Implicit Theories of Emotion: Affective and Social Outcomes Across a Major Life Transition

Abstract

The authors demonstrate that people differ systematically in their implicit theories of emotion: Some view emotions as fixed (entity theorists), whereas others view emotions as more malleable (incremental theorists). Using a longitudinal and multimethod design, the authors show that implicit theories of emotion, as distinct from intelligence, are linked to both emotional and social adjustment during the transition to college. Before entering college, individuals who held entity (vs. incremental) theories of emotion had lower emotion regulation self-efficacy and made less use of cognitive reappraisal (Part 1). Throughout their first academic term, entity theorists of emotion had less favorable emotion experiences and received decreasing social support from their new friends, as evidenced by weekly diaries (Part 2). By the end of freshman year, entity theorists of emotion had lower well-being, greater depressive symptoms, and lower social adjustment as indicated in both self- and peer-reports (Part 3). The emotional, but not the social, outcomes were partially mediated by individual differences in emotion regulation self-efficacy (Part 4). Together, these studies demonstrate that implicit theories of emotion can have important long-term implications for socioemotional functioning.

Full text

Implicit Theories of Emotion: Affective and Social Outcomes Across a
Major Life Transition
Maya Tamir
Boston College
Oliver P. John
University of California, Berkeley
Sanjay Srivastava
University of Oregon
James J. Gross
Stanford University
The authors demonstrate that people differ systematically in their implicit theories of emotion: Some
view emotions as fixed (entity theorists), whereas others view emotions as more malleable (incremental
theorists). Using a longitudinal and multimethod design, the authors show that implicit theories of
emotion, as distinct from intelligence, are linked to both emotional and social adjustment during the
transition to college. Before entering college, individuals who held entity (vs. incremental) theories of
emotion had lower emotion regulation self-efficacy and made less use of cognitive reappraisal (Part 1).
Throughout their first academic term, entity theorists of emotion had less favorable emotion experiences
and received decreasing social support from their new friends, as evidenced by weekly diaries (Part 2).
By the end of freshman year, entity theorists of emotion had lower well-being, greater depressive
symptoms, and lower social adjustment as indicated in both self- and peer-reports (Part 3). The
emotional, but not the social, outcomes were partially mediated by individual differences in emotion
regulation self-efficacy (Part 4). Together, these studies demonstrate that implicit theories of emotion can
have important long-term implications for socioemotional functioning.
Keywords: implicit beliefs, emotion regulation, well-being, social adjustment
The idea that individuals can control their emotions has had
currency since the dawn of recorded history. In the Old Testament,
one of the earliest messages God sent to Cain was that desire can
lead to sin and that people must govern their desire (Gen. 4:7).
Implicit in this injunction is the idea that people have the ability to
control their emotions. However, there have always been those
who have taken the opposite viewpoint, believing that emotions
are uncontrollable. For example, Shakespeare’s (1603/1992) Ham-
let described human behavior as shaped by higher powers, render-
ing attempts at controlling emotions futile. Similarly, Bedier
(1900) described Tristan and Isolde’s love as a carriage pulled by
wild horses that cannot be tamed, ultimately driving them to their
doom. These texts suggest that individuals differ in their beliefs
about the controllability of emotions. As in the case of Tristan and
Isolde, such texts also suggest that these beliefs can have important
long-term implications.
In the present study, we examine individual differences in
beliefs about the controllability of emotions and their implications
for emotional and social functioning. We first review research on
implicit theories (i.e., beliefs about the malleability of human
attributes). We then discuss how such theories might apply to
emotion, suggesting that people differ in the extent to which they
believe emotions can be controlled. We present a longitudinal
study to test whether the belief that emotions are malleable is
beneficial for emotional and social functioning, particularly during
a period of transition.
IMPLICIT THEORIES
Beliefs imply certain expectancies, which in turn, guide behav-
ior (Olson, Roese, & Zanna, 1996). In particular, beliefs about
controllability guide the way people construe their reality and
influence their motivation to engage in self-regulation (Bandura,
1986; Seligman, 1975). On the basis of these assumptions, Dweck
and her colleagues (for reviews, see Dweck, 1986, 1996, 1999;
Dweck, Chiu, & Hong, 1995a; Dweck & Leggett, 1988) have
studied beliefs about the malleability of personal attributes. Be-
cause these beliefs tend to be implied rather than explicitly held,
they have been referred to as implicit theories.
Dweck and her colleagues have distinguished between entity
theories and incremental theories. Individuals who hold incremen-
tal beliefs view attributes as malleable and controllable. Such
beliefs lead them to make flexible, contextual interpretations of
events. In the face of challenge, these beliefs promote assertive
attempts at self-regulation, increasing the chances of successful
behavior. On the other hand, individuals who hold entity beliefs
view attributes as fixed and impossible to control. Such beliefs
Maya Tamir, Department of Psychology, Boston College; Oliver P.
John, Department of Psychology, University of California, Berkeley; San-
jay Srivastava, Department of Psychology, University of Oregon; James J.
Gross, Department of Psychology, Stanford University.
This research was supported by National Institute of Mental Health
Grant MH58147.
Correspondence concerning this article should be addressed to Maya
Tamir, Department of Psychology, Boston College, McGuinn Hall, 140
Commonwealth Avenue, Chestnut Hill, MA 02467. E-mail:
tamirm@bc.edu
Journal of Personality and Social Psychology Copyright 2007 by the American Psychological Association
2007, Vol. 92, No. 4, 731–744 0022-3514/07/$12.00 DOI: 10.1037/0022-3514.92.4.731
731

often lead them to make stable, internal attributions. In the face of
challenge, these beliefs decrease the motivation to self-regulate,
leading to self-regulation failure (for a review, see Dweck, 1996).
Most research on implicit theories has focused on intelligence.
Implicit theories of intelligence have been associated with goals in
academic settings, responses to academic failure, and performance
outcomes (Dweck, 1986, 1996, 1999). Implicit theories, however,
are domain-specific. Theories regarding one personal attribute
(e.g., intelligence) might be largely independent of theories regard-
ing another personal attribute (e.g., moral character). Such domain
specificity has been demonstrated in research distinguishing im-
plicit theories of intelligence from implicit theories of morality
(Chiu, Dweck, Tong, & Fu, 1997; Dweck et al., 1995a) and
personality (Beer, 2002; Chiu, Hong, & Dweck, 1997; Hong, Chiu,
Dweck, & Sacks, 1997).
Implicit theories in a specific domain lead to outcomes within
that domain (Dweck, 1996). For example, incremental beliefs
about intelligence have been linked to academic performance
(Dweck, 1996; Dweck & Leggett, 1988), whereas incremental
beliefs about morality have been linked to judgments of moral
character (Chiu, Dweck, Tong, & Fu, 1997). In addition to such
intrapersonal implications, implicit theories can have interpersonal
implications. For example, Chiu and colleagues (Chiu, Hong, &
Dweck, 1997) have suggested that entity theorists judge others in
terms of stable dispositions, whereas incremental theorists are
more sensitive to situational determinants of behavior. In particu-
lar, they found that individuals who held entity theories of morality
viewed an immoral act as an indication of immoral character and,
therefore, tended to be more critical and more likely to punish
others (Chiu, Dweck, et al., 1997; Erdley & Dweck, 1993). Thus,
implicit theories might influence both personal and interpersonal
outcomes.
IMPLICIT THEORIES OF EMOTION
To date, research on implicit theories has focused primarily on
trait-like attributes, such as intelligence and personality. However,
people also might have implicit theories about more transient
attributes, such as emotion (for a related discussion on lay theories
of emotion, see Ben-Artzi & Mikulincer, 1996). The examination
of implicit theories of emotion is particularly important, given the
pervasive impact of emotion on human psychological and social
functioning.
Given that emotion and intelligence involve distinct domains,
we expected implicit theories of emotion to be distinct from those
of intelligence. We further expected emotions to be seen as some-
what more malleable than intelligence. This expectation flowed
from the fact that emotion and intelligence differ in their temporal
characteristics: Whereas intelligence is a more trait-like attribute,
emotions are transient states (e.g., Chaplin, John, & Goldberg,
1988).
Because implicit beliefs have domain-specific implications
(Dweck et al., 1995a), we expected implicit theories of emotion to
be linked to emotion-relevant outcomes. Existing research has
emphasized the link between implicit theories and domain-specific
outcomes related to self-regulation, intrapersonal, and interper-
sonal experiences (for reviews, see Dweck, 1999; Dweck et al.,
1995a). Building on this research, we sought to test whether
implicit theories of emotion are linked to emotion regulation,
emotional and social experience during a transition.
We chose to focus on a period of transition because we thought
individual differences in implicit theories should be particularly
influential during challenging transition periods (Henderson &
Dweck, 1990). This is because during transitions, preexisting
knowledge and beliefs guide the way individuals perceive and
respond to their environment. The transition to college is a life
transition, during which many students confront new personal
challenges and learn to cope with multiple demands (Cantor,
Norem, Niedenthal, Langston, & Brower, 1987). This transition
also disrupts existing networks of social support, separating indi-
viduals from high school friends and family and forcing them to
form new relationships (Shaver, Furman, & Buhrmester, 1985). If
implicit beliefs about emotion are linked to emotional and social
experiences, such relations should be evident during this important
transition period. Therefore, the present investigation examines
implicit theories of emotion during the transition to college.
IMPLICIT THEORIES OF EMOTION AND EMOTION
REGULATION
Implicit theories of intelligence are related to the regulation of
intellectual functioning. Individuals who view intelligence as fixed
are less likely to try to modify their intellectual functioning
(Dweck, 1986; Hong, Chiu, Dweck, Lin, & Wan, 1999). Following
the same logic, we propose that implicit theories of emotion might
be related to the regulation of emotion.
First, implicit theories of emotion should be related to emotion
regulation efficacy. Individuals who believe emotions are fixed
should be less likely to believe that they can actually modify their
own emotions. On the other hand, individuals who believe emo-
tions are malleable should be more likely to believe that they
possess the ability to control their emotions.
Second, individuals with entity theories of emotion should be
less likely to use antecedent (i.e., anticipatory) strategies of emo-
tion regulation. Cognitive reappraisal, for example, is an
antecedent-focused strategy that involves construing an emotion-
eliciting event in a way that changes its emotional impact before it
occurs (Gross, 1998; Gross & John, 2003). Individuals who view
emotion as fixed have little incentive to try to modify their emo-
tions by using antecedent strategies such as cognitive reappraisal.
On the other hand, individuals who view emotion as malleable
should try to actively modify their emotions by changing their
appraisal of emotion-eliciting events.
Third, associations between implicit theories of emotion and
emotion regulation strategies need not have a one-size-fits-all
quality. In its everyday usage, the term emotion often calls to mind
emotion experience rather than emotion expression. In fact, indi-
viduals frequently separate the experience of emotion from its
expression (e.g., Gross, John, & Richards, 2000). This, in part, is
because individuals are able to suppress or mask the expression of
their feelings in certain situations without modifying their emotion
experience (Gross, 1998; Gross et al., 2000). This suggests that
implicit theories of emotion might be focused on emotion experi-
ence and thus have little or no bearing on emotion expression. If
so, we would not expect implicit theories of emotion to have
similar associations with reappraisal and expressive suppression.
732
TAMIR, JOHN, SRIVASTAVA, AND GROSS

