Personality and Social Psychology

Bulletin [Link]

Self-Compassion Increases Self-Improvement Motivation

Juliana G. Breines and Serena Chen
Pers Soc Psychol Bull 2012 38: 1133 originally published online 29 May 2012
DOI: 10.1177/0146167212445599

The online version of this article can be found at:

[Link]

Published by:

[Link]

On behalf of:

Society for Personality and Social Psychology

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

Email Alerts: [Link]

Subscriptions: [Link]

Reprints: [Link]

Permissions: [Link]

>> Version of Record - Aug 3, 2012

OnlineFirst Version of Record - May 29, 2012

What is This?

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
445599
es and ChenPersonality and Social Psychology Bulletin
PSPXXX10.1177/0146167212445599Brein

Personality and Social

Self-Compassion Increases Psychology Bulletin

38(9) 1133­–1143
© 2012 by the Society for Personality
Self-Improvement Motivation and Social Psychology, Inc
Reprints and permission:
[Link]/[Link]
DOI: 10.1177/0146167212445599
[Link]

Juliana G. Breines1 and Serena Chen1

Abstract
Can treating oneself with compassion after making a mistake increase self-improvement motivation? In four experiments, the
authors examined the hypothesis that self-compassion motivates people to improve personal weaknesses, moral transgressions,
and test performance. Participants in a self-compassion condition, compared to a self-esteem control condition and either
no intervention or a positive distraction control condition, expressed greater incremental beliefs about a personal weakness
(Experiment 1); reported greater motivation to make amends and avoid repeating a recent moral transgression (Experiment
2); spent more time studying for a difficult test following an initial failure (Experiment 3); exhibited a preference for upward
social comparison after reflecting on a personal weakness (Experiment 4); and reported greater motivation to change the
weakness (Experiment 4). These findings suggest that, somewhat paradoxically, taking an accepting approach to personal
failure may make people more motivated to improve themselves.

Keywords
self-compassion, self-improvement, motivation, effort, self-esteem

Received August 6, 2011; revision accepted February 15, 2012

Self-compassion is seen as an especially adaptive self-attitude be compassionate toward yourself even if you don’t feel
due to its positive association with multiple aspects of psy- very good about yourself, just as you can be compassionate
chological well-being. Researchers have raised a concern, toward another person even if you disapprove of their
however, that self-compassion might at times increase com- actions. In addition, a self-compassionate approach avoids
placency and undermine motivation to correct mistakes both extremes of self-deprecation and self-enhancement
(e.g., Baker & McNulty, 2011). We propose that self-compassion characteristic of low and high self-esteem, respectively.
may actually increase self-improvement motivation given Research suggests that self-compassionate people react to
that it encourages people to confront their mistakes and lab-based stressors in more balanced ways, showing lower
weaknesses without either self-deprecation or defensive levels of negative affect as well as more realistic self-
self-enhancement. appraisals (Leary, Tate, Adams, Allen, & Hancock, 2007;
Neff, Kirkpatrick, & Rude, 2007). Self-compassion also
predicts greater self-worth stability and lower narcissism
Self-Compassion than self-esteem (Neff, 2003b; Neff & Vonk, 2009).
Self-compassion has been defined as a self-attitude that
involves treating oneself with warmth and understanding in
difficult times and recognizing that making mistakes is part Self-Compassion and Self-
of being human (Neff, 2003a). Self-compassion is associ- Improvement
ated with numerous aspects of well-being, including higher Prior research suggests that self-compassion is related to
levels of positive affect, optimism, and happiness (Neff, multiple aspects of well-being, but its role in self-improvement
Rude, & Kirkpatrick, 2007), lower levels of anxiety and motivation is less clear. Does self-compassion help people
depression (Neff, 2003b), and better romantic relationship
1
functioning, at least for women and conscientious men University of California, Berkeley, USA
(Baker & McNulty, 2011).
Corresponding Author:
Self-compassion has been distinguished from self- Juliana Breines, Department of Psychology, University of California, 3210
esteem conceptually and empirically. Unlike self-esteem, Tolman Hall, Berkeley, CA 94720-1650, USA
self-compassion is nonevaluative. In other words, you can Email: jbreines@[Link]

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
1134 Personality and Social Psychology Bulletin 38(9)

