Self-Compassion and Reactions to Unpleasant Self-Relevant Events: The Implications of Treating Oneself Kindly

Abstract and Figures

Five studies investigated the cognitive and emotional processes by which self-compassionate people deal with unpleasant life events. In the various studies, participants reported on negative events in their daily lives, responded to hypothetical scenarios, reacted to interpersonal feedback, rated their or others' videotaped performances in an awkward situation, and reflected on negative personal experiences. Results from Study 1 showed that self-compassion predicted emotional and cognitive reactions to negative events in everyday life, and Study 2 found that self-compassion buffered people against negative self-feelings when imagining distressing social events. In Study 3, self-compassion moderated negative emotions after receiving ambivalent feedback, particularly for participants who were low in self-esteem. Study 4 found that low-self-compassionate people undervalued their videotaped performances relative to observers. Study 5 experimentally induced a self-compassionate perspective and found that self-compassion leads people to acknowledge their role in negative events without feeling overwhelmed with negative emotions. In general, these studies suggest that self-compassion attenuates people's reactions to negative events in ways that are distinct from and, in some cases, more beneficial than self-esteem.
Content may be subject to copyright.
Self-Compassion and Reactions to Unpleasant Self-Relevant Events: The
Implications of Treating Oneself Kindly
Mark R. Leary and Eleanor B. Tate
Duke University
Claire E. Adams
Louisiana State University
Ashley Batts Allen
Duke University
Jessica Hancock
Wake Forest University
Five studies investigated the cognitive and emotional processes by which self-compassionate people deal
with unpleasant life events. In the various studies, participants reported on negative events in their daily
lives, responded to hypothetical scenarios, reacted to interpersonal feedback, rated their or others’
videotaped performances in an awkward situation, and reflected on negative personal experiences.
Results from Study 1 showed that self-compassion predicted emotional and cognitive reactions to
negative events in everyday life, and Study 2 found that self-compassion buffered people against negative
self-feelings when imagining distressing social events. In Study 3, self-compassion moderated negative
emotions after receiving ambivalent feedback, particularly for participants who were low in self-esteem.
Study 4 found that low-self-compassionate people undervalued their videotaped performances relative to
observers. Study 5 experimentally induced a self-compassionate perspective and found that self-
compassion leads people to acknowledge their role in negative events without feeling overwhelmed with
negative emotions. In general, these studies suggest that self-compassion attenuates people’s reactions to
negative events in ways that are distinct from and, in some cases, more beneficial than self-esteem.
Keywords: self-compassion, self-evaluation, self-esteem, self-criticism, coping
Psychologists have been interested for many years in factors that
promote coping, resilience, and subjective well-being (Bonanno,
2004; Diener, Lucas, & Oishi, 2002; Ryan & Deci, 2000; Ryff &
Singer, 2002). Some people roll with life’s punches, facing fail-
ures, losses, and problems with equanimity, whereas others react
maladaptively to unpleasant situations or, worse, exacerbate their
distress by ruminating excessively about life’s calamities, casti-
gating themselves for their shortcomings, and catastrophizing
about their problems (Leary, 2004).
Neff (2003a, 2003b) recently identified a previously unstudied
construct that may play an important role in how people deal with
life’s problems—self-compassion. According to Neff (2003a),
self-compassion involves “being open to and moved by one’s own
suffering, experiencing feelings of caring and kindness toward
oneself, taking an understanding, nonjudgmental attitude toward
one’s inadequacies and failures, and recognizing that one’s expe-
rience is part of the common human experience” (p. 224). Pre-
sumably, a person high in self-compassion sees his or her prob-
lems, weaknesses, and shortcomings accurately, yet reacts with
kindness and compassion rather than with self-criticism and harsh-
ness. Thus, self-compassion may buffer people against negative
events and engender positive self-feelings when life goes badly.
The process by which self-compassion protects people against
stressful events is presumably different from that of its more
familiar cousin, self-esteem. Whereas self-esteem is associated
with positive feelings about oneself and believing that one is
valued by others (Leary & MacDonald, 2003), self-compassion is
an orientation to care for oneself. Not surprisingly, self-
compassionate people tend to have high self-esteem (Neff, 2003b),
presumably because reacting kindly rather than critically toward
oneself promotes positive self-feelings. However, the positive self-
feelings that characterize self-compassionate people do not appear
to involve the hubris, narcissism, or self-enhancing illusions that
characterize many people who possess high self-esteem. Whereas
standard measures of trait self-esteem (e.g., Berger, 1952; Rosen-
berg, 1965) correlate positively with scores on the Narcissistic
Personality Inventory (Raskin & Hall, 1979), self-compassion
does not (Neff, 2003b). Furthermore, low self-compassion ac-
counts for a sizable portion of unique variance in depression and
anxiety, even with trait self-esteem partialed out, suggesting that
Mark R. Leary, Eleanor B. Tate, and Ashley Batts Allen, Department of
Psychology and Neuroscience, Duke University; Claire E. Adams, Depart-
ment of Psychology, Louisiana State University; Jessica Hancock, Depart-
ment of Psychology, Wake Forest University.
A portion of this research was conducted at Wake Forest University. We
thank Anna Jarrell and Jessica Tolbert for their assistance as researchers.
Correspondence concerning this article should be addresssed to Mark R.
Leary, Department of Psychology and Neuroscience, Soc/Psych 306, Flow-
ers Drive, Duke University, Durham, NC 27708. E-mail:
Journal of Personality and Social Psychology, 2007, Vol. 92, No. 5, 887–904
Copyright 2007 by the American Psychological Association 0022-3514/07/$12.00 DOI: 10.1037/0022-3514.92.5.887
self-compassion contributes to well-being in ways that are distinct
from those of trait self-esteem.
In addition to showing that self-compassion correlates with
psychological well-being, researchers have begun to examine the
moderating effects of self-compassion on people’s reactions to
negative events. Neff, Hsieh, and Dejitterat (2005) examined the
relationship between self-compassion and reactions to academic
failure. Among students who received an unsatisfactory midterm
grade, self-compassion correlated positively with the emotion-
focused coping strategies of acceptance and positive reinterpreta-
tion/growth and negatively with a focus on negative emotions and
avoidance-oriented coping. Furthermore, self-compassion was
positively associated with mastery orientation (being motivated by
curiosity and the desire to develop one’s skills) but negatively
associated with performance orientation (the motivation to defend
or enhance one’s self-worth). These studies showed that self-
compassion moderates reactions to real and potential failure, pos-
sibly by reducing the aversiveness of events that threaten self-
These early findings suggest that self-compassion holds promise
as an important and interesting construct that facilitates resilience
and coping, but with only a few previous studies on the topic,
many key questions have not been addressed. The five studies
reported in the present article were designed to examine four
overlapping issues regarding the nature of self-compassion. The
first was to document that people low versus high in self-
compassion respond to negative events as Neff’s (2003a, 2003b)
conceptualization predicts. Previous research on self-compassion
has relied primarily on correlational designs, highlighting the need
for both experience-sampling and experimental studies. Thus, we
used paradigms in which participants reported on unpleasant
events in their everyday lives (see Study 1), responded to stan-
dardized hypothetical scenarios (see Study 2), reacted to unpleas-
ant interpersonal feedback in a laboratory experiment (see Study
3), and performed an awkward and mildly embarrassing task (see
Study 4). Second, we examined the thoughts of low- and high-
self-compassionate individuals in these studies to understand the
cognitions that underlie a self-compassionate response to negative
events (see Studies 1, 2, 4, and 5). Understanding how self-
compassionate people think about themselves and their difficulties
may point the way to promoting self-compassion among those who
tend to treat themselves less kindly. Third, given the high corre-
lation between self-compassion and self-esteem (Neff, 2003b), we
examined the unique contributions of both measured and manip-
ulated self-compassion and self-esteem to people’s reactions (see
Studies 2, 3, and 5). Finally, to date, self-compassion has been
studied only as a traitlike, dispositional variable, raising the ques-
tion of whether it is possible to induce a transitory state of
self-compassion. In Study 5, we experimentally induced a self-
compassionate mindset to examine its effects, compare the effects
of state self-compassion with those of trait self-compassion, and
examine differences in the effects of inducing self-compassion
versus self-esteem.
Study 1
With the exception of Neff et al. (2005), self-compassion has not
been studied in the context of people’s reactions to real-life events.
Thus, although researchers know that self-compassion correlates with
self-report measures of emotion, life satisfaction, and well-being
(Neff, 2003b; Neff et al., 2005; Neff, Kirkpatrick, & Rude, in press),
they do not know much about how it relates to reactions to everyday
events. In Study 1, we examined this issue by asking participants to
report on the worst thing that happened to them on four occasions
over a 20-day period. Because the worst things that happen to people
on a daily basis are typically rather mundane and inconsequential, we
were able to examine how self-compassion plays out in the context of
the minor hassles, annoyances, inconveniences, and problems that
people regularly face.
Many of the problems that people confront on a daily basis are
clearly their own fault, arising from their own intended or unin-
tended actions. Others, of course, are not their fault, being caused
by other people or events beyond their control. Neff’s (2003a)
conceptualization of self-compassion suggests that self-
compassion should be equally effective in buffering people against
negative events regardless of whether the event was their fault, and
we examined whether self-compassion moderates people’s reac-
tions differently as a function of whether they perceive a negative
event to be their fault.
Participants were 59 male and 58 female undergraduate stu-
dents, ages 17–21, from introductory psychology classes who
participated to earn credit toward a course research participation
Mass testing. All participants completed the Self-Compassion
Scale (Neff, 2003b) in mass testing at the beginning of the semester.
The Self-Compassion Scale consists of 26 items that assess the three
dimensions that Neff (2003b) identified as components of self-
compassion: self-kindness (being kind and understanding toward one-
self rather than harshly self-critical), common humanity (viewing
one’s negative experiences as a normal part of the human condition),
and mindful acceptance (holding painful thoughts and feelings in
mindful awareness rather than overidentifying with them). Neff
(2003b) reported a Cronbach’s alpha coefficient of .92.
Introductory session. Participants reported in groups of about
20 for an introductory half-hour session in which the researcher
explained that the study was investigating how people react to
events that occur in their everyday lives. Participants were told
that, during the coming 3 weeks, they would receive four e-mail
messages instructing them to access a Web site where they would
complete a questionnaire about events that had occurred during the
previous 4 days.
Web-based questionnaire. Every fifth day after the introduc-
tory session, participants received an e-mail message providing a
hyperlink Internet address that allowed them to access the Web-
based questionnaire. When they accessed the Web site, they were
instructed to recall either the worst thing that had happened to
them during the previous 4 days that was their fault (fault condi-
tion) or the worst thing that had happened to them during the
previous 4 days that was not their fault (no-fault condition). Over
the course of the study, all participants were asked to report on two
events that were their fault and on two events that were not their
fault, with the order of conditions counterbalanced across partici-
After entering their student identification number, students re-
corded the date then described in two sentences or fewer the worst
thing that had happened during the past 4 days (that was or was not
their fault). The instructions noted that “the worst thing that
happened during the past four days may have been quite bad or it
could have been very minor.” After describing the event, partici-
pants indicated whether the event primarily involved (a) work or
school; (b) my family; (c) my friends or social life; (d) athletics,
recreation, or leisure; (e) my physical health; (f) a romantic rela-
tionship; or (g) other. They indicated when the event occurred as
well as who was affected by it (only me, other people but not me,
other people and me).
Participants then rated how “bad” the event was, the degree to
which they were responsible for the event, the degree to which
other people were responsible for the event, and “in the big scheme
of things, how important was this event to you?” Ratings were
made on a 6-point scale, where 1 not at all, 2 slightly, 3
somewhat, 4 moderately, 5 very, and 6 extremely. Partic-
ipants also rated how they felt in the situation using 20 affect-
relevant terms that were selected to assess sadness (sad, dejected,
down, depressed), anxiety (nervous, worried, anxious, fearful),
anger (irritated, angry, hostile, mad), and self-conscious emotions
(embarrassed, humiliated, guilty, ashamed).
Participants rated the degree to which they reacted in each of
several ways, also on 6-point scales. Some of these reactions were
hypothesized to relate to self-compassion: (a) “I tried to be kind to
myself,” (b) “I tried to make myself feel better,” (c) “I kept the
situation in perspective,” and (d) “I was really hard on myself.”
Other reactions were assumed to be independent of self-
compassion (e.g., “I expressed my emotions to let off steam,” “I
took steps to fix the problem in a positive way or I made plans to
do so,” “I wanted to spend time alone,” “I sought out the company
of others,” “I gave myself time to come to terms with it,” “I tried
to understand my emotions”).