In summary, we expected implicit theories of emotion to be
related to emotion regulation efficacy and the use of reappraisal.
However, we did not expect such associations with the use of
suppression. This is because we did not expect participants’ beliefs
about the malleability of emotion experience to influence their
attempts to modify the overt expressions of their emotion in
accordance with social or personal norms (e.g., Hochschild, 1979).
We tested these predictions in Part 1 of our investigation. In Part
1, we also measured individual differences in emotional intensity.
Because strong emotions are harder to control, individuals with
stronger emotional experiences might be less successful at emotion
regulation and view emotions as less amenable to control. There-
fore, in Part 1, we also tested whether the relations between
implicit theories of emotion and emotion regulation are evident
even when individual differences in emotional intensity are con-
trolled.
IMPLICIT THEORIES OF EMOTION AND
EMOTIONAL OUTCOMES
Implicit theories of intelligence influence academic perfor-
mance: Individuals who hold entity theories of intelligence tend to
perform worse academically (e.g., Dweck, 1999). Similarly, we
expected implicit theories of emotion to be related to emotional
experience, with entity (vs. incremental) theories associated with
less favorable emotion outcomes. If implicit theories of emotion
influence emotion experience, individuals who hold entity theories
should experience fewer positive and more negative emotions.
Such patterns of emotion experience might be established prior to
the transition into college and might be maintained as students go
through the transition. We tested these hypotheses in Part 2 of the
investigation, in which students provided weekly reports about
their emotional experiences throughout the first academic term.
IMPLICIT THEORIES OF EMOTION AND SOCIAL
OUTCOMES
During the first term in college, individuals typically need to
develop an entirely new social network. Being away from family
and childhood friends, students begin to develop new interpersonal
relationships (Shaver et al., 1985). During this period, beliefs
about emotion might influence the way individuals approach,
relate, and respond to others. Specifically, entity theories of emo-
tion might lead individuals to be less tolerant of the emotions of
others (see Chiu, Hong, & Dweck, 1997), gradually leading to
fewer and fewer supportive social relationships.
To the extent that implicit theories of emotion influence social
outcomes, the effects might not necessarily be evident immedi-
ately. Rather, as with relationships themselves, these patterns can
only develop over time, as people have a chance to know one
another and emerging friendships are put to the test. Therefore, one
might expect implicit theories of emotion to have some influence
on the development of novel relationships (i.e., involving college
friends) but not on well-established relationships (e.g., involving
parents). Individuals who hold entity (vs. incremental) theories of
emotion might have less control over their own emotions, or they
might cultivate more rigid attitudes toward the emotions of others.
We tested this possibility in Part 2 of the investigation, in which
students reported on the support they received from college friends
as well as from parents over the course of their first academic term
in college.
IMPLICIT THEORIES OF EMOTION AND LONG-
TERM OUTCOMES
One important question is whether implicit theories of emotion
are associated with longer term outcomes. One possibility is that
they are not. During the transition into college, implicit theories of
emotion might determine the way individuals react initially to
personal and social challenges. However, as individuals learn to
navigate their new environment, the effects of implicit theories of
emotion might wane or vanish altogether. A second possibility,
however, is that by helping shape an individual’s personal and
social environment, implicit theories of emotion might set in
motion an adjustment pattern that is maintained over time.
To assess whether the effects of implicit theories are temporary
or whether they persist, researchers must examine the effects of
implicit theories over relatively long periods of time (Dresel, 2001;
Fowler & Peterson, 1981). Because evidence for the long-term
implications of implicit theories is limited (for exceptions, see
Henderson & Dweck, 1990; Robins & Pals, 2002), we believe it is
essential to test whether implicit theories of emotion carry impor-
tant implications for the transition to college and whether these
implications persist over time.
In light of our earlier hypotheses, we expected incremental
theories of emotion to predict more favorable emotion experiences
and better psychological adjustment over time, as indicated by
greater well-being and fewer depressive symptoms. Furthermore,
we expected incremental theories of emotion to predict better
social outcomes even 1 year following the assessment of such
theories. Indeed, social adjustment and “fitting in” are important
indicators of successful adaptation to the college environment. In
fact, Astin (1993) has shown that social experiences are the stron-
gest determinants of students’ satisfaction with the overall college
experience. To test these predictions, we included self- and peer
reports of emotional and social functioning at the end of freshman
year in Part 3 of this investigation.
THE MEDIATING ROLE OF EMOTION
REGULATION SELF-EFFICACY
To the extent that implicit theories of emotion are associated
with important personal and social outcomes, it would be valuable
to begin to learn what processes mediate these effects. According
to the health belief model (Rosenstock, 1974; Rosenstock, Stre-
cher, & Becker, 1988), the general belief that health is malleable
leads to beliefs about one’s personal ability to change one’s own
health which, in turn, promote successful self-regulation. Indeed,
there is some evidence that perceived self-efficacy mediates the
impact of implicit beliefs on task performance (Tabernero &
Wood, 1999).
Using a similar logic, but now in the domain of emotion, we
expected that the effects of implicit theories of emotion would be
mediated by beliefs about emotion regulation self-efficacy. Al-
though emotion regulation can involve changing emotions in a
variety of ways (Gross, 1998), the vast majority of emotion reg-
ulation episodes involve attempts to decrease negative emotions
and to increase positive emotions (Gross, Richards, & John, 2006).
733
IMPLICIT THEORIES OF EMOTION

