Content uploaded by Veronika Job
Author content
All content in this area was uploaded by Veronika Job
Content may be subject to copyright.
Psychological Science
21(11) 1686 –1693
© The Author(s) 2010
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0956797610384745
http://pss.sagepub.com
Some of the most provocative and influential research of
the past decade has investigated the strength model of self-
control (e.g., Baumeister, Bratlavsky, Muraven, & Tice, 1998;
Baumeister, Vohs, & Tice, 2007). This model suggests that
acts of self-regulation consume a resource that is limited, leav-
ing people in a state of ego depletion and making them less
able to exert self-control on a subsequent task. The strength
model of self-control accounts for an impressive array of
empirical findings, including depletion effects on information
processing (Fischer, Greitemeyer, & Frey, 2008), intellectual
performance (Schmeichel, Vohs, & Baumeister, 2003), impres-
sion management (Vohs, Baumeister, & Ciarocco, 2005), and
violent responses to provocation by a partner (Finkel, DeWall,
Slotter, Oaten, & Foshee, 2009).
Some research, however, suggests that the exertion of self-
control does not invariably reduce subsequent self-control
(Moller, Deci, & Ryan, 2006; Tice, Baumeister, Shmueli, &
Muraven, 2007). For instance, people who are motivated by
incentives to control themselves may not show ego-depletion
effects (Muraven & Slessareva, 2003). Most relevant to the
present research are findings that expectancies about dimin-
ished self-control following exertion can moderate ego deple-
tion. In one study, some participants were told that performing
an effortful task (controlling their emotions) could improve
performance on a subsequent task (Martijn, Tenbült, Merckel-
bach, Dreezens, & de Vries, 2002). These participants showed
no decrease in subsequent self-control performance (squeez-
ing a handgrip).
Here, we ask a more general question: Does holding a
global theory that difficult tasks are energizing rather than
depleting prevent ego depletion and help people sustain self-
regulation? Specifically, we report a set of studies that tested
whether people’s implicit theories about self-control moderate
ego-depletion effects. Much research documents the effects of
implicit theories on self-related processes. These theories
include beliefs about the nature of human attributes, such as
whether intelligence and personality are fixed or malleable
(Blackwell, Trzesniewski, & Dweck, 2007; Dweck, 1999;
Molden & Dweck, 2006). In the context of self-regulation,
we propose that people differ in their implicit theories about
the availability and depletability of self-control resources
Corresponding Author:
Veronika Job, University of Zurich, Department of Psychology,
Binzmühlestrasse 14/6, 8050 Zürich, Switzerland
E-mail: v.job@psychologie.uzh.ch
Ego Depletion—Is It All in Your Head?
Implicit Theories About Willpower Affect
Self-Regulation
Veronika Job, Carol S. Dweck, and Gregory M. Walton
Stanford University
Abstract
Much recent research suggests that willpower—the capacity to exert self-control—is a limited resource that is depleted after
exertion. We propose that whether depletion takes place or not depends on a person’s belief about whether willpower is
a limited resource. Study 1 found that individual differences in lay theories about willpower moderate ego-depletion effects:
People who viewed the capacity for self-control as not limited did not show diminished self-control after a depleting experience.
Study 2 replicated the effect, manipulating lay theories about willpower. Study 3 addressed questions about the mechanism
underlying the effect. Study 4, a longitudinal field study, found that theories about willpower predict change in eating behavior,
procrastination, and self-regulated goal striving in depleting circumstances. Taken together, the findings suggest that reduced
self-control after a depleting task or during demanding periods may reflect people’s beliefs about the availability of willpower
rather than true resource depletion.
Keywords
implicit theories, self-control, self-regulation, ego depletion
Received 12/24/09; Revision accepted 4/17/10
Research Article
Ego Depletion—Is It All in Your Head? 1687
(or their “willpower”). Some people may think that self-
control is a limited resource, as described in the strength model
of self-control. Others may believe that self-control is not
limited and perhaps even that engaging in a strenuous task can
activate self-control resources. We call these the limited-
resource theory and the nonlimited-resource theory, respec-
tively. We suggest that these theories affect how well people
self-regulate when demands on self-control accumulate.