Participants then indicated the extent to which they had thought
each of several thoughts about the event (1 I did not think this
thought at all, 2 I thought this once, 3 I thought this a few
times, 4 I thought this several times, 5 I kept thinking this
thought): (a) I’ve had a really bad day—I need to do something
nice for myself; (b) I seem to have bigger problems than most
people do; (c) In comparison to other people, my life is really
screwed up; (d) Why do these things always happen to me?; (e)
This isn’t any worse than what lots of other people go through; (f)
I’m a loser. Finally, participants rated how well they thought they
handled the situation on a 6-point scale ranging from 1 (very
poorly)to6(very well) and the overall quality of the day on which
they experienced the negative event on a 6-point scale ranging
from 1 (very bad)to6(very good).
Preliminary Analyses
Cronbach’s alpha coefficient on our sample was .91, and the
mean of the self-compassion scores was 18.9 (SD 3.80). To
begin, we examined whether self-compassion was related to the
kinds of events that participants reported. Analyses of the catego-
ries of events that participants described revealed no relationship
between self-compassion and event category for any of the four
measurement periods (all ps .05). Overall, most of the reported
events involved work or school (38.6%), friends or social life
(24.1%), physical health (11.4%), and romantic relationships
(8.5%). Similarly, self-compassion scores were unrelated to
whether the negative event affected only the participant (54.9%),
only other people (2.0%), or both the participant and other people
(43.1%; ps .05).
Perhaps most important, self-compassion was unrelated to an-
swers to the questions “How bad was the thing that happened?”
(M 3.3) and “In the big scheme of things, how important was
this event to you?” (M 2.9; ps .05). Overall, participants who
scored low versus high in self-compassion reported events of
similar types and degrees of seriousness and importance. Thus, any
differences in thoughts and feelings are not likely to be because of
the fact that low- versus high-self-compassionate participants ex-
perienced or reported different kinds of events.
Hierarchical Linear Modeling
Data were analyzed using random coefficient modeling, specif-
ically hierarchical linear modeling (HLM; Bryk, Raudenbush, &
Congdon, 1998). Participants’ ratings of the negative events were
analyzed using the following (within-person) Level 1 model:
fault/no fault r
In these models, y
was a rating for a particular event (subscripted
i) for each participant (subscripted j),
was a random coefficient
(an intercept) representing the mean of y across all four events for
a particular participant,
was a random coefficient (a slope)
representing the effects of fault versus no-fault instructions on y,
and r
represented within-person residual or error. Thus, this
model estimated within-person ratings as a function of fault/no-
fault instructions.
Relationships between self-compassion and ratings of events
were analyzed using the following (between-person) Level 2 mod-
self-compassion u
self-compassion u
At Level 2,
were coefficients from the Level 1 model (i.e., the
intercept from each participant’s model),
represented the grand
mean of these coefficients,
represented the effects of self-
compassion scores, and u
represented error. Likewise,
coefficients from the Level 1 model (i.e., the slope of each par-
ticipant’s model),
represented the grand mean of these coeffi
represented the effects of self-compassion scores, and
represented error.
Manipulation check on fault/no-fault instructions. Before ex-
amining the relationship between self-compassion and the out-
come measures, it was important to ensure that participants prop-
erly reported events that were and were not their fault as instructed.
HLM conducted on answers to the question “To what extent were
you responsible for causing this event?” revealed a strong effect of
instructions (
3.56), t(432) 37.38, p .001. Participants
instructed to describe events that were their fault (M 5.1)
indicated that they were more responsible for the event than those
instructed to describe events that were not their fault (M 1.5).
Conversely, participants instructed to describe events that were
their fault indicated that other people were less responsible for the
event (M 2.1) than those instructed to describe events that were
not their fault (M 3.6,
⫽⫺1.48), t(432) ⫽⫺8.87, p .001.
Further support for the efficacy of the instructions is provided
by the fact that the fault/no-fault instructions were related to how
hard participants indicated they were on themselves after the event
(Ms 2.3 vs. 1.5,
0.81), t(432) 6.30, p .001; the degree
to which participants took steps to fix the problem (Ms 3.2 vs.
0.45), t(432) 3.13, p .002; and feelings of both
self-conscious emotions (Ms 7.2 vs. 5.4,
1.96), t(432)
7.16, p .001, and low self-esteem (Ms 9.2 vs. 5.5,
3.95), t(432) 13.04, p .001. Given that people who are at fault
for a negative event would be expected to be hard on themselves,
try to fix the problem, experience self-conscious emotions (such as
embarrassment or guilt), and have lower state self-esteem, these
patterns provide additional evidence that participants followed
instructions in reporting the proper type of events as instructed.
Self-compassion. As predicted, self-compassion was posi-
tively related to ratings of the items “I tried to be kind to myself”
and “I tried to make myself feel better” and negatively related to
ratings on “I was really hard on myself.” (See Table 1.) In addition,
participants high in self-compassion reported that they kept the
situation more in perspective. In contrast, self-compassion was not
related to reactions that had no conceptual relationship to self-
compassion—trying to take one’s mind off the problem, express-
ing emotions, trying to fix the problem, seeking out other people’s
company, or taking time to come to terms with the problem (see
Table 1). Thus, self-compassion was related, as expected, to par-
ticular reactions that involved treating oneself kindly or reacting
with equanimity but not to general reactions or coping tactics.
As predicted, self-compassion was consistently associated with
having fewer negative, pessimistic, and self-critical thoughts. As
seen in Table 1, the HLM analyses showed that self-compassion
was inversely related to believing that one has bigger problems in
life than most other people, wondering why “these things always
happen to me,” thinking that one’s life is more “screwed up” than
other people’s, and having the thought “I’m a loser.” In contrast,
self-compassion was positively associated with thinking that the
negative event was not worse than what many other people expe-
rience and unrelated to thinking “everyone has a bad day now and
As expected, self-compassion was negatively related to negative
feelings, specifically, anxiety, sadness, and self-conscious emo-
tions (see Table 1). The effect for anger, however, was not signif-
icant. Self-compassion was also positively related to participants’
ratings of how well they handled the difficult situation (
3.60), t(107) 37.38, p .001. Finally, high self-compassion was
associated with rating the day on which the negative event hap-
pened more positively (
0.07), t(107) 2.35, p .02. This
finding is particularly interesting given that self-compassion was
unrelated to ratings of the event itself, suggesting that negative
events taint people’s overall judgments of their days more if they
are low rather than high in self-compassion.
Moderating Effects of Self-Compassion on Reactions to
Fault and No-Fault Events
Self-compassion moderated participants’ reactions to events that
were versus were not their fault on only three items. First, an effect
Table 1
Hierarchical Linear Modeling Results for Self-Compassion in Study 1
I tried to be kind to myself. .05 1.90 .057
I tried to make myself feel better. .07 2.01 .044
I was really hard on myself. .06 2.85 .005
I kept the situation in perspective. .06 2.30 .020
I tried to do things to take my mind off the problem. .01 0.19 .85
I expressed my emotions to let off steam. .03 0.91 .36
I took steps to fix the problem or made plans to do so. .03 0.73 .47
I sought out the company of others. .03 1.06 .29
I gave myself time to come to terms with it. .01 0.19 .85
I seem to have bigger problems than most people do. .06 2.88 .005
I’m a loser. .02 2.09 .037
This isn’t any worse than what lots of other people go through. .05 2.34 .019
Why do these things always happen to me? .07 3.20 .002
In comparison to other people, my life is really screwed up. .05 2.69 .007
Everyone has a bad day now and then. .03 1.00 .32
Anxiety .22 2.18 .036
Self-conscious emotions .11 1.96 .050
Sadness .15 1.72 .083
was obtained on ratings of the item “I tried to be kind to myself”
0.07), t(107) 2.99, p .003. Calculation of conditional
regression equations revealed that self-compassion was positively
related to being kind to oneself for events that were one’s own
fault (B
.12, p .001) but not for events that were not one’s
fault (B
.04, ns). On the item “I tried to understand my
emotions,” self-compassion scores predicted trying to understand
one’s emotions when the event in question was one’s fault (B
.08, p .01) but not when it was not one’s fault (B
.02, ns).
Finally, ratings of self-conscious emotions (e.g., embarrassment,
shame, humiliation) were not moderated by self-compassion when
events were the participants’ fault (B
.02, ns), but self-
compassion was inversely related to such emotions in the case of
no-fault events (B
⫽⫺.10, p .05).
Self-compassion was consistently related to participants’ reac-
tions to the worst events that happened to them in theoretically
meaningful ways. In strong support of the Self-Compassion
Scale’s validity, self-compassionate participants more strongly
indicated that they tried to be kind to themselves and make
themselves feel better and less strongly indicated that they were
hard on themselves following negative events.
Many of the relationships between self-compassion and partic-
ipants’ thoughts emerged on items that are conceptually related to
the “common humanity” aspect of self-compassion—the recogni-
tion that problems are a normal part of life and that one’s own
difficulties are generally no worse than other people’s problems.
For example, self-compassion was inversely related to thinking
that one has bigger problems than other people and that one’s own
life is more screwed up than other people’s but positively related
to believing that the negative event was not any worse than what
other people experience.
Given these differences in the reported content of participants’
thoughts, it is not surprising that self-compassion also predicted
equanimity in the aftermath of the worst things that happened to
people. Self-compassion was associated with keeping the situation
in perspective, not endorsing the notion that one is a “loser,” lower
negative emotions, and later feeling that one handled the situation
better. As expected from Neff’s (2003a) conceptualization, self-
compassion did not generally moderate reactions to events differ-
ently that participants perceived were and were not their fault.
Only three items revealed such an effect. Self-compassion was
positively related to greater efforts to be kind to oneself and to
understand one’s emotions after events that participants reported
were their fault and inversely related to self-conscious emotions
after events that were not their fault.
Study 2
Although we saw no evidence in Study 1 that low- and high-
self-compassionate participants reported different kinds of nega-
tive events, the possibility remains that low- and high-self-
compassionate participants reacted differently to the events they
reported because they experienced, remembered, or reported dif-
ferent kinds of events. To control for this possibility, we asked
participants in Study 2 to respond to a common set of scenarios.
Although using scenarios sacrifices realism, it allowed us to ex-
amine differences in how low- and high-self-compassionate people
react to a common set of situations.
The second purpose of Study 2 was to compare the effects of
self-compassion with those of trait self-esteem and narcissism.
People who treat themselves in an understanding, kind, and com-
passionate manner should feel better about themselves than those
who are more self-critical and, thus, should score higher in self-
esteem. Indeed, previous research shows that self-compassion cor-
relates moderately with self-esteem (Neff, 2003b). However, self-
compassion should involve more than positive self-evaluations or
high self-esteem per se, reflecting a kinder and gentler approach to
problematic situations, and we should find that self-compassion
relates to various psychological outcomes in a fashion distinct
from self-esteem. Along these lines, Neff (2003b) showed that,
even with self-esteem scores partialed out, self-compassion ac-
counted for unique variance in depression, anxiety, and other
variables. Our goal was to extend her findings by exploring
whether self-compassion moderates emotions, thoughts, and be-
havioral inclinations to specific events differently than self-esteem.
Furthermore, because measures of self-esteem tend to correlate
with narcissism (Baumeister, Smart, & Boden, 1996; Neff, 2003b;
Neff, Kirkpatrick, Rude, & Dejitterat, 2004), we thought it impor-
tant to extend Neff’s (2003b) finding that self-compassion was
unrelated to narcissism by showing that the two constructs bear
distinct relationships to people’s reactions to particular situations.
Overall, we expected to find that self-compassion, unlike self-
esteem or narcissism, would be related to lower negative emotions,
more equanimous thoughts, and less reactivity in response to the
negative scenario events. Whereas self-compassion was predicted
to be associated with more balanced reactions to difficult situations
(Neff, 2003b), previous research has shown that people who are
high in trait self-esteem or narcissism sometimes react strongly to
events that threaten their ego or sense of control (Baumeister,
Campbell, Krueger, & Vohs, 2003; Baumeister et al., 1996; Heath-
erton & Vohs, 2000).
Participants were 123 students (70 men, 53 women, ages 18 –22)
from the psychology department subject pool who received re-
quired experimental participation credit in a course.
Self-compassion. The Self-Compassion Scale (Neff, 2003b),
described earlier, was administered.
Self-esteem. Participants completed Rosenberg’s (1965) Self-
Esteem Inventory, the most widely used measure of trait self-
esteem. This scale has good interitem reliability (␣⬎.85) and
strong evidence of validity (see Blascovich & Tomaka, 1991).
