self-control resources required.
SD = 1.44) than for strong (M = 1.60, SD = 1.28) argu-
ments, F(1, 116) = 25.49, p < .001. Importantly, there
Dependent Measures
were no self-control main effects or interactions for posi-
Attitude ratings. The composite attitude score was tive or negative thoughts, Fs < 1. Finally, total number of
subjected to a 2 (condition: depletion vs. control) × 2 thoughts (positive, negative, and irrelevant) showed no
(argument: strong vs. weak argument) ANOVA. The significant main effects or interaction.

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 427

In addition to a separate analysis of positive and neg-

ative thoughts, it is common practice to compute an A.
index of thought favorability by subtracting the number Thought
of negative thoughts from the number of positive Favorability
thoughts and dividing this difference by the total
.54*** .55***
number of thoughts listed. This index was also com-
puted and analyzed. Once again, only a main effect of
argument strength was found, such that there was .13
Argument
greater favorability toward strong (M = –0.09, SD = Strength
Attitude
0.38) than weak (M = –0.53, SD = 0.37) arguments, (.39**)
F(1, 116) = 40.26, p < .001. Importantly, self-control
had no effect on thought favorability, Fs < 1, indicating
that the thought suppression task did not lead partici-
pants to think more thoroughly about the message. B.

Thought
Mediational Analyses Favorability

For a more thorough examination of the findings, .48*** .66***

multiple mediational analyses were conducted. The cog-
nitive response approach to attitude change states that .46***
thought elaboration fully mediates the relationship Argument
Attitude
between argument strength and attitudes. As such, this Strength
(.67***)
pattern should clearly be seen in the control condition.
However, this pattern should not occur in the depletion Figure 2 Thought favorability as a mediator of argument strength
condition if something additional is influencing partici- and attitudes in control (A) and depletion (B) conditions.
NOTE: Path coefficients are represented as standardized betas. The
pants’ attitudes (i.e., self-control). If self-control is hav- coefficient in parentheses indicates the direct effect of argument
ing an influence on attitudes in this condition, thought strength on attitudes without mediators in the model.
elaboration should only partially mediate the relation- **p < .01. *** p < .001.
ship between argument strength and attitudes. This pat-
tern would suggest that self-control is having an effect The Sobel (1982) test of mediation was significant (z =
on attitudes above and beyond the typical cognitive 2.94, p < .01). This pattern indicates that for those in the
response effect. Thus, I predicted that thoughts would control condition, the effect of argument strength on atti-
completely mediate the effect of argument strength on tudes was fully mediated by thought favorability, sup-
attitudes in the control condition and only partially porting a cognitive response model of attitude change
mediate this relationship in the depletion condition. (Petty & Cacioppo, 1986a).
To assess mediation, the classic approach of regression- However, as can be seen in Part B of Figure 2, this
based causal models was used (Baron & Kenny, 1986). pattern was not evident for participants who were
The model was assessed for participants in both the deple- required to exert self-control. Once again, argument
tion and control conditions and the results are presented strength significantly influenced thought favorability,
in Figure 2. The data were first analyzed to determine β = .48, t(58) = 4.16, p < .001, such that stronger argu-
whether thought favorability mediated the effect of argu- ments led to more favorable thoughts. In turn, thought
ment strength on attitudes. As can be seen in Part A of favorability significantly influenced participants’ atti-
Figure 2, for those who were not required to exert self- tudes, β = .66, t(58) = 6.69, p < .001, such that greater
control, argument strength significantly influenced favorability led to less negative attitudes. However,
thought favorability, β = .54, t(58) = 4.85, p < .001, such when thought favorability and argument strength were
that stronger arguments led to more favorable thoughts. both added as predictors of attitudes, the direct effect
In turn, thought favorability significantly influenced par- between argument strength and attitudes remained sig-
ticipants’ attitudes, β = .55, t(58) = 5.03, p < .001, such nificant, only dropping from β = .67, t(58) = 6.82, p <
that greater favorability led to less negative attitudes .001, to β = .46, t(57) = 4.74, p < .001 (Sobel test: z =
toward the policy. Importantly, when thought favorabil- 3.08, p < .01). This mediational pattern indicates that
ity and argument strength were both added as predictors for those in the depletion condition, the effect of argu-
of attitudes, the direct effect between argument strength ment strength on attitudes was only partially mediated
and attitudes became nonsignificant, dropping from β = by thought favorability. Importantly, this pattern estab-
.39, t(58) = 3.23, p < .01, to β = .13, t(57) = 1.03, p = .31. lishes that in the condition where participants are made