Overview of the Studies
Three experiments and a longitudinal study tested the effect of
implicit theories about willpower on ego depletion. The first
studies, using a traditional ego-depletion paradigm, measured
(Study 1) and manipulated (Study 2) implicit theories to test the
hypothesis that implicit theories moderate ego depletion. Study
3 examined mechanisms involved in the findings from Studies
1 and 2. For example, one possibility based on the strength
model of self-control is that people given a nonlimited-resource
theory perform well on a postdepletion task because they
“overuse” their resources, whereas those given a limited-resource
theory conserve and replenish their resources (Baumeister &
Vohs, 2007). If so, on a third demanding task, people given the
limited-resource theory should perform better than those given
the nonlimited-resource theory (Muraven, Shmueli, & Burkley,
2006). Study 3 also tested whether, following a demanding
task, people with the nonlimited-resource theory felt less
exhausted than those with the limited-resource theory or expe-
rienced the same level of exhaustion but were less affected by
exhaustion. Finally, Study 4 examined the effect of implicit
theories on self-regulation during a time of high self-regulatory
demands (i.e., during students’ final exams).
Study 1
Study 1 investigated whether individual differences in implicit
theories about willpower moderate ego depletion.
Method
Participants. Sixty students (42 females, 18 males) partici-
pated in a “study on stimulus detection and cognitive process-
ing” in exchange for course credit or $10.
Materials and procedure. First, participants completed six
items assessing implicit theories about willpower, specifically,
their theories about the effects of mental exertion. So as not to
arouse suspicion, the measure was embedded among several
other implicit-theory measures (e.g., theories of personality
and of intelligence). Items included “After a strenuous mental
activity your energy is depleted and you must rest to get it
refueled again” (limited-resource theory) and “Your mental
stamina fuels itself; even after strenuous mental exertion you
can continue doing more of it” (nonlimited-resource theory).
Participants responded using a 6-point rating scale (1 =
strongly agree, 6 = strongly disagree). Table S1, in the Supple-
mental Material available online, presents the full scale. Items
were scored so that higher values represent greater agreement
with the limited-resource theory. The scale was reliable (α =
.89), so the scores on the six items were averaged (M = 4.13,
SD = 0.84).
Participants then completed a “stimulus detection task.” This
task was adopted from previous research to manipulate ego
depletion (Baumeister et al., 1998; Tice et al., 2007; Wheeler,
Briñol, & Hermann, 2007). It consisted of two parts, each lasting
5 min. First, all participants were instructed to cross out each e on
a page of typewritten text. This task establishes a behavioral pat-
tern. Second, on a second page, some participants (nondepleting
condition) were again instructed to cross out every e. Others
(depleting condition) were asked to follow complex rules that
sometimes required them to inhibit the previously established
response (e.g., not to cross out an e followed by a vowel).
Next, participants completed a standard measure of ego
depletion—a Stroop task (Gailliot et al., 2007; Inzlicht,
McKay, & Aronson, 2006; Webb & Sheeran, 2003). Color
words (red, green, yellow, and blue) appeared on a computer
screen in a font color that was either congruent or incongruent
with their meaning. Participants completed 48 trials (24 incon-
gruent). In each, they were instructed to press a key marked
with the color the word was written in. The Stroop task is a
widely used measure of self-control because on incongruent
trials, the meaning of the word interferes with naming its color
and has to be suppressed for accurate identification of the font
color. Previous research has found ego-depletion effects on
performance on incongruent Stroop trials and not on congru-
ent trials (Inzlicht & Gutsell, 2007). Therefore, the primary
outcome was accuracy on incongruent trials.1
Results and discussion
Accuracy on each incongruent Stroop trial was coded (correct =
0, incorrect = 1). We then fit a logistic curve for each partici-
pant using a logistic hierarchical linear model (HLM). HLM
allowed us to control for covariates at the trial level and thus
provides a more precise estimate of participants’ latent proba-
bility of responding accurately than analysis of variance or
regression would have. Participants were more accurate when
they took longer to respond, and they became more accurate as
they completed more trials. To control for extraneous variation
caused by speed-accuracy trade-offs and order effects, we
included reaction time and trial order as trial-level predictors
in each model.
Participant-level predictors were ego-depletion condition
(nondepleting = 0, depleting = 1), implicit theories about will-
power (centered), and their interaction term.2 There was a
main effect of ego-depletion condition, β = 0.36, odds ratio
(OR) = 1.44, t(1433) = 6.71, p < .01. Participants were more
likely to make mistakes on the Stroop task after the depleting
task than after the nondepleting task, a finding that replicates
past research. However, as predicted, this main effect was
1688 Job et al.
qualified by an interaction with implicit theories, β = 0.28,
OR = 1.32, t(1433) = 3.88, p < .01. As displayed in Figure 1,
only participants with a limited-resource theory (+1 SD) showed
the usual ego-depletion pattern, making more mistakes after
the depleting task. Participants with a nonlimited-resource
theory (–1 SD) showed no difference in accuracy between the
depleting and nondepleting conditions.