Indeed, emotion regulation self-efficacy has been associated with
favorable emotion experiences and lower depression (e.g., Catan-
zaro & Mearns, 1990). Therefore, considering emotional out-
comes, emotion regulation self-efficacy could be a promising
mediator.
What about social outcomes? On the one hand, emotion regu-
lation has some important social consequences (e.g., Gross &
John, 2003; Lopes, Salovey, Cote, & Beers, 2005). Therefore, one
possibility is that implicit theories of emotion lead to social out-
comes via their influence on emotion regulation self-efficacy. On
the other hand, social outcomes might depend less on beliefs the
individual has about him- or herself and more on interpersonal
beliefs and expectations about others that also arise from implicit
emotion theories. One alternative hypothesis, therefore, is that
implicit theories of emotion lead to social outcomes by mecha-
nisms that are distinct from emotion regulation self-efficacy (see
Chiu, Hong, & Dweck, 1997). In Part 4 of this investigation, we
began to address these possibilities.
THE PRESENT STUDY
To examine the links between implicit theories of emotion and
emotional and social outcomes during a time of transition, we drew
data from a longitudinal study of the transition to college life in a
sample of students at Stanford University. For conceptual clarity,
we have organized these data into four parts following a linear
timeline. As shown in Figure 1, we measured implicit theories
among students just before they entered college (Part 1). We then
monitored emotional and social experiences repeatedly as students
adjusted throughout their first academic term (Part 2). To test
whether implicit emotion theories had any persistent long-term
effects, we examined emotional and social adjustment at the end of
freshman year (Part 3). After testing the implications of implicit
theories of emotion in Parts 1–3, we began to explore one potential
mediating mechanism in Part 4, namely self-efficacy in emotion
regulation.
Method
Part 1: Relations to Emotion Regulation Before the
Transition to College
Participants
Data were available for 437 students (56% female, 44% male;
age: M ⫽ 18.22, SD ⫽ 0.66). The sample was composed of 59%
Caucasians, 29% Asians, 12% Latinos, 5% African Americans,
and 4% Native Americans, including those who identified them-
selves as multiracial. Various subsets of this sample participated in
subsequent parts of the investigation, as described in the following
paragraphs.
Implicit Theories
Implicit theories of intelligence. Beliefs about the fixed (vs.
malleable) nature of intelligence were assessed with the Implicit
Theories of Intelligence Scale (Dweck, 1999), which consists of
four statements about the fixed nature of intelligence (e.g., “To be
honest, people can’t really change how intelligent they are,” “Peo-
ple can learn new things, but they can’t really change their basic
intelligence”). Respondents rated the degree to which they agreed
with each statement on a 5-point scale (1 ⫽ strongly disagree,5⫽
strongly agree). Following Dweck (1999), ratings of all four items
were reversed and averaged to form an incremental theory scale
(␣⫽.95), so that, consistent with previous scoring procedures,
higher scores indicate incremental theories and lower scores indi-
cate entity theories of intelligence.
Implicit theories of emotion. To measure beliefs about the
malleable nature of emotion, we modified items from the Implicit
Part 4:
Mediation by
Emotion Regulation Self-Efficacy
Part 1:
Implicit Theories
and Relations to
Emotion Regulation
Part 2:
Weekly Dairies --
Emotional and Social
Outcomes
Part 3:
Long-Term Self
and Peer-Rated
Outcomes
Summer Fall Winter Spring Summer
Freshman year
Figure 1. Implicit theories of emotion and the transition to college: Timeline of Parts 1–4 of the investigation.
734
TAMIR, JOHN, SRIVASTAVA, AND GROSS

Theories of Intelligence Scale (Dweck, 1999) to refer to beliefs
about the fixed versus malleable nature of emotion. The scale
includes two incremental items, “Everyone can learn to control
their emotions,” “If they want to, people can change the emotions
that they have,” and two entity items, “No matter how hard they
try, people can’t really change the emotions that they have,” “The
truth is, people have very little control over their emotions.” To
compute an overall score, we reverse scored the two entity items
and then averaged across all items (␣⫽.75) so that, consistent
with beliefs about intelligence, higher scores indicate incremental
theories and lower scores indicate entity theories of emotion.
1
Emotion Regulation Self-Efficacy, Strategies, and
Emotional Intensity
To measure emotion regulation self-efficacy, participants read
10 scenarios regarding emotion-eliciting events drawn from daily
collegiate life (e.g., taking an important exam, running into a
former boyfriend or girlfriend) and rated how confident they were
in their ability to control their emotions in each scenario (␣⫽.82)
on a 0 –100 scale (0 ⫽ cannot do at all, 100 ⫽ certainly can do).
Individual differences in habitual use of the emotion regulation
strategies of cognitive reappraisal and expressive suppression were
assessed with the Emotion Regulation Questionnaire (Gross &
John, 2003). The questionnaire includes six items for cognitive
reappraisal (e.g., “I control my emotions by changing the way I
think about the situation I’m in”; ␣⫽.82) and four items for
expressive suppression (e.g., “I control my emotions by not ex-
pressing them”; ␣⫽.70). Finally, individual differences in emo-
tional intensity were measured with Gross and John’s (1998)
6-item scale (e.g., “I have strong emotions”; ␣⫽.74).
Part 2: Weekly Diaries of Emotional and Social
Outcomes
Participants completed up to 10 weekly assessments throughout
the fall quarter (M ⫽ 6), submitting their reports at the end of each
week through a specially designed Web site. Each week during the
fall quarter, participants rated their emotional experiences and the
degree of social support they received from friends and parents in
the past week. Positive emotions were measured with four items:
happy/pleased/contented, proud/accomplished/successful, cared
about/loved/connected, interested/intellectually engaged/stimu-
lated (␣⫽.78). Negative emotions were also measured with four
items: anxious/nervous, angry/irritated/pissed off, sad/depressed/
down, tired/fatigued (␣⫽.76). In both cases, participants rated
how much they felt each emotion in the past week on a 4-point
scale (0 ⫽ not at all,4⫽ extremely).
Social support was defined as the degree of emotional support
students received from their friends at Stanford. Parental support
was defined as the degree of emotional support students received
from their parents. Both ratings were made on a 7-point scale (1 ⫽
no support,7⫽ great support).
Weekly assessments were available for 236 students (57% female,
43% male) who participated in Part 1 of this investigation. As ex-
pected, attrition analyses indicated that students who completed the
weekly assessments did not differ significantly from students in our
core sample on the main variables of interest, including implicit
beliefs about emotion and emotion regulation self-efficacy.
Part 3: Long-Term Outcomes Reported by Self
and Peers
Self-Reported Outcomes: Positive and Negative Emotion,
Adjustment, and Social Functioning
Ratings were provided at the end of the final quarter of freshman
year. Participants rated how often in the past quarter they experi-
enced particular emotions on a 4-point scale (1 ⫽ rarely or none
of the time,4⫽ most or all of the time). Positive emotions were
measured by ratings of happiness and interest (␣⫽.71), and
negative emotions were measured by ratings of anxiety, anger, and
sadness (␣⫽.64). Psychological well-being across different do-
mains was measured by 9 items (␣⫽.69) that tap domains of
psychological well-being suggested by Ryff and Keyes (1995),
such as environmental mastery and personal growth. Depressive
symptoms were measured by the 10-item (␣⫽.83) version of the
Center of Epidemiological Studies Depression Scale (CES–D;
Andresen, Malmgren, Carter, & Patrick, 1994).
Following Astin (1993), social adjustment was defined for par-
ticipants as the degree to which they felt they had adjusted to
college and the degree to which they felt they fit in at Stanford;
participants rated both on a 7-point scale (1 ⫽ not very well,7⫽
very well), and the two ratings were averaged (␣⫽.80). Loneli-
ness was defined as feelings of “loneliness, isolation, or rejection,”
which participants rated on a 5-point scale (0 ⫽ not at all,4⫽
extremely).
These data were available for 364 (59% female, 41% male)
students who participated in Part 1 of this investigation. Attrition
analyses indicated that students who completed the end-of-year
assessment did not differ significantly from students in our core
sample on any of the variables of interest, including implicit
beliefs about emotion and emotion regulation self-efficacy.
Peer-Reported Outcomes: Positive and Negative Emotion,
Adjustment, and Social Functioning
Given time constraints, questions for peers were designed to be
brief but as similar as possible to those used in the self-reports. To
assess emotion in the past quarter, peers rated the target person’s
positive emotions (␣⫽.50) and negative emotions (␣⫽.66) by using
the same emotions used in the self-reports. Psychological well-being
was assessed by seven items (␣⫽.67) that tap the domains suggested
1
A reviewer wondered whether the two items on our measure that ask
about control of emotion and the two items that ask about change of
emotion might assess distinct aspects of implicit emotion beliefs. To find
out, we correlated the mean of the two “control” items with the mean of the
two “change” items and found these less reliable part scores were none-
theless strongly correlated (r ⫽ .61, p ⬍ .001). Even more important, the
pairwise interitem correlations across the two sets of items were as large as
those within each set; that is, the four items did not form two distinct
clusters but assessed a single underlying construct. Finally, to ensure that
the control and change items related to external variables in the same way
(as one would expect if they were tapping the same underlying construct),
we repeated all analyses reported in this article twice, once with the
aggregate of the two “control” items and once with the aggregate of the two
“change” items instead of the total score. The pattern of results obtained for
the total score was fully replicated by both of the two part scores, con-
firming that our measure assesses a single construct.
735
IMPLICIT THEORIES OF EMOTION