grow and learn from their mistakes, or does it make them 2011; Powers, Koestner, & Zuroff, 2007; Rice & Stuart,
complacent and unmotivated to change? 2010). For example, self-criticism is negatively associated
On one hand, self-compassion might hinder self- with goal progress due to its positive association with rumi-
improvement efforts by making people lazy and overly will- nation and procrastination (Powers et al., 2011). Examining
ing to let themselves “off the hook.” Researchers have both sides at once, Kim et al. (2010) found that both inflated
questioned whether self-compassion might interfere with and deflated self-assessments of academic performance led
self-improvement by curtailing self-criticism, a component of to self-handicapping and poor future performance.
perfectionism, given that perfectionism is positively associ- Realistic self-appraisal is not an easy task, however, espe-
ated with achievement (Baker & McNulty, 2011; Witcher, cially when the reality is harsh. We hypothesized that the
Alexander, Onwuebuzie, Collins, & Witcher, 2007). Other atmosphere of warmth and understanding generated by self-
research suggests, however, that the most adaptive forms of compassion may help people acknowledge a need for
perfectionism are those that involve high personal standards improvement without sinking into despair or feeling over-
but not self-criticism (Blatt, 1995; Grzegorek, Slaney, Franze, whelmed by anxiety about the possibility of failure (Neff,
& Rice, 2004; Rice & Stuart, 2010). Although self-compas- Hseih, & Dejitthirat, 2005). Research in the forgiveness lit-
sion is associated with lower self-criticism, it is not incompat- erature lends support to this idea, suggesting that forgiveness
ible with holding high personal standards (Neff, 2003b). generally reduces the likelihood of repeat offenses, in part
Research on self-forgiveness similarly suggests that because it helps the transgressor feel like relationship repair
being too easy on oneself can have costs, such as reduced is possible (Wallace, Exline, & Baumeister, 2008). We
empathy (Hall & Fincham, 2008) and reduced engagement hypothesized that the benefits of self-compassion may oper-
in reparative behaviors (Exline, Root, Yadavalli, Martin, & ate similarly, opening up the possibility of change and self-
Fisher, 2011). On the whole, however, self-forgiveness is improvement in a nonthreatening way.
seen as a healthy process, promoting repentance and other Finally, prior research has linked self-compassion to
prosocial behaviors, as long as self-forgivers take responsi- traits and behaviors that are related to self-improvement. In
bility for their actions and experience remorse (Fisher & the academic domain, trait self-compassion was found to
Exline, 2006). Leary et al. (2007) showed that inducing self- be positively associated with self-efficacy (Iskender, 2009)
compassion, which is related to self-forgiveness but is a and negatively associated with procrastination (Williams,
broader construct, led participants to take more, rather than Stark, & Foster, 2008). Self-compassionate individuals
less, responsibility for their role in a negative event. Overall, have also been found to hold greater mastery as opposed to
then, although some have raised the possibility that self- performance goals, a relationship mediated by self-com-
compassion can be detrimental to personal growth, there are passionate participants’ lower fear of failure and greater
reasons to suspect this might not be the case. perceived competence (Neff et al., 2005). In addition, as
In fact, we hypothesized that self-compassion may actu- mentioned previously, a lab experiment showed that par-
ally facilitate self-improvement. Why would this be? Self- ticipants who were induced to feel self-compassion as
compassion is associated with realistic self-appraisal (Leary opposed to self-esteem were more willing to acknowledge
et al., 2007), which in turn predicts growth-related outcomes their role in a negative event (Leary et al., 2007, Study 5).
(Kim, Chiu, & Zou, 2010). In this way, self-compassion In the health domain, one study showed that trait self-compas-
stands in contrast to self-enhancement, a common response sion is positively correlated with intrinsic exercise motiva-
to self-threatening information (Campbell & Sedikides, tion (Magnus, Kowalski, & McHugh, 2010), and inducing
1999; Sedikides, Gaertner, & Toguchi, 2003). Although self- self-compassion has been shown to (a) attenuate overeating
enhancement may benefit certain aspects of well-being (e.g., after a dieting “failure” among participants prone to disor-
Taylor & Brown, 1988), research suggests that it is less help- dered eating behaviors (Adams & Leary, 2007) and (b)
ful when it comes to learning and growth (e.g., Baumeister, reduce smoking for a subset of participants (Kelly, Zuroff,
Campbell, Kreuger, & Vohs, 2003; Kim et al., 2010; Kwan, Foa, & Gilbert, 2009). In romantic relationships, trait self-
John, Robin, & Kuang, 2008; Robins & Beer, 2001; Tice & compassion is positively correlated with motivation to correct
Baumeister, 1990). For example, Kwan et al. (2008) found interpersonal mistakes among men high in conscientious-
that self-enhancement was associated with greater defen- ness and among women (Baker & McNulty, 2011, Studies
siveness, lower resilience, and poor academic performance. 1 and 3).
It should be noted, however, that evidence for the effect of
self-enhancement on growth is mixed, as some suggest that
it may be beneficial (e.g., Alicke & Sedikides, 2009). The Present Experiments
Responding to self-image threats with self-deprecation, Research suggests that self-compassion is associated with
on the other hand, seems to be no better for self-improvement holding realistic self-appraisals (Leary et al., 2007), which in
(e.g., Fisher & Exline, 2006; Lyubomirsky, Tucker, Caldwell, turn have been linked to self-improvement motivation.
& Berg, 1999; Powers, Koestner, Lacaille, Kwan, & Zuroff, Building on this work, as well as a recent thread of findings

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
Breines and Chen 1135

suggesting that self-compassion is associated with self- & Dweck, 1978, 1980; Elliott & Dweck, 1988; Rhodewalt,
improvement-related traits and behaviors, the present series 1994; Robins & Pals, 2002).
of studies systematically tested the hypothesis that self-
compassion increases self-improvement motivation. Whereas
the majority of past work bearing on the link between self- Method
compassion and self-improvement has been either correla- Participants. A total of 69 undergraduates (52% female) par-
tional or has focused on a single domain (e.g., smoking ticipated for course credit. Of them, 1 participant was excluded
behavior), in the current studies we experimentally manipu- from the analyses for noncompliance with the manipulation
lated self-compassion and examined its effects on self- instructions, leaving a final sample of 68. Fifty-one percent of
improvement across multiple domains, including domains participants identified as Asian American, 16% as European
that have not yet been assessed experimentally or at all. American, 13% as Latino/a, and 10% as “Other.” Four partici-
Finally, because self-compassion and self-esteem are closely pants’ demographics information was missing. Participants
related both conceptually and empirically, we included a self- ranged in age from 18 to 43 years (M = 20.9, SD = 4.1).
esteem control condition across experiments to assess the Procedure. Participants were told that the purpose of the
unique effects of self-compassion—a feature of the current study was to understand the different ways that people think
research absent in nearly all prior work in this area (for an about personal weaknesses. Participants first identified, in
exception, see Leary et al., 2007, Study 5). one to two sentences, what they considered to be their big-
More concretely, across four experiments, we tested the gest weakness or shortcoming. Because self-compassion is
hypothesis that self-compassion increases the belief that most relevant in situations that elicit feelings of shame and
shortcomings can be changed (Experiment 1), the desire to self-criticism (Neff, 2003a), the instructions emphasized
make amends and avoid repeating a moral transgression that participants should select a weakness that made them
(Experiment 2), effort in studying for a subsequent test fol- feel bad about themselves.
lowing an initial failure (Experiment 3), preference for Next, participants were randomly assigned to one of three
upward social comparison (Experiment 4), and motivation conditions. In the self-compassion reflection condition, par-
to improve a personal weakness (Experiment 4). In ticipants were instructed to write for 3 min in response to the
Experiments 1, 2, and 4, self-compassion was induced by following prompt: “Imagine that you are talking to yourself
having participants write a paragraph to themselves express- about this weakness from a compassionate and understanding
ing kindness and understanding regarding their transgres- perspective. What would you say?” (adapted from Leary
sion or weakness. In Experiment 3, self-compassion was et al., 2007, Study 5). In the first control condition, a self-
induced via a statement embedded in the instructions fol- esteem reflection, participants were instructed to “Imagine
lowing an initial test. Because previous research suggests that you are talking to yourself about this weakness from a
that positive mood can increase self-improvement motiva- perspective of validating your positive (rather than negative)
tion (e.g., Fishbach & Labroo, 2007; Raghunathan & Trope, qualities.” The self-esteem reflection instructions were simi-
2002), we controlled for the effects of post-manipualtion lar to the self-compassion instructions in that they asked par-
positive affect. Across experiments, we expected greater ticipants to address themselves and to reflect on the weakness
self-improvement motivation among participants in the self- they identified, but the emphasis was on self-validation rather
compassion condition compared to participants in the two than self-compassion. This allowed us to ensure that our
control conditions (self-esteem and either no intervention or manipulation was isolating something specific to self-
positive distraction). compassion, not just positive self-talk in general. In a second
control condition, participants did not receive any reflection
instructions after identifying a personal weakness.
Experiment 1 After completing the manipulation, participants filled out
In the first experiment, we examined whether self-compassion a brief measure of positive affect. Following the stem,
increases a mind-set that is theoretically and empirically “Right now, how much do you feel . . . ,” positive affect was
related to self-improvement: incremental beliefs. Incremental assessed with three items: content, sad (reverse), and upset
beliefs entail a view that some aspect of the self (e.g., intel- (reverse; α = .69; M = 5.28, SD = 1.15). Ratings were made
ligence, personality, or in this case a personal weakness) is using a 7-point scale (1 = not at all, 7 = a lot).
changeable as opposed to fixed and immutable (Dweck, Finally, participants spent 5 min responding in writing to
Chiu, & Hong, 1995; Dweck & Leggett, 1988). A large two prompts. They were asked to describe (a) whether
body of research suggests that incremental beliefs are posi- they’ve ever done anything to change their weakness and
tively related to growth-related behaviors. For example, (b) where they think it comes from. Two independent coders
people who believe that their intelligence level can be rated the degree to which these statements contained evi-
changed are more likely to seek out challenging tasks and dence of incremental beliefs or the belief that their weak-
to respond to academic failures in constructive ways, adopt- ness was malleable and could be changed, as opposed to
ing learning goals and striving for self-improvement (Diener seeing it as fixed and unchangeable (Dweck, 1999).