Narcissism. The 40-item Narcissistic Personality Inventory
(Raskin & Hall, 1979) assesses seven components of narcissism:
authority, exhibitionism, superiority, vanity, self-sufficiency, enti-
tlement, and exploitativeness. It has high internal consistency (␣⫽
.83) and acceptable validity (Raskin & Terry, 1988) and has been
used in hundreds of studies.
Scenarios. Participants read three hypothetical scenarios in-
volving (a) getting a poor grade on an important test, (b) being
responsible for losing an athletic competition for their team, and
(c) forgetting their part while performing on stage, causing a
musical or dramatic performance to grind to an embarrassing halt.
Participants were asked to imagine themselves in each situation as
vividly as possible and then answer questions about each one.
First, participants were asked “How good or bad would you feel
about what happened?” (1 extremely bad, 12 extremely
good). They then rated the degree to which they would experience
20 feelings that were selected to assess five emotions: sadness
(sad, dejected, down, depressed), anxiety (nervous, tense, worried,
anxious), anger (angry, irritated, mad, hostile), embarrassment
(embarrassed, humiliated, disgraced, ashamed), and feelings of
incompetence (incompetent, worthless, stupid, self-conscious).
Ratings were made on 7-point scales (1 not at all, 4 moder-
ately, 7 extremely).
Participants were then asked to indicate how likely they would
be to react in seven ways if they found themselves in the situation
described (1 not at all, 2 slightly, 3 moderately, 4 very,
5 extremely): (a) remain relatively calm and unflustered, (b)
overreact, (c) experience strong emotions but not get carried away
with them, (d) have no emotional reaction whatsoever, (e) take the
situation in stride, (f) leave the situation quickly in order to deal
with my emotions, and (g) replay the situation in my mind for a
long time afterwards.
Finally, participants rated how likely they would be to think
each of seven thoughts (1 not at all, 2 slightly, 3 moder-
ately, 4 very, 5 extremely): (a) This is awful! (b) Everybody
goofs up now and then, (c) In the long run, this really doesn’t
matter, (d) I am such a loser, (e) I wish I could die, (f) This is sort
of funny, and (g) I should have expected this would happen.
Participants completed the Self-Compassion Scale (Neff,
2003b), Self-Esteem Inventory (Rosenberg, 1965), and Narcissis-
tic Personality Inventory (Raskin & Hall, 1979) during a mass
testing session. Several weeks later, they returned in small groups
to respond to the scenarios.
Internal reliabilities were acceptable for self-compassion (␣⫽
.73), self-esteem (␣⫽.87), and narcissism (␣⫽.82). The mean
self-compassion score was 18.2 (SD 3.47), the mean self-esteem
score was 39.7 (SD 6.62), and the mean for narcissism was 57.2
(SD 6.34). The correlations among self-compassion, self-
esteem, and narcissism were similar to those obtained by Neff
(2003b): self-compassion with self-esteem, r .58; self-
compassion with narcissism, r .18; and self-esteem with narcis-
sism, r .37 (all ps .05). When both self-esteem and narcissism
were entered as predictors of self-compassion in a multiple regres-
sion analysis, self-esteem remained a significant predictor (sr
.55, p .001), but narcissism did not (sr
⫽⫺.03, p .60).
The same two-stage strategy was used in all analyses. First,
Pearson correlations were calculated separately between self-
compassion, self-esteem, and narcissism on one hand, and the
outcome variables on the other. Then, a simultaneous multiple
regression analysis was conducted in which self-compassion, self-
esteem, and narcissism were entered and their semipartial corre-
lations with the outcome variables examined.
The five emotion scales (sadness, anxiety, anger, embarrass-
ment, negative self-feelings) each possessed acceptable reliability
(s .75), so items within each scale were summed. Principal-
axis factor analyses conducted on these five emotion scores sep-
arately for each scenario revealed a single factor on which all five
scales loaded greater than .40. In light of this, the five emotion
scales were summed to create a single negative affect score for
each scenario. Table 2 presents both zero-order and semipartial
correlations between self-compassion, self-esteem, and narcissism
and reactions to the three scenarios. Inspection of the boldface
semipartial correlations in Table 2 reveals that self-compassion
accounted for significantly more unique variance in negative affect
than self-esteem or narcissism. Whereas self-compassion ac-
counted for unique variance in emotion on all three scenarios,
self-esteem did not account for unique variance on any scenario,
and narcissism accounted for unique variance on only one.
A principal-axis factor analysis was performed on participants’
reports of what they would think in each of the situations. This
analysis revealed four factors that accounted for 63.5% of the
variance and reflected catastrophizing (e.g., “This is awful”), per-
sonalizing (e.g., “I am such a loser”; “I wish I could die”),
equanimity (e.g., “Everybody goofs up now and then”; “In the long
run, this doesn’t really matter”), and humor (e.g., “This is sort of
funny”). Items loading highly on each factor were summed (reli-
abilities for multi-item scales ranged from .67 to .79). As seen in
the third column of Table 2, only self-compassion predicted unique
variance in catastrophizing (one scenario), personalizing (all three
scenarios), and equanimity (all three scenarios). In contrast, self-
esteem predicted no unique variance in any of the thoughts, and
narcissism predicted only humor (one scenario).
Ratings of the seven behavioral reactions (e.g., I would remain
relatively calm and unflustered; I would overreact) were highly
correlated, with interitem reliabilities in excess of .70 for all three
scenarios. Thus, we reverse scored highly reactive statements and
summed the seven reactions to create an index of behavioral
equanimity. As seen in Table 2, self-compassion uniquely pre-
dicted behavioral equanimity for all three scenarios. In contrast,
narcissism predicted equanimity in only one scenario, and self-
esteem predicted equanimity in none.
Self-compassion displayed distinct patterns of associations with
most measures, patterns that mirrored those obtained in reactions
to real events in Study 1. Self-compassion uniquely predicted
emotional reactions to all three scenarios as well as thoughts that
reflected less catastrophizing, less personalizing, and greater equa-
nimity. Self-compassion was also associated with less extreme
behavioral inclinations in response to the imagined events.
As noted, we should not be surprised that people who treat
themselves with understanding and kindness feel better about
themselves than people who treat themselves critically and
harshly, and, as expected, the correlation between self-compassion
and self-esteem was high, replicating Neff (2003b). Even so,
self-compassion related to measures of emotional, cognitive, and
behavioral reactivity independently of self-esteem, and when self-
compassion and self-esteem were entered simultaneously into the
regression analyses, self-compassion accounted for unique vari-
ance in the outcome variables, whereas self-esteem did not. More
important, had we measured only self-esteem, we would have
undoubtedly concluded that all of these effects were because of
self-esteem, unaware that self-compassion may be the important
factor at work. These data raise the possibility that many concom-
itants of trait self-esteem identified in previous research may be a
function of self-compassion rather than self-esteem. We return
later to a general discussion of similarities and differences between
self-compassion and self-esteem.
Study 3
Thus far, we have seen that self-compassion predicts people’s
reactions to both real-life events (see Study 1) and hypothetical
situations involving failure, loss, and humiliation (see Study 2).
We extended the findings of the first two studies in Study 3 by
examining low- and high-self-compassionate people’s reactions to
an actual unpleasant interpersonal event, one in which they re-
ceived ambivalent feedback from another individual.
Being negatively evaluated or rejected by other people is an
inherently aversive experience, resulting in negative emotions
(such as hurt feelings, sadness, and anger), diminished self-esteem,
and a loss of perceived control (Leary, Koch, & Hechenbleikner,
2001; Williams & Zadro, 2005), and even ambivalent or neutral
evaluations from other people can be quite distressing (Leary,
Haupt, Strausser, & Chokel, 1998). Both the conceptualization of
self-compassion (Neff, 2003a) and our previous findings suggest
that self-compassion may buffer people against some of the dis-
tress that interpersonal ambivalence induces. People who treat
themselves kindly after receiving less positive reactions from
others than they desire, recognize the universality of rejection in
the human experience, and hold their emotions in mindful aware-
ness should find such episodes less distressing than those who are
low in self-compassion. They should also have less negative re-
actions toward those who do not evaluate them as they desire. In
Study 3, participants disclosed information about themselves to
another individual and then received either positive or neutral
feedback. Then, participants’ emotional reactions and judgments
of the other person were assessed.
In light of the high correlation between self-compassion and
trait self-esteem, we were interested in examining whether self-
compassion and self-esteem differentially moderate reactions to
feedback. The results from Neff (2003b) and Study 2 suggested
that self-compassion has distinct effects from self-esteem, but we
thought it was important to document that self-compassion and
self-esteem moderate people’s reactions to an actual event in
different ways. Furthermore, we were interested in examining the
possibility that self-compassion may moderate the relationship
between trait self-esteem and reactions to interpersonal feedback.
Previous research suggests that people with low self-esteem are
more reactive to the self-relevant implications of negative feed-
back than people with high self-esteem, whose positive self-views
and beliefs about their social acceptability help to buffer them
against negative evaluations (Campbell & Lavallee, 1993). If so,
then one may expect that being high in self-compassion would
provide a greater benefit to people who are low versus high in
self-esteem because a self-compassionate orientation may allow
low-self-esteem people to maintain equanimity in the face of
unflattering evaluations.
Sixty-six undergraduate psychology students (36 men and 30
women, ages 18–22) completed the Self-Compassion Scale (Neff,
2003b) and the Self-Esteem Inventory (Rosenberg, 1965) in a mass
testing session. The mean self-compassion score was 19.5 (SD
3.48), and the mean self-esteem score was 39.8 (SD 6.02).
Table 2
Comparative Effects of Self-Compassion, Self-Esteem, and Narcissism in Study 2
Failed important test Lost game for team Forgot part on stage
S-comp S-est Narciss S-comp S-est Narciss S-comp S-est Narciss
Total negative affect .34
.06 .09 .19
.12 .04 .17
.10 .18
Catastrophizing .25
.14 .06 .19
.09 .03 .18 .08 .07
.04 .11 .17 .03 .00 .16 .04 .05
Personalizing .42
.17 .36
.08 .08 .31
.08 .05 .27
.02 .16
Equanimity .46
.09 .38
.17 .36
.05 .03 .28
.03 .09 .23
.06 .16
Humorous .11 .07 .17
.19 .23
.13 .12 .18
.09 .04 .16 .15 .01 .18
.08 .00 .15
Behavioral equanimity .40
.05 .10 .31
.04 .11 .28
.07 .28
Note. The upper number in each cell is the Pearson correlation. The lower number is the semipartial correlation with the influence of the other two
predictors removed. Significant semipartial correlations are presented in boldface type. S-comp self-compassion; S-est self-esteem; Narciss
p .05.
p .01.
When participants arrived for the experiment several weeks
later, they were told that the study was investigating one-way
video interactions. They were led to believe that they would
introduce themselves on a video link that projected into another
room, where another participant (the “observer”) would ostensibly
watch and provide feedback about their introduction.
Video introduction. Participants were told to talk about them-
selves while facing a video camera for 3 min. To facilitate their
introduction, they were given a list of topics used in previous
research using a similar paradigm (B. Pontari, personal communi-
cation, September 24, 2004)—the participant’s hometown, major,
likes and dislikes about college, hobbies, interests, future plans,
things in life that are personally important, and miscellaneous
information. Participants were told to continue talking until the
researcher returned. The researcher turned on the video camera,
which appeared to be connected by cables to the adjacent room,
and left. After 3 min, the researcher returned, turned off the video
camera, and asked the participants to complete a questionnaire
about their introduction.
Feedback manipulation. The researcher left again and re-
turned with an envelope ostensibly containing feedback from the
observer who had watched the participant’s introduction. This
feedback consisted of the observer’s ratings on six bipolar scales
(socially unskilled–socially skilled, unfriendly–friendly, unlikable–
likable, cold–warm, unintelligent–intelligent, and immature–
mature), each of which was given on a 7-point scale (with 7 being
the most positive rating). Participants who were randomly assigned
to the positive-feedback condition received ratings of 5s, 6s, and
7s, whereas participants in the neutral feedback condition received
ratings of 3s, 4s, and 5s.
Neutral feedback was used rather than negative feedback to
minimize participants’ discomfort and because people typically
view neutral feedback as negative (Leary et al., 1998). Pilot testing
confirmed that the neutral feedback was viewed significantly more
negatively than the positive feedback. In a pilot study, 12 partic-
ipants were given either the positive- or the neutral feedback sheet
and rated the extent to which they would feel happy, angry, and
good about themselves if they received these ratings from another
person. Participants indicated that the neutral feedback would
make them feel less happy (Ms 3.3 vs. 9.3), more angry (Ms
7.7 vs. 3.5), and worse about themselves (Ms 4.2 vs. 10.0) than
the positive feedback (all ps .001).