by Ryff and Keyes (1995). To measure depressive symptoms, peers
rated the degree to which the target felt depressed in the past quarter
on a 5-point scale (0 ⫽ not at all,4⫽ extremely). Finally, social
adjustment was defined for the peers as the degree to which the target
person seemed “settled in at Stanford,” and loneliness was defined as
the degree to which the target was “lonely and isolated” during the
past quarter; the peers rated both social indices on a 5-point scale (0 ⫽
not at all,4⫽ extremely).
Ratings from 2 or 3 knowledgeable peers were available for 163
of the participants who provided self-reported outcomes in Part 3
of the investigation. In total, students nominated 382 peers (63%
female, 37% male) who knew the target students well. Participants
who provided ratings from 2–3 peers did not differ significantly
from those who did not provide ratings on all the variables of
interest that were assessed in Part 1 of this investigation.
Part 4: Emotion Regulation Self-Efficacy as a Mediator
To test whether emotion regulation self-efficacy mediated the
effects of implicit theories of emotion on emotional and social
functioning, we followed the steps outlined by Baron and Kenny
(1986) and by Krull and MacKinnon (1999, 2001). First, in Part 1,
we tested whether implicit theories of emotion (i.e., the predictor)
were related to emotion regulation self-efficacy (i.e., the hypoth-
esized mediator). Second, in Parts 2 and 3, we tested whether
implicit theories of emotion were linked to emotional and social
outcome variables. Third, in Part 4, we tested whether emotion
regulation self-efficacy was associated with the outcomes de-
scribed in Parts 2 and 3 in regression analyses that also included
implicit theories of emotion.
Full mediation was indicated when implicit theories of emotion
no longer predicted an outcome variable after controlling for
emotion regulation self-efficacy. Partial mediation was indicated
when the association between implicit theories of emotion and an
outcome variable was reduced substantially after controlling for
emotion regulation self-efficacy. The significance of the mediation
effects were evaluated with Sobel (1982) tests, as recommended by
Krull and MacKinnon (1999, 2001).
Results
Part 1: Relations to Emotion Regulation Before the
Transition to College
Implicit Theories of Emotion and Intelligence
The means and standard deviations of individual differences in
implicit theories of emotion as well as their intercorrelations are
presented in Table 1. As shown in the table and consistent with
prior work on implicit theories of intelligence (Dweck et al.,
1995a), participants in our sample, on average, tended to view
intelligence as somewhat malleable (M ⫽ 3.17). As expected,
participants tended to view emotions as more malleable (M ⫽
3.33) than intelligence, as indicated by a paired-samples t test,
t(435) ⫽ 2.71, p ⬍ .01.
The correlation between implicit theories of intelligence and
emotion was, as expected, positive but modest in size, indicating
that the two are related but clearly separable from one another.
Because implicit theories of emotion and intelligence were signif-
icantly correlated, we controlled for implicit theories of intelli-
gence in subsequent analyses to ensure independent effects. As in
prior work, implicit theories of emotion and intelligence were not
consistently related to either gender or ethnicity.
Emotion Regulation and Implicit Theories of Emotion and
Intelligence
Table 1 presents the correlations among implicit theories of
emotion and intelligence and emotion regulation variables. Con-
sistent with our formulation, incremental theories of emotion were
associated with greater emotion regulation self-efficacy. Self-
efficacy was also positively related to incremental theories of
intelligence. However, the association between emotion regulation
self-efficacy and incremental theories of emotion remained signif-
icant and essentially unchanged when the effects of implicit the-
ories of intelligence were controlled ( pr ⫽ .21, p ⬍ .05). On the
other hand, implicit theories of intelligence were no longer signif-
icantly correlated with emotion regulation self-efficacy after par-
tialling out implicit theories of emotion ( pr ⫽ .08, ns).
As expected, incremental theories of emotion, but not intelli-
gence, were positively associated with habitual use of cognitive
reappraisal. Implicit theories of emotion, however, were not re-
lated to expressive suppression. These associations remained un-
changed even after we controlled for implicit theories of intelli-
gence (cognitive reappraisal: pr ⫽ .34, p ⬍ .05; expressive
suppression: pr ⫽ .04, ns).
Finally, individuals with more intense emotional experiences
might be more likely to develop entity theories of emotion. Indeed,
in our sample, incremental theories of emotion were negatively
associated with emotional intensity (r ⫽⫺.21, p ⬍ .05). To
examine whether emotional intensity might drive the link between
implicit theories of emotion and emotion regulation, we partialled
out emotional intensity. As expected, all the effects reported in
Table 1 remained significant and were essentially unchanged in
effect sizes (i.e., emotion regulation self-efficacy: pr ⫽ .20; cog-
nitive reappraisal: pr ⫽ .35; expressive suppression: pr ⫽⫺.02).
Thus, the associations between implicit theories of emotion and
individual differences in emotion regulation could not be ex-
plained simply by underlying differences in the intensity of the
individual’s emotional experiences.
Table 1
Cross-Sectional Correlations Among Incremental Theories of
Emotion and Intelligence and Emotion Regulation Variables at
the Beginning of the Longitudinal Study (Part 1)
Variable 1 2 3 4 5
Incremental theories
1. Emotion —
2. Intelligence .27
*
—
Emotion regulation
3. Self-efficacy .24
*
.14
*
—
4. Reappraisal .35
*
.07 .21
*
—
4. Suppression .04 ⫺.01 .00 ⫺.03 —
*
p ⬍ .05.
736
TAMIR, JOHN, SRIVASTAVA, AND GROSS

The results of Part 1 indicate that students entered college
holding rather different beliefs about the malleable nature of
emotions. Moreover, those who viewed emotions as malleable (vs.
fixed) felt a greater sense of self-efficacy in regulating their
emotions and used cognitive reappraisal more frequently to do so.
In Part 2, we examine the links between implicit theories and
students’ emotional and social experiences during their first aca-
demic term in college.
Part 2: Weekly Diaries of Emotional and Social
Outcomes
Data Analysis
Individual growth modeling is a type of multilevel modeling
(also known as mixed or hierarchical linear modeling) that esti-
mates individual-level trajectories as well as the sample-level
trajectory (Singer & Willett, 2003). The analysis of multilevel data
is accomplished by using one- and two-level models. The Level 1
model estimates variations in an outcome variable (e.g., positive
emotion) as a function of time (e.g., week in the academic quarter).
Such models include an intercept (␲
0i
) that describes an individ
-
ual’s average level on the outcome variable across time and a
time-varying covariate (␲
1i
) that captures the magnitude of time-
dependent changes. “Week in the term” was the variable that
represented time in our models. Following Singer and Willett
(2003), week was centered at the sample mean (5.5) across all
measurement occasions (see also Kreft, DeLeeuw, & Aiken,
1995). The Level 1 model took the following form:
Y
ij
⫽ ␲
0i
⫹ ␲
1i
共WEEK ⫺ 5.5兲 ⫹ ε
ij
. (1)
The Level 2 model estimates individual differences in the Level 1
parameters as a function of between-subjects variables. Given that
implicit theories of emotion and intelligence were correlated in
Part 1, implicit theories of intelligence were included as a control
in all Level 2 models. Scores on implicit theories were centered
(see Singer & Willett, 2003). Level 2 models took the following
form:
␲
0i
⫽ ␥
01EMOTIONSi
⫹ ␥
02INTELLIGENCEi
⫹ ␨
0i
.
␲
1i
⫽ ␥
10
⫹ ␥
11EMOTIONSi
⫹ ␥
12INTELLIGENCEi
⫹ ␨
1i
. (2)
We tested four separate linear mixed models, one for each depen-
dent measure, as described in the following paragraphs.
2
All
models were tested with the linear mixed models procedure in
SPSS 11.5 (SPSS, Inc., 2002).
Positive and Negative Emotion Experiences
We examined a linear mixed model in which positive emotions
were entered as the outcome variable, with centered week, implicit
theories of emotion, and implicit theories of intelligence as pre-
dictors. As expected, implicit theories of emotion were a signifi-
cant predictor of positive emotions (␥⫽.11), t(238) ⫽ 2.60, p ⬍
.05. To illustrate this relationship, we computed estimated values
separately for individuals high (⫹1SD) and low (⫺1SD) in incre-
mental theories of emotion (Aiken & West, 1991), with all other
variables set to their mean level (i.e., zero). As shown in Figure 2,
individuals who held entity (vs. incremental) theories of emotion
experienced fewer positive emotions throughout the term. This
was the case even after controlling for the effects of theories of
intelligence. Implicit theories of emotion did not interact with
week, t(188) ⬍ 1, indicating that the effect on positive emotion
experience was consistent throughout the first academic term.
In a second linear mixed model, negative emotions were entered
as the outcome variable. Implicit theories of emotion were again a
significant predictor of negative emotions (␥⫽⫺.12), t(234) ⫽
2.45, p ⬍ .05. As shown in Figure 2, individuals who held entity
(vs. incremental) theories of emotion experienced more negative
emotions throughout the academic term. This effect was found
even after controlling for implicit theories of intelligence. Again,
implicit theories of emotion did not interact with week, t(193) ⬍
1, indicating that the effect on negative emotions was consistent
throughout the academic term.
Social Support From College Friends and Parental
Support
We ran an additional model, with support from college friends
as the outcome variable. The Week ⫻ Implicit Theories of Emo-
tion interaction was significant (␥⫽.05), t(180) ⫽ 2.66, p ⬍ .05,
indicating that the change trajectory of emotional support over the
course of the academic term varied as a function of individual
differences in implicit theories of emotion. To illustrate this tra-
jectory, we computed separate estimated values for the first and
last week in the term, separately for individuals high (⫹1SD) and
low (⫺1SD) in incremental theories of emotion (Aiken & West,
1991). As shown in Figure 3, as time went by, individuals who
held entity theories of emotion received less and less support from
their new college friends. On the other hand, as individuals with
incremental theories of emotion progressed through the quarter,
they received gradually increasing support from their new college
friends.
In a separate model, we examined support from parents as the
outcome variable. As expected, there were no significant effects
for implicit theories of emotion. To test whether implicit theories
of emotion still affected social support from college friends even
when weekly parental support was controlled, we ran an additional
model in which parental support was also centered as a Level 1
predictor. The Week ⫻ Implicit Theories of Emotion interaction
remained significant, with the same effect size (␥⫽.05), t(181) ⫽
2.66, p ⬍ .05. This finding indicates that implicit theories of
emotion were relevant to social support in new college relation-
ships, but not to all personal relationships, such as those that had
been established prior to the beginning of college.
Summary and Limitations
The results of Part 2 indicate that early in the transition to
college, implicit theories of emotion influence the emerging tra-
jectory of emotional and social experiences. Over the course of
several months, individuals who held entity theories of emotion,
but not intelligence, experienced more negative and fewer positive
emotions. Furthermore, as time went by, they received decreasing
2
The pattern of findings remained the same when the effects of gender
and interactions with gender were included in the Level 2 models.
737
IMPLICIT THEORIES OF EMOTION