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
428 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

to exert self-control, there is an additional effect on

their attitudes beyond that of elaboration. Thought
The preceding results indicate that the effect of argu- Favorability
ment strength on attitudes is fully mediated by thought .03 .51***
favorability but that this is not the case when participants
exert self-control, suggesting that self-control has an
influence independent of thoughts. To explore the effects .26*
of self-control on these relationships, a second media- Self-Control Attitude
tional analysis was conducted to determine whether the (.28*)
relationship between self-control exertion and attitudes
Figure 3 Thought favorability as a mediator of self-control and atti-
was mediated by thought favorability. Because the effect tudes in the strong arguments condition.
of self-control exertion on attitudes is thought to occur NOTE: Path coefficients are represented as standardized betas. The
independent of cognitive responses, I predicted that coefficient in parentheses indicates the direct effect of self-control on
attitudes without mediators in the model.
thought favorability would not mediate this direct effect. *p < .05. ***p < .001.
The relationship among self-control, thought favora-
bility, and attitudes was separately assessed for partici-
pants in the strong and weak argument conditions. Study 4 also tested the alternative explanation that
However, given that the self-control manipulation was thought suppression led to thought rebound, which
not found to affect attitudes regarding weak messages, caused greater elaboration of the persuasive message.
only the analysis for the strong arguments is reported.6 Contrary to a rebound account, thought suppression did
These results are presented in Figure 3. As can be seen, not influence participants’ thoughts (for or against) the
when participants were exposed to strong arguments, persuasive message. Furthermore, the mediational analy-
thought favorability had a significant effect on atti- ses demonstrated that the effect of self-control on attitude
tudes, β = .51, t(58) = 4.49, p < .001, such that report- change in the strong arguments condition was not medi-
ing greater thought favorability led to more favorable ated by thought favorability. Taken together, these results
attitudes. Importantly, when thought favorability and indicate that thought suppression did not lead to greater
self-control were both added as predictors of attitudes, message elaboration. Importantly, the results of this study
the direct effect between self-control and attitudes demonstrate that self-control depletion increases persua-
remained significant, only dropping from β = .28, t(58) = sion in a manner that does not influence the quantity or
2.23, p < .05, to β = .26, t(57) = 2.43, p < .05. The Sobel favorability of attitude-relevant thoughts.
test of mediation was not significant (z = .26, p > .80). Overall, this study provides further evidence that
In other words, after controlling for thought favorabil- self-control plays an integral role in resistance to per-
ity, self-control exertion still had a significant direct suasion, particularly in the presence of strong persua-
effect on attitudes. Consistent with a self-control model sive arguments. In addition, this study revealed an
of resistance to persuasion, the direct effect of self-con- important distinction between effortful and noneffort-
trol on attitude change was independent of participants’ ful forms of resistance. Although many forms of resis-
cognitive responses. tance require self-control resources, some forms are less
effortful and therefore do not consume or require these
Discussion resources. This distinction between forms of resistance
that do or do not require self-control resources is an
The results of Study 4 clarify the role self-control important addition to the persuasion literature and a
plays in resistance to persuasion. These results revealed fruitful area for future research.
that resources are involved in the resistance of strong
but not weak arguments. For participants who read
strong arguments, those who exerted self-control on a GENERAL DISCUSSION
previous task were more favorable toward the message
than those who did not exert self-control. This result The present studies provide evidence that one must
suggests that when individuals have their self-control possess self-control resources to effectively resist persua-
resources intact they are better able to successfully resist sion. Successful resistance leads to an expenditure of these
strong arguments. Conversely, for participants who resources (Study 1), and depletion of these resources leads
read weak arguments, there was no influence of self- to diminished resistance (Studies 2 and 3), particularly in
control exertion. This is likely because weak arguments response to strong arguments (Study 4). Importantly, the
are easily disregarded, do not tax resources, and thus effect of self-control on persuasion was demonstrated
are unaffected by self-control exertion. across these studies using different self-control manipulations