Final questionnaire. After viewing the feedback, participants
completed a questionnaire that assessed reactions to the feedback,
state self-esteem, attributions for the feedback, and ratings of the
observer. First, participants rated the degree to which the feedback
was positive or negative (1 very negative; 12 very positive),
the extent to which their introduction gave the observer an accurate
understanding of them (1 not at all; 12 extremely), and how
much they thought the observer liked them (1 not at all; 12
Participants then rated themselves on 5-point emotion scales
that were designed to assess happiness (happy, cheerful, delighted,
pleased), sadness (down, depressed, sad, dejected), anger (irri-
tated, annoyed, mad, angry), and anxiety (anxious, tense, uneasy,
nervous) ranging from 1 (not at all)to5(extremely). Next,
participants rated the degree to which six factors may have caused
the observer’s reactions to them: (a) what I said about myself, (b)
the other person’s mood, (c) the other person’s tendency to be an
accepting versus a rejecting person, (d) my own social skill and
verbal ability, (e) my personality, and (f) the other person did not
get an accurate impression of what I am really like. Ratings were
made on 12-point scales with five equally spaced scale labels (not
a reason, a small reason, a moderate reason, a big reason, the
main reason). Finally, participants rated the degree to which they
thought they would like the observer (1 not at all; 12
extremely). Participants were thoroughly debriefed about the pro-
cedure of the study. They were assured that they were not actually
observed and that the feedback they received had been created for
the experiment in advance.
Data were analyzed with a series of three-step hierarchical
multiple regression analyses. Main effects of self-compassion
(zero centered), trait self-esteem (zero centered), and feedback
condition (dummy coded—0, 1) were entered on Step 1, followed
by the three two-way interactions on Step 2 and the three-way
interaction on Step 3. Significant interactions were decomposed by
testing simple slopes (Aiken & West, 1991), although we present
predicted condition means to facilitate descriptions of feedback
main effects. Data were examined for adherence to assumptions of
normality, homoscedasticity, and outliers.
Feedback Manipulation Check
A main effect of condition on ratings of the feedback showed
that the feedback manipulation was successful. Participants who
received positive feedback rated the other person’s impressions as
significantly more positive (M 9.9, SD 0.99) than those who
received neutral feedback (M 5.3, SD 0.97), t(65) 18.67,
.85, p .001. In addition, participants in the positive-
feedback condition indicated that the observer liked them more
(positive condition M 7.8, SD 1.48; neutral condition M
4.3, SD 1.32), t(65) 10.40, sr
.62, p .001.
Self-compassion moderated participants’ ratings of the feed-
back, however. The Feedback Self-Compassion interaction,
t(65) ⫽⫺1.99, sr
.01, p .05, showed that participants high
in self-compassion rated the neutral feedback as more positive than
participants low in self-compassion (b .12), t(65) 2.26, p
.03. However, self-compassion was not related to ratings of the
positive feedback (b ⫽⫺.02), t(65) ⫽⫺0.41, p .68. More
important, both low- and high-self-compassion participants per-
ceived the neutral feedback as more negative than the positive
feedback ( p .01).
Reactions to the Feedback
A main effect of feedback, t(64) 6.36, sr
.39, p .001, as
well as a three-way interaction of feedback, self-compassion, and
self-esteem, t(64) 2.29, sr
.05, p .03, were obtained on
answers to the question “To what extent did your introduction give
the other person an accurate understanding of who you are as a
person?” For participants low in self-esteem, the Feedback
Self-Compassion simple interaction showed that higher self-
compassion was associated with lower accuracy ratings of neutral
feedback but higher accuracy ratings of positive feedback (see
Figure 1). Put differently, participants who were low in both
self-compassion and self-esteem essentially accepted the neutral
feedback, but participants who were low in self-esteem but high in
self-compassion questioned it. In contrast, in the positive-feedback
condition, the highest accuracy ratings were made by participants
with low self-esteem and high self-compassion. Together, these
two patterns suggest that participants with low self-esteem cut
themselves more slack—in terms of accepting the validity of
positive feedback and recognizing that neutral feedback may not
be accurate—when they were high in self-compassion.
Emotional Reactions
The emotion measures had adequate internal consistency: anger,
␣⫽.80; anxiety, ␣⫽.89; sadness, ␣⫽.89. A factor analysis
conducted on the emotion scores revealed a single factor (eigen-
value 2.10) that accounted for 69.6% of the variance, so the
three scores were summed to create a measure of negative affect
(␣⫽.78). This analysis revealed a main effect of self-compassion
(but not self-esteem), t(65) ⫽⫺3.02, sr
.10, p .01, and a
Self-Compassion Self-Esteem interaction, t(64) 2.52, sr
.09, p .01, that were qualified by a significant interaction of
Feedback Self-Compassion Self-Esteem, t(63) 2.09, sr
.05, p .04. As seen in Figure 2, among participants who received
positive feedback, the Self-Compassion Self-Esteem simple
interaction was not significant ( p .20). However, in the neutral
feedback condition, a strong simple interaction emerged. The
greatest negative affect was reported by participants who scored
low in both self-compassion and self-esteem. However, negative
affect was lower with increasing self-compassion. As a result, the
lowest negative affect was expressed by participants with low
self-esteem and high self-compassion.
Attributions for Feedback
Self-compassion was also related to participants’ attributions for
the other person’s evaluations. A Feedback Self-Compassion
Condition interaction was obtained for the degree to which partic-
ipants attributed the feedback they received to their own person-
ality, t(65) ⫽⫺2.14, sr
.05, p .04. As seen in Figure 3a
participants who were low in self-compassion gave quite different
ratings for positive versus neutral feedback, whereas the ratings for
participants high in self-compassion were more similar. Ratings on
this item were significantly different between positive and neutral
conditions at 1 standard deviation below the self-compassion mean
(B 2.69), t(65) 3.45, p .001, but not at 1 standard deviation
above the self-compassion mean ( p .20). Participants low in
self-compassion were the most likely to make what appeared to be
defensive attributions, indicating that positive feedback, but not
negative feedback, was caused by their own personality.
A significant Feedback Self-Esteem interaction was also
obtained for this item, but the pattern was precisely opposite to that
of the Self-Compassion Feedback effect just described, t(65)
2.39, sr
.07, p .02. Specifically, participants high in self-
esteem indicated that positive feedback reflected their personality
more and that neutral feedback reflected their personality less (see
Figure 1. Accuracy of other person’s understanding. Solid triangles
positive/low (Lo) self-esteem (SE); solid xs positive/high (Hi) SE; solid
squares neutral/high SE; solid diamonds neutral/low SE.
Figure 2. Negative affect. Solid xs neutral/high (Hi) self-esteem (SE);
solid triangles positive/high (Hi) SE; solid squares positive/low (Lo)
SE; solid squares neutral/low SE.
Figure 3. Attributions to “My personality” for (a) self-compassion and
(b) self-esteem. Solid squares positive feedback; solid diamonds
neutral feedback.
Figure 3b). Ratings on this item were significantly different be-
tween positive- and neutral feedback conditions at 1 standard
deviation above the self-esteem mean (B 1.66), t(65) 2.13,
p .04, but not at 1 standard deviation below the mean of
self-esteem ( p .20).
Main effects of both self-compassion and self-esteem emerged
for the extent to which participants attributed the other person’s
evaluation to what they said about themselves in their video
introduction, but the effects were in opposite directions. Self-
esteem was inversely related to attributions to what they said about
themselves, t(65) ⫽⫺2.12, sr
.06, p .04, whereas self-
compassion was positively related to these attributions, t(65)
1.71, sr
.04, p .09. No significant effects were obtained on
the other attribution items.
Ratings of the Observer
Participants who received positive feedback understandably in-
dicated that they liked the observer more (M 8.2, SD 1.24)
than participants who received neutral feedback (M 5.3, SD
2.03), t(65) 7.44, sr
.45, p .001. In addition, participants
high in self-compassion indicated that they liked the observer more
than those lower in self-compassion, t(65) 2.13, sr
.04, p
.04, and a marginal Self-Compassion Feedback Condition in-
teraction suggests that this effect was stronger in the neutral
feedback condition, t(65) ⫽⫺1.70, sr
.002, p .10. Tests of
simple slopes revealed that self-compassion was associated with
greater liking for the observer in the neutral (B .28), t(64)
3.39, p .001, but not positive-feedback condition (B .04),
t(64) 0.56, p .58. Furthermore, the relationship between
feedback and ratings of liking was strongest for participants low in
self-compassion (B 3.80), t(64) 7.07, p .001. When rating
how much they liked the observer, participants higher in self-
compassion seem to have been less influenced by the rater’s
evaluation of them than participants lower in self-compassion.
Self-compassion was related to participants’ perceptions of the
feedback, attributions, emotional reactions, and ratings of other
people. Not surprisingly, self-compassion moderated reactions pri-
marily to the neutral feedback. Furthermore, participants low in
self-compassion reacted more differently to neutral than to positive
evaluations than participants high in self-compassion. For exam-
ple, participants low in self-compassion attributed positive feed-
back more to themselves but neutral feedback less to themselves,
but people high in self-compassion reacted more similarly to
positive and neutral feedback. These patterns suggest not only that
self-compassion buffers people against the psychological impact
of negative events but also that it more generally attenuates reac-
tions to both positive and negative events. This effect may occur
because two of the three components of self-compassion—
common humanity and mindful acceptance— have implications
for how people respond to good fortune as well as to bad. Recog-
nizing that success, like failure, is experienced by everyone, people
who are high in self-compassion may be less affected by it.
Similarly, maintaining a degree of mindful psychological distance
from one’s emotions may temper emotional reactions to success,
interpersonal acceptance, and other positive events.
This study provided further evidence for the discriminant valid-
ity of self-compassion vis-a`-vis trait self-esteem, as different pat-
terns were obtained for self-compassion and self-esteem. Self-
compassion, but not self-esteem, was associated with lower
negative affect, and high self-esteem was associated with stronger
differential reactions to positive versus neutral feedback. Whereas
participants higher in self-esteem were less likely to attribute
neutral feedback to themselves than those lower in self-esteem,
self-compassion was associated with indicating that feedback was
caused by one’s own personality. As suggested earlier, high self-
esteem may be associated with defensive reactions to neutral
feedback, whereas high self-compassion is not. Rather than deny-
ing that they may have caused negative life outcomes, self-
compassionate people may assume some level of personal respon-
sibility. However, even when assuming personal responsibility for
negative events, self-compassionate people do not seem to rumi-
nate about possible negative implications, as evidenced by lower
negative affect.
In addition, in some cases, self-compassion moderated reactions
to feedback differently, depending on whether participants were
high or low in self-esteem. Participants who were low in both
self-esteem and self-compassion had the most negative reactions to
the neutral feedback, but high self-compassion attenuated these
effects. Among participants with low self-esteem who received
neutral feedback, those with higher self-compassion were less
upset and more accepting of the feedback. Among participants
with low self-esteem, higher levels of self-compassion were also
associated with lower negative affect. Self-compassion may be
beneficial when coping with negative interpersonal events, and a
self-compassionate mindset may be particularly important for peo-
ple with low self-esteem.
Study 4
The first three studies showed clearly that self-compassionate
people react with greater equanimity to difficult situations than
people who are lower in self-compassion. Neff’s (2003b) concep-
tualization of self-compassion suggests that these effects arise
from differences in how low- versus high-self-compassion indi-
viduals think about negative events with respect to self-kindness,
mindfulness, and common humanity. Specifically, self-
compassionate people are assumed to be able to maintain emo-
tional equanimity while seeing themselves accurately (without
either self-enhancing or self-deprecating) because they compas-
sionately recognize their own imperfect humanity.
However, at least two other explanations may be offered. First,
self-compassionate people may treat themselves well because they
are more competent than people low in self-compassion (either
more capable in general or with respect to solving problems that
arise). Thus, they may react less strongly to negative events
because they know that they are good at dealing with whatever
happens. If individual differences in self-compassion spring from
real behavioral differences rather than (or in addition to) private
self-attitudes, then we should find that high-self-compassion indi-
viduals are rated more positively than low-self-compassion indi-
viduals by outside observers. However, finding that high-self-
compassion participants rate themselves more positively than
observers do would suggest that self-compassion lies in the mind
of the individual rather than in his or her objective actions and that
self-compassionate people display a self-enhancing bias similar to
that shown by people who are high in self-esteem (Robins & Beer,
A second possibility is that people who score high in self-
compassion have a generalized positive orientation toward every-
one. Thus, self-compassion may reflect a broad, positive interper-
sonal orientation that happens to include the individual him- or
herself. If people high in self-compassion are globally positive
rather than specifically compassionate toward themselves, then we
should find that they rate others more positively than people low in
self-compassion do.