levels of support from their new college friends. On the other hand,
individuals who held incremental theories of emotion experienced
fewer negative, and more positive, emotions throughout the quar-
ter. In addition, as the quarter progressed and their new college
friends got to know them better, they received gradually and
continuously increasing support. It is important to note that these
effects were not applicable to all interpersonal relationships (e.g.,
those involving parents); instead, they were unique to new rela-
tionships with friends in college.
Part 2 also has several limitations. First, the temporal scope was
limited in that participants were followed only through the end of
the first quarter of their freshman year. One important goal of Part
3 of the investigation, therefore, was to assess whether these
implicit theories would predict outcomes even at the end of fresh-
man year. Second, the considerable time commitment required of
participants in continuous sampling procedures puts constraints on
sample size (Scollon, Kim-Prieto, & Diener, 2003). Although our
sample was large enough to test our hypotheses, it is important to
demonstrate that the effects emerge again in a relatively larger
sample. In Part 3, therefore, we examine the implications of
implicit theories of emotion with a larger sample of participants at
the end of freshman year.
Finally, in Part 2, individuals reported on their own emotional and
social experiences. Although such weekly reports provide a direct
window into individuals’ personal experience, it is important to
broaden their ecological validity by examining peer-report evidence
as well. Well-acquainted peers who have observed the individual in a
wide range of emotion-eliciting situations and social settings can
provide an important additional perspective on the individual’s emo-
tional experiences and social adjustment. Peer reports reflect behav-
ioral trends the individual has manifested over time and across situ-
ations. Reports by peers are particularly important when assessing the
individual’s social functioning because such functioning involves
relationships with the peers themselves (see also Cairns & Green,
1979). Therefore, in Part 3, we also test whether findings for self-
reported outcomes are replicated in peer ratings.
Part 3: Long-Term Outcomes Reported by
Self and Peers
Predicting Self-Reported Outcomes From Implicit
Theories of Emotion and Intelligence
Table 2 presents the longitudinal correlations of incremental
theories of emotion and intelligence with self-reported emotional
Figure 2. Estimated negative and positive emotion experience as a function of incremental versus entity
theories of emotion (Part 2).
Figure 3. Estimated social support from college friends as a function of
incremental versus entity theories of emotion and week in the academic
term (Part 2).
738
TAMIR, JOHN, SRIVASTAVA, AND GROSS

and social outcomes. Consistent with the results in Part 2, incre-
mental theories of emotion, measured before coming to Stanford,
were linked to more positive emotions and fewer negative emo-
tions at the end of freshman year. In terms of broader well-being,
incremental emotion theories predicted greater psychological well-
being and lower rates of depression. Finally, in terms of social
outcomes, incremental theories of emotion predicted better social
adjustment and less loneliness. Unlike incremental theories of
emotion, incremental theories of intelligence were not related to
any of these outcomes. Moreover, as shown in Table 2, the effects
of implicit theories of emotion remained essentially unchanged—
even when the effects of implicit theories of intelligence were
controlled. These findings indicate that cognitive beliefs about the
malleability of emotion prior to college can foretell emotional and
social adjustment to college almost 1 year later.
Predicting Peer-Reported Outcomes From Implicit
Theories of Emotion and Intelligence
Table 3 presents the longitudinal correlations of the implicit
theories with peer-reported outcomes. Replicating the pattern
found with self-reports, peers reported that individuals with incre-
mental theories of emotion had more positive emotions and less
negative emotions. Peers also rated individuals with incremental
theories of emotion as greater in psychological well-being and as
less depressed. Corroborating self-reports, peers rated individuals
holding incremental theories of emotion as socially better adjusted
and as less lonely during the last quarter of freshman year. As with
self-reports, peer-rated outcomes were not related to implicit the-
ories of intelligence, and the correlations with implicit theories of
emotion remained essentially unchanged when controlling for im-
plicit theories of intelligence.
Overall, across outcome variables, the pattern of findings in
reports by knowledgeable peers closely paralleled those in the
self-reports. Such consistency substantially buttresses our confi-
dence in the present findings and, by implication, further corrob-
orates the findings in Part 2. Implicit theories of emotion that were
held before entering college predicted critical psychological and
social outcomes nearly 1 year later, regardless of data source.
Part 4: Emotion Regulation Self-Efficacy as a Mediator
Mediation of Self-Reported Emotional Outcomes
To test for mediation of emotional outcomes, we conducted six
mediation tests, two tests for emotion experiences during the first
quarter and four tests at the end of freshman year. Overall, emotion
regulation self-efficacy significantly mediated five of these emo-
tion and well-being outcomes.
Immediate emotion effects during the first quarter. We first
tested whether emotion regulation self-efficacy mediated the ef-
fects of implicit theories of emotion on negative or positive emo-
tions (Part 2). Specifically, we ran linear mixed models that
included implicit theories of emotion and emotion regulation self-
efficacy as Level 2 predictors and negative emotion or positive
emotion as the outcome variable. Higher levels of emotion regu-
lation self-efficacy predicted less negative emotion (␥⫽⫺.01),
t(230) ⫽ 5.19, p ⬍ .05, as did implicit theories of emotion. As
expected, Sobel tests for mediation indicated that the effect of
implicit theories of emotion on negative emotion was significantly
reduced (Sobel’s z ⫽ 3.98, p ⬍ .05). Emotion regulation self-
efficacy did not significantly predict positive emotion, but the
effect was in the right direction (␥⫽.005), t(239) ⫽ 1.89, p ⫽ .06.
The Sobel test in this case was not significant (Sobel’s z ⫽ 1.62,
ns).
Long-term emotion effects. To test for mediation of outcomes
at the end of freshman year (Part 3), we first conducted regression
analyses predicting negative emotion, positive emotion, psycho-
logical well-being, and depression reported at the end of freshman
year from both implicit theories of emotion and emotion regulation
self-efficacy. Emotion regulation self-efficacy predicted negative
emotions (␤⫽⫺.10, p ⬍ .05), positive emotions (␤⫽.19, p ⬍
.05), psychological well-being (␤⫽.25, p ⬍ .05), and depression
(␤⫽⫺.22, p ⬍ .05) at the end of freshman year. In all these cases,
implicit theories of emotion remained a significant predictor when
Table 2
Predicting Self-Reported Emotional and Social Outcomes at the
End of Freshman Year: Longitudinal Correlations With
Incremental Theories of Emotion and Intelligence (Part 3)
Outcome variable
Incremental theories
Emotion Intelligence
Emotion
(partialling out
intelligence)
Emotional and well-being
outcomes
Positive emotions .16
*
.04 .14
*
Negative emotions ⫺.18
*
⫺.04 ⫺.16
*
Well-being .24
*
.04 .20
*
Depression ⫺.15
*
⫺.03 ⫺.13
*
Social outcomes
Social adjustment .12
*
⫺.06 .13
*
Loneliness ⫺.16
*
⫺.01 ⫺.14
*
*
p ⬍ .05.
Table 3
Predicting Peer-Reported Emotional and Social Outcomes at the
End of Freshman Year: Longitudinal Correlations With
Incremental Theories of Emotion and Intelligence (Part 3)
Outcome variable
Incremental theory
Emotion Intelligence
Emotion
(partialling
out
intelligence)
Emotion and well-being
outcomes
Positive emotions .19
†
.10 .18
†
Negative emotions ⫺.15
†
.05 ⫺.17
†
Well-being .16
†
.05 .15
†
Depression ⫺.14
†
.08 ⫺.15
†
Social outcomes
Social adjustment .24
†
.03 .24
†
Loneliness ⫺.17
†
.06 ⫺.18
†
†
p ⬍ .05, one-tailed.
739
IMPLICIT THEORIES OF EMOTION

