Adolescents' implicit theories predict desire for vengeance after peer conflicts: Correlational and experimental evidence

Abstract

Why do some adolescents respond to interpersonal conflicts vengefully, whereas others seek more positive solutions? Three studies investigated the role of implicit theories of personality in predicting violent or vengeful responses to peer conflicts among adolescents in Grades 9 and 10. They showed that a greater belief that traits are fixed (an entity theory) predicted a stronger desire for revenge after a variety of recalled peer conflicts (Study 1) and after a hypothetical conflict that specifically involved bullying (Study 2). Study 3 experimentally induced a belief in the potential for change (an incremental theory), which resulted in a reduced desire to seek revenge. This effect was mediated by changes in bad-person attributions about the perpetrators, feelings of shame and hatred, and the belief that vengeful ideation is an effective emotion-regulation strategy. Together, the findings illuminate the social-cognitive processes underlying reactions to conflict and suggest potential avenues for reducing violent retaliation in adolescents.

Full text

Adolescents’ Implicit Theories Predict Desire for Vengeance After Peer
Conflicts: Correlational and Experimental Evidence
David S. Yeager
Stanford University Kali H. Trzesniewski
University of California, Davis
Kirsi Tirri
University of Helsinki Petri Nokelainen
University of Tampere
Carol S. Dweck
Stanford University
Why do some adolescents respond to interpersonal conflicts vengefully, whereas others seek more positive
solutions? Three studies investigated the role of implicit theories of personality in predicting violent or
vengeful responses to peer conflicts among adolescents in Grades 9 and 10. They showed that a greater belief
that traits are fixed (an entity theory) predicted a stronger desire for revenge after a variety of recalled peer
conflicts (Study 1) and after a hypothetical conflict that specifically involved bullying (Study 2). Study 3
experimentally induced a belief in the potential for change (an incremental theory), which resulted in a reduced
desire to seek revenge. This effect was mediated by changes in bad-person attributions about the perpetrators,
feelings of shame and hatred, and the belief that vengeful ideation is an effective emotion-regulation strategy.
Together, the findings illuminate the social–cognitive processes underlying reactions to conflict and suggest
potential avenues for reducing violent retaliation in adolescents.
Keywords: implicit theories, aggression, victimization, shame, emotion regulation
Kenneth Dodge recently argued that when public figures use
fixed labels (such as super predator or morally defective)to
describe aggressive youths, it leads people to focus on extreme
punishment instead of on prevention or rehabilitation (Dodge,
2008). Would this same analysis apply to adolescents’ own reac-
tions to their conflicts with peers?
Past research has indeed suggested that a belief in fixed traits
(an entity theory of personality) predicts a greater focus on pun-
ishment for a wrongdoer (Chiu, Dweck, Tong, & Fu, 1997; Erdley
& Dweck, 1993; Gervey, Chiu, Hong, & Dweck, 1999; Loeb &
Dweck, 1994; cf. Giles, 2003). In contrast, a belief in the potential
for change (an incremental theory) predicts a greater focus on
negotiation, education, forgiveness, and rehabilitation among chil-
dren and adults (Chiu, Dweck, Tong, & Fu, 1997; Gervey et al.,
1999; Haselhuhn, Schweitzer, & Wood, 2010; Loeb & Dweck,
1994). Yet, no previous research has shown that adolescents’
implicit theories about people’s potential for change will predict
their vengeful responses to interpersonal conflicts. Therefore, in
the present research, we investigated whether adolescents who
endorse more of an entity theory would give more vengeful re-
sponses to recalled and hypothetical conflicts. To deepen our
theoretical understanding of these effects, we also explored, for the
first time, the cognitive and emotional mediators of this process.
1
Adolescent Responses to Victimization
Recent incidents of planned violence in American and Finnish
high schools have turned the public’s eye toward issues of coping
1
We use the term conflict in this article as an umbrella term to refer to
conflicts between acquaintances that involve aggression or victimization or
that could be the precursors to aggressive retaliation. We do not mean to
suggest that we are investigating all types of social conflict between
adolescents or that all conflicts involve aggression or victimization.
This article was published Online First May 23, 2011.
David S. Yeager, School of Education and Department of Psychology,
Stanford University; Kali H. Trzesniewski, Department of Human and Com-
munity Development, University of California, Davis; Kirsi Tirri, Department
of Teacher Education, University of Helsinki, Helsinki, Finland; Petri Noke-
lainen, School of Education, University of Tampere, Tampere, Finland; Carol
S. Dweck, Department of Psychology, Stanford University.
This research was supported in part by a dissertation support grant from the
Stanford University School of Education, a research training grant from the
Spencer Foundation, and a grant from the Thrive Foundation for Youth. Some
of the data were presented at the 2009 annual convention for the Association
for Psychological Science and the 2009 annual meeting of the American
Education Research Association. We are grateful to the following people for
their assistance: Christel Anderson, Patrinia Sandles, Alan Weyland, Allison
Master, Whitney Worthen, Rebecca Johnson, Yana Galperin, Adriana Miu,
Joseph Lester, Matthew Williams, Kate Belden, and Jason Singer. We would
also like to thank S. Shirley Feldman, Ann Porteus, Deborah Stipek, William
Damon, Melanie Killen, Kristin Pauker, Chris Bryan, and members of the
Dweck–Walton lab for their helpful suggestions.
Correspondence concerning this article should be addressed to David S.
Yeager, 271 Jordan Hall, School of Education and Department of Psychol-
ogy, Stanford University, Stanford, CA 94305. E-mail: dyeager@
stanford.edu
Developmental Psychology © 2011 American Psychological Association
2011, Vol. 47, No. 4, 1090–1107 0012-1649/11/$12.00 DOI: 10.1037/a0023769
1090

with peer conflicts in general and peer victimization in particular.
Although such highly violent responses are rare, it is quite com-
mon for adolescents to be victims of peer aggression in school
(Kaufman et al., 1999; Nishina & Juvonen, 2005; Olweus, 1993;
for reviews, see Berger, 2007; Swearer, Espelage, Vaillancourt, &
Hymel, 2010) and to subsequently desire revenge (e.g., Juvonen,
Graham, & Schuster, 2003; Kaltiala-Heino, Rimpela¨, Marttunen,
Rimpela¨, & Rantanen, 1999; Olweus, 1993; Reijntjes et al., in
press; Sourander et al., 2007; Swearer, Song, Cary, Eagle, &
Mickelson, 2001; Thomaes, Bushman, Stegge, & Olthof, 2008).
Thus, it is critical that researchers improve the understanding of
why some adolescents fantasize about revenge after their peer
conflicts, especially those involving victimization, whereas others
seek different solutions.
Implicit Theories Predict Responses to Setbacks
Previous research on implicit theories has illuminated their
role in determining how people cope with setbacks in the
academic domain (Blackwell, Trzesniewski, & Dweck, 2007;
Hong, Chiu, Dweck, Lin, & Wan, 1999; Robins & Pals, 2002)
and, more directly relevant here, the social domain (Beer, 2002;
Erdley, Cain, Loomis, Dumas-Hines, & Dweck, 1997; Loeb &
Dweck, 1994). Previous research has also investigated how
children and adults evaluate others who fail or transgress (Chiu,
Dweck, Tong, & Fu, 1997; Erdley & Dweck, 1993; Gervey et
al., 1999; Haselhuhn et al., 2010; Levy & Dweck, 1999; Levy,
Stroessner, & Dweck, 1998; see also Plaks, Stroessner, Dweck,
& Sherman, 2001).
For example, Erdley et al. (1997) found that when fifth-grade
students experienced social rejection, those who held more of an
entity theory about their personality were more likely to view their
social rejection as evidence of their inability to make friends,
making attributions about themselves such as “maybe I’m not a
likable person.” As a result, those with an entity theory held goals
that predicted defensive and helpless responses to failure. Simi-
larly, Rudolph (2010) found that peer-victimized children who
held more of an entity theory reported more depressive symptoms
than peer-victimized children who held more of an incremental
theory.
Implicit theories have also been shown to predict how people
respond to wrongdoing on the part of others. Specifically,
research has shown that both children (Erdley & Dweck, 1993;
Levy & Dweck, 1999) and college students (Chiu, Dweck,
Tong, & Fu, 1997; Chiu, Hong, & Dweck, 1997; Gervey et al.,
1999; Levy et al., 1998) who hold an entity theory about others’
personalities are more judgmental and punitive when they eval-
uate a wrongdoer. That is, after learning about a transgression,
they are more likely to indict the target’s moral character (Chiu,
Dweck, Tong, & Fu, 1997; Chiu, Hong, & Dweck, 1997; Erdley
& Dweck, 1993; Levy & Dweck, 1999) and to levy harsher
punishment than are those holding an incremental theory (Chiu,
Dweck, Tong, & Fu, 1997; Erdley & Dweck, 1993; Loeb &
Dweck, 1994; Miller, Burgoon, & Hall, 2007). In contrast,
those with an incremental theory are more likely to suggest
discussion, negotiation, or education (Chiu, Dweck, Tong, &
Fu, 1997; Kray & Haselhuhn, 2007; Loeb & Dweck, 1994; see
also Haselhuhn et al., 2010).
In summary, theory and research suggest that when students
hold an implicit theory that people’s traits are fixed and not
changeable, they are more likely to see wrongdoing as stemming
from permanent negative qualities of the wrongdoers and to see
their own social failures as resulting from personal deficiencies (cf.
Giles, 2003). In contrast, when students think of people in terms of
malleable qualities—ones that can be developed through effort or
experience—they are more likely to view transgressions as reme-
diable and their own social setbacks as less diagnostic of their
future. We sought to extend this past theory and research by
examining the relationship between implicit theories and vengeful
responses to interpersonal conflicts and by examining the media-
tors of this relationship.
Limitations of Previous Research
The previous research on implicit theories has several limita-
tions. First, as noted, it has never been shown that adolescents with
more of an entity theory desire to respond more vengefully after
peer conflict than those with more of an incremental theory. Erdley
and Dweck (1993) and other investigators (e.g., Chiu, Dweck,
Tong, & Fu, 1997; Gervey et al., 1999; Levy & Dweck, 1999) have
focused on the role of implicit theories in third-party observers’
judgments of transgressors but not on judgments and reactions
made by parties involved in the conflict. Additionally, although
Erdley et al. (1997) focused on implicit theories and responses to
social rejection, they explored only self-relevant attributions and
reactions and not judgments of and reactions to the transgressor at
the same time. In the one published study to document the rela-
tionship between implicit theories and the desire for personal
vengeance (Chiu, Dweck, Tong, & Fu, 1997, Study 4), the trans-
gressor was a college professor, not a peer. Because a norm
violation by an adult may differ considerably from peer bullying or
victimization, it remains unclear whether implicit theories would
be related to high school adolescents’ interpersonal conflicts. We
therefore tested this possibility more extensively in the present
research.
Second, no study has systematically explored mediators of the
relation between implicit theories and desire for vengeance. To
fully understand this relationship and, eventually, to design effec-
tive interventions, we must know how and why implicit theories
might lead to a desire for revenge. Thus, in the present research,
we propose to test the role of attributions and emotions in shaping
implicit theories’ effects on responses to conflict.
Third, most previous studies have used correlational methods,
and no experiment to date has been conducted to show the causal
role of implicit theories in shaping adolescents’ responses to
conflict. Thus, it is unclear whether it is possible to change an
adolescent’s implicit theories of personality and whether doing so
could reduce the desire for vengeance and increase the desire to
use more prosocial solutions.
Fourth, the generalizability across populations and age groups of
implicit theories’ effects on responses to conflict has not been ade-
quately tested. Many previous studies have been conducted on sam-
ples of mostly White, Midwestern grade-school students (e.g., Erdley
& Dweck, 1993; Erdley et al., 1997; Rudolph, 2010) or on college
students attending elite universities (e.g., Chiu, Dweck, Tong, & Fu,
1997; Haselhuhn et al., 2010; Kammrath & Dweck, 2006; Loeb &
Dweck, 1994). No similar studies have been conducted on samples of
1091
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