To test these two explanations, participants were videotaped
while performing an awkward and mildly embarrassing task. Then
each videotape was rated either by the participant him- or herself
or by another participant. In this way, we were able to compare the
thoughts and feelings of low- and high-self-compassion individu-
als while evaluating themselves versus other people. Finding that
participants high in self-compassion are judged more positively by
observers than participants low in self-compassion would suggest
that self-compassionate people react with greater equanimity be-
cause they are better able to handle difficult situations that arise.
However, if observers rate low- and high-self-compassion partic-
ipants similarly, then it would seem that self-compassion does not
emerge from actual behavioral differences. Furthermore, finding
that self-compassionate participants rate other people’s perfor-
mances more positively than low-self-compassionate participants
do would suggest that self-compassion may involve a broad pos-
itive, optimistic, or compassionate orientation rather than a spe-
cific positive self-focused attitude.
In addition to testing differences in how low- and high-self-
compassion people rate themselves versus others, in Study 4, we
also examined participants’ emotional comfort in a potentially
awkward and embarrassing situation. As noted, emotional equa-
nimity is considered a primary feature of self-compassion, so
assessing emotional upset or calm in a real situation provides an
additional test of Neff’s (2003b) conceptualization of self-
Participants were 48 male and 54 female participants, ages
18 –21, from the introductory psychology subject pool who re-
ceived required experimental participation credit for their course.
Participants completed the Self-Compassion Scale (Neff,
2003b) in a mass testing session (M 19.7, SD 3.01, ␣⫽.78).
Several weeks later, participants were tested individually by one of
three female researchers. The researcher seated the participant at a
desk in a lab room and explained that the participant would be
videotaped while performing a task and then would watch and
evaluate either his or her own videotaped performance or the
videotape made by a previous participant. The informed consent
form asked for the participant’s permission to show his or her
videotape to other participants if needed.
The researcher started a video camera, situated 6 feet (1.83 m)
from the participant, as well as an audiotape containing the in-
structions and left the room. The audiotape instructed the partici-
pant to look into the camera and make up a children’s story that
began “Once upon a time, there was a little bear. . . . After a
minute had passed, the researcher returned and randomly assigned
the participant to watch either the videotape he or she had just
made or a videotape made previously by another participant of his
or her own gender. One third of the participants (n 34) rated
their own videotape. The other two thirds of the participants also
rated one of these 34 videotapes, subject to the constraint that each
tape was rated by both a low- and a high-self-compassion individ-
ual (based on a median split, Mdn 19) of the same gender as the
person who recorded the tape. Thus, over the course of the study,
each of the 34 target videotapes was rated three times— by the
participant who had made the tape, by a low-self-compassionate
participant, and by a high-self-compassionate participant.
First, participants rated how they (if rating their own tape) or the
other person (if rating another participant’s tape) appeared on the
videotape on nine adjectives—awkward, competent, confident,
attractive, nervous, foolish, creative, likable, and reasonable. Rat-
ings were made on 7-point scales with three scale labels (1 not
at all, 4 moderately, 7 extremely). Participants also rated how
they felt while watching the videotape on eight emotions—
relaxed, embarrassed, happy, irritable, sad, nervous, proud, and
peaceful on 7-point scales (1 not at all, 4 moderately, 7
extremely). Finally, they provided their overall evaluation of their
own or the other person’s story on a 12-point scale with five
equally spaced labels (very bad, somewhat bad, neither bad nor
good, somewhat good, and very good) and rated how good or bad
they felt while watching the videotape (1 very bad, 12 very
good). After completing the final questionnaire, participants were
debriefed and dismissed.
Because each of the 34 targets’ videotapes was rated three
times— by the participant him- or herself, by a low-self-
compassionate participant, and by a high self-compassionate par-
ticipant—analyses were conducted in a manner that accounted for
the nonindependence of the data. Repeated measures multiple
regression analyses were conducted using targets’ self-compassion
scores, rater type (self, low-self-compassion other, high-self-
compassion other), and their interaction as predictors. In this
analysis, self-compassion was treated as a continuous between-
subjects predictor (zero centered), and rater was treated as a
categorical within-subjects predictor with three levels (dummy
coded). This analysis was conducted using the general linear
model procedure on SPSS and is equivalent to a between-within
regression analysis that treated rater as a within-subjects factor
(Pedhazur, 1997). Significant interactions were decomposed via
tests of simple slopes.
Ratings of Performance on the Task
A multiple regression analysis performed on participants’ rat-
ings of the story revealed a significant Self-Compassion Rater
interaction, F(2, 124) 2.27, p .05. As can be seen in Figure 4,
tests of simple slopes revealed that target self-compassion was
related to ratings of the story when participants rated themselves
(B 0.34), t(33) 2.32, p .03, but was unrelated to ratings
made by either low- or high-self-compassionate observers ( ps
.40). In addition, differences between self-ratings and observer
ratings emerged only when targets were low in self-compassion.
Given that other low- and high-self-compassionate raters detected
no difference in the quality of the stories told by low- and high-
self-compassion targets, it seems that low-self-compassion targets
rated their own performance unduly harshly.
The internal reliability of the ratings of the target (e.g., awk-
ward, competent, confident) was high (Cronbach’s ␣⬎.70), so we
summed the nine items to create an overall evaluation of the target
(after reverse scoring negatively valenced items). Analyses re-
vealed a significant two-way interaction that was virtually identi-
cal to that obtained on ratings of the story, F(4, 124) 2.62, p
.04, sr
.09. As with ratings of the story (see Figure 4), tests of
simple slopes revealed that self-compassion was unrelated to either
low- or high-self-compassion observers’ evaluations of the target
( ps .85). However, when targets rated themselves, self-
compassion was related to the positivity of these self-ratings (B
0.17), t(33) 2.69, p .01.
Emotional Reactions
Separate measures of positive and negative emotion were cal-
culated by summing the relevant emotion ratings. Analysis of the
sum of the four positive emotions (relaxed, happy, proud, peaceful;
all s .70) revealed a two-way Self-Compassion Rater
interaction, F(1, 31) 3.70, p .06. As seen in Figure 5, target
self-compassion was related to positive emotions only when par-
ticipants watched themselves (B .18), t(33) 2.43, p .05.
Effects for low- and high-self-compassion observers were not
significant ( ps .80). Analysis of the sum of the negative emo-
tions (embarrassed, irritable, sad, nervous; all s .70) revealed
only a main effect of target self-compassion, showing that self-
compassion was inversely related to negative emotion, F(1, 31)
7.12, p .01, sr
.18. This finding suggests that raters felt
greater negative affect watching low-self-compassion targets, pos-
sibly because low-self-compassion targets subtly communicated
their own discomfort to the raters.
Participants’ overall ratings of how good versus bad they felt
while watching the videotape revealed a significant Target Self-
Compassion Rater interaction, F(2, 60) 3.18, p .05, sr
.09. When participants watched themselves, their self-compassion
scores were positively related to how good they felt (B .36),
t(33) 2.58, p .02. However, when participants rated another
individual, target self-compassion scores did not predict ratings for
either low- or high-self-compassionate individuals (Bs ⫽⫺.07 and
.04, ps .50).
The results of Study 4 provide five general conclusions about
the nature of self-compassion. First, as expected, participants’
reactions to their own videotaped performances differed as a
function of self-compassion. Participants who were low in self-
compassion evaluated their answers less favorably and rated their
personal characteristics (as observed on the videotape) less posi-
tively. They also felt worse while watching the tape compared with
participants who were high in self-compassion.
Second, these effects were primarily attributed to the fact that
low-self-compassionate participants undervalued their perfor-
mances relative to observers. Whereas self-compassionate partic-
ipants’ ratings were similar to observers’ ratings, low-self-
compassionate participants rated themselves significantly less
positively than the observers did. This pattern is notably different
than that obtained in previous studies of trait self-esteem, which
generally show that high self-esteem is associated with inflated
evaluations of one’s performance and characteristics relative to
observers’ ratings (Robins & Beer, 2001). In contrast to the self-
enhancing tendencies of people who are high in self-esteem, those
who are high in self-compassion appear to judge themselves as
others do.
Third, the differences in low- versus high-self-compassionate
participants’ self-evaluations and emotional reactions did not ap-
pear to be based on real differences in their performances. Al-
though low-self-compassionate targets rated their performances
less positively than high-self-compassionate targets did, observers
did not judge low- and high-self-compassionate targets differently.
This pattern suggests that the highly self-compassionate partici-
pants had more accurate perceptions of themselves than less self-
compassionate participants, although the possibility exists that less
self-compassionate participants were aware of negative aspects of
their behavior or characteristics that were not obvious to observers.
Fourth, we saw no evidence to suggest that self-compassion
predicted participants’ ratings of other people. Low- and high-self-
compassion observers did not differ in their evaluations of targets’
Figure 4. Rating of story. Solid squares low-self-compassion (SC)
rater; solid diamonds self-rating; solid triangles high-self-compassion
Figure 5. Positive emotion. Solid diamonds self-rating; solid trian-
gles high-self-compassion (SC) rater; solid squares low-self-
compassion rater.
answers or personal characteristics. This finding does not bear on
the issue of whether self-compassionate people treat others more
compassionately (see Neff, 2003a; Salzberg, 1997) but does sug-
gest that they do not make generally more positive evaluations of
Finally, the results showed that self-compassion predicted pos-
itive affect when participants watched their own videotape but not
when they watched others’ tapes. This finding supports Neff’s
(2003b) hypothesis that self-compassion is related to equanimity in
unpleasant, stressful, and awkward situations. It also provides
additional evidence that self-compassion is distinct from more
general feelings of compassion toward others.
Study 5
The first four studies documented a number of psychological
concomitants of trait self-compassion and showed that self-
compassion bears a distinctly different relationship to thought,
behavior, and emotion than do self-esteem and narcissism. In light
of the findings from the four previous studies, the purpose of Study
5 was threefold. First, Study 5 was designed to examine how
self-compassion moderates people’s reactions to remembered life
events. Studies 1, 3, and 4 documented the relationship between
self-compassion and contemporaneous reactions to immediate
events, and Study 2 examined how low- versus high-self-
compassionate individuals respond to hypothetical situations. In
Study 5, participants recalled a previous failure, rejection, or loss
that made them feel badly about themselves and answered ques-
tions about it.
The second purpose of Study 5 was to determine whether a
self-compassionate perspective could be experimentally induced.
To induce self-compassion, some participants responded in writing
to prompts that were designed to lead them to think about the
negative event in ways that tapped into the three components of
self-compassion identified by Neff (2003a)—self-kindness, com-
mon humanity, and mindful acceptance.
Third, Study 5 compared the effects of this self-compassion
induction with a self-esteem induction. Studies 2 and 3 showed
that, at the trait level, self-compassion and self-esteem have dis-
tinct relationships to various outcomes. The question examined
here was whether transitory inductions of state self-compassion
versus state self-esteem would reveal similar differences. Thus,
whereas some participants responded in writing to self-compassion
prompts, others responded to prompts that were designed to protect
or bolster their self-esteem—reminding them of their positive
characteristics, leading them to explain the negative event in a way
that lowered their personal responsibility for it, and asking them to
explain why the event does not reflect on them personally. These
three prompts were based on research dealing with the tactics that
people use to protect their self-esteem, such as self-affirmation,
self-serving attributions, and egotistical reframing (see Blaine &
Crocker, 1993; Greenwald, 1980; Steele, 1988; Taylor & Brown,
1988), as well as interventions designed to raise people’s self-
esteem in clinical settings (see Mruk, 1995). A writing-only con-
dition was also included in the design to control for the possibility
that merely writing about negative events in a self-disclosing
manner may reduce negative emotions (Pennebaker, Colder, &
Sharp, 1990). We predicted not only that participants scoring low
versus high in trait self-compassion would react differently to the
remembered event (with self-compassionate people responding
more positively) but also that the induction of state self-
compassion would attenuate negative reactions to the event com-
pared with the self-esteem induction and control conditions.
Participants were 54 male and 61 female undergraduate stu-
dents, ages 17–22, from the psychology subject pool. They re-
ceived required experimental participation credit for their partici-
pation. Experimental sessions were conducted by one of three
female researchers.
At the beginning of the semester, participants completed the
Self-Compassion Scale (Neff, 2003b) and Rosenberg Self-Esteem
Inventory (Rosenberg, 1965), described earlier. Several weeks
later, participants reported for the experiment, where they were
tested individually. Initial instructions stated that participants
would write about a negative event from their past and answer
questions about it. After signing an informed consent form, par-
ticipants were asked to “. . . think about a negative event that you
experienced in high school or college that made you feel badly
about yourself—something that involved failure, humiliation, or
Participants were asked to describe the event and then provide
details regarding what led up to the event, who was present,
precisely what happened, and how they felt and behaved at the
time. After writing about the event, participants were randomly
assigned to one of four experimental conditions (a) self-
compassion induction (n 29), (b) self-esteem induction (n
31), (c) writing control (n 28), or (d) true control (n 27).