To analyze the interaction, we conducted separate HLM
models for participants with a limited-resource theory and
those with a nonlimited-resource theory.3 These analyses con-
firmed that the difference between the depleting and the non-
depleting condition was significant for participants with a
limited-resource theory, β = 0.63, OR = 1.88, t(739) = 8.27,
p < .01, and nonsignificant for participants with a nonlimited-
resource theory, β = 0.04, OR = 1.04, t < 1.
The results support the hypothesis that implicit theories
about willpower moderate ego depletion. Only participants
with a limited-resource theory showed ego depletion. Partici-
pants with a nonlimited-resource theory showed no difference
between the depleting and nondepleting conditions.4
Study 2
In Study 2, we manipulated implicit theories about willpower
to test their causal effect.
Method
Participants. Forty-six students (27 females, 19 males) par-
ticipated in exchange for course credit or $10.
Procedure. First, we manipulated implicit theories about will-
power. Participants completed a biased questionnaire contain-
ing nine items formulated to foster agreement with either a
limited-resource theory (e.g., “Working on a strenuous mental
task can make you feel tired such that you need a break before
accomplishing a new task”) or a nonlimited-resource theory
(e.g., “Sometimes, working on a strenuous mental task can
make you feel energized for further challenging activities”).
Participants responded on a 6-point scale (1 = strongly agree,
6 = strongly disagree; α = .84). One-sample t tests comparing
the mean in each condition with the scale’s midpoint (3.50)
indicated that participants agreed with the suggested theory in
both the limited-resource-theory condition (M = 2.27, SD = 0.69),
t(23) = –8.74, p < .01, and the nonlimited-resource-theory
condition (M = 2.80, SD = 0.68), t(21) = –4.78, p < .01.
The rest of the study was identical to Study 1. Participants
completed what was described as a stimulus detection task
(which contained the ego-depletion manipulation) and then
the Stroop task.
Results and discussion
We ran a logistic HLM model with accuracy of responses on
incongruent Stroop trials as the dependent variable (correct = 0,
false = 1), again controlling for reaction time and order. Pre-
dictor variables were ego-depletion condition (nondepleting =
0, depleting = 1), implicit-theory condition (nonlimited-resource
theory = 0, limited-resource theory = 1), and their interac-
tion. As predicted, the interaction was significant, β = 1.15,
OR = 3.17, t(1097) = 8.47, p < .01. As displayed in Figure 2,
only participants led to adopt the limited-resource theory
showed ego depletion, making more mistakes after the
depleting task than after the nondepleting task. The opposite
pattern emerged for participants in the nonlimited-resource-
theory condition. Separate analyses of the two groups found
that the difference between the depleting and nondepleting
conditions was significant (though in opposite directions) in
both the limited-resource-theory condition, β = 0.72, OR =
2.06, t(571) = 10.53, p < .01, and the nonlimited-resource-
theory condition, β = –0.42, OR = 0.66, t(523) = –4.35,
p < .01.
The results show that manipulated theories about willpower
as either a limited or a nonlimited resource moderate ego
depletion, confirming that the moderating role of implicit the-
ories about willpower is causal. Interestingly, participants who
had been induced to hold a nonlimited-resource theory per-
formed worse after the nondepleting task than after the deplet-
ing task. It is intriguing to speculate that they were “depleted”
by boredom rather than by self-control, though this effect
needs to be replicated in future research.
Study 3
Study 3 was designed to test whether the findings of Study 2
would be replicated and to shed light on possible mechanisms
0
.02
.04
.06
.08
.10
Nondepleting Depleting
Depletion Condition
Probability of a Mistake
Nonlimited-Resource Theory
Limited-Resource Theory
Fig. 1. Results from Study 1: probability of making a mistake on incongruent
trials of the Stroop task as a function of ego-depletion condition and implicit
theories about willpower. The limited-resource-theory group represents
participants 1 standard deviation above the mean on the implicit-theories
measure. The nonlimited-resource-theory group represents participants 1
standard deviation below the mean on the implicit-theories measure.