adolescents from diverse socioeconomic backgrounds attending high
schools in the United States or other countries.
In our research, we included a sample of participants from Finland,
home of much research on adolescents’ social cognitions after vic-
timization or exclusion (e.g., Ojanen, Grönroos, & Salmivalli, 2005;
Peets, Hodges, & Salmivalli, 2008; Salmivalli, Ojanen, Haanpa¨a¨, &
Peets, 2005; Sourander et al., 2007). Although the results of studies of
such populations are often presumed to generalize to urban American
populations, this assumption has almost never been directly tested.
Finland has one of the highest adolescent suicide rates in the world—
over twice that found in the United States (Lahti et al., 2006; Statistics
Finland, 2009). Thus, Finnish adolescents with more of an entity
theory might have reacted to social challenges with more self-directed
aggression as opposed to other-directed aggression. Therefore, we
sought to explicitly test the generalizability of our findings by exam-
ining a socioeconomically diverse but racially homogeneous Finnish
population and also a racially diverse U.S. population.
Recent developmental research has suggested that high school
adolescents may be at a critical point in their understanding of others,
particularly in their understanding about whether others can change.
Specifically, research has suggested that relative to children or
younger adolescents, high school students are coming to hold more
fixed beliefs about aggressive peers (Killen, Crystal, & Watanabe,
2002), are less likely to try to intervene to change bullies (Trach,
Hymel, Waterhouse, & Neale, 2010), and are more likely to rely on
information they have about a peer’s past transgressions when decid-
ing whether the peer is guilty of a subsequent transgression (Killen,
Kelly, Richardson, & Jampol, 2010).
In this light, it is interesting to note that the transition to high
school is also associated with an increased frequency of extreme
reactions to victimization, in the form of suicide or other-directed
violence (Brent, Baugher, Bridge, Chen, & Chiappetta, 1999;
Leary, Kowalski, Smith, & Phillips, 2003; Vossekuil, Fein, Reddy,
Borum, & Modzeleski, 2002; for a review, see Borum, Cornell,
Modzeleski, & Jimerson, 2010). Thus, it may be especially im-
portant to understand the role of entity beliefs in predicting more
extreme reactions to victimization among adolescents in this age
group.
Overall, then, the present investigation is unique (a) in examin-
ing, with diverse groups of high school adolescents, the relation-
ship between implicit theories and the desire for personal ven-
geance after peer conflicts, (b) in testing the hypothesis that
implicit theories of personality play a causal role in the desire for
revenge, and (c) in proposing and testing a mediational model
explaining this relationship.
To further explain why adolescents with a greater entity theory
might be more likely to desire vengeance, we now turn to the
hypothesized mediators of this effect: thoughts and emotions about
the transgressors and thoughts and emotions about the self.
Mediators of Implicit Theory Effects on Responses to
Conflict
Bad-Person Attributions
Implicit theories have been found to be associated with adoles-
cents’ attributions about what caused a transgressing peer to act
that way. Adolescents who endorse more of an entity theory may
be more likely to interpret a perpetrator’s actions as evidence that
the offender is a bad person. Specifically, because previous re-
search has documented that an entity theory is related to a greater
use of global character traits when judging others (Chiu, Hong, &
Dweck, 1997; Erdley & Dweck, 1993; Plaks et al., 2001) and to a
greater endorsement of punishment, as opposed to education or
rehabilitation, for a transgressor (Chiu, Dweck, Tong, & Fu, 1997;
Erdley & Dweck, 1993; Gervey et al., 1999; Loeb & Dweck,
1994), we predicted that adolescents with more of an entity theory
would be more inclined to conclude that the offender was charac-
terologically a bad person, to hate the offender, and, because of
this, to desire revenge.
Note that the fixed-trait or bad-person attributions investigated
in the present study differ from the attributions of hostile intent
that are often investigated in studies of adolescents’ social cogni-
tions after conflict (e.g., Crick & Dodge, 1994; Dodge, Coie, &
Lynam, 2006; Killen et al., 2010; McGlothlin & Killen, 2006,
2010; Peets, Hodges, Kikas, & Salmivalli, 2007; Peets et al.,
2008). Studies of hostile intent have mostly investigated children’s
and adolescents’ attributions of intent after an ambiguous event,
whereas our research is concerned with victims’ attributions about
the kind of person a perpetrator is, regardless of the ambiguity of
that event. Indeed, much of the present research focuses on inci-
dents in which the intent is clearly negative, as is the case with
many incidents of bullying (Olweus, 1993).
Bad Feelings About the Self
In addition to thinking about the transgressors as bad people,
adolescents holding more of an entity theory may also experience
more negative feelings about themselves after conflicts, and these
feelings may also contribute to more vengeful responses. That is,
for those who hold an entity theory, even a minor incident—like
being ignored or teased—can be loaded with emotional signifi-
cance. In an entity framework, such conflicts are not small events
that will soon pass; they are diagnostic, lasting symptoms of one’s
own permanent failings as a person: Does it mean I’m not likable?
That I’m a loser? Under these circumstances, peer conflicts—
especially those involving victimization or rejection—can be
highly charged events, resulting in strong negative emotions. In-
deed, previous studies have shown that those who hold more of an
entity theory blame uncontrollable and stable characteristics about
themselves after social conflicts (Erdley et al., 1997) and that
people who engage in these kinds of attributions report more
negative feelings about themselves after victimization (Graham,
Bellmore, & Mize, 2006; Graham & Juvonen, 1998; see also
Janoff-Bulman, 1979).
Shame, rather than another negative emotion, such as sadness, is
particularly likely to be felt by those with more of an entity theory
because of the focus on fixed traits as an explanation for social
failures. Shame occurs after blaming a negative event on one’s
self, as opposed to one’s behavior (Tangney, Miller, Flicker, &
Barlow, 1996; Tangney, Stuewig, & Mashek, 2007), and “shame is
considered the more painful emotion because one’s core self—not
simply one’s behavior—is at stake” (Tangney et al., 2007, p. 346).
Why might these negative attributions and feelings about the
self lead to aggressive responses to conflicts? Leading theories of
aggression contend that negative feelings resulting from peer con-
flict are especially likely to result in retaliation when they are
accompanied by the belief that one is unlikely to be reincluded
1092 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