Self-compassion induction. Participants in the self-
compassion condition responded to three prompts that were de-
signed to lead them to think about the event in a self-
compassionate manner (see Neff, 2003a). The first prompt,
designed to focus on the common humanity element of self-
compassion, asked participants to list ways in which other people
also experience similar events. The second prompt asked partici-
pants to write a paragraph expressing understanding, kindness, and
concern to themselves in the same way that they might express
concern to a friend who had undergone the experience, thereby
focusing on the self-kindness component of self-compassion. To
induce a mindful perspective on the event, the third prompt in-
structed participants to describe their feelings about the event in an
objective and unemotional fashion.
Self-esteem induction. Participants in the self-esteem condi-
tion responded to three prompts geared toward leading them to feel
good about themselves. The first item prompted them to “write
down your positive characteristics—indications that you are com-
petent and valuable.” The second prompt asked participants to
write a paragraph explaining how what had happened was not
entirely their fault and to interpret the event in a way that made
them feel better about themselves. The third prompt instructed
them to describe “why the event does not really indicate anything
about the kind of person you are.”
Writing control condition. The writing control condition was
included to control for the fact that merely writing about negative
events can change how people feel. The instructions were modeled
after those used in previous studies on the effects of self-disclosure
on emotion (Pennebaker et al., 1990). Participants were instructed
to “really let go” and explore their deepest emotions as they wrote
about the event.
Control condition. After describing the negative event and
their feelings at the time, participants in the control condition
completed the dependent measures.
Dependent measures. After completing one of the four condi-
tions, participants rated how they felt on 16 items that were
selected to assess four emotions: happiness (happy, cheerful, de-
lighted, pleased), sadness (down, depressed, sad, dejected), anger
(irritated, annoyed, mad, angry), and anxiety (anxious, tense,
uneasy, nervous). Participants rated themselves on each adjective
on a 7-point scale ranging from 1 (not at all)to7(extremely).
Participants then rated the degree to which the event they
described was caused by (a) other people, (b) something they did,
(c) bad luck, and (d) the kind of person they are (their personality,
abilities, attitudes, character, and so on); they also indicated how
much the event was their fault. Each question was answered on a
12-point scale with five scale labels (not at all, slightly, moder-
ately, very, completely). Participants also rated the event they
described on a 12-point scale ranging from 1 (very good)to12
(very bad), and, to directly assess the effect of the manipulation on
perceptions of common humanity, participants answered the ques-
tion, “Overall, how similar to or different from most other people
are you?” ranging from 1 (very different)to12(very similar).
Trait self-compassion scores (M 18.46, SD 3.51) and trait
self-esteem scores (M 39.92, SD 6.20) were obtained for each
participant. Cronbach’s alpha coefficient revealed acceptable reli-
ability for both the Self-Compassion Scale (␣⫽.81) and the
Self-Esteem Scale (␣⫽.87). As before, self-compassion corre-
lated with trait self-esteem (r .56, p .01). To eliminate the
confounding effects of trait self-esteem, the effects of trait self-
esteem were partialed out on all analyses. Specifically, we con-
ducted hierarchical multiple regression analyses in which trait
self-esteem was entered on Step 1, followed by experimental
condition (dummy coded via three dummy variables), self-
compassion (zero centered), and the Condition Self-Compassion
Because participants chose their own negative event, the possi-
bility exists that the events chosen by low- versus high-self-
compassion participants differed in some systematic way. How-
ever, analyses revealed no effects of self-compassion, the
experimental manipulation, or their interaction on ratings of how
“bad” the event was ( ps .20).
The four items measuring each emotion were summed to pro-
vide measures of anger (␣⫽.92), sadness (␣⫽.90), anxiety (␣⫽
.84), and happiness (␣⫽.86). Because a principal-axis factor
analysis revealed that these four scores formed a single factor,
these four emotion measures were summed (after reverse scoring
happiness). After partialing out trait self-esteem (which accounted
for 7.7% of the variance), the regression analysis revealed only a
main effect of the experimental manipulation on this variable, F(3,
110) 2.77, p .05, sr
.065. Post hoc tests revealed that
participants in the self-compassion condition (M 46.5) reported
significantly lower negative affect than participants in the self-
esteem (M 54.0), disclosure (M 53.5), or control (M 58.6)
conditions ( ps .05), which did not differ from each other.
Attributions and Ratings
Only a main effect of the experimental manipulation was ob-
tained on how much participants thought that the negative event
was caused by the kind of person they were, F(3, 110) 3.11, p
.05, sr
.08. Participants in the self-compassion condition (M
6.9) thought that the negative event was caused more by the kind
of person they were than participants in the self-esteem (M 4.4),
self-disclosure (M 5.8), and control conditions (M 5.5; ps
.05), which did not differ significantly from each other ( ps .05).
Neither experimental condition nor trait self-compassion were
related to participants’ ratings of how much the event was caused
by other people, by something they did, or by bad luck, or on
ratings of how much the event was their fault.
A main effect of condition was obtained on participants’ ratings
of how similar to most other people they were, F(3, 110) 3.17,
p .05, sr
.08, that was qualified by a significant Condition
Trait Self-Compassion interaction, F(3, 106) 2.77, p .05,
.07. The main effect showed that participants in the self-
compassion condition (M 7.8) thought that they were signifi-
cantly more similar to other people than participants in the self-
disclosure (M 6.0) or control conditions (M 6.7; ps .05).
The Condition Self-Compassion interaction is shown in Figure
6. Tests of simple slopes revealed that the regression line for the
self-compassion condition significantly differed from zero (B
0.34), t(106) 2.01, p .05, whereas the slopes for the
self-esteem, writing, and control conditions did not. The pattern
suggests that the self-compassion induction markedly increased
perceptions of similarity for participants who were low in trait
Figure 6. Rating of similarity to others. The regression line for the
self-compassion induction significantly differs from zero, but the slopes for
the self-esteem, writing, and control conditions do not. Solid squares
self-esteem condition; xs control condition; solid diamonds self-
compassion condition; solid triangles writing condition.
In Study 5, we successfully induced state self-compassion by
leading participants to think about personally relevant events in
terms of self-kindness, common humanity, and mindful acceptance
(Neff, 2003a, 2003b). Participants in the self-compassion condi-
tion reported lower negative affect than those in any other condi-
tion. They also indicated that the event was more likely to be
caused by the kind of person they were than participants in other
The latter finding is intriguing in that the self-compassion con-
dition led participants to report less negative affect even as they
acknowledged that their personal characteristics played a larger
role in causing the negative event in the first place. To follow up
on this finding, we calculated within-cell correlations between
personal attributions and the emotion composite for each experi-
mental condition. In the control condition, attributing the event to
“the kind of person you are” correlated positively with negative
emotions (r .18, p .05), but in the self-compassion condition,
the correlation between taking responsibility and negative emotion
was virtually zero (r ⫽⫺.03). Although only suggestive, these
findings indicate that inducing self-compassion may decouple the
relationship between taking responsibility and experiencing nega-
tive affect. The self-compassion induction allowed participants to
acknowledge that they were the kind of people who made mis-
takes, yet they did not feel badly about something that is a common
experience. As a result, they were less defensive and distressed, as
shown by the lower ratings of negative emotions. This pattern may
alleviate the concern that self-compassion could lead to compla-
cency or shirking of responsibility (Neff, 2003a). Apparently,
self-compassion does not undermine people’s willingness to ac-
cept responsibility for their actions and, in fact, may promote it.
This possibility deserves future research attention.
For participants who were low in self-compassion, the self-
compassion induction increased perceptions of similarity to other
people. However, people who were high in self-compassion were
not affected by the self-compassion induction on this measure.
Thus, the self-compassion induction seemed to be particularly
effective (especially in increasing perceptions of common human-
ity) for people who were lower in self-compassion.
As expected, the self-compassion and self-esteem manipulations
had different effects on participants’ reactions. As noted, partici-
pants in the self-compassion induction condition expressed lower
negative emotions than those in the self-esteem induction condi-
tion, and they also believed that they were more similar to other
people. Finally, as noted, participants in the self-compassion con-
dition rated their character or personality as playing a greater role
in the negative event than participants in the self-esteem condition.
Given the multifaceted nature of the experimental manipulation,
we cannot with any certainty identify the key element that created
these differences. However, one possibility may involve the fact
that the self-compassion and self-esteem conditions differed in the
degree to which they encouraged participants to focus on them-
selves. Whereas the self-esteem condition focused participants on
their positive characteristics and on thinking about why the neg-
ative event did not reflect on them, the self-compassion condition
implicitly led participants to accept the negative event but then to
counteract its negative effects via self-kindness, common human-
ity, and mindful acceptance. Even though the prompts in the
self-esteem condition may have affirmed participants’ positive
self-images, they nonetheless kept the participants focused on
themselves instead of widening their perspective to include other
people, which may explain why the self-esteem induction did not
yield the same benefits as self-compassion.
General Discussion
These five studies show that self-compassion is an important
construct that moderates reactions to distressing situations involv-
ing failure, rejection, embarrassment, and other negative events.
Self-compassion was associated with lower negative emotions in
the face of real, remembered, and imagined events and with
patterns of thoughts that generally facilitate people’s ability to
cope with negative events. Particularly interesting was the will-
ingness of self-compassionate people to accept responsibility for
their role in negative events. Participants high in self-compassion
were more likely to believe that their own personality resulted in
receiving a mediocre evaluation (see Study 3) and to attribute the
cause of negative events to the kind of person they are (see Study
5). Even so, they were less likely to ruminate about unpleasant
evaluations (see Study 3) or experience negative affect when
confronted with their mistakes (see Study 5). Together, these
findings suggest that self-compassionate people more readily ac-
cept undesirable aspects of their character and behavior than
people low in self-compassion without obsessing about them,
becoming defensive, or feeling badly.
Clearly, self-compassion buffers people against the impact of
negative events, but why? The present studies suggest three pos-
sible processes that may underlie differences between low- and
high-self-compassion individuals. First, Study 4 showed not only
that highly self-compassionate people judged themselves less
harshly than those low in self-compassion but also that their
self-evaluations and emotional reactions were based more on their
actual performance (as assessed by observers) than those of less
self-compassionate people, who judged their performances more
unfavorably than observers did. These results suggest that self-
compassionate people may be more accurate in their self-
evaluations, possibly because their self-judgments are less tainted
by either catastrophizing self-criticism, on one hand, or defensive
self-enhancement, on the other. Additional research is needed to
examine this hypothesis directly.
Second, the results suggest that self-compassionate people’s
self-evaluations may not depend as strongly on their outcomes as
those of less self-compassionate people, presumably because they
respond in a kind and accepting manner toward themselves
whether things go well or badly. In contrast, people low in self-
compassion may feel good about themselves primarily when life
treats them well, for example, after they have succeeded or re-
ceived positive feedback. If high self-compassion is associated
with lower reactivity to external events, then we should find that
the self-esteem of highly self-compassionate individuals is more
stable over situations and time. Given that stable self-esteem is
associated with more adaptive responses to failure and rejection
than unstable self-esteem (Greenier et al., 1999; Kernis, Cornell,
Sun, Berry, & Harlow, 1993), the question may be raised whether
the benefits of stable self-esteem stem from stability per se or from
the fact that people with stable self-esteem tend to be high in
Third, people high in self-compassion appear to cognize about
negative events in ways that reduce their impact. The present
studies showed that self-compassionate participants had more self-
relevant thoughts that reflected self-kindness, common humanity,
and mindful acceptance.
Self-Compassion Versus Self-Esteem
As noted earlier, self-compassionate people should feel better
about themselves than low self-compassionates do, yet the high
correlation between self-compassion and trait self-esteem raises
questions regarding the discriminant validity of these constructs
(Neff, 2003a; Neff et al., 2004, 2005). The present research pro-
vided additional evidence that self-compassion is related to
thoughts, feelings, and behavior differently than self-esteem. Self-
compassion, but not self-esteem, was uniquely related to lower
negative affect and, in Study 3, self-compassion, but not self-
esteem, was related to more favorable ratings of other people and
lower negative affect after receiving unflattering feedback.