emotion regulation self-efficacy was included in the regression.
However, subsequent Sobel tests indicated that in all cases the
effects of implicit theories of emotion were significantly reduced
(Sobel’s zs ⫽ 3.89, 2.38, 4.48, and 3.82, respectively, ps ⬍ .05).
Taken together, these findings suggest that emotion regulation
self-efficacy partially mediated the effects of implicit theories of
emotion on negative emotion, positive emotion, psychological
well-being, and depression at the end of freshman year.
Mediation of Self-Reported Social Outcomes
None of the social outcomes were mediated by individual dif-
ferences in emotion regulation self-efficacy. Emotion regulation
self-efficacy predicted neither the effects of implicit theories of
emotion on social support during the first academic term (Part 2)
nor the loneliness or social adjustment at the end of freshman year
(Part 3). Moreover, none of the Sobel tests for mediation were
significant in these cases.
Mediation of Peer-Reported Emotional and Social
Outcomes
The mediation analyses above demonstrate that emotion regu-
lation self-efficacy mediated the link between implicit emotion
theories and emotional, but not social, outcomes. Emotional and
social outcomes at the end of freshman year were also reported by
peers, although the reliabilities of peer reports were generally
lower because of the smaller number of items included. Therefore,
to test whether the same mediation pattern found with self-reported
outcomes was replicated with peer-reported outcomes, we first
created an aggregated peer-reported emotion outcome by reverse
scoring peer-reported negative emotions and depression and aver-
aging the reverse scores with positive emotions and well-being so
that higher scores on the emotion aggregate reflect more positive
affect. As expected, the resulting measure was substantially more
reliable than each of its components (␣⫽.81). Similarly, we
created an aggregated score of peer-reported social outcomes by
averaging across social adjustment and reverse-scored loneliness
(␣⫽.74).
To test for mediation of peer-reported emotional outcomes, we
first conducted regression analyses predicting peer-reported emo-
tional outcomes from both implicit theories of emotion and emo-
tion regulation self-efficacy. Emotion regulation self-efficacy pre-
dicted emotional outcomes (␤⫽.19, p ⬍ .05). However, implicit
theories of emotion were no longer a significant predictor when
emotion regulation self-efficacy was included in the regression. A
subsequent Sobel test indicated that the effects of implicit theories
of emotion were significantly reduced (Sobel’s z ⫽ 2.08, p ⬍ .05).
The above mediation analysis was repeated with the social
outcome measure. As with self-reported social outcomes, peer-
reported social outcomes were not mediated by individual differ-
ences in emotion regulation self-efficacy. Emotion regulation self-
efficacy did not predict the effects of implicit theories of emotion
on peer-reported social outcomes, and the respective Sobel test for
mediation was not significant. Thus, as with the self-reported
outcomes, emotion regulation self-efficacy mediated peer-reported
emotional, but not social, outcomes. Such findings provide impor-
tant support to the validity of our findings.
Summary
Emotion regulation self-efficacy generally mediated the effects
of implicit emotion theories on emotional outcomes but not on
social outcomes. In general, this pattern was consistent over time:
It was obtained when examining outcomes during the first term
(i.e., in Part 2) as well as at the end of freshman year (i.e., in Part
3). Such mediational patterns were found for both self-reported
and peer-reported outcomes. Nevertheless, all mediation paths
were partial, suggesting that emotion regulation self-efficacy is
likely only one of several potential mediators.
Discussion
Individuals differ in their implicit theories of emotion. Almost
40% of the students in our sample believed emotions to be rela-
tively fixed (i.e., scored above the scale midpoint), whereas the
rest believed them to be relatively malleable. These beliefs had
important implications for emotion experience, social functioning,
and general adaptation during a major life transition, namely the
transition to college.
Before they entered college, individuals with entity (vs. incre-
mental) theories of emotion reported using cognitive reappraisal
less frequently to control their emotions and had a weaker sense of
emotion regulation self-efficacy (Part 1). Then, during the initial
period of the transition to college, implicit beliefs about emotion
helped launch experiential trajectories (Part 2). In particular, indi-
viduals with entity (vs. incremental) theories of emotion experi-
enced fewer positive and more negative emotions. In addition, as
the term progressed, these individuals received less and less social
support from college friends, whereas those who held incremental
theories of emotion received increasing social support. The emo-
tional and social benefits that accrued during the first academic
term were sustained all the way until the end of freshman year. At
the end of the year, individuals who had held entity (vs. incremen-
tal) theories of emotion before entering college had less favorable
emotion experiences, lower well-being, greater depression, more
loneliness, and poorer social adjustment as was evident in both self
and peer-reports (Part 3).
How do beliefs about the controllability of emotion lead to
increased depression and loneliness almost a year following their
assessment? In Part 4, we provided preliminary findings that
suggest that emotion regulation self-efficacy might serve as the
link between implicit emotion theories and their emotional out-
comes. Emotion regulation self-efficacy, however, did not mediate
the link between implicit theories of emotion and their social
outcomes.
Beliefs About Emotion and Their Implications
Our findings demonstrate that beliefs about the controllability of
emotion are linked to important emotional and social outcomes. By
demonstrating such links, the present findings extend prior re-
search on cognitive beliefs in emotional experience (e.g., Ben-
Artzi & Mikulincer, 1995; Robinson & Clore, 2002). Prior studies
have examined general beliefs about the nature of emotion, exam-
ining beliefs about controllability, predictability, stability, and
other attributes of emotion (e.g., Ben-Artzi & Mikulincer, 1996;
Mayer & Gaschke, 1988). Other research has linked emotional
740
TAMIR, JOHN, SRIVASTAVA, AND GROSS

outcomes to self-beliefs about personal abilities. For example,
Beck (1996) suggested that emotional disorders arise from dys-
functional self-beliefs. Furthermore, Wells and Matthews (1996)
suggested that maladaptive assumptions about one’s ability to
change can contribute to pathological outcomes.
The present findings extend such prior approaches in several
important ways. First, some of the prior research has examined
patterns of beliefs about emotion other than focusing on the
implications of beliefs about malleability, per se, as distinct from
beliefs about stability and predictability (e.g., Ben-Artzi & Miku-
lincer, 1995). Second, with respect to beliefs about malleability,
prior research has focused primarily on self-specific beliefs. How-
ever, general beliefs likely precede self-specific beliefs. According
to the health belief model (Rosenstock, 1974; Rosenstock et al.,
1988), general beliefs about the malleability of health conditions
are assumed to be a necessary precursor of subsequent beliefs
about self-efficacy. In the emotion domain, the belief that emo-
tions are malleable could be an important precursor of beliefs
about emotion-specific capabilities in the self and in others. Con-
sistent with these assumptions, we found that emotion regulation
self-efficacy mediated the link between implicit theories of emo-
tion and emotional outcomes.
Third, whereas self-beliefs are about the self, general beliefs
have implications for both the way a person views him- or herself
and the way a person perceives and responds to others. By explor-
ing general beliefs about emotions, we were able to demonstrate
important consequences in both the intrapersonal and the interper-
sonal domains. A final point is that previous research on self-
beliefs and emotional outcomes has focused exclusively on their
maladaptive implications. Our findings, on the other hand, suggest
that general beliefs about emotion can have both adaptive and
maladaptive implications. Our research is, therefore, of relevance
to psychologists interested in unhealthy as well as in healthy
functioning.
The present findings also encourage the exploration of other
beliefs about emotion, including beliefs about mixed attributes
such as emotional intelligence. In this respect, it is worth noting
that emotion, unlike intelligence, is transient and variable. Build-
ing on prior work on implicit theories, in this investigation we
examined beliefs about the controllability of emotion. It might also
be useful, in the future, to explore beliefs about the stability of
emotion, as well as other attributes.
Exploring Underlying Mechanisms
We found that the links between implicit emotion theories and
emotional experience were partially mediated by emotion regula-
tion self-efficacy. Emotion regulation self-efficacy, however, is
only one of many factors that could mediate the link between
implicit theories of emotion and emotional experience. If the belief
that emotions are malleable is a prerequisite for beliefs about one’s
personal ability to regulate one’s emotions, it could underlie a host
of emotion regulation processes, any one of which could influence
emotion either separately or in combination. Future research
should examine other emotion regulation variables as possible
mediators and test whether some are more prominent than others.
The present investigation provides initial support for one pos-
sible mediation model. Clearly, however, there are many other
mediation models that need to be tested. In future studies, it might
be useful to assess implicit theories of emotion, potential media-
tors, and relevant outcomes in each of several assessments that are
spread over time. Doing so would minimize any assessment-
specific effects as well as enable tests of reversed causal arrows
(e.g., might the link between implicit theories of emotion and
self-efficacy in emotion regulation be mediated by emotion expe-
rience?).
In the present investigation, although emotion regulation self-
efficacy mediated the link between implicit emotion theories and
emotional experiences, it did not mediate their links to social
functioning, and this was true for outcome variables from both
Parts 2 and 3. This dissociation was replicated with both self-
reported outcomes and peer-reported outcomes, attesting to the
reliability and validity of the findings. This replicated dissociation
between emotional and social effects raises the intriguing possi-
bility that different pathways might underlie the intra- and the
interpersonal implications of implicit theories of emotion.
An important question for future research, therefore, concerns
the factors that mediate the social outcomes of implicit theories of
emotion. We suggest that the mediators of personal and intrapsy-
chic outcomes (e.g., emotions, subjective well-being) involve be-
liefs about the self (e.g., emotion regulation self-efficacy), and our
findings begin to build some support for this hypothesis. Con-
versely, mediators of social outcomes (e.g., social support) might
involve beliefs about others (e.g., interpersonal trust). Exploring
such possible mediators is an important task for future research.
Ruling Out Alternative Hypotheses
In the present investigation, we were able to rule out some
alternative accounts of the associations between implicit emotion
theories and adaptive functioning. First, although implicit theories
are assumed to be domain specific (Dweck, 1999), they are clearly
based on similar worldviews (Dweck et al., 1995a). Therefore, one
alternative account is that the effects of implicit theories of emo-
tion are driven by the nature of implicit theories more generally or,
specifically, by the well-studied beliefs about the malleability of
intelligence. Our findings show little support for this hypothesis.
Although these two kinds of implicit theories were moderately
correlated (Part 1), the present effects were not driven by implicit
theories of intelligence but instead were specifically and uniquely
the result of implicit theories of emotion, as we were able to show
consistently in our investigation.
A second alternative account centers on a third variable that
potentially underlies implicit emotion theories, emotion regulation,
and emotional experience as well. One such variable is emotional
intensity. People with very strong emotions might be less able to
regulate their emotions and might therefore view emotions as
generally more difficult for people to control. As demonstrated in
Part 1, however, the correlation between implicit theories of emo-
tion and emotional intensity was small. Even more important, the
associations between implicit emotion theories and all emotion
regulation variables remained unchanged even after controlling for
emotional intensity.
Finally, prior research has linked emotion regulation concur-
rently to both emotional and social outcomes (e.g., Gross & John,
2003). Thus, it is possible that individual differences in emotion
regulation, such as in emotion regulation self-efficacy, underlie
both implicit theories of emotion and emotional and social conse-
741
IMPLICIT THEORIES OF EMOTION