(Williams, 2009). For example, in one experiment, college stu-
dents who were given feedback that their traits would lead them to
live their lives alone—somewhat analogous to being in an entity
framework—behaved more aggressively in what they believed
was a peer interaction after being made to feel bad through social
rejection (Twenge, Baumeister, Tice, & Stucke, 2001). Using other
methodologies, researchers have found convergent results among
adolescents (Reijntjes et al., in press). Therefore, we propose that
adolescents who hold more of an entity theory feel more ashamed
by victimization and, rather than seeking to improve their relation-
ships with a transgressor, are more likely to desire revenge.
The Present Investigation
In Studies 1 and 2, we measured adolescents’ beliefs about
whether specific types of people, such as bullies, victims, winners,
or losers, can change. In Study 1, we tested whether an entity
theory was correlated with a greater desire for vengeance after a
recalled conflict. In addition, we began to test two mechanisms
underlying the key effect: the attribution that the perpetrator was a
bad person and negative feelings about the self. We administered
identical measures to a countrywide sample from Finland and to a
diverse group of students from low-income families attending an
urban public high school in Oakland, California. We explicitly
tested whether the effects found were the same for participants
who differed in racial/ethnic group, culture, and socioeconomic
status. Thus, Study 1’s samples provided a strict test of general-
izability.
In Study 2, we sought to test our hypothesis using a hypothetical
scenario involving bullying. Study 1’s procedure was designed to
elicit a broad array of recalled conflicts with acquaintances. Yet,
not all conflicts involve aggression or victimization. Therefore, in
Study 2, we narrowed our focus to a conflict that included peer
victimization.
In Study 3, we used an experiment to test whether implicit
theories could be changed and whether promoting an incremental
theory might reduce vengeful responses to victimization and pro-
mote more positive ones. In this experiment, participants were
randomly assigned to read either an article supporting an incre-
mental theory or a control article. In addition to testing whether the
incremental theory reduced the desire for vengeance, we also
tested whether an expanded set of potential mediating mechanisms
that measured more precise emotions and beliefs might more fully
explain the predicted reduction in the desire for vengeance.
Study 1
Methods
Participants.
Finnish subsample. This subsample consisted of 219 Finnish
adolescents in Grades 9 and 10, ranging in age from 14 to 16 years.
Of the participants, 47% were girls, 95% described themselves as
White, and 94% reported that their parents spoke Finnish at home.
Participants were from a variety of socioeconomic backgrounds;
22% reported that their mothers held a graduate degree, 10% a
college degree, 36% some college education, and the rest (44%) a
high school degree or less. Some (n⫽16) did not provide data on
at least one major measure, so their responses were multiply
imputed.
U.S. subsample. The U.S. subsample was a diverse group of
138 students in Grade 9 (56%) and Grade 10 (44%), who ranged
in age from 14 to 16 years. Among participants, 21% were Black/
African American, 36% were Asian or Asian American, 35% were
Hispanic/Latino, 4% were White, and 4% indicated another race/
ethnicity; 42% were boys, and 58% were girls. Nearly all received
free or reduced price lunch, and only 10% reported mothers who
had a college degree or higher. Some participants (n⫽16) did not
provide data on at least one dependent measure, and so their
responses were imputed as described earlier.
Procedure.
Finnish subsample. Six schools in Finland, sampled from
various regions across the country, were invited to participate in
this study, and all agreed. Three were rural schools, and three were
urban or suburban schools. All schools were from different parts of
the country with varying levels of socioeconomic status. After
consent and assent were obtained, surveys were administered
during school hours to a subset of classrooms in each school. In
this and in subsequent studies, students were assured of the con-
fidentiality of their responses and were told that raw data would
never be shared with anyone at the school. Measures were pro-
vided in Finnish. They were translated from an English version by
a research professor who was a native Finnish speaker and were
back-translated into English to check the quality of the translation.
Response rates varied by classroom within each school, but all
were above 70%. After the surveys were completed, students
were thanked for their time but were not yet told the specific
hypotheses of the study because they were later invited to
participate in Study 3.
U.S. subsample. All of the students in Grades 9 and 10 at a
low-income, urban public charter school in Oakland were invited
to participate, and consent and assent were obtained from 78%.
The survey was administered during school hours and took about
30 min to complete. The instructions and debriefing scripts read by
the teachers to the participants were identical to those read to the
Finnish subsample. After completing the survey, participants were
thanked and debriefed.
Measures.
Implicit theories. Six items measuring an entity theory about
bullies, victims, winners, and losers were written specifically for
this study. These were as follows: “Bullies and victims are types of
people that really can’t be changed.” “Everyone is either a winner
or a loser in life.” “You can’t change people who are jerks in
school.” “Bullies can try acting nice, but deep down they’re just
bullies.” “There are two types of people: Bullies and their vic-
tims.” Participants were asked to indicate their agreement or dis-
agreement on a 6-point scale (1 ⫽strongly disagree,6⫽strongly
agree). In an exploratory factor analysis of these items in which we
used maximum likelihood estimation, the first unrotated factor
accounted for 84% of the variance, and all six items loaded on this
factor at .59 or above. We tested for measurement invariance
across subsamples by conducting a multigroup confirmatory factor
analysis and constraining the factor loadings from the manifest
indicators of the latent implicit theories variable to be equal in the
two samples. We found that this did not significantly reduce model
fit, ⌬␹
2
(5) ⫽7.17, ns, suggesting that the underlying construct is
the same across subsamples. Having concluded that the six items
1093
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

measured a single construct in both subsamples, these items were
averaged and combined into a single scale, with higher values
corresponding to more of an entity theory (␣⫽.82).
2
In a study
conducted with an ethnically diverse sample in the United States,
we found that responses to the implicit theories measure were
relatively stable over a 2-month period (from Time 1 to Time 2,
r⫽.51, N⫽194, p⬍.05).
Desire for vengeance. To measure the desire for vengeance
after an actual, recalled conflict, we asked participants to “tell us
about a time that an acquaintance upset you in the past month or
so” either in school or outside of school. We then asked partici-
pants to rate how much they “felt like doing” each of seven options
on a 5-point scale (1 ⫽not at all,5⫽a great deal). The items
were as follows: “hurting this person,” “never forgiving them,”
“trying to get back at them in any way I could,” “finding a way to
punish this person,” “dreaming about a way to give them what they
deserve,” “wishing that somebody would hurt them,” and “imag-
ining them getting hurt.” Participants also rated several positively
worded filler items. To test whether the seven vengeful items were
measuring a single construct, we used maximum likelihood esti-
mation to conduct an exploratory factor analysis and found that the
seven items had loadings of .66 or above on a single unrotated
factor that accounted for 91% of the variance. We used the method
described earlier to test whether the construct of desire for ven-
geance was invariant across the subsamples, and we found that it
was, ⌬␹
2
(6) ⫽7.95, ns. The final measure was an unweighted
average of the seven vengeful items, with higher values corre-
sponding to a greater desire for vengeance (␣⫽.92).
We also coded the types of incidents respondents recalled. Two
independent coders, who were unaware of participants’ other re-
sponses, coded participants’ descriptions of the conflicts into cat-
egories, on the basis of a list of common peer conflicts described
by Olweus (1993). The most frequently named categories were
verbal aggression and insults (22.7%), spreading rumors/lies
(19.8%), and shunning/ignoring (16.1%). About one fourth of
participants wrote about more minor conflicts that were not clearly
victimization (24.9%; e.g., “He looked at me wrong”; Cohen’s ␬⬎
.85). Participants were also asked how long ago the incident
occurred (open-ended, coded in days since the incident happened).
After answering these relatively negative questions, participants
were given an extensive set of positively worded filler questions
about laughing with friends, eating ice cream, and having goals in
life.
We conducted numerous pilot studies with high school students
in Finland and in the United States to test the reliability and
validity of our measure of the desire for vengeance. In one study
conducted with students in Grades 9 and 10, the desire for ven-
geance after a recalled incident was significantly correlated with
the same desire for vengeance after a hypothetical bullying sce-
nario, r⫽.57, N⫽283, p⬍.05 (this scenario is described in
Study 2). Next, in two different diverse urban samples of U.S.
students in Grades 9 and 10, the desire for vengeance after the
recalled incident was significantly correlated with a measure of
anger rumination (Peled & Moretti, 2007), r⫽.33, N⫽127, and
r⫽.56, N⫽57.
We next examined what types of aggressive behavior were
associated with the desire for vengeance. Because we conceptual-
ized the desire for vengeance as including both an immediate
response of angry retaliation and a more delayed, planned, and
deliberate processes of rehearsing or enacting vengeful fantasies to
recover one’s status or restore justice, we predicted that this desire
would be associated with both reactive and proactive aggression.
Therefore, it is not surprising that in an additional U.S. pilot
sample, the desire for vengeance was equally correlated with both
self-reported reactive aggression (e.g., “gotten angry when others
threaten you”) and proactive aggression (“threaten or bully oth-
ers”; Raine et al., 2006), rs⫽.50, .51, N⫽33, ps⬍.05,
respectively.
3
Thus, our measure of the desire for vengeance seems
to capture both the desire to retaliate reactively following a prov-
ocation and a willingness to proactively plan and calculate an
aggressive act.
Bad-person attributions. To test whether the effect of an
entity theory on the desire for vengeance toward an acquaintance
was mediated, at least in part, by its effect on bad-person attribu-
tions (a global negative belief about the moral character of the
acquaintance; Weiner, 1986, 1995), we asked participants to an-
swer “What do you think about this acquaintance?” on a fully
labeled 7-point scale (1 ⫽an extremely good person,4⫽really
can’t say if they are good or bad, 7⫽an extremely bad person).
Bad feelings about the self. To test whether the effect of an
entity theory on the desire for vengeance was also mediated by its
effect on bad feelings about the self, we asked participants, “How
badly did you feel about yourself after this incident?” (1 ⫽not at
all bad,5⫽extremely bad). In a pilot study, we found that this
item was significantly correlated with shame following a recalled
conflict with an acquaintance, r⫽.56, N⫽164, p⬍.05.
Descriptive statistics, correlations, and reliabilities for these survey
measures are reported in Table 1.
Participant characteristics. As noted, participants were asked
to indicate the highest level of education their mothers had re-
ceived (1 ⫽less than high school, 2⫽high school degree, 3⫽
some college, 4⫽college degree, 5⫽graduate degree [master’s,
doctor, lawyer, or PhD]) and what grades they usually made in
school (0 ⫽D average or below, .3366 ⫽C average, .6633 ⫽B
average, 1⫽A average).
We explored the possibility that more frequently victimized
adolescents would be more likely to believe that people cannot
change and would also be more likely to recall conflicts about
which they harbored vengeful thoughts. To control for the poten-
tial influence of victimization on both an entity theory and the
desire for vengeance, participants were also asked to indicate how
often they were victims of 12 forms of peer aggression, such as
“called mean or hurtful names,” “the subject of false rumors or
lies,” “had mean things sent by e-mail from lots of people,” and
2
Note that in some previous research (Dweck, Chiu, & Hong, 1995)
implicit theories were analyzed dichotomously. However, some more
recent guidelines (MacCallum, Zhang, Preacher, & Rucker, 2002) suggest
that such measured variables are best analyzed continuously. Therefore, we
analyzed implicit theories continuously whenever they were measured.
3
As in some previous investigations (Raine et al., 2006), we also
examined whether desire for vengeance was more strongly related to one
or the other type of aggression, controlling for the alternative type in a
regression. In these regressions, we found that the coefficient predicting the
effect of desire for vengeance on proactive aggression was not significantly
different from the coefficient predicting reactive aggression, F(1, 29) ⫽
0.09, ns, suggesting that desire for vengeance was equally predictive of
both forms of aggression.
1094 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