Self-compassion and self-esteem were also differentially related
to the degree to which participants attributed negative outcomes to
themselves. Participants high in self-esteem tended to attribute
unflattering evaluations and negative events less to themselves
than people low in self-esteem. In contrast, people high in self-
compassion did the opposite, attributing unfavorable events more
to their own personality. Thus, one difference between self-
compassion and self-esteem might involve defensiveness. When
people high in self-esteem encounter negative life events, they
sometimes engage in self-serving biases (e.g., downward social
comparisons, egotistical attributions), presumably because such
reactions help to make them feel better about themselves (Blaine &
Crocker, 1993; Taylor & Brown, 1988). In contrast, self-
compassionate people may assume personal responsibility while
simultaneously being kind to themselves.
Both of these tactics can help people to avoid negative self-
feelings, but self-compassion may have an advantage that self-
enhancement does not. As many theorists have noted, people often
maintain positive self-feelings through self-serving illusions
(Blaine & Crocker, 1993; Leary, 2004; Murray, Holmes, & Grif-
fin, 1996; Taylor & Brown, 1988). However, such illusions are
often maintained tenuously, girded by a certain degree of self-
deception and in spite of other people’s unflattering judgments.
Not only do self-serving illusions compromise people’s ability to
see themselves and other people accurately (Leary, 2004, 2007),
but they are sometimes impossible to sustain in the face of incon-
trovertible evidence of one’s shortcomings. In contrast, a self-
compassionate orientation requires no illusions or defensiveness
but rather involves a clear perception of one’s characteristics, both
good and bad. Given that it may be more beneficial to recognize
rather than deny one’s shortcomings, while remaining kind and
understanding toward oneself, self-compassion may foster accu-
rate perceptions and, thus, more effective behavior.
Recently, Gilbert (2005; Gilbert & Irons, 2005) has contrasted
self-compassion and self-esteem with reference to their relation-
ships to biopsychosocial systems that mediate reactions to threat.
He suggested that self-compassion deactivates the “threat system”
(associated with insecurity and defensiveness) and activates the
“self-soothing system” (associated with feelings of safety),
whereas self-esteem is associated with evaluations of one’s rela-
tive superiority and social rank. Thus, self-compassion promotes
emotional regulation, but self-esteem does not. Our studies do not
address whether two fundamentally different systems are involved
in self-compassion versus self-esteem, but the findings are consis-
tent with the idea that self-compassion and self-esteem have quite
different effects.
A few findings from the present studies show that self-
compassion is particularly beneficial for people who are low in
self-esteem, suggesting that people who treat themselves compas-
sionately in spite of unflattering self-evaluations fare as well, if not
better, than those with high self-esteem. This pattern may emerge
from the fact that it may be better to have nondefensive (i.e.,
self-compassionate) low self-esteem than defensive high self-
esteem. This consideration raises the possibility that self-
compassion may help to distinguish people with true or authentic
self-esteem from those with defensive self-esteem (Schneider &
Turkat, 1975). People high in both self-compassion and self-
esteem might be more likely to have “optimal” self-esteem that is
characterized, among other things, by stable feelings of self-worth
that do not fluctuate greatly over time (Kernis, 2003). As a result,
people with optimal self-esteem are less defensive and do not need
to improve their self-feelings through downward social compari-
son and other self-enhancing tactics.
Perhaps the most striking finding vis-a`-vis the comparison of
self-compassion and self-esteem was that self-compassion rather
than self-esteem accounted for more of the unique variance in
outcomes that are typically regarded as correlates of self-esteem. A
great deal of previous research has shown that trait self-esteem
predicts almost all negative emotions, including anxiety, depres-
sion, embarrassment, and hurt feelings (Baumeister et al., 2003;
Leary, Schreindorfer, & Haupt, 1995). However, the fact that
simultaneous regression analyses showed that these relationships
were more clearly associated with self-compassion than with self-
esteem raises the possibility that what are widely regarded as
self-esteem effects may, in fact, be better explained in terms of
self-compassion. Clearly, this possibility warrants a great deal of
research attention.
Avenues for Future Research
Although the present studies provide glimpses of the thoughts
that are associated with a self-compassionate perspective, addi-
tional research is needed to describe the cognitions and self-
evaluations of self-compassionate people more fully. For example,
it is unclear from our studies whether people who are high in
self-compassion engage in self-evaluation less than people who are
low in self-compassion or whether they self-evaluate just as fre-
quently but maintain a reasonably positive self-view by accepting
themselves in spite of their shortcomings (see Chamberlain &
Haaga, 2001a, 2001b).
Although self-compassion appears to have certain benefits, fu-
ture research should explore the possibility that it may also have
drawbacks. For example, it is important to consider whether self-
compassionate people are sometimes complacent or lazy, forgiving
themselves for their mistakes and not taking action to prevent
future ones. Previous research has suggested that this is not the
case, showing that self-compassion is not related to the stringency
of people’s personal standards and that students high in self-
compassion were more likely to react to academic failure by
viewing it as a chance for growth, acceptance, and positive rein-
terpretation (Neff et al., 2005). Self-compassion correlated nega-
tively with denial or mental disengagement, indicating a mindset
of confronting current negative outcomes rather than avoiding
them. In line with this finding, Studies 3 and 5 revealed that
self-compassionate people took responsibility for their actions.
Even so, an interesting question for future research is precisely
why self-compassion does not lead to indifference or compla-
Third, although Study 5 was not designed to help participants
with a particular personal problem, the results suggest that foster-
ing a self-compassion mindset may be beneficial in clinical set-
tings with clients who are excessively self-critical. Indeed, re-
searchers have begun to explore the psychotherapeutic benefits of
self-compassion (Gilbert & Irons, 2005) as well as examine the
role of self-compassion in the effectiveness of other treatments,
such as mindfulness-based stress reduction (Shapiro, Astin,
Bishop, & Cordova, 2005). Rather than focusing on changing
clients’ self-evaluations, as many cognitive–behavioral ap-
proaches do, self-compassion changes people’s relationships to
their self-evaluations. In clinical settings, clients with low self-
esteem may be better served by interventions that foster self-
compassion rather than self-esteem. If clients learn to feel better
about themselves but continue to castigate themselves when they
fail or make mistakes, then they will remain unable to cope
nondefensively with their difficulties. And, perhaps most impor-
tantly, it should be easier to teach people with low self-esteem to
be self-compassionate than to raise their self-esteem (Swann,
1997). Although clinical psychologists have begun to explore the
clinical benefits of training people to be self-compassionate (Gil-
bert & Irons, 2005; Marlatt, 2002; Wiser & Telch, 1999), a great
deal of research is needed on the basic processes involved in
Finally, future research should examine how self-compassionate
people respond to serious real-life events. The situations that
participants confronted in our studies were relatively mild in
comparison to many of the hardships people encounter in daily
life. Our research shows that being self-compassionate helps peo-
ple deal with real and imagined negative events, but the question
remains of the degree to which treating oneself kindly buffers
people against the impact of severe challenges in everyday life.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. Thousand Oaks, CA: Sage.
Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003).
Does high self-esteem cause better performance, interpersonal success,
happiness, or healthier lifestyles? Psychological Science in the Public
Interest, 4, 1–44.
Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened
egotism to violence and aggression: The dark side of high self-esteem.
Psychological Review, 103, 5–33.
Berger, E. M. (1952). The relation between expressed acceptance of self
and expressed acceptance of others. Journal of Abnormal Psychology,
47, 561–571.
Blaine, B., & Crocker, J. (1993). Self-esteem and self-serving biases in
reactions to positive and negative events: An integrative review. In R. F.
Baumeister (Ed.), Self-esteem: The puzzle of low self-regard (pp. 55–
85). New York: Plenum Press.
Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P.
Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of
personality and social psychological attitudes (pp. 115–160). San Di-
ego, CA: Academic Press.
Bonanno, G. A. (2004). Loss, trauma, and human resilience. American
Psychologist, 59, 20–28.
Bryk, A., Raudenbush, S., & Congdon, R. (1998). HLM. Chicago: SSI
Campbell, J. D., & Lavallee, L. F. (1993). Who am I? The role of
self-concept confusion in understanding the behavior of people with low
self-esteem. In R. F. Baumeister (Ed.), Self-esteem: The puzzle of low
self-regard (pp. 3–20). New York: Plenum Press.
Chamberlain, J. M., & Haaga, D. A. F. (2001a). Unconditional self-
acceptance and psychological health. Journal of Rational-Emotive &
Cognitive-Behavior Therapy, 19, 163–176.
Chamberlain, J. M., & Haaga, D. A. F. (2001b). Unconditional self-
acceptance and responses to negative feedback. Journal of Rational-
Emotive & Cognitive-Behavior Therapy, 19, 177–189.
Diener, E., Lucas, R. E., & Oishi, S. (2002). Subjective well-being: The
science of happiness and life satisfaction. In C. R. Snyder & S. J. Lopez
(Eds.), The handbook of positive psychology (pp. 63–73). Oxford, En-
gland: Oxford University Press.
Gilbert, P. (2005). Compassion and cruelty: A biopsychosocial approach.
In P. Gilbert (Ed.), Compassion: Conceptualisations, research and use
in psychotherapy (pp. 9–74). London: Routledge.
Gilbert, P., & Irons, C. (2005). Therapies for shame and self-attacking,
using cognitive, behavioural, emotional imagery, and compassionate
mind training. In P. Gilbert (Ed.), Compassion: Conceptualisations,
research, and use in psychotherapy (pp. 263–325). London: Routledge.
Greenier, K. D., Kernis, M. H., Whisenhunt, C. R., Waschull, S. B., Berry,
A. J., Herlocker, C. E., & Abend, T. (1999). Individual differences in
reactivity to daily events: Examining the roles of stability and level of
self-esteem. Journal of Personality, 67, 185–208.
Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of
personal history. American Psychologist, 35, 603–618.
Heatherton, T. F., & Vohs, K. D. (2000). Interpersonal evaluations follow-
ing threats to self: Role of self-esteem. Journal of Personality and Social
Psychology, 78, 725–736.
Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem.
Psychological Inquiry, 14, 1–26.
Kernis, M. H., Cornell, D. P., Sun, C., Berry, A., & Harlow, T. (1993).
There’s more to self-esteem than whether it’s high or low: The impor-
tance of stability of self-esteem. Journal of Personality and Social
Psychology, 65, 1190–1204.
Leary, M. R. (2004). The curse of the self: Self-awareness, egotism, and the
quality of human life. New York: Oxford University Press.
Leary, M. R. (2007). Motivational and emotional aspects of the self.
Annual Review of Psychology, 58, 317–344.
Leary, M. R., Haupt, A. L., Strausser, K. S., & Chokel, J. T. (1998).
Calibrating the sociometer: The relationship between interpersonal ap-
praisals and state self-esteem. Journal of Personality and Social Psy-
chology, 74, 1290–1299.
Leary, M. R., Koch, E., & Hechenbleikner, N. (2001). Emotional responses
to interpersonal rejection. In M. R. Leary (Ed.), Interpersonal rejection
(pp. 145–166). New York: Oxford University Press.
Leary, M. R., & MacDonald, G. (2003). Individual differences in self-
esteem: A review and theoretical integration. In M. R. Leary & J. P.
Tangney (Eds.), Handbook of self and identity (pp. 401– 418). New
York: Guilford Press.
Leary, M. R., Schreindorfer, L. S., & Haupt, A. L. (1995). The role of
self-esteem in emotional and behavioral problems: Why is low self-
esteem dysfunctional? Journal of Social and Clinical Psychology, 14,
Marlatt, G. A. (2002). Buddhist philosophy and the treatment of addictive
behavior. Cognitive and Behavioral Practice, 9, 44–50.
Mruk, C. (1995). Self-esteem: Research, theory, and practice. New York:
Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The benefits of
positive illusions: Idealization and the construction of satisfaction in
close relationships. Journal of Personality and Social Psychology, 70,
79 –98.
Neff, K. (2003a). The development and validation of a scale to measure
self-compassion. Self and Identity, 2, 223–250.
Neff, K. (2003b). Self-compassion: An alternative conceptualization of a
healthy attitude toward oneself. Self and Identity, 2, 85–102.
Neff, K., Hsieh, Y., & Dejitterat, K. (2005). Self-compassion, achievement
goals, and coping with academic failure. Self and Identity, 4, 263–287.
Neff, K. D., Kirkpatrick, K., & Rude, S. S. (in press). Self-compassion and
its link to adaptive psychological functioning. Journal of Research in
Neff, K., Kirkpatrick, K. L., Rude, S. S., & Dejitterat, K. (2004). The link
between self-compassion and psychological well-being. Unpublished
manuscript, University of Texas at Austin.
Pedhazur, E. J. (1997). Multiple regression in behavioral research. Fort
Worth, TX: Holt, Rinehart & Winston.