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 429

and different persuasive messages. Overall, these experi- takes some resource-dependent work to implement
ments offer strong evidence that self-control depletion those arguments into active resistance. Identifying when
reduces resistance to persuasion. Together, these studies self-control does and does not influence attitudinal cog-
suggest a cyclical relationship between self-control nitions is a worthy endeavor for future research, but it
resources and resistance to persuasion. When one resists is beyond the scope of this article.
persuasive appeals, this act can expend self-control The distinction between the influence of self-control on
resources, thereby leaving one depleted and susceptible to counterarguments and its influence independent of
future persuasive influences. thoughts also offers unique predictions for future work. If
But this cycle is not inevitable. Although some forms depletion disrupts the generation of counterarguments
of resistance to persuasion are effortful, it appears that (Wheeler et al., 2007), depletion effects should occur in con-
others are not. The results of Study 4 show that indi- ditions where counterarguments are likely to occur. For
viduals are able to resist weak efforts routinely without example, counterarguments are more likely to occur for
becoming depleted or vulnerable to future influence. In strongly held preexisting attitudes, such as attitudes toward
fact, many forms of resistance that have been studied in abortion (Petty & Krosnick, 1995). Thus, depletion effects
the persuasion literature appear to be noneffortful, such would be stronger with firmly held attitudes than with new
as source derogation (Tannenbaum, Macauley, & or weakly held attitudes. The present theory, based on the
Norris, 1966) and avoidance of inconsistent informa- self-regulatory literature, predicts the opposite pattern.
tion (Frey, 1986). However, when individuals are Because self-control acts as a limited resource, we are likely
unable to easily dismiss or avoid persuasive messages in to conserve it to use as needed (Muraven et al., 2006).
this manner, self-control resources are needed to suc- When we encounter situations that are important to us
cessfully combat these influences. (e.g., beliefs about abortion) we often choose to devote
This pattern of results bears some similarity to Gilbert’s those resources to that endeavor (Muraven & Slessareva,
(1991) research on automatic acceptance. Gilbert suggests 2003). Thus, depletion effects would be stronger for weakly
that people automatically accept information to be true, held or newly formed attitudes than for firmly held atti-
and the subsequent evaluation and then rejection of infor- tudes because people would put all their resources into
mation are more effortful than acceptance. This may be defending these self-important beliefs.
the case for persuasive messages as well—people may Although depletion may influence thoughts, the pre-
automatically accept a persuasive request and then must sent work is important because it demonstrates that
exert effort to evaluate and then reject the message. changes in thoughts are not a necessary requirement for
Although this may be the case, Gilbert’s analysis focused attitude change to occur. Even when the exertion of
on cognitive resources (e.g., distraction, cognitive load) self-control does not affect message-relevant thoughts,
rather than regulatory resources. Thus, it is likely that it still can have a direct effect on persuasion nonethe-
Gilbert’s effects do not completely dovetail with the less. Thus, the present studies offer the first demonstra-
processes described here. Nevertheless, the role that auto- tion of the effects of self-control depletion on attitude
matic processes play in the relationship between self- change, independent of thoughts.
control and persuasion is a worthy area for future study. The present research is also important because it
One significant contribution the present findings offer offers one of the first demonstrations connecting the
is the demonstration that the self-regulation process, vis- self-regulation literature with the attitude literature.
à-vis depletion, affects attitude change independent of Other topics in our field have recently benefited from a
elaboration. The results of Study 4, particularly the self-regulation perspective, including stereotypes (von
mediational analyses, clearly show that the effect of Hippel, Silver, & Lynch, 2000), aggression (DeWall,
depletion on attitudes is not caused by changes in cog- Baumeister, Stillman, & Gailliot, 2007), and self-
nitive responses. That is not to say that depletion may presentation (Vohs, Baumeister, & Ciarocco, 2005).
never influence attitude-relevant thoughts; recent The present research clearly shows that a regulatory
research in independent labs has shown such an effect perspective can facilitate in the understanding of atti-
can occur under certain circumstances (e.g, Gitter, tude change by offering a unique way of viewing per-
Maner, & Tice, 2006). Wheeler et al. (2007) recently suasion and by providing a burgeoning area for future
showed that depletion can impair the ability to generate research (see also Cesario, Grant, & Higgins, 2004; Lee
counterarguments, and this in turn influenced attitudes, & Aaker, 2004; Wheeler et al., 2007).
but only in response to weak messages. Wheeler et al. sug-
gest that depletion merely moderates the counterargument When Do We Resist?
process. The present work shows that self-regulation plays
a more fundamental role in the resistance process— The self-control model provides an account for why
even when counterarguments have been generated, it people are less able to resist multiple persuasive