“were hit, slapped, or pushed.” The 12 kinds of victimization were
adapted from an existing measure of bullying (Olweus, 1993), and
the word bullying was not used. Participants rated how often they
experienced each category (1 ⫽never,5⫽all the time), and these
ratings were combined into a single scale (␣⫽.91) by taking their
unweighted average, with higher values corresponding to more
victimization.
Results and Discussion
Analysis plan. To test our primary hypothesis about the
effects of the implicit theories on the desire for vengeance, we built
a latent variable structural model, using Amos 17 (Arbuckle,
2006).
4
We next expanded this model to test for the effects of our
potential mediators. We then used multigroup analyses to test
whether the mediation model was the same across different groups
of participants (i.e., subsample, sex, and victimization). We did
this by constraining the paths of interest to be equal across groups
and then tested whether these constraints significantly reduced
model fit by examining the changes in the chi-square fit statistic.
To assess the impact of each individual mediator, we used the
multiple mediation macro from Preacher and Hayes (2008). Note
that a cutoff of p⬍.05 was used to determine the significance for
all statistics throughout the studies.
Does an entity theory predict a greater desire for ven-
geance? In a latent variable model combining both subsamples,
Finnish and American adolescents holding more of an entity
theory desired more vengeance after their recalled peer conflicts,
as predicted, ␤⫽.43, (unstandardized B⫽.51; 95% CI [.36, .66],
p⬍.05; also see Table 1 for a comparable zero-order correlation).
This model provided adequate fit to the data, ␹
2
(60) ⫽129.208,
p⬍.05, ␹
2
/df ⫽2.153, comparative fit index [CFI] ⫽.968,
root-mean-square error of approximation [RMSEA] ⫽.057. Con-
trolling for time since the incident (both linear and log-
transformed), levels of victimization, or the type of conflict did not
influence the relationship between implicit theories and the desire
for vengeance. Thus, we found support for our primary hypothesis
that holding more of an entity theory would be related to a greater
desire for vengeance after a variety of interpersonal conflicts.
Why do adolescents with more of an entity theory have a
desire for vengeance? As predicted, in the combined sample of
Finnish and American youths, an entity theory was significantly
related to bad-person attributions and bad feelings about the self
after the recalled peer conflict, ␤⫽.17 and ␤⫽.20, respectively
(see also column 1 of Table 1), and these variables were signifi-
cantly related to the desire for vengeance, ␤⫽.40 and ␤⫽.17,
respectively, (see also column 2 of Table 1). To formally test for
mediation by these variables, we built the multiple-group (Finland
vs. United States) latent variable structural model presented in
Figure 1, ␹
2
(170) ⫽233.146, p⬍.05, ␹
2
/df ⫽1.37, CFI ⫽.973,
RMSEA ⫽.032.
Tests of generalizability of effects of implicit theories. First,
we tested the generalizability of the effects of implicit theories
across subsamples. To do so, we compared our final model, which
constrained each of the paths between focal variables in Figure 1
to be equal across subsamples, to a model in which none of these
five paths were constrained across subsamples, ␹
2
(165) ⫽
231.369, p⬍.05, ␹
2
/df ⫽1.40, CFI ⫽.973, RMSEA ⫽.034. Our
final model did not have a significantly lower model fit than the
unconstrained model, ⌬␹
2
(5) ⫽1.78, ns, suggesting that the
strength of associations between implicit theories, bad person
attributions, bad feelings about the self, and the desire for ven-
geance were not different in Finnish versus U.S. adolescents. The
same null findings emerged from tests for moderation by sex in a
multigroup analysis, ⌬␹
2
(5) ⫽1.90, ns, and by victimization status
(dichotomized at the median, 1.5), ⌬␹
2
(5) ⫽4.2, ns
5
Tests of mediation. In our final model, the association of an
entity theory with the desire for vengeance was reduced from ␤⫽
.43 (in a model with no mediators) to ␤⫽.32 (B⫽.21; 95% CI
[.13, .29], p⬍.05) when these two mediators were entered into the
model (Figure 1). This was a significant reduction in effect size
(indirect effect B⫽.08; 95% CI [.03, .15], p⬍.05, ␤⫽.11). Of
interest, the main effect of an entity theory on the desire for
vengeance was still significant when controlling for these media-
tors, suggesting that there was additional variance in this relation-
ship that was not yet accounted for.
We next tested whether the significant reduction in effect size
was due to one or both of the possible mediators, using Preacher
and Hayes’s (2008) macro. We found that both were significant
mediators. The reduction in the effect of an entity theory on the
desire for vengeance due to bad-person attributions was significant
(B⫽.05; 95% CI [.01, .11], p⬍.05, ␤⫽.07). The reduction in
effect size due to bad feelings about the self was also significant
(B⫽.03; 95% CI [.01, .07], p⬍.05, ␤⫽.04). Thus, an entity
theory exerted indirect effects on the desire for vengeance partially
through both bad-person attributions and bad feelings about the
self.
We also explored the alternative hypothesis that these findings
were obtained because those with an entity theory had less of the
cognitive ability required to make the perhaps more complex
situational attributions about transgressors, as opposed to trait or
4
All of the findings reported with latent variable structural models were
replicated with least squares regressions, and these results are available on
request from the first author.
5
In this study and in Study 2, the null findings regarding moderation by
participant characteristics were replicated when conducting regressions
that tested the significance of Participant Characteristic ⫻Implicit Theo-
ries interaction effects on the dependent measures or mediators, controlling
for main effects of these variables. In additional regression analyses, we
also found similar null findings regarding moderation by victimization
when victimization was treated as a continuous variable as opposed to a
dichotomous variable.
Table 1
Descriptive Statistics and Correlations for Study 1 Measures
Measure MSD␣Range 1 2 3 4
1. Entity theory (bullies and
victims) 3.21 1.06 .82 1–6 —
2. Desire for vengeance 2.87 0.94 .92 1–7 .38
ⴱ
—
3. Bad-person attributions 3.92 1.83 — 1–7 .19
ⴱ
.49
ⴱ
—
4. Bad feelings about the self 2.60 1.36 — 1–5 .18
ⴱ
.34
ⴱ
.26
ⴱ
—
Note. ␣⫽internal consistency reliability. Rating scale for Measure 1:
1⫽strongly disagree,6⫽strongly agree; Measure 2: 1 ⫽not at all,7⫽
an extreme amount; Measure 3: 1 ⫽an extremely good person,7⫽an
extremely bad person; Measure 4: 1 ⫽not at all bad,5⫽extremely bad.
ⴱ
p⬍.05.
1095
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