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
1136 Personality and Social Psychology Bulletin 38(9)

Although the prompts did not directly ask about this aspect a moral transgression. Examining self-compassion for a
of the weakness, they provided an opportunity for partici- moral transgression allowed us to assess whether the bene-
pants to make statements regarding the malleability of their fits of self-compassion would generalize to a potentially
weakness (e.g., “It’s just inborn, there’s nothing I can do”— more severe and ego-threatening context. It is important to
low incremental; “With hard work I know I can change note that participants in this study were instructed to think
it”—high incremental). A score of 0 indicated an absence of about a transgression that made them feel guilt and remorse.
incremental beliefs, 1 indicated some evidence of incremen- It is unlikely that self-compassion would be either relevant
tal beliefs, and 2 indicated strong evidence of incremental or helpful for a transgression that a person does not feel
beliefs. The two coders’ ratings were averaged to create a bad about.
composite score (α = .53; M =.65, SD = .48).1 At the end of
the experiment, participants completed demographic ques-
tions and a suspicion probe and then were debriefed. Method
Participants. A total of 100 undergraduates (66% female)
participated in an online survey for course credit. Of them,
Results and Discussion one was excluded for noncompliance with the transgression
Most of the weaknesses that participants selected involved identification instructions, and eight were excluded because
social difficulties (e.g., lack of confidence, social anxiety, they indicated in a questionnaire at the end of the study that
shyness, insecurity in relationships). No one selected a they had clicked responses at random, leaving a final sam-
weakness that could not hypothetically be changed in some ple of 91. Thirty-three percent of of participants identified
way or an undesired physical attribute. In response to the as Asian American, 37% as European American, 11% as
suspicion probe, no one guessed the hypothesis that view- Latino/a, and the remainder as “Other.” Participants ranged
ing a personal weakness in a certain way might affect one’s in age from 18 to 33 years (M = 20.6, SD = 2.6).
belief in its malleability. Neither gender nor ethnicity inter- Procedure. Participants were told that the purpose of the
acted with condition to predict self-compassion in this or study was to understand how people think about different
any of the subsequent studies, so these variables will not be kinds of personal events. Participants first described a
discussed further. recent moral transgression. The specific instructions read
Incremental beliefs were significantly different across as follows:
conditions, F(2, 65) = 3.64, p < .05, ηp2 =.10, with a greater
number of these beliefs expressed in the self-compassion “Please recall a time in your recent past when you did some-
condition (M = .89, SD = .55, n = 22) compared to the self- thing you felt was wrong and as a result experienced guilt,
esteem condition (M = .58, SD = .44, n = 26) and no inter- remorse, and regret. For example, this could be a time when
vention condition (M = .55, SD = .36, n = 20). Follow-up you cheated on a significant another or on an exam, betrayed
contrasts between self-compassion and self-esteem, ηp2 =.09, and a friend’s trust, or did something else that was harmful or
between self-compassion and no intervention, ηp2 =.12 , were potentially harmful to yourself or others. Try to think of the
significant (ps < .05). Thus, as predicted, participants who most recent experience of this kind that made you feel bad
were instructed to take a self-compassionate approach about yourself and still makes you feel bad when you think
toward a personal weakness viewed that weakness as more about it.”
changeable than participants instructed to take a self-esteem-
enhancing approach or no approach.2 Participants were informed that their responses would be
Positive affect did not differ significantly across condi- anonymous.
tions (p > .7), and when we controlled for the effect of this Next, participants were randomly assigned to one of three
variable on incremental beliefs, the effect of condition remained conditions: self-compassion, self-esteem control, or positive
significant (p < .05). Thus, the effect of self-compassion on distraction control. The self-compassion and self-esteem
self-improvement motivation could not be explained by instructions were similar to those used in Experiment 1. The
differences in positive affect. self-compassion instructions were worded as follows:

“In this section, you will be asked to reflect back on the

Experiment 2 event you wrote about from a compassionate perspective. In
Experiment 1’s results suggest that self-compassion helps the space below, please write a paragraph to yourself (as if
participants view personal weaknesses as more changeable, you are addressing yourself) expressing kindness and under-
but it does not bear directly on motivation to change. In standing regarding the event you described above.”
Experiment 2, we examined the effects of self-compassion
on transgression-related self-improvement motivation, such The self-esteem instructions were worded as follows:
as desire to make amends and commitment to avoid repeating “In the space below, please write a paragraph describing