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
430 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

attempts that occur over time (e.g., Knowles & Linn, be due to the handgrip being a “heavy tension” model. As such, only
the data from the male participants were used in the analysis. Females
2004). It also suggests that people may be more vulner- did not significantly differ in attitudes across the self-control condi-
able to persuasion later in the day because the expendi- tions, t(18) = .42, p = .68.
tures of self-control resources might accumulate as the 4. Attitude certainty was also measured (Tormala & Petty, 2002).
It was found that self-control exertion did not significantly affect par-
day progresses. Furthermore, people who are actively ticipants’ attitude certainty, t(20) = .64, p = .53. This was also found
pursuing self-improvement (e.g., dieting, quitting smok- to be the case in Studies 3 and 4.
ing) may be particularly vulnerable to persuasive influ- 5. Strong arguments produced more favorable attitudes than weak
arguments in the control condition, F(1, 116) = 12.07, p < .01, and
ences. Finally, it seems likely that those with chronically the depletion condition, F(1, 116) = 39.34, p < .001.
low self-control resources should be most persuadable, 6. Only the path from thought favorability to attitudes reached
whereas those with chronically high self-control should significance for the weak arguments condition, β = .43, t(58) = 3.67,
p = .001.
be able to resist persuasion even in the face of self-control
demands (Tangney, Baumeister, & Boone, 2004).
Just as this model offers insights into the factors that REFERENCES
might diminish resistance, it also offers suggestions on
how to increase it. The two primary ways of increasing Baron, R. S. (2000). Arousal, capacity, and intense indoctrination.
resistance to persuasion are forewarning (Freedman & Personality and Social Psychology Review, 4, 238-254.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator vari-
Sears, 1965) and inoculation (McGuire, 1964). Both able distinction in social psychological research: Conceptual,
techniques work by sensitizing an individual to the strategic, and statistical considerations. Journal of Personality and
arguments that might appear in an upcoming appeal. A Social Psychology, 51, 1173-1182.
Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M.
self-control approach to resistance differs from these in (1998). Ego-depletion: Is the active self a limited resource? Journal
that resistance is raised through actions that are unre- of Personality and Social Psychology, 74, 1252-1265.
lated to the approaching arguments. By engaging in self- Baumeister, R. F., & Vohs, K. D. (2005). Handbook of self-regulation
research. New York: Guilford.
control “exercises,” such as regulating one’s mood or Cacioppo, J. T., & Petty, R. E. (1981). Social psychological proce-
improving one’s posture, people may develop a reser- dures for cognitive response assessment: The thought listing tech-
voir of stronger self-control resources to draw on nique. In T. Merluzzi, C. Glass, & M. Genest (Eds.), Cognitive
assessment (pp. 309-342). New York: Guilford.
(Muraven, Baumeister, & Tice, 1999) and thus may Cesario, J., Grant, H., & Higgins, E. T. (2004). Regulatory fit and