bad-person attributions. We did this by controlling for two differ-
ent proxies for cognitive ability: grade point average and mother’s
education level. In past research, these proxies have consistently
predicted IQ (e.g., Bee et al., 1982; Sirin, 2005; Walker, Green-
wood, Hart, & Carta, 1994). Neither of these controls reduced the
effect of the entity theory on the desire for vengeance or either of
the mediators, consistent with the idea that implicit theories of
personality are not a by-product of cognitive ability or exposure to
more sophisticated thinking.
Summary. Overall, Study 1 found support for our hypothesis
that an entity theory would significantly predict the desire for
vengeance after a recalled incident of peer conflict. Adolescents
who endorsed the idea that people cannot change held more
festering resentment, in that they wished that extreme harm and
punishment would befall the peers who had transgressed against
them. Study 1 also began to build a theoretical account that is
consistent with previous work that has focused on the role of
attributions and negative feelings in response to social challenges
(Chiu, Hong, & Dweck, 1997; Erdley & Dweck, 1993; Graham &
Juvonen, 1998; Thomaes et al., 2008). This study extended previ-
ous work by showing that an entity theory was associated with
both self-relevant feelings and other-relevant social cognitions.
Previous research on attributions has posited that people tend to
make either internal or external attributions after an event (Weiner,
1995). Yet, our research suggests that both internal and external
trait attributions may result from the same overall entity theory
(see also Erdley & Dweck, 1993; Erdley et al., 1997).
Study 1 also showed that the processes relating an entity
theory to the desire for vengeance were replicated when the
participants were a sample of students in Finland and when they
were a diverse group of students attending an urban public high
school in the United States. An explicit test of the differences in
effect sizes between the two samples did not support a cultural
difference hypothesis. Thus, the present study not only demon-
strated that the same social–cognitive process occurred in two
highly different samples—one in Finland and one in the United
States—but also showed that the magnitudes of these associa-
tions were not different.
Study 2
Although Study 1 provided convergent results across sub-
samples, these results were limited by the lack of control over the
incidents that participants reacted to, because each adolescent
wrote about a different personal incident. For this reason, we could
not rule out the possibility that Study 1’s results were obtained
because those with more of an entity theory actually experienced
or recalled more serious events—even though our coding of these
events did not detect such differences. Therefore, in Study 2 we
tested whether an entity theory predicted the desire for vengeance
after an experimenter-controlled hypothetical incident of direct
victimization. In addition, because this study’s sample was racially
and ethnically diverse, we were able to test whether the relation-
ship between an entity theory and the desire for vengeance varied
across racial/ethnic groups.
Methods
Participants. Participants were a diverse group of students in
Grades 9 and 10 (N⫽314) attending a public high school in San
Jose, California. Of the participants, 57% were Hispanic/Latino,
10% were Asian/Asian American, 9% were Black/African Amer-
ican, 17% were White, non-Hispanic, and 7% indicated another
race/ethnicity; 55% were boys and 45% were girls. Nearly all
participants received free or reduced-price lunch. Among parents,
4% had a graduate degree, 18% had a college degree, 26% had
Figure 1. Multiple partial mediation model in Study 1: Entity theory exerts indirect effects on the desire for
vengeance through bad-person attributions and bad feelings about the self. Numbers are standardized coeffi-
cients. Circles represent latent variables, and squares represent manifest variables. For indirect effect of entity
theory on vengeance, ␤⫽.11, p⬍.05. All coefficients shown were significant, p⬍.05. For this multigroup
model (United States vs. Finland), parameters for the paths depicted were constrained across groups.
1096 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

some college education (but no degree), and the rest had a high
school degree or less. Some students (n⫽17) did not complete at
least one of the measures on the survey; therefore, their responses
were multiply-imputed as described earlier.
Procedure. All of the students in Grades 9 and 10 in six
classes were invited to participate in this study, and 75% provided
parental consent and assented. Participants completed the survey
during school hours. After completing the study, students were
thanked for their time and debriefed.
Measures. Implicit theories measures identical to those used
in Study 1 were administered (M⫽2.76, SD ⫽1.06, ␣⫽.83).
The survey also included a measure of current victimization (␣⫽
.89) and of the desire for vengeance after a hypothetical bullying
scenario.
Participants read a hypothetical incident of bullying. Then, to
elicit a more authentic emotional response to the scenario, they
wrote an essay describing how they would feel and what they
would feel like doing if they experienced this incident. The stim-
ulus scenario was as follows:
One day you found out that two acquaintances from your grade saw
you make a fool out of yourself at school. At first you tried to ignore
them, but over the next few weeks they called you a “loser” every time
they saw you. Now they make fun of what you’re wearing and they
spread rumors about you to your friends. It keeps happening every
single day. EVERY day. Finally, now they’re threatening to make fun
of you on MySpace and Facebook.
After reading the scenario and writing a brief essay in response
to the prompt “If the story had actually happened to you, how
would you feel?” participants were asked “How much would you
feel like doing each of these things to the people who did those
things to you?” Participants then responded to the same seven
items measuring the desire for vengeance used in Study 1 (M⫽
2.93, SD ⫽1.91, range: 1–7; ␣⫽.94). Last, they answered some
positively worded filler items and were asked to write a short essay
describing “what makes you really happy in life.”
Results and Discussion
An entity theory was significantly related to a greater desire for
vengeance after the hypothetical bullying scenario in a latent
variable structural model, ␤⫽.33 (B⫽.53; 95% CI [.31, .75], p⬍
.05), and the model provided adequate fit to the data, ␹
2
(44) ⫽
98.425, p⬍.05, ␹
2
/df ⫽2.237, CFI ⫽.977, RMSEA ⫽.062.
Next, in separate multigroup analyses, this association was not
significantly moderated by victimization status (dichotomized at
the median, 1.67), ⌬␹
2
(1) ⫽0.01, ns, or by participant sex,
⌬␹
2
(1) ⫽0.90, ns. To test for differences in this relationship
across racial/ethnic groups, we built a structural model with five
groups (Black, Latino, Asian, White/non-Hispanic, and other race/
ethnicity). We then constrained the path from an entity theory to
the desire for vengeance to be equal across all five groups, and this
did not reduce model fit, ⌬␹
2
(4) ⫽2.7, ns, suggesting no moder-
ation by race/ethnicity. We also tested whether each group indi-
vidually differed from the White/non-Hispanic group, and none of
these constraints significantly reduced model fit, all ⌬␹
2
s(1) ⬍2,
ns. Thus, an entity theory was related to more vengeful desires
even when participants responded to an identical scenario of direct
victimization, and this was not moderated by sex, previous vic-
timization, or race/ethnicity.
Study 3
Studies 1 and 2 documented that implicit theories have mean-
ingful associations with a desire for vengeance after both peer
conflicts in general and victimization in particular, and they began
to map out the mediational pathways explaining this relationship.
However, those studies were limited in several ways.
First, they did not test whether changing implicit theories would
have the effect of reducing the desire for vengeance. Indeed, no
previous study has done so. Such an experiment is especially
important for translating the present research into an intervention
in the future. Therefore, in the present study, we tested whether
experimentally manipulating implicit theories would reduce the
desire for vengeance.
In Study 1, the particular negative emotions felt after victimiza-
tion were not specified; thus, it is unclear exactly which emotions
led to the effects we observed. Therefore, in the present study, we
sought to better specify the emotional mediators. To do so, we
measured hatred for the perpetrator and feelings of shame and
sadness about the self. Because leading accounts of shame suggest
that it arises from threats to one’s core self, whereas sadness is not
considered a self-conscious emotion (Tangney et al., 2007), we
expected that shame, and not sadness, would mediate the link
between implicit theories and vengeance.
In Study 1 we found that bad-person attributions and bad feel-
ings about the self only partially mediated the effect of an implicit
theory. Thus, we sought to extend our theoretical explanation of
why negative emotions might lead to vengeance by exploring a
new potential mediator. Some have suggested that people aggress
in part because they are motivated to regulate the negative emo-
tions that they have (Baumeister, Smart, & Boden, 1996; Bush-
man, Baumeister, & Phillips, 2001; Thomaes et al., 2008).
Baumeister et al. (1996) argued that “aggression can be regarded
as a crude technique of affect regulation . . . to avoid certain
negative emotional states such as shame, dejection, sadness and
disappointment with oneself (pp. 10–11),” and numerous investi-
gations have pointed to the importance of anticipated positive
emotional states in motivating aggression (for a review, see
Baumeister, Vohs, DeWall, & Zhang, 2007). In this view, the
belief that one will feel better is a motivational antecedent to
the desire to aggress and is a distinct construct. We therefore
conducted a preliminary investigation of this issue. We assessed
adolescents’ beliefs about whether imagining vengeance is a good
way to regulate negative emotions, and we determined whether
this would help further explain the relation between an entity
theory and the desire for vengeance.
Finally, we included more positive and prosocial reactions to the
conflicts and added more positive attributions to the attributional
measures. This allowed us to investigate whether learning an
incremental theory would lead participants to seek to educate
transgressors, as proposed by previous research with college stu-
dents (Chiu, Dweck, Tong, & Fu, 1997; Rattan & Dweck, 2010).
We also examined whether an incremental theory might reduce
trait attributions and increase alternative attributions, such as the
importance of the situation or a transgressor’s immaturity as
causes of bullying (cf. Chiu, Hong, & Dweck, 1997).
1097
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