Ego Depletion—Is It All in Your Head? 1689
underlying the observed effect. First, we tested whether par-
ticipants with a nonlimited-resource theory “overuse” their
resources on the task following the depletion manipulation,
leaving them depleted for a third task (see Muraven et al.,
2006). Therefore, we assessed performance on two successive
tasks after the depletion manipulation—Stroop performance
and IQ performance. Second, we examined participants’ sub-
jective experience of exhaustion. We tested (a) whether
implicit theories about willpower changed the degree to which
the initial self-control task was experienced as exhausting and
(b) whether implicit theories changed the degree to which the
subjective experience of the task as exhausting undermined
subsequent performance.
Method
Participants. Seventy-seven students (53 females, 24 males)
participated in exchange for course credit or $10.
Procedure. Participants were randomly assigned to complete
one of the two versions of the biased questionnaire used in
Study 2. Next, they completed the e-crossing task containing
the depletion manipulation. This task was followed by a ques-
tion assessing subjective exhaustion: “How exhausting was
the stimulus detection task for you?” (1 = not at all, 9 = very
much). Participants then completed the Stroop task. Finally,
they completed eight challenging IQ problems, as previous
research has shown that intellectual performance is particu-
larly sensitive to ego depletion (Schmeichel et al., 2003). In
each problem, participants were given 20 s to select which of
five figures best fit in a series of figures.5
Results and discussion
We ran the same logistic HLM model described in Study 2 on
Stroop performance. The interaction between implicit-theory
condition and ego-depletion condition was significant, β =
0.25, OR = 1.29, t(1842) = 2.62, p < .01. As displayed in
Figure 3a, participants in the limited-resource-theory condi-
tion exhibited ego depletion, making more mistakes after the
depleting task than after the nondepleting task. There was no
difference between ego-depletion conditions for participants
in the nonlimited-resource-theory condition. Separate analy-
ses for the two implicit-theory groups found that the difference
between the nondepleting and depleting conditions was sig-
nificant in the limited-resource-theory condition, β = 0.59,
OR = 1.81, t(979) = 8.59, p < .01, but not in the nonlimited-
resource-theory condition, t(859) < 1.
Analysis of IQ performance yielded the same pattern. The
interaction between implicit-theory condition and ego-depletion
condition was significant, β = 0.48, OR = 1.62, t(610) = 2.81,
p < .01. As displayed in Figure 3b, participants in the limited-
resource-theory condition made more mistakes after the
depleting task than after the nondepleting task, β = 0.33, OR =
1.39, t(324) = 2.85, p < .01. The performance of participants in
the nonlimited-resource-theory condition did not vary by
depletion condition, t(284) < 1.
Next, we examined subjective exhaustion. First, we tested
whether theories about willpower affected participants’ expe-
rience of exhaustion. We conducted a 2 (implicit-theory condi-
tion) × 2 (ego-depletion condition) analysis of variance. Only
the main effect of ego-depletion condition was significant,
F(1, 76) = 8.17, p < .01, η2 = .10. Participants experienced the
e-crossing task as more exhausting in the depleting condition
(M = 4.59, SD = 2.29) than in the nondepleting condition (M =
3.31, SD = 2.03). Neither the main effect of implicit-theory
condition nor the interaction was significant, Fs < 1. Thus, the
induced theory of willpower did not affect the degree to which
participants experienced the e-crossing task as exhausting.
Second, we tested whether theories about willpower moder-
ated the relationship between felt exhaustion and subsequent
performance. We ran the same logistic HLM model on Stroop
performance as initially, but replacing ego-depletion condition
with self-reported exhaustion (centered). The interaction
between self-reported exhaustion and implicit-theory condition
was significant, β = 0.17, OR = 1.19, t(1842) = 3.58, p < .01
(see Fig. 3c). Separate analysis for each implicit-theory condi-
tion showed that greater self-reported exhaustion predicted
more mistakes in the limited-resource-theory condition, β = 0.09,
OR = 1.09, t(979) = 6.18, p < .01, but not in the nonlimited-
resource-theory condition, t(859) < 1. We conducted the same
analyses on IQ performance. Again, the interaction was signifi-
cant, β = 0.22, OR = 1.25, t(610) = 2.67, p < .01.