become more able to resist persuasion even after sus- persuasion: Transfer from “feeling right.” Journal of Personality
tained self-control demands. Current work in my lab is and Social Psychology, 86, 388-404.
Cialdini, R. B. (2001). Influence: Science and practice (4th ed.).
exploring these issues. Boston: Allyn & Bacon.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
Conclusion DeWall, C. N., Baumeister, R. F., Stillman, T. F., & Gailliot, M. T.
(2007). Violence restrained: Effects of self-regulation and its deple-
The current research offers a clear demonstration of tion on aggression. Journal of Experimental Social Psychology,
the connection between self-control and resistance to per- 43, 62-76.
Eagly, A. H., & Chaiken, E. (1998). Attitude structure and function.
suasion. It points out that a person must work to thwart In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook
persuasive appeals and that, without the means of doing of social psychology (Vol. 1, 4th ed., pp. 269-322). New York:
so, one can become susceptible to influence. Like erosion McGraw-Hill.
Freedman, J. L., & Sears, D. O. (1965). Warning, distraction, and
wearing away at a piece of stone, people can be worn resistance to influence. Journal of Personality and Social
down over time and succumb to influences that they Psychology, 1, 262-266.
would have otherwise been able to resist. Indeed, this Frey, D. (1986). Recent research on selective exposure to information.
Advances in Experimental Social Psychology, 19, 41-80.
effect was intuited by Leonardo da Vinci long ago: That Gilbert, D. T. (1991). How mental systems believe. American
it is easier to resist at the beginning than at the end. Psychologist, 46, 107-119.
Gitter, S., Maner, J., & Tice, D. (2006, January). Going where the
green light leads you: Self-regulatory failure and argument gener-
ation. Poster session presented at the annual meeting of the Society
NOTES of Personality and Social Psychology, Palm Springs, CA.
Hunter, E. (1960). Brainwashing: From Pavlov to Powers. New York:
1. Thirty-three participants exceeded this maximum time limit of The Bookmailer.
10 min. Insko, C. A., Lind, E. A., & LaTour, S. (1976). Persuasion, recall, and
2. Across all four studies, none of the correlations between the thoughts. Representative Research in Social Psychology, 7, 66-78.
task perceptions items and the outcome variables reached significance. Insko, C. A., Turnbull, W., & Yandell, B. (1974). Facilitative and
3. For a self-control manipulation to be effective, the individual inhibiting effects of distraction on attitude change. Sociometry, 37,
must exert self-control for a sustained period of time. Female partici- 508-528.
pants were also included in this study; however, most women were Knowles, E. S., & Linn, J. A. (Eds.). (2004). Resistance and
unable to maintain their grip for very long. On average, the women persuasion. Mahwah, NJ: Lawrence Erlbaum.
held for little over 30 s (M = 44.18 s), whereas the men were able to Lee, A. Y., & Aaker, J. L. (2004). Bringing the frame into focus: The
hold for over 2 min (M = 125.78 s), t(18) = –2.34, p < .05. Thus, the influence of regulatory fit on processing fluency and persuasion.
self-control manipulation was only effective for men. This may likely Journal of Personality and Social Psychology, 86, 205-218.