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
Breines and Chen 1137

your positive qualities. For example, what personal attri- Experiment 3

butes and accomplishments are you proud of?” Participants
in the positive distraction control condition were asked to The results of Experiments 1 and 2 indicated that self-compassion
describe a hobby they enjoyed. We used this new control increases self-improvement motivation (Experiment 2) and
condition so that we could compare the effects of self- related mind-sets (i.e., incremental beliefs; Experiment 1). In
compassion with that of a positive distraction, which could Experiment 3, we examined whether self-compassion follow-
confer similar mood benefits. Following the manipulation, ing a failure experience would increase self-improvement
positive affect (α = .75; M = 4.93, SD = 1.08) was assessed behavior. Self-improvement behavior was operationalized as
with the same items used in Experiment 1, with the addition time spent studying difficult vocabulary words for a second
of the item “happy.” test following the initial failure. Time spent working on a dif-
Participants next filled out a questionnaire assessing ficult task is often used to measure persistence and other self-
their desire to make amends and their commitment to not improvement-related behaviors (e.g., Di Paula & Campbell,
repeat the transgression in the future. On a scale ranging 2002; Williams & DeSteno, 2008). In this experiment, we did
from 1 (strongly disagree) to 7 (strongly agree), participants not want participants to know that the initial test was intended
rated the extent to which they agreed with the following to make them feel bad about themselves, so we used a more
items regarding their transgression: “I am committed to not subtle self-compassion prime embedded in the instructions that
repeating this behavior (or anything like it) again,” “I will followed the first test.
do my best to never do something like this again,” “I wish I
could go back and erase what happened,” “I want to con-
fess” (if applicable), “I want to apologize” (if applicable), Method
“Realistically, it is likely that I will do something like this Participants. A total of 103 undergraduates (62% female)
again in the future” (reverse), and “I feel no need to make participated for course credit. Of them, 17 were excluded
amends” (reverse). The scale was internally consistent (α = .75; because they indicated that they were not fluent in English
M = 4.49, SD = 1.12).3 (fluency in English was important for this experiment because
At the end of the experiment, participants completed it focused on verbal test performance), leaving a final sample
demographic questions and a suspicion probe and then were of 86. Forty-seven percent of participants identified as Asian
debriefed. American, 21% as European American, 12% as Latino/a, and
the remainder as “Other.” Participants ranged in age from 18
to 23 (M = 19.77, SD = 1.27).
Results and Discussion Procedure. Participants were told that the purpose of the
The majority of participants’ transgressions involved roman- study was to understand the relationship between test per-
tic infidelity, academic misconduct, dishonesty, betrayal of formance and personality. Participants first took a difficult
trust, or hurting someone they cared about. A few participants test (a 10-item version of a Graduate Record Examination
mentioned illegal behavior. In response to the suspicion [GRE] antonyms test, pilot tested for difficulty), received
probe, no one guessed the hypothesis that viewing a trans- the correct answers, and then had an opportunity to study a
gression in a certain way might affect one’s motivation to list of words and definitions that would be on a subsequent
improve oneself or make amends. 10-item antonyms test. Participants could study these words
Self-improvement motivation differed significantly for as long as they wanted, and study time was used as a
across conditions, F(2, 88) = 4.38, p < .05, ηp2 =.09 , with measure of improvement motivation. Immediately preced-
higher motivation reported in the self-compassion condition ing the presentation of the study words, participants read an
(M = 4.97, SD = 1.04, n = 31) compared to the self-esteem instructions screen that contained the manipulation. Partici-
condition (M = 4.11, SD = 1.27, n = 29) and positive distrac- pants randomly assigned to the self-compassion condition
tion condition (M = 4.36, SD = 1.18, n = 31). Follow-up saw an additional statement embedded in the instructions
contrasts between self-compassion and self-esteem, ηp2 =.13, that read,
and between self-compassion and positive distraction,
ηp2 =.07, were significant (ps < .05). Thus, as predicted, par- “If you had difficulty with the test you just took, you’re not
ticipants who were instructed to take a self-compassionate alone. It’s common for students to have difficulty with tests
approach toward a transgression that they felt bad about like this. If you feel bad about how you did, try not to be too
were more motivated to make amends and avoid repeating hard on yourself.”
the transgression in the future than participants instructed to
take a self-esteem-focused approach or to engage in a posi- This method of manipulating self-compassion was
tive distraction task. adapted from Adams and Leary (2007). Participants
Positive affect did not differ across conditions (p > .4), assigned to the self-esteem control condition saw an addi-
and when we controlled for this variable, the effect of condi- tional statement that read, “If you had difficulty with the test
tion on self-improvement remained significant (p < .05). you just took, try not to feel bad about yourself—you must

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
1138 Personality and Social Psychology Bulletin 38(9)

be intelligent if you got into Berkeley.” Participants in the to mind-set (Experiment 1) and motivation (Experiment 2).
no intervention control condition saw no additional In Experiment 4, we wanted to extend these results to a dif-
statements. ferent behavioral domain, social comparison preferences.
Following the manipulation, participants filled out a Research suggests that choosing to engage in upward social
brief questionnaire about their perceptions of the first test comparison (or more specifically, to seek contact with those
(i.e., difficulty, perceived performance) and their expecta- who seem better off than the self) tends to reflect self-
tions for how well they would perform on the second test. improvement motives (Collins, 1996; Taylor & Lobel, 1989).
Finally, participants completed the second 10-item test. The Whereas downward comparison can increase self-esteem,
words used on both tests were pilot tested to ensure that upward contact can provide hope, information, and inspira-
most undergraduates would find them difficult and perform tion (e.g., Taylor & Lobel, 1989; van den Borne, Pruyn, & van
poorly (i.e., experience failure). At the end of the experi- den Heuvel, 1987). For Experiment 4, we recruited partici-
ment, participants completed demographic questions and a pants from across the United States to increase the generaliz-
suspicion probe and then were debriefed. ability of our findings.