Finally, we examined whether the altered relationship
between exhaustion and performance mediated the effect of
implicit theories on ego depletion (see Fig. 4). We ran logistic
HLM models with Stroop and IQ performance as dependent
0
.02
.04
.06
.08
.10
.12
Nondepleting Depleting
Probability of a Mistake
Nonlimited-Resource Theory
Limited-Resource Theory
Depletion Condition
Fig. 2. Results from Study 2: probability of making a mistake on incongruent
trials of the Stroop task as a function of ego-depletion condition and implicit-
theory condition.
1690 Job et al.
variables. Predictors were the two experimental conditions
(implicit-theory and ego-depletion conditions), their interac-
tion, self-reported exhaustion, and the interaction between
self-reported exhaustion and implicit-theory condition. For
accuracy on the Stroop task, the interaction between self-
reported exhaustion and implicit-theory condition remained
significant, β = 0.18, OR = 1.20, t(1840) = 3.55, p < .01, but
the interaction between ego-depletion condition and implicit-
theory condition was no longer significant, β = 0.14, OR = 1.15,
t(1840) = 1.56, p > .10, Sobel test: z = 2.20, p < .05. Similarly,
for accuracy on the IQ problems, only the interaction between
self-reported exhaustion and implicit-theory condition remained
significant when both self-reported exhaustion and ego-depletion
condition were included in the analysis, β = 0.19, OR = 1.21,
t(609) = 2.15, p < .05. The results suggest that self-reported
exhaustion in interaction with the induced resource theory medi-
ated the Ego-Depletion Condition × Implicit-Theory Condition
effect on both Stroop performance and IQ performance.
In sum, Study 3 yielded no evidence that a nonlimited-
resource theory led participants to overuse self-control
a
0
.01
.02
.03
.04
.05
.06
.07
.08
Probability of a Mistake: Stroop Task
Probability of a Mistake: Stroop Task
b
c
.20
.25
.30
.35
.40
.45
Probability of a Mistake: IQ Task
0
.01
.02
.03
.04
.05
.06
.07
.08
Low High
Self-Reported Exhaustion
Nondepleting Depleting
Depletion Condition
Nondepleting Depleting
Depletion Condition
Nonlimited-Resource Theory Limited-Resource Theory
Fig. 3. Results from Study 3: probability of making a mistake on (a) incongruent trials of the Stroop task and (b) the IQ task as a function of ego-
depletion condition and implicit-theory condition and (c) probability of making a mistake on incongruent trials of the Stroop task as a function of
experienced exhaustion and implicit-theory condition. Low and high exhaustion represent participants 1 standard deviation below and above the mean,
respectively.
Ego Depletion—Is It All in Your Head? 1691
resources. After a depletion manipulation, participants in the
nonlimited-resource-theory condition showed no evidence of
resource depletion even on a series of tasks. Further, theories
about willpower did not affect the experience of the depleting
task as exhausting. Instead, these theories affected the rela-
tionship between the experience of exhaustion and subsequent
performance. People with a nonlimited-resource theory expe-
rienced the depleting task as just as exhausting as those with a
limited-resource theory, but for them, exhaustion did not
undermine subsequent performance.
Study 4
Studies 1 through 3 showed that measured and induced theo-
ries about willpower as a limited versus nonlimited resource
moderate ego depletion in a classic laboratory paradigm. Study
4 examined the effect of implicit theories about willpower on
people’s everyday self-regulation and goal striving. Given
our previous findings, we hypothesized that the nonlimited-
resource theory, compared with the limited-resource theory,
would predict better self-regulation during times of height-
ened stress and self-regulatory demands. Therefore, we tracked
college students across three time points, the last of which was
during final exams. We expected implicit theories about
willpower at the second time point to predict self-regulation
during final exams, but we did not expect implicit theories
to predict self-regulation at the prior time points, when self-
regulatory demands were lower.
Method
Participants and procedure. An initial Web questionnaire
was completed by 101 undergraduates in April, at the begin-
ning of the academic quarter (Time 1, or T1). Of these partici-
pants, a subsample of 44 completed measures at the second
time point, in May (T2). In the critical comparison, 41 of those
44 participants (30 women, 11 men) also completed measures
at the third time point, during final exams in the first week of
June (T3). Participants in the final sample did not differ on any
measure from participants who completed measures at T1 but
did not continue in the study.