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.
 Burkley / SELF-CONTROL AND PERSUASION 431

Lifton, R. J. (1961). Thought reform and the psychology of totalism. Taylor, K. (2004). Brainwashing: The science of thought control.
New York: Norton. Oxford, UK: Oxford University Press.
Macrae, C. N., Bodenhausen, G. V., Milne, A. B., & Jetten, J. (1994). Tormala, Z. L., Brinol, P., & Petty, R. E. (2006). When credibility
Out of mind but back in sight: Stereotypes on the rebound. attacks: The reverse impact of source credibility on persuasion.
Journal of Personality and Social Psychology, 67, 808-817. Journal of Experimental Social Psychology, 42, 684-691.
McGuire, W. J. (1964). Inducing resistance to persuasion: Some contem- Tormala, Z. L., & Petty, R. E. (2002). What doesn’t kill me makes me
porary approaches. In L. Berkowitz (Ed.), Advances in experimental stronger: The effects of resisting persuasion on attitude certainty.
social psychology (Vol. 1, pp. 191-229). New York: Academic Press. Journal of Personality and Social Psychology, 83, 1298-1313.
Muraven, M., & Baumeister, R. F. (2000). Self-regulation and deple- Twenge, J. M., & Baumeister, R. F. (2002). Self-control: A limited yet
tion of limited resources: Does self-control resemble a muscle? renewable resource. In Y. Kashima, M. Foddy, & M. Platow
Psychological Bulletin, 126, 247-259. (Eds.), Self and identity: Personal, social, and symbolic (pp. 57-70).
Muraven, M., Baumeister, B. F., & Tice, D. M. (1999). Longitudinal Mahwah, NJ: Lawrence Erlbaum.
improvement of self-regulation through practice: Building self- Vohs, K. D., Baumeister, R.F., & Ciarocco, N. (2005). Self-regulation
control strength through repeated exercise. Journal of Social and self-presentation: Regulatory resource depletion impairs
Psychology, 139, 446-457. impression management and effortful self-presentation depletes
Muraven, M., Shmueli, D., & Burkley, E. (2006). Conserving self- regulatory resources. Journal of Personality and Social Psychology,
control strength. Journal of Personality and Social Psychology, 91, 88, 632-657.
524-537. Vohs, K. D., & Heatherton, T. F. (2000). Self-regulatory failure: A
Muraven, M., & Slessareva, E. (2003). Mechanisms of self-control resource-depletion approach. Psychological Science, 11, 249-254.
failure: Motivation and limited resources. Personality and Social von Hippel, W., Silver, L. A., & Lynch, M. E. (2000). Stereotyping
Psychology Bulletin, 29, 894-906. against your will: The role of inhibitory ability in stereotyping and
Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self-control prejudice among the elderly. Personality and Social Psychology
as a limited resource: Regulatory depletion patterns. Journal of Bulletin, 26, 523-532.
Personality and Social Psychology, 74, 774-778. Wegner, D. M. (1992). You can’t always think what you want:
Petty, R. E., & Cacioppo, J. T. (1986a). The elaboration likelihood Problems in the suppression of unwanted thoughts. In M. Zanna
model of persuasion. In L. Berkowitz (Ed.), Advances in experi- (Ed.), Advances in experimental social psychology (Vol. 25,
mental social psychology (Vol. 19, pp. 123-205). New York: pp. 193-225). San Diego, CA: Academic Press.
Academic Press. Wegner, D. M., Schneider, D., Carter, S. R., & White, T. L. (1987).
Petty, R. E., & Cacioppo, J. T. (1986b). Communication and persua- Paradoxical effects of thought suppression. Journal of Personality
sion: Central and peripheral routes to attitude change. New York: and Social Psychology, 53, 5-13.
Springer-Verlag. Wheeler, S. C., Briñol, P., & Hermann, A. D. (2007). Resistance to
Petty, R. E., Cacioppo, J. T., & Goldman, R. (1981). Personal involvement persuasion as self-regulation: Ego-depletion and its effects on atti-
as a determinant of argument-based persuasion. Journal of Personality tude change processes. Journal of Experimental Social Psychology,
and Social Psychology, 41, 847-855. 43, 150-156.
Petty, R. E., & Krosnick, J. A. (Eds.). (1995). Attitude strength: Zimbardo, P., Snyder, M., Thomas, J., Gold, A., & Gurwitz, S.
Antecedents and consequences. Mahwah, NJ: Lawrence Erlbaum. (1970). Modifying the impact of persuasive communications with
Sobel, M. E. (1982). Asymptotic intervals for indirect effects in struc- external distraction. Journal of Personality and Social Psychology,
tural equations models. In S. Leinhart (Ed.), Sociological method- 16, 669-680.
ology 1982 (pp. 290-312). San Francisco: Jossey-Bass.
Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High self- Received May 7, 2007
control predicts good adjustment, less pathology, better grades, Revision accepted September 13, 2007
and interpersonal success. Journal of Personality, 72, 271-324.
Tannenbaum, P. H., Macauley, J. R., & Norris, E. L. (1966).
Principle of congruity and reduction of persuasion. Journal of
Personality and Social Psychology, 3, 233-238.

Downloaded from http://psp.sagepub.com at OKLAHOMA STATE UNIV on February 21, 2008

© 2008 Society for Personality and Social Psychology, Inc.. All rights reserved. Not for commercial use or unauthorized distribution.