Methods
Participants. Participants were a subset of the Finnish stu-
dents in Grades 9 and 10 from Study 1. Two hundred and two
students took the online experiment in their school’s computer lab.
Nearly all of the participants were White, and they ranged in age
from 14 to 16. There were nearly equal numbers of boys (52%) and
girls (48%). Some participants (n⫽15) did not complete the
dependent measure of desire for vengeance. Therefore, they were
excluded from the study. For those who completed the vengeance
measure but did not provide responses to other questions, missing
values were multiply imputed, yielding a final sample of N⫽187
for all analyses.
Procedure. The six schools from various regions in Finland
that provided the participants for the Study 1 sample were invited
to administer this experiment to the Study 1 participants through
the Internet. This allowed us to test whether the results of the
experiment were moderated by Study 1’s baseline measures.
All of the schools and 92% of the students agreed to participate.
Consent was obtained during Study 1, and assent was again ob-
tained for Study 3. On average, 20.3 days elapsed between partic-
ipation in Studies 1 and 3 (range ⫽7 to 37). Participants were not
told that the Study 1 and Study 3 surveys would be related, nor
were they given information that would reveal any experimental
hypotheses after Study 1.
The survey administration software randomly assigned partici-
pants to either a control article group or to an incremental article
group. As a part of an online “reading comprehension task,”
participants in the incremental group read an article, purportedly
from Psychology Today, concluding that people can change who
they are. The style of the article was modeled after teen self-help
articles that are popular in magazines aimed at adolescents. In the
incremental article, a protagonist (whose sex was matched to the
participant’s sex) was bullied in school in a manner similar to
the scenario used in Study 2: The protagonist did something
embarrassing, and then some peers created a mean nickname that
they circulated to many people and threatened to post on the
Internet. Next, the protagonist received advice from a peer, a
soccer coach, and a school counselor. Each provided the protag-
onist with the message that people (both the protagonist and the
bullies) could change the kind of person they were, for example,
“People’s characteristics are changeable and can be influenced
over time.” Only beliefs about change were discussed in the
article. The incremental article did not endorse one course of
action or another (i.e., vengeance vs. forgiveness), and participants
were never told what specific attribution to make to explain their
victimization or what emotion they should or should not feel.
Participants in the control group were given an article with an
identical story about the protagonist being victimized but without
any mention of whether people can change. We did not assign
control participants to read an entity article because of ethical
considerations. To equate the experience as much as possible
across the groups and to support the cover story of a reading
comprehension task, all participants wrote detailed essays summa-
rizing their respective articles (“Please summarize the article in
your own words in 3–5 sentences. What happened? What was the
main idea?”) and communicating the thoughts and feelings they
had about the peer victimization scenario (“Imagine that all the
events in the article actually happened to you. How would you
feel?”) before completing any of the dependent measures. Finally,
participants completed the dependent measures and manipulation
checks. At the end of the study, participants answered another
series of positively valenced filler items to counteract the relatively
negative nature of the dependent measures. They were debriefed
and thanked for their participation, and then they returned to class.
Dependent measures. Means, standard deviations, and cor-
relations for the measures used in the present study are presented
in Table 2.
Manipulation checks. To test whether the article had the
intended effect on implicit theories, two kinds of manipulation
check items were included at the very end of the survey. The first
set of items asked participants to rate three statements about
whether people could change: “How much do you think other
people’s personalities can change?” “How much do you think
other people can change the kind of person they are?” “How much
do you think bullies can change the way they are?” Ratings were
made on a 5-point scale (1 ⫽not at all,5⫽a great deal). These
three items were collapsed into a single scale (␣⫽.77) with higher
scores corresponding to a more incremental theory.
Next, on the basis of findings from previous research (Erdley &
Dweck, 1993), another manipulation check asked participants to
make predictions about the consistency of traits for a new person.
We reasoned that if participants’ implicit theories had become
more incremental, they would make fewer dispositional attribu-
tions about targets and, therefore, would expect them to behave
less consistently. We measured trait consistency predictions by
administering an item similar to those used in previous research
(Chiu, Hong, & Dweck, 1997; Kunda & Nisbett, 1986) to measure
individual differences in the tendency to expect that people always
act in the same way. This item first asked participants to “Imagine
that one day you observed two students at your school, Mary and
Alicia, and on that day you noticed that Mary was MEANER than
Alicia in school.” Next, participants answered the following ques-
tion: “How likely is it that Mary would also be MEANER than
Alicia in a very different situation OUTSIDE of school?” on a
fully labeled 5-point scale (1 ⫽not at all likely,5⫽extremely
likely).
Desire for vengeance. We measured participants’ desire for
vengeance with the same rating-scale items as we used in Study 1
for the recalled conflict and in Study 2 for the hypothetical con-
flict. We administered these items after participants had read the
bullying article, imagined that the events in the article had hap-
pened to them, and described their feelings in an essay format. The
seven items measuring the desire for vengeance were again reli-
able in this sample (␣⫽.90). In support of the validity of this
measure, we found that the desire for vengeance after a recalled
conflict at baseline (i.e., from Study 1) was strongly correlated
with the desire for vengeance following the bullying scenario, r⫽
.68, p⬍.05. In addition to these relatively negative items, partic-
ipants also rated neutral or prosocial items, such as “just ignoring
the person,” “forgetting about it and getting over it,” “forgiving
them eventually,” “helping them see that what they did was
wrong,” and “helping them act better in the future.” Neither set of
items formed a common factor; therefore, each item was analyzed
individually.
As a forced-choice measure of vengeful intentions, participants
were asked to indicate which one of the 12 options listed (i.e., the
seven vengeful items and the five prosocial or neutral items) they
1098 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

would “be most likely to ACTUALLY do.” Responses to this
question were converted to a dichotomous item of vengeful versus
prosocial or neutral responses to conflict (1 ⫽vengeful,0⫽
prosocial or neutral;M⫽.39, SD ⫽.49).
Bad-person attributions. We expanded the Study 1 measure
by allowing participants to choose one attribution from among
several, using a forced-choice question, as opposed to rating only
the belief that the acquaintance was a bad person. The new
question first asked, “What do you think is probably the biggest
reason why the two people treated the main character that way?”
Respondents then chose among several attributions, some of which
were characterological, including “They’re just bad people” and
“They’re bullies and they were doing what bullies do.” Others
were situational, environmental, or developmental attributions,
including “They’re trying to feel better about themselves,” “They
have problems at home,” or “They’re going through an ‘immature’
phase in life.” Participants could also provide “some other reason”
and then write their own attributions in an open-ended question;
fewer than 7% did so. The forced-choice and open-ended re-
sponses were subsequently recoded for analyses (1 ⫽character-
ological bad-person attributions,0⫽other attributions;␬⬎.85
for open-ended attributions).
Emotions. To examine specific emotions, participants indi-
cated to what extent they would feel sadness and to what extent
they would feel shame. In addition, in the present study, we
separated the attribution that the bullies were bad people from the
emotion of hatred for the bullies. Thus, participants were first
asked, “How much would you feel each of the following ways?”
and then were asked to rate these three emotions—hatred for the
transgressor, shame, and sadness—on a fully labeled 7-point scale
(1 ⫽not at all,7⫽an extreme amount).
Vengeance as an emotion-regulation strategy. To begin to
explore whether imagining vengeance might function as an
emotion-regulation strategy, we asked participants to rate how
much better they would feel if they did each of the following:
“Imagining them getting punished or hurt” or “Planning out some
revenge against them.” Responses were given on a fully labeled
7-point scale (1 ⫽not at all better, maybe worse,7⫽extremely
better). These two items were averaged, with higher values corre-
sponding to a greater belief that vengeance would make one feel
better.
Baseline entity theory and other moderators. Participants’
scores on the entity theory scale reported in Study 1 were linked to
their Study 3 data. Other Study 1 measures were also linked to
participants’ Study 3 data, including ratings of the frequency of
various bullying actions toward them, desire for vengeance toward
an acquaintance after a conflict (measured about 3 weeks earlier in
Study 1), and sociodemographic variables such as sex, mother’s
education, grade level and age.
Results and Discussion
Preliminary analyses.
Effectiveness of randomization. Randomization was effec-
tive. Assignment to the incremental or control group was indepen-
dent of baseline entity theory in Study 1, t(185) ⫽⫺0.72, ns,
desire for vengeance in Study 1, t(185) ⫽0.70, ns, frequency of
being bullied in school (treated continuously), t(185) ⫽0.58, ns,
sex, ␹
2
(1, N⫽187) ⫽0.84, ns, mother’s education, ␹
2
(5, N⫽
187) ⫽0.52, ns, grade level, ␹
2
(1, N⫽187) ⫽0.82, ns, and
school attended, ␹
2
(6, N⫽187) ⫽1.88, ns.
Manipulation checks. We found that the incremental article
led participants to believe that people could change, relative to
those who read the control article: for the incremental article, M⫽
3.19, SD ⫽0.92; for the control article, M⫽2.80, SD ⫽0.88;
t(185) ⫽2.89, p⬍.05, d⫽.43.
6
The incremental theory also
reduced participants’ predictions of people’s behavioral consis-
tency in a scenario that was unrelated to the stimulus scenario,
relative to the control group: for the incremental theory, M⫽2.79,
SD ⫽0.93; for the control theory, M⫽3.09, SD ⫽0.94; t(185) ⫽
2.83, p⬍.05, d⫽.42.
7
This result is consistent with the conclu-
sion that participants’ beliefs that people are fixed were reduced by
the incremental article, because it reduced participants’ predictions
7
Throughout our discussion of Study 3’s results, we use the term
reduced to refer to the difference between the incremental group and the
control group, or the potential outcome for the incremental participants had
they not received the treatment.
6
Although the difference between the means for this manipulation check
may appear modest, the standardized effect size was in fact moderate. It is
also important to note that the manipulation check showed a substantive
difference between conditions. The control group had an average score
below the midpoint, whereas the incremental paragraph group had an
average above the midpoint, suggesting they held more of an incremental
theory. In any case, a modest change in implicit theories would work
against finding the hypothesized effect on vengeance.
Table 2
Descriptive Statistics and Correlations for Study 3 Measures
Measure MSD␣Range 1 2 3 4 5 6 7
1. Incremental theory article 0.46 0.50 — 0–1 —
2. Desire for vengeance 2.90 1.28 .89 1–7 ⫺.23
ⴱ
—
3. Bad-person attributions 0.29 0.46 — 0–1 ⫺.19
ⴱ
.12 —
4. Hatred 4.49 1.90 — 1–7 ⫺.17
ⴱ
.54
ⴱ
.08 —
5. Shame 3.89 1.95 — 1–7 ⫺.15
ⴱ
.24
ⴱ
⫺.09 .35
ⴱ
—
6. Sadness 4.10 1.85 — 1–7 ⫺.03 .18
ⴱ
⫺.15
ⴱ
.45
ⴱ
.64
ⴱ
—
7. Vengeance as an emotion-regulation strategy 2.85 1.62 .83 1–7 ⫺.14
ⴱ
.62
ⴱ
.23
ⴱ
.47
ⴱ
.06 .02 —
Note. Rating scale for Measures 1 and 2: 1 ⫽not at all,7⫽an extreme amount; Measure 3: 1 ⫽bad-person attributions, 2⫽other attributions; Measures
4, 5, and 6: 1 ⫽not at all,7⫽an extreme amount; Measure 7: 1 ⫽not at all better, maybe worse,7⫽extremely better.
ⴱ
p⬍.05.
1099
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