Measures. The same measures were assessed at each time
point. First, we assessed individual differences in implicit the-
ories about willpower using 12 items: the 6 items used previ-
ously plus 6 items that assessed resistance to temptation as a
further aspect of self-control (see Table S1). Items were coded
so that higher values represent agreement with a limited-
resource theory. The scale was internally reliable, α(T1) = .77,
α(T2) = .86, α(T3) = .89, and showed high reliability over time
(test-retest rs > .77).
Second, we assessed participants’ everyday efforts at self-
regulation by examining reported consumption of unhealthy
foods and reported procrastination. Participants were asked
how often in the previous week they had consumed several
high-fat or high-sugar foods and drinks. They were also asked
how often they had engaged in various nonacademic activities
rather than studying (e.g., “How often did you watch TV
instead of studying?”). Responses were made on 7-point scales
(1 = never, 7 = two or more times per day).
Third, we assessed self-regulation with respect to a per-
sonal goal using a procedure developed by Brunstein, Schul-
theiss, and Grässmann (1998). At T1, participants listed a
personal goal that involved challenge and achievement. This
goal was presented to participants at each time point, and they
were asked how well they had regulated themselves in pursu-
ing it (five items, e.g., “I was often not in the mood to do
something for this goal”; 1 = strongly disagree, 5 = strongly
agree), α(T1) = .69, α(T2) = .86, α(T3) = .81.
Results and discussion
The hypothesis that implicit theories about willpower affect
self-regulation when demands on self-regulation are high
implies that a limited-resource theory at T2 should predict
worse self-regulation at T3. To test this hypothesis, we
regressed self-regulation variables at T3 on implicit theories at
T2, controlling for self-regulation at T2. These analyses
revealed that a limited-resource theory at T2 predicted worse
self-regulation on all three measures at the stressful time point,
T3—consumption of unhealthy food: b = 0.41, ΔR2 = .16,
ΔF(1, 38) = 11.76, p < .01; procrastination rather than study-
ing: b = 0.29, ΔR2 = .08, ΔF(1, 38) = 8.11, p < .01; and
E-Crossing Task:
Depleting vs. Nondepleting
Probability of a Mistake:
Stroop Task
Perceived
Exhaustion
(β = .25*)
β = .14
× Theory
× Theory
β = .18*
β = .31*
Fig. 4. Analysis of self-reported exhaustion as a mediator of the interactive effect of ego-depletion condition
and implicit-theory condition on Stroop performance. Asterisks indicate significant coefficients (*p < .01). In
the bottom path, the beta in parentheses refers to the analysis without the mediator.
1692 Job et al.
self-regulation with respect to personal goal striving: b =
–0.27, ΔR2 = .06, ΔF(1, 38) = 5.80, p < .05. A limited-resource
theory at T2 also predicted lower scores on a composite of all
three self-regulation measures at T3 (created by reverse-scoring
personal-goal self-regulation and then standardizing and aver-
aging the three indicators), b = 0.51, ΔR2 = .20, ΔF(1, 38) =
24.71, p < .001. The more participants agreed with a limited-
resource theory at T2, the more they reported eating unhealthy
food, procrastinating, and self-regulating ineffectively while
pursuing an important goal at T3.
Next, we tested the reverse causal relationship—from self-
regulation at T2 to implicit theories at T3. Implicit theories at
T3 were regressed on T2 self-regulation, controlling for T2
implicit theories. There was no significant relationship
between any T2 self-regulatory variable and T3 implicit theo-
ries, ΔFs(1, 38) < 1.30.
The same analyses were repeated with the same sample using
self-regulation variables and implicit theories at T1 and T2. As
predicted, no relationship in either direction was significant.
The results support the hypothesis that the nonlimited-
resource theory of willpower, compared with the limited-
resource theory, predicts better self-control during periods of
heightened stress and self-regulatory demands. Of course, the
results do not imply that a nonlimited-resource theory will
always produce better self-regulation. In times of low stress,
the limited-resource theory could prove superior (see Study 2).
General Discussion
In a classic laboratory paradigm, only people who thought of
or who were led to think of willpower as a limited resource
showed ego depletion (Studies 1–3). By contrast, for people
who had or were led to adopt a nonlimited-resource theory, a
demanding initial task did not undermine subsequent perfor-
mance. In one study, the demanding task actually raised their
subsequent performance. Further, Study 4 showed that the
more people held a limited-resource theory, the poorer was
their self-regulation in the real world when demands on self-
regulation were high.