Results and Discussion Method

In response to the suspicion probe, no one guessed the Participants. A total of 78 adults from across the United
hypothesis that viewing test failure in a certain way might States (81% female) participated in an online survey for a
affect study time or effort. Participants scored an average of chance at receiving a raffle gift certificate. Of them, two
4 out of 10 (40%) on the initial test (SD = 1.96), indicating participants were excluded because they discussed a mental
that most participants performed poorly. Scores ranged from illness as their personal weakness, two were excluded
0 to 9 (though only one participant scored a 9 and excluding because they guessed the exact hypothesis (one was in the
them did not change the results). As predicted, study time for self-compassion condition and one in the self-esteem condi-
the second test differed significantly across conditions, tion), and one participant was excluded because he or she
F(2, 83) = 3.12, p < .05, ηp2 =.07 , with longer times in the was presented with all three manipulations due to a techni-
self-compassion condition (M = 306.5 s, SD = 236.95 s, cal glitch, leaving a final total of 73 participants. Fifty-two
n = 29) compared to the self-esteem condition (M = 229.9 s, percent of participants identified as Asian American, 15% as
SD = 119.15 s, n = 27) and no intervention condition European American, 6% as Latino/a, 4% as Native Ameri-
(M = 203.2 s, SD = 101.09 s, n = 30). Follow-up contrasts can, and the remainder as “Other.” Participants ranged in
between self-compassion and self-esteem were marginal age from 18 to 62 (M = 32.3, SD = 12.4).
(p = .085, ηp2 =.04) and were significant between self-com- Procedure. Participants were told that the purpose of the
passion and no intervention (p < .05, ηp2 =.08). Neither initial study was to understand the different ways that people think
test performance (actual or perceived), perceptions of test about personal weaknesses. As in Experiment 1, partici-
difficulty, nor performance expectations differed significantly pants first identified a personal weakness (this time, how-
across conditions (ps > .6), and controlling for these variables ever, they were explicitly instructed to select a weakness
did not reduce the effect of condition on study time (ps < .05). that could hypothetically be changed) and then were ran-
Study time was positively correlated with performance domly assigned to one of three conditions: self-compassion,
on the second test (r = .37, p < .01). Performance on the self-esteem control, or positive distraction control. The self-
second test did not differ significantly across conditions, p > .6, compassion instructions were worded as follows:
but the pattern of results suggested that participants in the
self-compassion condition scored slightly higher (M = 7.4, “In this next section, we would like you to write a paragraph
SD = 2.06) than those in each of the control conditions (M = 6.9 to yourself expressing compassion and understanding
and 7.0, SD = 2.21 and 1.95). Thus, although in this experi- regarding the personal weakness you described above. In
ment self-compassion did not directly lead to improved per- other words, try to take a caring and concerned approach,
formance, it did increase study time, which in turn predicted rather than a critical one.”
higher test scores. To the extent that sustained self-improvement
motivation leads to better performance over time (e.g., The self-esteem instructions were worded as follows:
Duckworth, Peterson, Matthews, & Kelly, 2007), repeatedly
taking a self-compassionate approach to failure should ulti- “In this next section, we would like you to write a paragraph
mately lead to better performance through its effect on self- to yourself describing aspects of yourself that you consider
improvement motivation. to be positive qualities, in contrast to the aspect you
described above. In other words, try to focus on the positives
and the things that you are proud of.”
Experiment 4
The results of Experiment 3 indicated that self-compassion In the positive distraction condition, participants were asked
influenced self-improvement behavior (study time) in addition to describe a hobby they enjoyed doing in their spare time.

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
Breines and Chen 1139

Following the manipulation, positive affect was assessed with Table 1. Social Comparison Preferences by Condition—
the single item “content” (M = 3.10, SD = 1.18). Ratings were Experiment 4
made using a 5-point scale (1 = not at all, 5 = a lot). Upward Lateral Downward
Although the manipulation was face-valid (i.e., partici-
pants were asked to write to themselves in a compassionate Self-compassion 15 1 2
way), we included a brief measure of state self-compassion Self-esteem control 19 10 1
to serve as a manipulation check (α = .77; M = 4.17, SD = .83). Positive distraction control 11 12 2
The scale included the following four items: “I’m being
understanding toward myself,” “I’m treating myself with
caring and kindness,” “I’m trying to take a balanced view of
things,” and “I see my weakness as part of being human” self-compassion condition were the most likely to engage in
(adapted from Neff, 2003b). upward social comparison as opposed to downward or lateral
Next, for the social comparison preferences measure, par- social comparison. That is, they were more likely to select a
ticipants were asked to imagine that they had the opportunity hypothetical interaction partner who had successfully worked
to interact socially with one of three people: (a) “Someone through a similar weakness (as opposed to someone who had
who has had a personal weakness similar to your own but a similar or worse weakness), compared to participants in the
who has worked successfully at overcoming it” (upward), control conditions, χ2(2, n = 73) = 6.91, p < .05, ηp2 =.06.
(b) “Someone who has a personal weakness similar to your When we limited the sample to examine contrasts between
own” (lateral), or (c) “Someone who has a personal weak- conditions, the comparison between self-compassion and posi-
ness similar to but more severe than your own” (downward). tive distraction was significant, χ2(2, n = 43) = 6.77, p < .01,
We designed this task so that the upward social comparison ηp2 =.10 and the comparison between self-compassion and
option would be more directly related to self-improvement self-esteem was in the expected direction, although not sig-
regarding the weakness. nificant, χ2(2, n = 48) = 2.18, p = .14, ηp2 =.03. Frequencies are
Finally, participants filled out a self-report measure of presented in Table 1. Supplementary analyses revealed that the
motivation to improve their weakness that included the fol- observed frequencies of upward compared with non-upward
lowing seven items: “I want to learn and improve myself,” “I within the self-compassion condition, χ 2(1) = 8.00, p <
want to find opportunities that will challenge me and help .01, but not within the other two conditions (ps > .1), differed
me grow as a person,” “I feel capable of making positive significantly from expected values (set at equal proportions
changes,” “I would like to discover new strategies for within each cell relative to total cell size).
improving myself,” “I feel confident that I can make positive Self-reported self-improvement motivation differed sig-
changes,” “It’s up to me whether or not I continue to have nificantly across conditions, F(2, 70) = 4.00, p < .05, ηp2 =.11 ,
this weakness,” and “I don’t think there is much I can do to with higher motivation in the self-compassion condition
change this weakness” (reverse). Ratings were made using a (M = 4.37, SD = .43) compared with the self-esteem condi-
7-point scale (1 = strongly disagree, 7 = strongly agree). The tion (M = 4.00, SD = .48) and positive distraction condition
scale was internally consistent (α = .80; M = 4.06, SD = .58). (M = 3.90, SD = .66). Follow-up contrasts between self-
compassion and self-esteem, and between self-compassion
and positive distraction, were significant (ps < .05). Thus,
Results and Discussion conceptually replicating Experiment 2’s results, participants
As in Experiment 1, most of the weaknesses that participants who were instructed to take a self-compassionate approach
selected involved social difficulties. No one selected a weak- toward a personal weakness were more motivated to change
ness that could not hypothetically be changed in some way. that weakness than participants instructed to take a self-
The self-compassion manipulation check differed signifi- esteem-enhancing approach or to engage in a positive dis-
cantly across conditions, F(2, 70) = 3.33, p < .05, ηp2 =.10 , with traction task.5
higher scores on the state self-compassion scale in the self- Finally, positive affect did not differ significantly across
compassion condition (M = 4.58, SD = .87, n = 19) com- conditions (p > .1), and when we controlled for the effect of
pared to the self-esteem condition (M = 3.98, SD = .84, this variable on self-improvement motivation, the effect of
n = 32) and positive distraction condition (M = 4.10, SD = .63, the condition remained significant (p < .05).
n = 27). The follow-up contrast between self-compassion
and self-esteem was marginal (p = .057; ηp2 =.13) and
between self-compassion and positive distraction was sig- General Discussion
nificant (p < .05, ηp2 =.16). Across four experiments, we found support for the hypoth-
The majority of social comparison preferences were esis that responding to a moral transgression, personal
either upward (62%) or lateral (31.5%).4 Because we were weakness, or test failure with self-compassion subsequently
specifically interested in the tendency to make upward com- makes people more motivated to improve themselves and
parisons, we focused the analysis on upward versus non-upward their performance. These experiments are among the first to
(i.e., lateral or downward) comparisons. Participants in the show that self-compassion leads to increased self-improvement