that behaviors such as meanness would be manifested consistently
across very different situations.
Did an incremental theory reduce the desire for vengeance?
Recall that the primary objective of the present research was to test
whether learning an incremental theory about oneself and about
bullies would reduce the desire to take revenge on perpetrators. We
found that this change in mind-set did, in fact, reduce vengeance. The
incremental article significantly reduced ratings of the desire for
vengeance: for the incremental article, M⫽2.56, SD ⫽1.19; for the
control article, M⫽3.16, SD ⫽1.25; t(185) ⫽3.23, p⬍.05, d⫽.48.
The incremental article did not affect the ratings of neutral or proso-
cial strategies. This may be because our article did not explain that
they personally could play a role in setting the process of personality
change in motion, which has been found to be a key component in
boosting the motivation to confront or educate a transgressor (e.g.,
Rattan & Dweck, 2010). However, for the vengeful intentions mea-
sure, the incremental article substantially cut the proportion of partic-
ipants saying that they would be most likely to use vengeance as
opposed to neutral or prosocial strategies: for the incremental article,
M⫽32%;for the control article, M⫽53%; logistic B⫽⫺.81, Z⫽
⫺2.63, p⬍.05, odds ratio ⫽.44.
8
None of these effects were moderated by sex, age, frequency of
being bullied (dichotomized at the median or treated continuously), or
desire for vengeance toward an acquaintance reported 3 weeks earlier.
That is, in separate regressions predicting desire for vengeance with a
variable indicating condition, the moderator, and the Condition ⫻
Moderator interaction, all interactions were nonsignificant. In addi-
tion, controlling for these variables did not change the size or signif-
icance of the effect of the incremental theory. Thus, the incremental
article, with its endorsement of change and growth, reduced students’
ratings of their desire for vengeance and increased their choice of
neutral/prosocial versus vengeful responses. Moreover, it did so sim-
ilarly across groups of participants.
Why did learning an incremental theory reduce vengeance?
The second objective of the present research was to continue to
build a theory of why an incremental theory might affect responses
to conflict. We predicted that a shift in mind-set toward a more
malleable view of personality would reduce bad-person attribu-
tions, shame, hatred, and vengeful emotion regulation and that
these, in turn, would reduce the actual desire to use strategies such
as imagining or planning revenge. Next, we present results dem-
onstrating the direct effect of the incremental theory article on each
of the hypothesized mediators and then the tests of mediation.
Bad-person attributions. Participants learning an incremental
theory were half as likely to make bad-person attributions: for the
incremental theory, M⫽19%; for the control theory, M⫽38%;
logistic B⫽⫺.87, Z⫽⫺2.38, p⬍.05, odds ratio ⫽.42. Instead,
those who learned an incremental theory explained the bullies’
actions with other, more malleable factors—such as the bullies’
wanting to feel better, having a bad home life, or going through an
immature phase in life. Thus, the incremental theory made the
participants see the bullies as people with their own motivations—
however flawed those motivations may be—as opposed to simply
placing a pejorative label on them.
Emotions. The incremental article, as hypothesized, signifi-
cantly reduced feelings of hatred for the perpetrators: for the
incremental article, M⫽4.11, SD ⫽1.85; for the control article,
M⫽4.76, SD ⫽1.89; t(185) ⫽2.38, p⬍.05, d⫽.35. The
incremental article also significantly reduced feelings of shame
toward the self: for the incremental article, M⫽3.51, SD ⫽1.78;
for the control article, M⫽4.14, SD ⫽2.04; t(185) ⫽2.25, p⬍
.05, d⫽.31. It did not, however, reduce feelings of sadness after
the bullying scenario: for the incremental article, M⫽4.04, SD ⫽
1.70; for the control article, M⫽4.14, SD ⫽2.00; t(185) ⫽0.37,
ns, This latter result was also predicted, because sadness can result
from many sources (e.g., sympathy) and is not an emotion that is
expected to stem primarily from an entity theory, as hatred and
shame are (Tangney et al., 2007). Hence, sadness was not included
in the mediational model discussed later.
Vengeance as an emotion-regulation strategy. The incre-
mental article, in addition, significantly reduced how much better
participants thought they would feel after imagining revenge: for
the incremental article, M⫽2.58, SD ⫽1.42; for the control
article, M⫽3.05, SD ⫽1.74; t(185) ⫽2.02, p⬍.05, d⫽.29.
Mediational model. To formally test for mediation by these
variables, we first used Preacher and Hayes’s (2008) macro to test
the model presented in Figure 2, which included all four mediators.
We found that the effect of the incremental theory on vengeance
was reduced from ␤⫽⫺.23, p⬍.05, in a model with no
mediators to ␤⫽⫺.10 (B⫽⫺.17, 95% CI [⫺.46, .12], ns)inthe
model presented in Figure 2, suggesting full mediation by these
variables (indirect B⫽⫺.42, 95% CI [⫺.72, ⫺.15], p⬍.05, ␤⫽
⫺.13). Of interest, when this same analysis was conducted without
including emotion regulation as a mediator, only partial mediation
was found. Thus, by expanding our set of mediators to include
emotion regulation, and by specifying the negative emotions felt
after victimization (shame and hatred), we were able to fully
account for the effects of implicit theories on the desire for
vengeance.
However, when we built the latent variable structural model
presented in Figure 2, we found that it was a very poor fit to the
data,␹
2
(49) ⫽143.75, ␹
2
/df ⫽3.04, p⬍.05, CFI ⫽.884,
RMSEA ⫽.108, suggesting that there is a more complicated chain
of mediation involving these variables. Yet, the present study’s
design, with its concurrent self-report measures, does not allow for
firm conclusions regarding the temporal ordering of the mediators.
Therefore, we briefly describe an exploratory analysis that is
suggestive of one possible order, and we propose it as a guide for
future investigations.
Specifically, we propose that emotion regulation is a later step
in the path from an incremental theory to vengeance, with attri-
butions and emotions as antecedents. That is, judgments and
emotions may need to be generated before they are regulated.
On the basis of this theory, we tested the model presented in
Figure 3. We found that it provided adequate fit to the data,
␹
2
(48) ⫽81.821, p⬍.05, ␹
2
/df ⫽1.705, CFI ⫽.961, RMSEA ⫽
.064, and yielded a significant indirect effect of the incremental
theory on vengeance (B⫽⫺.32, 95% CI [⫺.55, ⫺.15], p⬍.05,
␤⫽⫺.13). In addition, the direct effect of the incremental theory
8
To test the robustness of this result, we also recoded prosocial and
neutral responses into separate categories (0 ⫽vengeful,1⫽neutral,2⫽
prosocial) and analyzed the data with a multinomial logistic regression in
which neutral and prosocial responses were each separately compared with
vengeful responses. We found a similar pattern of results: An incremental
theory significantly predicted the desire to use prosocial or neutral re-
sponses relative to vengeful responses.
1100 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