According to the strength model of self-control, motiva-
tional factors that counteract ego-depletion effects (e.g., incen-
tives or expectancies) may do so because motivation can
compensate for a lack of self-regulatory strength to some
degree (Baumeister & Vohs, 2007; Muraven et al., 2006). It is
argued that this motivation can lead people to expend more of
the depleted psychological resource, leaving less available for
subsequent tasks. This process did not account for the effects
of implicit theories about willpower. In Study 3, people led to
adopt a nonlimited-resource theory performed better than peo-
ple with a limited-resource theory not only on the task imme-
diately following the depleting task (the Stroop task), but also
on a third task (IQ problems).
Study 3 also suggests a mechanism for ego depletion
and for how implicit theories sustain self-control. Perceived
exhaustion mediated the effects of the depletion manipulation
in the limited-resource condition. This finding is consistent
with research showing that depletion effects are better pre-
dicted by people’s perception of depletion than by an actual
depletion experience (Clarkson, Hirt, Jia, & Alexander, 2010).
The present research suggests that implicit theories changed
how people responded given their level of felt exhaustion on
the initial task. People led to adopt a limited-resource theory
performed worse the more exhausted they felt. But for people
led to adopt a nonlimited-resource theory, there was no rela-
tionship between perceived exhaustion and subsequent perfor-
mance. For them, exhaustion was not a sign to reduce effort.
Taken together, the results suggest that in some cases, ego
depletion may result not from a true lack of resources after an
exhausting task, but from people’s beliefs about their resources.
We do not question that biological resources contribute to suc-
cessful self-control (Gailliot & Baumeister, 2007; Gailliot
et al., 2007). But these resources may be less limited than is
commonly supposed. A key direction for future research is to
examine how top-down processes (e.g., theories about will-
power) and bottom-up processes (e.g., the availability of glu-
cose) interact to affect self-control.
Psychological research has the power to shape how people
think about themselves (Herman, 1996). People who learn
about the strength model of self-control may conclude that
they are at the mercy of a fixed, physiological process that
limits their willpower. It is important that people understand
that their own beliefs about willpower as a limited or nonlim-
ited resource can affect their self-regulation. It is also impor-
tant that psychologists appreciate the impact of powerful and
widely shared lay theories about the self and distinguish their
effects from seemingly immutable biologically driven processes.
Acknowledgments
We thank Katharina Bernecker, Alena Friedrich, Yousuf Haq,
Krishna Savani, and Rebecca Wheeler for invaluable assistance and
Geoffrey Cohen and the Dweck-Walton Lab for helpful comments.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Funding
This research was supported by a Swiss National Science Foundation
fellowship (PBZHP1-123313) to Veronika Job.
Supplemental Material
Additional supporting information may be found at http://pss.sagepub
.com/content/by/supplemental-data
Notes
1. In all three studies using the Stroop task, the described pattern of
results held when we combined incongruent and congruent trials but,
as expected, not when only congruent trials were analyzed.
Ego Depletion—Is It All in Your Head? 1693
2. Age (centered) was included as a covariate on the participant level
in all three studies using the Stroop task, given research showing that
it relates to Stroop performance. All analyses used the population-
average model.
3. These two groups were created using a median split to allow for
the calculation of contrasts within HLM.
4. A possible alternative explanation is that people with a nonlimited-
resource theory have better self-control than people with a limited-
resource theory. However, a pilot study (N = 65) did not find a negative
relationship between a limited-resource theory and trait self-control
(Schwarzer, Diehl, & Schmitz, 1999), r = .17, p > .20.
5. To test whether effort took on a different (positive vs. negative)
meaning for the two implicit-theory groups, we administered a brief
word categorization task following the depletion manipulation. The
task did not yield clear results and so is not discussed further.
References
Baumeister, R.F., Bratlavsky, E., Muraven, M., & Tice, D.M. (1998).
Ego depletion: Is the active self a limited resource? Journal of
Personality and Social Psychology, 74, 1252–1265.
Baumeister, R.F., & Vohs, K.D. (2007). Self-regulation, ego deple-
tion, and motivation. Social and Personality Psychology Com-
pass, 1, 115–128.
Baumeister, R.F., Vohs, K.D., & Tice, D.M. (2007). The strength
model of self-control. Current Directions in Psychological Sci-
ence, 16, 351–355.
Blackwell, L.S., Trzesniewski, K.H., & Dweck, C.S. (2007). Implicit
theories of intelligence predict achievement across an adolescent
transition: A longitudinal study and an intervention. Child Devel-
opment, 78, 246–263.