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
1140 Personality and Social Psychology Bulletin 38(9)

motivation and does so across a range of domains and was marginal or approaching significance. Overall, the pat-
populations. Self-compassion is unique in that it provides a tern of results suggests that the effects of self-compassion and
safe and nonjudgmental context to confront negative aspects self-esteem, though both strong predictors of many aspects of
of the self and strive to better them. Unlike other approaches well-being, may differ when it comes to self-improvement
to failure that tend to undermine personal growth by motivation. Prior research indicates that in some cases posi-
encouraging inflated or deflated self-assessments (e.g., Kim tive illusions can improve health and well-being (e.g., Taylor
et al., 2010) rather than realistic self-appraisals (Leary et al., & Brown, 1988), but in the context of learning and growth, its
2007), our results suggest that self-compassion is a more self-protective function may be less useful, as accurate self-
effective method of motivating change. appraisal is often necessary to motivate change (e.g., Kim
The results of Experiments 1 and 4 indicated that partici- et al., 2010). Like self-esteem, self-compassion may serve as
pants who took a self-compassionate approach to a personal a buffer against debilitating self-criticism, but unlike self-
weakness, compared with those in a self-esteem, no interven- esteem, it does not require inflated self-evaluations that may
tion (Experiment 1), or positive distraction (Experiment 4) con- hinder self-improvement motivation.
trol condition, viewed their weakness as more malleable Theoretical accounts of self-compassion have empha-
(Experiment 1), reported greater self-improvement motiva- sized the important differences between self-compassion
tion (Experiment 4), and were particularly likely to choose and self-esteem particularly when it comes to personal
interaction partners who had successfully worked through a growth (Neff, 2003a), and some studies have controlled for
similar weakness (i.e., upward social comparison; Experiment 4). trait self-esteem (e.g., Leary et al., 2007; Neff, 2003b; Neff
Experiment 2 showed that participants who reflected on & Vonk, 2009). With the exception of one study (Leary et al.,
a recent moral transgression with self-compassion, com- 2007, Study 5), however, the present experiments are the
pared with those in a self-esteem or positive distraction con- first to systematically include a self-esteem control condi-
trol condition, consequently reported more self-improvement tion, allowing us to determine whether self-compassion and
motivation, such as desire to make amends and commitment self-esteem do indeed have distinct effects in this context.
to not repeat a similar transgression in the future. This find- The self-esteem manipulations used in Experiments 2 and 4
ing parallels research showing that interpersonal forgive- are similar to self-affirmation manipulations that involve
ness generally reduces the likelihood of repeat offenses affirming personal attributes (e.g. Cohen, Aronson, &
(Wallace, Baumeister, & Exline, 2008) and suggests that Steele, 2000; Steele, Spencer, & Lynch, 1993), which allow
self-compassion may similarly help people see that their our results to generalize to this widely studied and often
moral failings do not have to define them. However, future beneficial form of self-validation. It is unclear, however,
research would be needed to examine whether this increased how self-compassion might compare with affirming a per-
motivation to improve translates into changes in behavior sonal value, another form of self-affirmation.
over time. Furthermore, it would be important to examine These experiments examined the influence of self-
the effects of self-compassion in the context of transgres- compassion in the context of a salient personal weakness,
sions that are more severe than those typical of undergradu- transgression, or failure, but they did not address the ques-
ates, such as extreme acts of cruelty and violence. tion of whether self-compassion motivates people to actively
Experiment 3 showed that a subtle reminder to be self- seek out negative information about themselves when it is
compassionate about a lab-based test failure, compared with not salient to begin with. Although self-compassion is pri-
a self-esteem reminder or no reminder, led participants to marily relevant in response to salient negative self-aspects,
spend more time studying for a subsequent test, a commonly making it difficult to experimentally examine this hypothe-
used measure of effort and persistence (e.g., Di Paula & sis, future research could assess downstream consequences
Campbell, 2002; Williams & DeSteno, 2008). Study time for information-seeking behavior in a separate domain or
was positively correlated with test performance, indicating could employ correlational methods.
that self-compassion may indirectly increase performance These findings extend the self-compassion literature by
through its effect on study time, though a direct effect was clarifying the unique role of self-compassion in promoting
only trending in the predicted direction. Thus, as noted, our self-improvement, suggesting that it may be beneficial across
results suggest that taking a self-compassionate approach to a diverse range of domains. In addition, they address a cen-
failure may increase performance over time to the extent that tral question regarding self-improvement—namely, how to
it increases effort (e.g., Duckworth et al., 2007). confront shortcomings without becoming paralyzed by harsh
In all four studies, self-compassion was significantly more self-criticism on one hand or by defensive self-enhancement
helpful than engaging in positive distraction (Experiments 2 on the other. Self-compassion may be one solution.
and 4) or doing nothing (Experiments 1 and 3). In Experiments Furthermore, these findings have implications for enhanc-
1 and 2, and for the self-report self-improvement scale in ing coping skills in educational settings. Self-compassion
Experiment 4, self-compassion was significantly more help- may help students respond to failure in a way that facilitates
ful then self-esteem. For the study time (Experiment 3) and growth and improvement without leading to debilitating
social comparison (Experiment 4) measures, this difference negative affect. Further research is needed to determine