quences. By showing that the link between implicit theories of
emotion and social outcomes was not mediated by emotion regu-
lation self-efficacy, we were able to rule out this and similar
alternative accounts. In other words, beliefs about the malleable
nature of emotion were associated with emotion regulation self-
efficacy, but they also had unique and important implications to
social functioning for which emotion regulation self-efficacy could
not account.
Limitations and Future Directions
Research on implicit theories joins a growing body of research
on the role of beliefs in self-regulation (e.g., Beck, 1996; Rosen-
stock, 1974; Wells & Matthews, 1996). Within this framework,
our findings link implicit beliefs about emotion to emotion regu-
lation, to emotion experience, and to social functioning. Consistent
with recent evidence (e.g., Gross & John, 2003; Mikulincer,
Shaver, & Pereg, 2003; Tamir, 2005), our findings emphasize the
importance of individual differences in emotion regulation. In
doing so, the present findings might help explain why some people
are less successful than others at modifying their emotions. For
example, individuals who hold entity theories of emotion might be
less aware of, and open to, the possibility of changing their
emotions and, therefore, less responsive to attempts by close others
or by health professionals to modify maladaptive emotion patterns.
Exploring such pragmatic implications and ways to devise appro-
priate interventions is an important task for future research.
In our investigation, implicit theories of emotion were positively
associated with emotion regulation self-efficacy and cognitive
reappraisal. However, it is important to acknowledge that such
links might not necessarily generalize to all emotion regulation
variables. In this respect, in Part 1 of this investigation, we found
that implicit theories of emotion were related to cognitive reap-
praisal, which concerns the modification of subsequent emotion
experience, but were not related to expressive suppression, which
concerns the modification of overt emotion expression. Although
such findings are consistent with existing evidence for the disso-
ciation between reappraisal and suppression (e.g., Gross & John,
2003), they suggest that individuals might differ in their implicit
beliefs about the experience versus the expression of emotion.
Future research, therefore, should directly examine whether indi-
viduals have distinct beliefs about the malleable nature of emotion
experience compared with emotion expression.
An important advantage of the present investigation is its lon-
gitudinal, multimethod design. These features, however, come at a
price. Specifically, our sample size was often limited and peer-
reported outcomes had relatively low reliability because of the
limited number of items that were included in the assessments.
Such limitations might also account for the relatively small effect
sizes. In the future, it is important to examine the implications of
implicit theories of emotion through the use of multiple items in
the assessments as well as larger sample sizes.
The longitudinal design of the present investigation allowed us
to test certain hypotheses about the real-life implications of im-
plicit theories of emotion as they evolved over time. One limitation
of this design, however, is that it cannot provide direct support for
causal accounts. Although causal conclusions are premature, our
findings so far are certainly consistent with the view that implicit
theories of emotion exert a causal influence. This view is also
consistent with empirical evidence for the causal role of implicit
theories in guiding self-regulation (see Dweck, 1999; Dweck et al.,
1995a) and with leading models of behavioral change (e.g., Rosen-
stock, 1974). Future research should test the causal hypothesis
more directly, for example, by experimentally manipulating beliefs
about the controllability of emotion and by testing the effects of
such manipulations on subsequent emotion regulation and experi-
ence.
The present investigation demonstrates that implicit theories of
emotion as measured prior to the transition to college were asso-
ciated with noteworthy differences in adjustment to college. How-
ever, the present investigation did not allow us to fully test the
extent to which implicit theories of emotion predicted changes in
adjustment during (vs. prior to) college. One exception, however,
concerns the impact of implicit theories of emotion on the devel-
opment of social relationships. By examining the nature of these
relationships as they were being created, we were able to demon-
strate that incremental theories of emotion contributed to increas-
ing social support during the first academic term. However, be-
cause we did not include measures of emotional and social
adjustment in Part 1 of the investigation, we were unable to test
other models that control for baseline levels. By assessing predic-
tors and outcomes at every longitudinal assessment, future re-
search should examine whether implicit theories of emotion pre-
dict levels of adjustment, changes in adjustment, or both.
Consistent with findings in other domains, the present findings
demonstrate positive implications to incremental beliefs about
emotion. It is important to emphasize, however, that we subscribe
to the idea that all beliefs have their costs and benefits. Some
potential costs of incremental theories are discussed in Dweck et
al. (1995b). Building on this discussion, it might be possible, for
instance, that incremental theories of emotion are costly for indi-
viduals who find it difficult to come to terms with their own
limitations. If a student believes that emotions are malleable yet
repeatedly fails in altering her emotions, the student might end up
feeling very incompetent. Examining the potential costs and ben-
efits of implicit theories of emotion and the extent to which they
vary as a function of stable dispositions are important tasks for
future research.
Finally, given the implications of implicit theories of emotion,
one important topic for future research concerns the precursors and
the developmental trajectory of such theories. In other words,
where do implicit theories of emotion come from and how do they
develop? On the basis of the analysis of the origins and develop-
ment of implicit theories provided in Dweck et al. (1995b), it is
plausible that both temperamental differences as well as socializa-
tion practices can contribute to the development of implicit theo-
ries of emotion. For instance, might individuals who often expe-
rience unpleasant emotion (e.g., depression) be more prone to
develop entity theories of emotion? In addition, might child-
rearing practices that encourage emotion regulation through con-
structive responses promote the development of incremental the-
ories of emotion? These and other potential antecedents remain to
be explored.
Conclusions
“He is able who thinks he is able,” noted Buddha, acknowledg-
ing the directive power of belief. The present investigation dem-
742
TAMIR, JOHN, SRIVASTAVA, AND GROSS