Brunstein, J.C., Schultheiss, O.C., & Grässmann, R. (1998). Personal
goals and emotional well-being: The moderating role of motive
dispositions. Journal of Personality and Social Psychology, 75,
494–508.
Clarkson, J.J., Hirt, E.R., Jia, L., & Alexander, M.B. (2010). When
perception is more than reality: The effects of perceived versus
actual resource depletion on self-regulatory behavior. Journal of
Personality and Social Psychology, 98, 29–46.
Dweck, C.S. (1999). Self-theories: Their role in motivation, person-
ality and development. Philadelphia, PA: Taylor & Francis.
Finkel, E.J., DeWall, C.N., Slotter, E.B., Oaten, M., & Foshee, V.A.
(2009). Self-regulatory failure and intimate partner violence
perpetration. Journal of Personality and Social Psychology, 97,
483–499.
Fischer, P., Greitemeyer, T., & Frey, D. (2008). Self-regulation
and selective exposure: The impact of depleted self-regulation
resources on confirmatory information processing. Journal of
Personality and Social Psychology, 94, 382–395.
Gailliot, M.T., & Baumeister, R.F. (2007). The physiology of will-
power: Linking blood glucose to self-control. Personality and
Social Psychology Review, 11, 303–327.
Gailliot, M.T., Baumeister, R.F., DeWall, C.N., Maner, J.K., Plant,
E.A., Tice, D.M., et al. (2007). Self-control relies on glucose as a
limited energy source: Willpower is more than a metaphor. Jour-
nal of Personality and Social Psychology, 92, 325–336.
Herman, E. (1996). The romance of American psychology: Political
culture in the age of experts. Berkeley, CA: University of Cali-
fornia Press.
Inzlicht, M., & Gutsell, J.N. (2007). Running on empty: Neural sig-
nals for self-control failure. Psychological Science, 18, 933–937.
Inzlicht, M., McKay, L., & Aronson, J. (2006). Stigma as ego deple-
tion: How being the target of prejudice affects self-control. Psy-
chological Science, 17, 262–269.
Martijn, C., Tenbült, P., Merckelbach, H., Dreezens, E., & de Vries,
N.K. (2002). Getting a grip on ourselves: Challenging expecta-
tions about loss of energy after self-control. Social Cognition, 20,
441–460.
Molden, D.C., & Dweck, C.S. (2006). Finding “meaning” in psychol-
ogy: A lay theories approach to self-regulation, social perception,
and social development. American Psychologist, 61, 192–203.
Moller, A.C., Deci, E.L., & Ryan, R.M. (2006). Choice and ego-
depletion: The moderating role of autonomy. Personality and
Social Psychology Bulletin, 32, 1024–1036.
Muraven, M., Shmueli, D., & Burkley, E. (2006). Conserving self-
control strength. Journal of Personality and Social Psychology,
91, 524–537.
Muraven, M., & Slessareva, E. (2003). Mechanism of self-control
failure: Motivation and limited resource. Personality and Social
Psychology Bulletin, 29, 894–906.
Schmeichel, B.J., Vohs, K.D., & Baumeister, R.F. (2003). Intellectual
performance and ego depletion: Role of the self in logical reason-
ing and other information processing. Journal of Personality and
Social Psychology, 85, 33–46.
Schwarzer, R., Diehl, M., & Schmitz, G.S. (1999). Self-Regulation
Scale. Retrieved October 28, 2006, from http://www.fu-berlin.de/
gesund/skalen
Tice, D.M., Baumeister, R.F., Shmueli, D., & Muraven, M. (2007).
Restoring the self: Positive affect helps improve self-regulation
following ego depletion. Journal of Experimental Social Psy-
chology, 43, 379–384.
Vohs, K.D., Baumeister, R.F., & Ciarocco, N.J. (2005). Self-regulation
and self-presentation: Regulatory resource depletion impairs
impression management and effortful self-presentation depletes
regulatory resources. Journal of Personality and Social Psychol-
ogy, 88, 632–657.
Webb, T.L., & Sheeran, P. (2003). Can implementation intentions
help to overcome ego-depletion? Journal of Experimental Social
Psychology, 53, 5–13.
Wheeler, S.C., Briñol, P., & Hermann, A.D. (2007). Resistance to
persuasion as self-regulation: Ego-depletion and its effects on
attitude change processes. Journal of Experimental Social Psy-
chology, 43, 150–156.