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
Breines and Chen 1141

whether self-compassion leads to sustained changes in behav- 4. T he low number of downward preferences may be due to the
ior over time, however. For example, would repeatedly fact that participants were asked what kind of person they
responding with self-compassion to failure and poor perfor- would most want to interact with, not evaluate themselves
mance lead students to develop better study habits and per- against. Previous research suggests that downward comparisons
form better over the course of a semester? are more desirable in evaluative contexts, whereas upward com-
In sum, the present findings suggest that self-compassion parisons tend to be preferred in the context of information and
may indeed represent a useful alternative to other common affiliation (Taylor & Lobel, 1989).
responses to failure and personal weakness. Resolving to 5. A mediation test conducted following the guidelines described
make changes can be scary, as roadblocks and setbacks are by Preacher and Hayes (2004, 2008) revealed that scores on the
inevitable along the way. From a self-compassionate per- state self-compassion manipulation check fully mediated the
spective, however, there is less to fear. effect of the self-compassion condition, compared with the self-
esteem condition, on self-improvement motivation. State self-
Acknowledgment compassion was not a candidate for mediation of the contrast
We would like to thank the members of the Self, Identity and between the self-compassion and positive distraction condition,
Relationships Lab at the University of California, Berkeley, for however, because this effect was not fully reduced by the intro-
their comments. We would also like to thank Bernie Wong, duction of the potential mediator.
Chellsee Hill, Clarissa Tu, and Linzi Nelson for assistance with
coding and data collection. References
Adams, C. E., & Leary, M. R. (2007). Promoting self-compassionate
Declaration of Conflicting Interests attitudes toward eating among restrictive and guilty eaters. Jour-
The authors declared no potential conflicts of interest with nal of Social & Clinical Psychology, 26, 1120-1144.
respect to the research, authorship, and/or publication of this Alicke, M., & Sedikides, C. (2009). Self-enhancement and self-
article. protection: What they are and what they do. European Review
of Social Psychology, 20, 1-48.
Funding Baker, L. R., & McNulty, J. K. (2011). Self-compassion and rela-
The authors disclosed receipt of the following financial support for tionship maintenance: The moderating roles of conscientious-
the research, authorship, and/or publication of this article: This ness and gender. Journal of Personality and Social Psychology,
work was supported by a National Science Foundation Graduate 100, 853-873.
Research Fellowship awarded to Juliana G. Breines. Baumeister, R., Campbell, J. D., Kreuger, J. I., & Vohs, K. D.
(2003). Does high self-esteem cause better performance, inter-
Notes personal success, happiness, or healthier lifestyles? Psychologi-
1. T o examine whether participants in the self-esteem condition cal Science in the Public Interest, 4, 1-44.
might have focused less on their weakness during the reflec- Blatt, S. J. (1995). The destructiveness of perfectionism: Implica-
tion task, despite the instructions in both conditions to address tions for the treatment of depression. American Psychologist,
oneself regarding the weakness, we coded the open-ended 50, 1003-1020.
responses for number of references to the weakness. Average Campbell, W. K., & Sedikides, C. (1999). Self-threat magnifies the
number of references was slightly higher in the self-compas- self-serving bias: A meta-analytic integration. Review of Gen-
sion condition compared with the self-esteem condition, but the eral Psychology, 3, 23-43.
difference was nonsignificant. Most critically, controlling for Cohen, G. L., Aronson, J., & Steele, C. M. (2000). When beliefs
this frequency variable did not reduce the effect of condition yield to evidence: Reducing biased evaluation by affirm-
on incremental beliefs, suggesting that the effect was likely not ing the self. Personality and Social Psychology Bulletin, 26,
driven by differential focus on the weakness. 1151-1164.
2. We acknowledge that the interrater reliability was low. It has Collins, R. L. (1996). For better or worse: The impact of upward
been argued that alphas less than .70 should not necessarily be social comparison on self-evaluation. Psychological Bulletin,
discounted, however, when the pattern of results is generaliz- 119, 51-69.
able across coders (John & Benet-Martinez, 2000), which was Diener, C. I., & Dweck, C. S. (1978). An analysis of learned help-
the case in this experiment. Thus, although the low reliability is lessness: Continuous changes in performance, strategy, and
clearly a limitation, we believe that the results are still mean- achievement cognitions following failure. Journal of Person-
ingful, especially when considered in light of the four sets of ality and Social Psychology, 36, 451-462.
findings as a whole. Diener, C. I., & Dweck, C. S. (1980). An analysis of learned help-
3. For the apology and confession items, a “not applicable” lessness: II. The processing of success. Journal of Personality
response option was included, coded as system missing. Out and Social Psychology, 39, 940-952.
of 90 participants, 10 selected this option for apology and 16 Di Paula, A., & Campbell, J. D. (2002). Self-esteem and persistence
selected it for confession. Their scores on these items were not in the face of failure. Journal of Personality and Social Psy-
included in the self-improvement composite. chology, 83, 711-724.

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
1142 Personality and Social Psychology Bulletin 38(9)

Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. Lyubomirsky, S., Tucker, K. L., Caldwell, N. D., & Berg, K.
(2007). GRIT: Perseverance and passion for long-term (1999). Why ruminators are poor problem solvers: Clues from
goals. Journal of Personality and Social Psychology, 92, the phenomenology of dysphoric rumination. Journal of Per-
1087-1101. sonality and Social Psychology, 77, 1041-1060.
Dweck, C. S. (1999). Self-theories: Their role in motivation, Magnus, C. M. R., Kowalski, K. C., & McHugh, T. L. F. (2010).
personality, and development. Philadelphia, PA: Psychology The role of self-compassion in women’s self-determined
Press. motives to exercise and exercise-related outcomes. Self and
Dweck, C. S., Chiu, C., & Hong, Y. (1995). Implicit theories and Identity, 9, 363-382.
their role in judgments and reactions: A world from two per- Neff, K. D. (2003a). Self-compassion: An alternative conceptual-
spectives. Psychological Inquiry, 6, 267-285. ization of a healthy attitude toward oneself. Self and Identity, 2,
Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to 85-102.
motivation and personality. Psychological Review, 95, 256-273. Neff, K. D. (2003b). Development and validation of a scale to mea-
Elliott, E. S., & Dweck, C. S. (1988). Goals: An approach to moti- sure self-compassion. Self and Identity, 2, 223-250.
vation and achievement. Journal of Personality and Social Psy- Neff, K. D., Hseih, Y., & Dejitthirat, K. (2005). Self-compassion,
chology, 54, 5-12. achievement goals, and coping with academic failure. Self and
Exline, J. J., Root, B. L., Yadavalli, S., Martin, A. M., & Fisher, M. L. Identity, 4, 263-287.
(2011). Reparative behaviors and self-forgiveness: Effects of a Neff, K. D., Kirkpatrick, K., & Rude, S. S. (2007). Self-compassion
laboratory-based exercise. Self and Identity, 10, 101-126. and its link to adaptive psychological functioning. Journal of
Fishbach, A., & Labroo, A. A. (2007). Be better or be merry: How Research in Personality, 41, 139-154.
mood affects self-control. Journal of Personality and Social Neff, K. D., Rude, S. S., & Kirkpatrick, K. (2007). An examina-
Psychology, 93, 158-173. tion of self-compassion in relation to positive psychological
Fisher, M. L., & Exline, J. J. (2006). Self-forgiveness versus excusing: functioning and personality traits. Journal of Research in Per-
The roles of remorse, effort, and acceptance of responsibility. Self sonality, 41, 908-916.
and Identity, 5, 127-146. Neff, K. D., & Vonk, R. (2009). Self-compassion versus global self-
Grzegorek, J. L., Slaney, R. B., Franze, S., & Rice, K. G. (2004). esteem: Two different ways of relating to oneself. Journal of
Self-criticism, dependency, self-esteem, and grade point average Personality, 77, 23-50.
satisfaction among clusters of perfectionists and non-perfectionists. Powers, T. A., Koestner, R., Lacaille, N., Kwan, L., & Zuroff, D. C.
Journal of Counseling Psychology, 51, 192-200. (2011). Self-criticism, motivation, and goal progress of athletes
Hall, J. H., & Fincham, F. D. (2008). The temporal course of self- and musicians: A prospective study. Personality and Individual
forgiveness. Journal of Social & Clinical Psychology, 27, Differences, 47, 279-283.
174-202. Powers, T. A., Koestner, R., & Zuroff, D. C. (2007). Self-criticism,
Iskender, M. (2009). The relationship between self-compassion, goal motivation, and goal progress. Journal of Social & Clinical
self-efficacy, and control beliefs about learning in Turkish uni- Psychology, 26, 826-840.
versity students. Social Behavior and Personality, 37, 711-720. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures
John, O.P. & Benet-Martínez, V. (2000). Measurement, scale con- for estimating indirect effects in simple mediational models.
struction, and reliability. In H.T. Reis and C.M. Judd (Eds.), Behaviour Research Methods Instruments and Computers, 36,
Handbook of research methods in social and personality psy- 717-731.
chology (pp. 339-369). New York, NY: Cambridge University Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resam-
Press. pling strategies for assessing and comparing indirect effects
Kelly, A. C., Zuroff, D. C., Foa, C. L., & Gilbert, P. (2009). Who in multiple mediator models. Behavior Research Methods,
benefits from training in self-compassionate self-regulation? A 40, 879-891.
study of smoking reduction. Journal of Social & Clinical Psy- Raghunathan, R., & Trope, Y. (2002). Walking the tightrope
chology, 29, 727-755. between feeling good and being accurate: Mood as a resource
Kim, Y., Chiu, C., & Zou, Z. (2010). Know thyself: Misperceptions in processing persuasive messages. Journal of Personality and
of actual performance undermine achievement motivation, future Social Psychology, 83, 510-525.
performance, and subjective well-being. Journal of Personality Rhodewalt, F. (1994). Conceptions of ability, achievement goals,
and Social Psychology, 99, 395-409. and individual differences in self-handicapping behavior: On
Kwan, V. S., John, O., Robin, R., & Kuang, L. (2008). Conceptu- the application of implicit theories. Journal of Personality, 62,
alizing and assessing self enhancement bias: A componential 67-85.
approach. Journal of Personality and Social Psychology, 94, Rice, K. G., & Stuart, J. (2010). Differentiating adaptive and mal-
1062-1077. adaptive perfectionism on the MMPI-2 and MIPS revised.
Leary, M. R., Tate, E. B., Adams, C. E., Allen, A. B., & Hancock, J. Journal of Personality Assessment, 92, 158-167.
(2007). Self-compassion and reactions to unpleasant self-relevant Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self:
events: The implications of treating oneself kindly. Journal of Short-term benefits and long-term costs. Journal of Personality
Personality and Social Psychology, 92, 887-904. and Social Psychology, 80, 340-352.

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014
Breines and Chen 1143

Robins, R. W., & Pals, J. L. (2002). Implicit self-theories in the van den Borne, H. W., Pruyn, J. F. A., & van den Huevel, W. J. A.
academic domain: Implications for goal orientation, attribu- (1987). Effects of contacts between cancer patients on their
tions, affect, and self-esteem change. Self and Identity, 1, psychosocial problems. Patient Education and Counseling, 9,
313-336. 33-51.
Sedikides, C., Gaertner, L., & Toguchi, Y. (2003). Pancultural self- Wallace, H. M., Exline, J. J., & Baumeister, R. F. (2008). Inter-
enhancement. Journal of Personality and Social Psychology, personal consequences of forgiveness: Does forgiveness deter
84, 60-79. or encourage repeat offenses? Journal of Experimental Social
Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image Psychology, 44, 453-460.
resilience and dissonance: The role of affirmational Williams, J. G., Stark, S. K., & Foster, E. E. (2008). Start today or the
resources. Journal of Personality and Social Psychology, very last day? The relationships among self-compassion, moti-
64, 885-896. vation, and procrastination. American Journal of Psychological
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A Research, 4, 37-44.
social psychological perspective on mental health. Psychologi- Williams, L. A., & DeSteno, D. (2008). Pride and perseverance:
cal Bulletin, 103, 193-210. The motivational role of pride. Journal of Personality and
Taylor, S. E., & Lobel, M. (1989). Social comparison activity under Social Psychology, 94, 1007-1017.
threat: Downward evaluation and upward contacts. Psychologi- Witcher, L. A., Alexander, E. S., Onwuebuzie, A. J., Collins, K. M.,
cal Review, 96, 569-575. & Witcher, A. E. (2007). The relationship between psychology
Tice, D. M., & Baumeister, R. F. (1990). Self-esteem, self-handicapping, students’ levels of perfectionism and achievement in a gradu-
and self-presentation: The strategy of inadequate practice. ate-level research methodology course. Personality and Indi-
Journal of Personality, 58, 443-464. vidual Differences, 43, 1396-1405.

Downloaded from [Link] at UNIV OF UTAH SALT LAKE CITY on June 25, 2014