onstrates that people differ in their beliefs about the malleability of
emotions. Consistent with Buddha’s insight, those who believe
that people can control their emotions are themselves more effec-
tive in controlling their emotions and enjoy long-term emotional
and social benefits. By examining implicit beliefs in the emotion
domain, the present findings shed important new light on the
cognitive precursors of emotion regulation.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. Thousand Oaks, CA: Sage.
Andresen, E. M., Malmgren, J. A., Carter, W. B., & Patrick, D. B. (1994).
Screening for depression in well older adults: Evaluation of a short form
of the CES–D. American Journal of Preventive Medicine, 10, 77–84.
Astin, A. W. (1993). What matters in college? Four critical years revisited.
San Francisco, CA: Jossey-Bass.
Bandura, A. (1986). Social foundations of thought and action: A social
cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51, 1173–1182.
Beck, A. T. (1996). Beyond belief: A theory of modes, personality, and
psychopathology. In P. M. Salkovskis (Ed.), Frontiers of cognitive
therapy (pp. 1–25). New York: Guilford Press.
Bedier, J. (1900). Le roman de Tristan et Iseult [The romance of Tristan
and Isolde]. Paris: Paris L’edition D’art.
Beer, J. S. (2002). Implicit self-theories of shyness. Journal of Personality
and Social Psychology, 83, 1009 –1024.
Ben-Artzi, E., & Mikulincer, M. (1995). Lay theories of emotion: I.
Conceptualization and measurement. Imagination, Cognition and Per-
sonality, 15, 249–271.
Ben-Artzi, E., & Mikulincer, M. (1996). Lay theories of emotion: IV.
Reactions to negative and positive emotional episodes. Imagination,
Cognition, and Personality, 16, 89 –113.
Cairns, R. B., & Green, J. A. (1979). How to assess personality and social
patterns: Observations or ratings. In R. B. Cairns (Ed.), The analysis of
social interactions: Methods, issues, and illustrations (pp. 209–226).
Hillsdale, NJ: Erlbaum.
Cantor, N., Norem, J. K., Niedenthal, P. M., Langston, C. A., & Brower,
A. M. (1987). Life tasks and cognitive strategies in a life transition.
Journal of Personality and Social Psychology, 53, 1178 –1191.
Catanzaro, S. J., & Mearns, J. (1990). Measuring generalized expectancies
for negative mood regulation: Initial scale development and implica-
tions. Journal of Personality Assessment, 54, 546 –563.
Chaplin, W. F., John, O. P., & Goldberg, L. R. (1988). Conceptions of
states and traits: Dimensional attributes with ideals as prototypes. Jour-
nal of Personality and Social Psychology, 54, 541–557.
Chiu, C. Y., Dweck, C. S., Tong, J. Y. Y., & Fu, J. H. Y. (1997). Implicit
theories and conceptions of morality. Journal of Personality and Social
Psychology, 73, 923–940.
Chiu, C. Y., Hong, Y. Y., & Dweck, C. S. (1997). Lay dispositionism and
implicit theories of personality. Journal of Personality and Social Psy-
chology, 73, 19–30.
Dresel, M. (2001). A longitudinal analysis of Dweck’s motivation-process-
model in the classroom. Psychologische Beitrage, 43, 129 –152.
Dweck, C. S. (1986). Motivational processes affecting learning. American
Psychologist, 41, 1040–1048.
Dweck, C. S. (1996). Implicit theories as organizers of goals and behavior.
In P. M. Gollwitzer & J. A. Bargh (Eds.), The psychology of action:
Linking cognition and motivation to behavior (pp. 69–90). New York:
Guilford Press.
Dweck, C. S. (1999). Self-theories: Their role in motivation, personality,
and development. New York: Psychology Press.
Dweck, C. S., Chiu, C. Y., & Hong, Y. Y. (1995a). Implicit theories and
their role in judgments and reactions: A world from two perspectives.
Psychological Inquiry, 6, 267–285.
Dweck, C. S., Chiu, C. Y., & Hong, Y. Y. (1995b). Implicit theories:
Elaboration and extension of the model. Psychological Inquiry, 6, 322–
333.
Dweck, C. S., & Leggett, E. L. (1988). A social– cognitive approach to
motivation and personality. Psychological Review, 95, 256 –273.
Erdley, C. A., & Dweck, C. S. (1993). Children’s implicit personality
theories as predictors of their social judgments. Child Development, 64,
863– 878.
Fowler, J. W., & Peterson, P. L. (1981). Increasing reading persistence and
altering attributional style of learned helpless children. Journal of Edu-
cational Psychology, 73, 251–260.
Gross, J. J. (1998). The emerging field of emotion regulation: An integra-
tive review. Review of General Psychology, 2, 271–299.
Gross, J. J., & John, O. P. (1998). Mapping the domain of expressivity:
Multimethod evidence for a hierarchical model. Journal of Personality
and Social Psychology, 74, 170 –191.
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion
regulation processes: Implications for affect, relationships, and well-
being. Journal of Personality and Social Psychology, 85, 348–362.
Gross, J. J., John, O. P., & Richards, J. M. (2000). The dissociation of
emotion expression from emotion experience: A personality perspective.
Personality and Social Psychology Bulletin, 26, 712–726.
Gross, J. J., Richards, J. M., & John, O. P. (2006). Emotion regulation in
everyday life. In D. K. Snyder, J. A. Simpson, & J. N. Hughes (Eds.),
Emotion regulation in couples and families: Pathways to dysfunction
and health (pp. 13–35). Washington, DC: American Psychological As-
sociation.
Henderson, V. L., & Dweck, C. S. (1990). Motivation and achievement. In
S. Feldman & G. R. Elliot (Eds.), At the threshold: The developing
adolescent (pp. 308–329). Cambridge, MA: Harvard University Press.
Hochschild, A. R. (1979). Emotion work, feeling rules, and social struc-
ture. American Journal of Sociology, 85, 551–575.
Hong, Y. Y., Chiu, C. Y., Dweck, C. S., Lin, D. M. S., & Wan, W. (1999).
Implicit theories, attributions, and coping: A meaning system approach.
Journal of Personality and Social Psychology, 77, 588 –599.
Hong, Y. Y., Chiu, C. Y., Dweck, C. S., & Sacks, R. (1997). Implicit
theories and evaluative processes in person cognition. Journal of Exper-
imental Social Psychology, 33, 296 –323.
Kreft, I. G. G., DeLeeuw, J., & Aiken, L. S. (1995). Variable centering in
hierarchical linear models: Model parameterization, estimation, and in-
terpretation. Multivariate Behavioral Research, 30, 1–21.
Krull, J. L., & MacKinnon, D. P. (1999). Multi-level mediation modeling
in group-based intervention studies. Evaluation Review, 23, 418 – 444.
Krull, J. L., & MacKinnon, D. P. (2001). Multi-level modeling of individ-
ual and group level mediated effects. Multivariate Behavioral Research,
36, 249–277.
Lopes, P. N., Salovey, P., Cote, S., & Beers, M. (2005). Emotion regulation
ability and the quality of social interaction. Emotion, 5, 113–118.
Mayer, J. D., & Gaschke, Y. N. (1988). The experience and meta-
experience of mood. Journal of Personality and Social Psychology, 55,
102–111.
Mikulincer, M., Shaver, P. R., & Pereg, D. (2003). Attachment theory and
affect regulation: The dynamics, development, and cognitive conse-
quences of attachment-related strategies. Motivation and Emotion, 27,
77–102.
Olson, J. M., Roese, N. J., & Zanna, M. P. (1996). Expectancies. In E. T.
Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of
basic principles (pp. 211–238). New York: Guilford Press.
Robins, R. W., & Pals, J. L. (2002). Implicit self-theories in the academic
743
IMPLICIT THEORIES OF EMOTION

domain: Implications for goal orientation, attributions, affect, and self-
esteem change. Self and Identity, 1, 313–336.
Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evidence for
an accessibility model of emotional self-report. Psychological Bulletin,
128, 934–960.
Rosenstock, I. M. (1974). Historical origins of the health belief model.
Health Education Monographs, 2, 328 –335.
Rosenstock, I. M., Strecher, V. J., & Becker, M. H. (1988). Social learning
theory and the health belief model. Health Education Quarterly, 15,
175–183.
Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological
well-being revisited. Journal of Personality and Social Psychology, 69,
719 –727.
Scollon, C. N., Kim-Prieto, C., & Diener, E. (2003). Experience sampling:
Promises and pitfalls, strengths and weaknesses. Journal of Happiness
Studies, 4, 5–34.
Seligman, M. E. P. (1975). Helplessness: On depression, development, and
death. San Francisco: Freeman.
Shakespeare, W. (1992). Hamlet: Prince of Denmark. Washington, DC:
The Folger Shakespeare Library. (Original work published 1603)
Shaver, P. R., Furman, W., & Buhrmester, D. (1985). Transition to college:
Network changes, social skills, and loneliness. In S. Duck & D. Perlman
(Eds.), Understanding personal relationships: An interdisciplinary ap-
proach (pp. 193–219). Thousand Oaks, CA: Sage.
Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis:
Modeling change and event occurrence. New York: Oxford University
Press.
Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural
equations models. In S. Leinhart (Ed.), Sociological methodology (pp.
290 –312). San Francisco, CA: Jossey-Bass.
SPSS, Inc. (2002). SPSS (Version 11.5) [Computer software]. Chicago:
Author.
Tabernero, C., & Wood, R. E. (1999). Implicit theories versus the social
construal of ability in self-regulation and performance on a complex
task. Organizational Behavior and Human Decision Processes, 78,
104 –127.
Tamir, M. (2005). Don’t worry, be happy? Neuroticism, trait-consistent
affect regulation, and performance. Journal of Personality and Social
Psychology, 89, 449–461.
Wells, A., & Matthews, G. (1996). Modeling cognition in emotional
disorder: The S-REF model. Behavior Research and Therapy, 34, 881–
888.
Received November 11, 2005
Revision received October 23, 2006
Accepted November 2, 2006 䡲
744
TAMIR, JOHN, SRIVASTAVA, AND GROSS