Pennebaker, J. W., Colder, M., & Sharp, L. K. (1990). Accelerating the
coping process. Journal of Personality and Social Psychology, 58,
528 –537.
Raskin, R., & Hall, C. S. (1979). A narcissistic personality inventory.
Psychological Reports, 45, 590.
Raskin, R., & Terry, H. (1988). A principal-components analysis of the
Narcissistic Personality Inventory and further evidence of its construct
validity. Journal of Personality and Social Psychology, 54, 890 –902.
Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self:
Short-term benefits and long-term costs. Journal of Personality and
Social Psychology, 80, 340–352.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
NJ: Princeton University Press.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the
facilitation of intrinsic motivation, social development, and well-being.
American Psychologist, 55, 68–78.
Ryff, C. D., & Singer, B. (2002). From social structure to biology:
Integrative science in pursuit of human health and well-being. In C. R.
Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp.
541–555). New York: Oxford University Press.
Salzberg, S. (1997). Lovingkindness: The revolutionary art of happiness.
Boston: Shambala.
Schneider, D. J., & Turkat, D. (1975). Self-presentation following success
or failure: Defensive self-esteem models. Journal of Personality, 43,
Shapiro, S. L., Astin, J. A., Bishop, S. R., & Cordova, M. (2005).
Mindfulness-based stress reduction for health care professionals: Results
from a randomized trial. International Journal of Stress Management,
12,164 –176.
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the
integrity of the self. In L. Berkowitz (Ed.), Advances in experimental
social psychology (Vol. 21, pp. 261–302). New York: Academic Press.
Swann, W. B., Jr. (1997). The trouble with change: Self-verification and
allegiance to the self. Psychological Science, 8, 177–180.
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social
psychological perspective on mental health. Psychological Bulletin, 103,
Williams, K. D., & Zadro, L. (2005). Ostracism: The indiscriminate early
detection system. In K. D. Williams, J. P. Forgas, & W. von Hippel
(Eds.), The social outcast: Ostracism, social exclusion, rejection, and
bullying (pp. 19–34). New York: Psychology Press.
Wiser, S., & Telch, C. F. (1999). Dialectical behavior therapy for binge
eating disorder. Journal of Clinical Psychology, 55, 755–768.
Received August 16, 2006
Revision received September 5, 2006
Accepted September 25, 2006
... An alternative way to relate to the self is via selfcompassion. Like self-esteem, self-compassion involves one's self-perception, but, in comparison to self-esteem, self-compassion leads individuals to be less defensive and concerned about self-image threats (Breines & Chen, 2012;Leary et al., 2007). Interestingly, empirical investigations on the role of individual-level self-compassion in the expression of negative outgroup attitudes are limited (Fuochi et al., 2018;Verhaeghen & Aikman, 2020). ...
... Self-compassion is an adaptive self-concept that is associated with reduced defensiveness to self-images threats (Breines & Chen, 2012;Leary et al., 2007), so it stands to reason that high self-compassion individuals would exhibit lower levels of negative outgroup attitudes. Our results support this hypothesis when self-compassion is rooted in common humanity, which conceptually replicates Fuochi et al. (2018). ...
... This also explains why high self-esteem, but not high selfcompassion, is part of a narcissistic personality profile (Locke, 2009;Neff, 2003b). Second, negative self-relevant information and experiences threaten self-esteem, but they have no such role in self-compassion (Breines & Chen, 2012;Leary et al., 2007). Individuals with high self-compassion rooted in common humanity view negative personal characteristics and experiences as normal human experiences and, thus, are not likely to engage in defensive actions such as outgroup derogation. ...
Full-text available
Self-compassion, rooted in common humanity, self-kindness, and mindfulness, is an adaptive self-concept that assuages defensiveness to self-image threats. We hypothesized that self-compassion would buffer the need to express negative intergroup attitudes and that this relation would be explained by compassion for others. In a preregistered study, participants (N = 163) with stronger self-compassion rooted in common humanity, but not self-kindness or mindfulness, expressed less negative attitudes toward outgroups than those with lower self-compassion rooted in common humanity. Moreover, this relation persisted even after controlling for self-esteem, a construct related to but distinct from self-compassion. Finally, compassion for others mediated the relation between self-compassion and intergroup attitudes. These findings support the positive and unique role of individual-level self-compassion in intergroup relations.
... Trait self-compassion has been linked with lower levels of self-reported stress, neuroticism, and depression, and higher levels of life satisfaction, social connectedness, and subjective well-being (Leary et al., 2007;Neely et al., 2009;Neff, 2003b;Neff et al., 2007). Meta-analyses have found strong associations among higher levels of self-compassion, lower levels of anxiety, and higher levels of overall psychological well-being (MacBeth & Gumley, 2012;Marsh et al., 2018;Zessin et al., 2015). ...
... Following the audio recording, participants were given 5 minutes to respond to one of two writing prompts (both adapted from Breines & Chen, 2012;Leary et al., 2007). Participants in the self-compassion training condition wrote a response to the prompt: "Imagine a weakness that you have. ...
Full-text available
The present study investigated links between self-compassion and responses to social stress. Participants (N = 102) were randomly assigned to a self-compassion training or a comparison condition and engaged in the Trier Social Stress Test for Groups (TSST-G). Measures of trait self-compassion, subjective perceptions of stress, and salivary cortisol were collected. Participants with higher trait self-compassion had significantly lower subjective and cortisol responses to stress during the TSST-G than did participants with lower trait self-compassion. Participants in the self-compassion training condition did not have significantly lower responses to stress. Results suggest that trait self-compassion is linked with subjective and physiological responses to a social-evaluative stressor. Implications for trait self-compassion and self-compassion training on subjective and physiological responses to stress are discussed.
... According to the emotional regulation model of selfcompassion (Finlay-Jones et al., 2015), self-compassion can decrease depressive symptoms by helping adolescents cope with adverse situations. Evidence exists that individuals with high self-compassion reported less negative emotion when their autonomy or competence was frustrated (Leary et al., 2007). In addition, Wong and Mark (2013) found that self-compassionate individuals were likely to take a balanced and broad perspective of the circumstances to cope with failures and unattainable goals, instead of overidentifying with the failures and fixating on unachievable goals. ...
Full-text available
According to self-determination theory (SDT), when needs are chronically obstructed, individuals may use compensatory behaviors to temporarily relieve negative emotions aroused by need frustration. This study tested whether basic psychological need frustration is associated with non-suicidal self-injury (NSSI) as a compensatory behavior. In a moderated mediation model, we also tested whether depression acted as a mediator in this association, and whether self-compassion moderated the indirect effect of depression. 1691 Chinese adolescents (ages 11–18; 55.3% girls) filled out questionnaires regarding basic psychological need frustration, depression, self-compassion, and NSSI. The results revealed that depression mediated the relationship between basic psychological need frustration and adolescent NSSI. Self-compassion moderated the indirect effect. Specifically, the indirect path was only significant for adolescents with low self-compassion. These findings explain the underlying mechanism of NSSI from a new perspective of SDT, which is of significance for the prevention and treatment of NSSI.
... Neff (2003) also revealed that self-compassion and self-esteem are equivalent predictors of happiness, optimism and positively impact mental health. For instance, students with high levels of compassion and self-esteem experience fewer negative emotions when receiving bad feedback (Leary et al., 2007). This means that self-compassion could be positively correlated with self-esteem as both contribute to reducing negative feelings (Donald et al., 2018; Reilly et al., 2014). ...
Full-text available
Empathy is a critical element of subjective well-being and is an important personality trait for undergraduate students. To improve undergraduate students' empathy, the current study examines the relationship between self-compassion and empathy and the mediating role of self-esteem underlying this relationship. Participants were six hundred and twenty-two (320 males and 302 females) students from ve Vietnamese universities, aged 18-21 years (M age = 19.5; SD age = 0.95 years), who completed self-compassion scale (SCS), empathy scale in adults (BES-A), and self-esteem scale of Toulouse (ETES).The results indicated that (1) self-compassion was positively associated with empathy; (2) and self-esteem mediated the relationship between the two variables. Therefore, enhancing undergraduate students' self-compassion may be an e cacious way to improve undergraduate students' empathy. However, additional study is needed to elucidate the roles of self-compassion in the educational context.
... It is characterized by increased self-kindness and less self-judgment, increased common humanity, and less isolation, as well as greater mindfulness and less overindulgence in negative thoughts and emotions (Neff, 2003a). Individuals with high levels of self-compassion can alleviate negative emotion caused by painful events (Leary et al., 2007). While buffering negative emotions, self-compassion is not content with the status quo but encourages individuals to constantly improve themselves in difficult situations. ...
Full-text available
Focusing on the domain of self-compassion, this study explored the promotion mechanism of online learning behavioral engagement (OLBE) of international students in China under COVID-19. Positive emotion and self-improvement motivation were selected as mediators. Participants were 606 international students from 8 countries who were studying online in their own countries due to the international travel restriction of COVID-19. Results showed positive emotion and self-improvement motivation completely mediated self-compassionate mindfulness (SCM) and OLBE of international students. Positive emotion and self-improvement partially mediated SCM and OLBE of international students respectively. Students with higher SCM engage with online learning more in that they possess more positive emotion and self-improvement motivation. This study suggested that SCM may facilitate OLBE via positive emotion and effective self-improvement motivation.
... This emphasizes the importance of having a healthy condition for the health conditions for the most appropriate development of the self and requires intervention and prevention in childhood to assist correct the negative outcomes caused by toxic situations such as child maltreatment and/or exposure to violence. High self-esteem, self-compassion, self-efficacy, and perceived internal point of control have all been established to promote resilience and be protective against the negative consequences of child maltreatment and other life stressors (Afifi & MacMillan, 2011;Bolger & Patterson, 2001;Cicchetti & Rogosch, 1997;Kort-Butler, 2010;Lamoureux et al., 2012;Leary et al., 2007;Vettese et al., 2011). Bolger and Patterson (2001) attest that when individuals perceived themselves as having an internal point of control, this was related with reducing internal problems (i.e., anxiety, depression, etc.). ...
This study assessed the competencies of the classroom teachers as counselor as the basis for the development of Humanities and Social Science students’ self-esteem. Through descriptive correlational research design, it investigated the level of competencies of classroom teacher-counselor into behavioral competencies and work-related competencies, and the level of student’s self-esteem into identity, appreciation, acceptance, self-confidence, pride, humility and selfishness. It involved 147 Grade 11 and Grade 12 Humanities and Social Sciences students of one private school in the Philippines during the academic year 2021-2022. The findings revealed that there is a significant relationship between the competencies of classroom teacher-counselors and the self-esteem of senior high school students. Therefore, the null hypothesis posited in the study is not sustained. In this light, the study posits that the competence of the teachers as counselors are significant in the development of students’ self-esteem. Therefore, students may directly share their concerns with their teacher-counselor dealing with personal issues affecting their general well-being and academic progress. Teachers need to continue to support the learners through counseling and guidance leading to their academic success. Similarly, strengthening the competencies of teachers–counselors reassure the students’ mental health.
Full-text available
: Compassion and mindfulness as a basis for psychological evaluation, in order to evaluate and activate the desired variable. Considering the fact that in the research background we found out the effect of the educational package on the desired component of the component as effective instruction, by the decentralization of the components Negative determination. For this purpose, 297 of all women under the sponsorship of a charity in Rudbar city based on the Morgan table in the year (1396) who had a depression diagnosis based on a score of at least 12 in the Beck Depression Test, were selected by random sampling of target audiences They were . In a intervention group, 15 (one) and a control group of 15 (single) were replaced. The present study was a semi experimental design with pretest-posttest design with control group. The tool was used by Self-employed Scheer Questionnaire (Scherer, Maddox, Mercury, Prentice, 1982). Multivariate analysis of variance (MANOVA) was used to analyze the data. The results show the significant effect of this educational package on self-efficacy in depressed female head of household. Research hypotheses were confirmed at the significant level (p <0.0001).
Full-text available
Alcohol use disorder is a condition that develops as a result of problematic alcohol use. The study examined the relationship between social anxiety and self-compassion among persons with alcohol use disorder in the three main psychiatric hospitals and an alcohol rehabilitation centre in Ghana. The study employed a correlational design and used purposive sampling in obtaining its participants. Sixty participants were involved in the study and Frequencies and Pearson's product-moment correlation coefficient were used to analyse the data. The findings revealed that social anxiety was highly prevalent among participants and moderate level of self-compassion was observed. A highly significant inverse relationship between social anxiety and self-compassion was also found. The sample size and the use of an adapted version of the self-compassion scale are considered limitations for the study. Also, persons with alcohol use disorders in the study were addicted to at least one other psychoactive substance and this was also considered a limitation of the study.