on vengeance was reduced to being no longer significant (B⫽
⫺.27, 95% CI [⫺.56, .01], p⫽.056, ␤⫽⫺.11). In this model, an
incremental theory reduced participants’ feelings of shame after
victimization, which in turn reduced their hatred for the bullies
(indirect effect of incremental theory on hatred, B⫽⫺.19, 95% CI
[⫺.45, ⫺.01], p⬍.05, ␤⫽⫺.05).
9
In addition, the incremental
theory reduced bad-person attributions, which in turn reduced their
vengeful emotion-regulation strategies (total indirect effect of in-
cremental theory on emotion regulation, B⫽⫺.39, 95% CI [⫺.69,
⫺.14], p⬍.05, ␤⫽⫺.12). Finally, vengeful emotion-regulation
strategies significantly mediated the effect of both hatred and
bad-person attributions on the desire for vengeance (indirect effect
of bad-person attributions on vengeance, B⫽.25, 95% CI [.06,
.52], p⬍.05, ␤⫽.09; indirect effect of hatred on vengeance, B⫽
.15, 95% CI [.09, .23], p⬍.05, ␤⫽.22; see Figure 3).
10
It is interesting to note that the effect of the incremental
theory on hatred was mediated by feelings of shame. This
account is consistent with established sociological theories on
shame, which suggests that shame arises from a worry about
one’s permanent status in the community. In this view, shamed
victims are motivated to derogate and hate those who threaten
that status (Scheff, 1994; Scheff & Retzinger, 1991; see also
Gilbert & Andrews, 1998). This finding also resonates with
psychological research on hatred. For example, Fitness and
Fletcher (1993) found that hatred, relative to anger, more fre-
quently followed from incidents of humiliation. In addition,
experimental investigations have shown that leading adoles-
cents to feel ashamed can increase their motivation to behave
aggressively (Reijntjes et al., in press; Thomaes et al., 2008).
The present study’s findings contribute to this research by
showing that an incremental theory can reduce shame and
subsequently hatred by changing the meaning of peer victim-
ization, so that the conflict is perceived as a less permanent (and
hence less humiliating) threat to a victim’s status.
Internal replication of the causal effect. Last, we attempted
to internally replicate our experiment’s primary findings within
our control group by examining whether changes over time in
implicit theories from Study 1’s survey to the present study’s
survey significantly predicted changes in the desire for vengeance.
To do so, we first created a change in implicit theories measure by
subtracting the zscore for the implicit theories at baseline from the
zscore for the implicit theories manipulation check in the present
study, and we coded it so that higher values corresponded to more
of an incremental theory. We then used this change score to predict
changes in desire for vengeance, as indexed by the zscore for
baseline vengeance subtracted from the zscore for the desire for
vengeance in the present study, so that higher values corresponded
to more vengeance. In this regression analysis, changes toward
holding more of an incremental theory significantly predicted
decreases in desire for vengeance, B⫽⫺.34, SE ⫽.14, p⬍.05,
␤⫽⫺.24, replicating our experiment’s main finding with a
different methodology.
Summary. Overall, these results suggest that when adoles-
cents are taught that people can change, they are not as likely to
think that bullies are simply bad people, they do not feel as
ashamed after victimization, they do not hate the bullies as
much, and they are less motivated to imagine vengeance to
regulate their negative emotions. In turn, they are less likely to
say they would like to respond to the conflict in an aggressive,
vengeful way.
Our overall hypothesis that implicit theories led to the differ-
ences we observed in this study was supported by an internal
replication with short-term longitudinal associations and by the
finding that the article changed implicit theories, predictions of
trait consistency, and the key mediators that we found in our
previous studies.
General Discussion
Adolescents frequently experience peer conflicts, ranging from
minor slights to outright victimization, and they respond to those
conflicts in various ways. Sometimes adolescents seek prosocial
ways to handle the conflict and move on. Other times, adolescents
hold enduring grudges toward their transgressors and wish them
harm. What predicts which type of response adolescents prefer? In
the present research, we propose that the meaning adolescents
make of their social conflicts, arising from their implicit theories of
personality, can shed light on how they cope with social setbacks.
Three studies found that implicit theories can predict how ado-
lescents in high school respond to conflicts with their peers. In
Study 1, we found that an entity theory, or a belief in fixed traits,
was related to a greater desire for vengeance when adolescents
recalled recent conflicts in their lives. In Study 2, we replicated the
results, using a vivid hypothetical incident of victimization. In an
experiment, Study 3 showed that learning an incremental theory,
or a belief in the potential for change, could reduce the desire for
revenge after hypothetical peer victimization.
The relation between implicit theories and the desire for ven-
geance was consistent and robust. This association was clear
across very different samples: a socioeconomically diverse, mostly
9
The total indirect effect of shame on desire for vengeance was B⫽.11,
95% CI [.06, .17], p⬍.05, ␤⫽.18, and the indirect effect of shame on
vengeful emotion regulation was B⫽.13, 95% CI [.07, .21], p⬍.05,
␤⫽.15.
10
Because this mediational analysis was exploratory, we built alterna-
tive models in which we changed the order of the mediators or added/
removed paths. We tested whether the fit of these models was lower or
higher by comparing the ␹
2
/df, CFI, RMSEA, Akaike information crite-
rion, and Bayesian information criterion fit statistics relative to the theo-
retical model in Figure 3. Our final model had better fit with regard to each
of these five statistics relative to any alternative model we tested, although
the magnitude of these differences was often quite small. For example,
model fit was lower when the desire for vengeance was not the final
variable in the model, when bad-person attributions were placed after
emotion regulation, and when shame was placed after either hatred or
emotion regulation. Because the model in Figure 3 provided better fit to the
data than all other models tested, and because theory led us to predict that
emotion regulation would come after the emotions were generated, we
have presented the model shown in Figure 3. It should be noted that the
final model in Figure 3, which includes ordering of the mediators, fit the
data similarly to a model with covariances between the mediators and no
ordering, ␹
2
(46) ⫽79.330, p⬍.05, ␹
2
/df ⫽1.725, CFI ⫽.961, RMSEA ⫽
.065. We chose a final model that ordered the mediators because, as noted
earlier, we thought of attributions and emotions as prior processes to the
evaluation of strategies for regulating those emotions, and the ordering of
the mediators provides more direction for future empirical and theoretical
studies.
1101
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

White sample from Finland and two racially and ethnically diverse
low-income samples from urban schools in the United States Of
interest, when we explicitly tested the difference in effect size
between the homogeneous Finnish subsample and the diverse U.S.
subsample in Study 1, we found no evidence that the effect of
implicit theories on adolescents’ desire for vengeance (or on the
mediators of that relationship) was stronger or weaker in either
subsample. Because few studies have directly tested the general-
Figure 2. Preliminary multiple mediation model in Study 3.
Figure 3. Path model leading from an incremental theory to a reduced desire for vengeance in Study 3.
Numbers are standardized coefficients. Circles represent latent variables, and squares represent manifest
variables. Bad-person attributions and the incremental manipulation were categorical variables; all others were
continuous. Nonsignificant paths and indicators of latent variables were removed from the figure for ease of
presentation.
ⴱ
p⬍.05.
1102 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

izability of the social–cognitive processes documented among
Finnish youths, this study yielded reassuring findings about the
generality of these processes.
Implicit theories were associated with adolescents’ desire for
vengeance above and beyond the effects of many other factors
(such as sex, levels of peer victimization, or race/ethnicity). Im-
plicit theories were related to vengeance after a broad array of
social conflicts (Study 1) and, more specifically, after hypothetical
peer victimization and aggression (Studies 2 and 3). Changes in
implicit theories over a few weeks toward more of an incremental
theory were associated with a reduction in the desire for vengeance
over the same period (Study 3). Moreover, when implicit theories
were experimentally changed, an incremental theory played a
causal role in reducing the vengeance a victim desired (Study 3).
No previous experiment with children or adults has shown that
vengeful responses to peer victimization could be reduced by
learning an incremental theory.
Our demonstration of the causal role of implicit theories of
personality can have implications for intervention. Recent inci-
dents of planned school violence in America, Finland, and other
countries have been found to be related to experiences of peer
victimization in school and fantasies of revenge afterward (Leary
et al., 2003; Punama¨ki, Nokelainen, Marttunen, & Tirri, in press;
Vossekuil et al., 2002; for a review, see Borum et al., 2010).
Notably, peer victimization does not appear to have a direct effect
on adolescents’ adjustment problems but rather an indirect one that
works through their social cognitions. In past research, when a
maladaptive pattern of social cognitions did not result from peer
rejection, later violence was less likely to ensue (see Dodge,
Greenberg, Malone, & the Conduct Problems Prevention Research
Group, 2008; Pettit, Lansford, Malone, Dodge, & Bates, 2010).
Thus, the present research suggests that interventions targeting one
particular psychological factor—beliefs about whether people can
change—may perhaps buffer peer-victimized adolescents from the
desire to use vengeance as a solution.
Relation of Implicit Theories to Attributions and
Emotions
This research also illuminates why and how implicit theories
affect the desire for vengeance after a conflict by exploring several
mediators. Study 1 showed that, when considering a past conflict,
adolescents who held an entity theory were more likely to judge
the perpetrator to be a bad person and to harbor negative feelings
about themselves. These then predicted the desire to use vengeance
as a solution. This study was unique in examining both self-
relevant and other-relevant processes as mediators in the path
leading from an entity theory to a desire for revenge. Study 3,
which changed implicit theories, found support for similar medi-
ators and showed that the negative emotions of hatred and shame
played a key role in promoting vengeance. Study 3 also extended
the theoretical model by adding a measure that tapped a belief in
vengeful thoughts as a good way to regulate emotions. This, too,
was found to partially explain the impact of implicit theories on
vengeful responses, consistent with theories of the role of antici-
pated emotional states in motivating aggression (Baumeister et al.,
2007). Taken together, our mediators fully accounted for the effect
of learning an incremental theory on reduced endorsement of
vengeful responses. Although future research is still needed to
understand the order and manner in which these mediators operate,
our studies suggest that adolescents’ implicit theories of person-
ality predict their responses to social conflicts with acquaintances
because they predict their attributions, emotions, and emotion-
regulation strategies.
Depicting the transgressor. Our findings in Studies 1 and 3
regarding bad-person attributions suggest that believing people can
be permanently labeled with descriptors like bully may make it
more likely that one will summarily dismiss transgressors as bad
people. Because “bad” people may be seen as deserving more
punishment, those with more of an entity theory may feel more
justified in their endorsement of revenge toward the transgressors.
An interesting future direction is to investigate the role of dehu-
manization and moral disengagement (Bandura, 1999) in the pro-
cess just described. For example, adolescent victims of bullying
who also hold an entity theory may find it easier to think of
transgressors in terms of fixed labels, instead of in terms of a
bundle of complex human motivations. Once the victim has de-
humanized the transgressor with a pejorative label, it may be easier
for the victim to disengage from the guilt that may be normally
associated with desiring that someone suffer.
Next, previous research on attribution theory (Weiner, 1986,
1995) has focused on the role of stable characterological attribu-
tions in leading to more punitive evaluations of transgressors. The
results of Study 3, which showed that characterological attribu-
tions could be reduced by an incremental theory, added to this
perspective by showing that implicit theories may be an antecedent
step in this social–cognitive process (see also Blackwell et al.,
2007, Study 1, for related results in the academic domain).
Attributions of intent. Previous groundbreaking and highly
important research has shown the role of biased attributions of
hostile intent in predicting aggression (Hudley & Graham, 1993;
for reviews, see Crick & Dodge, 1994; Dodge et al., 2006). This
research has shown that adolescents who chronically perceive
hostile intent when a perpetrator’s motivation is ambiguous are
more likely to respond aggressively. However, the focus on per-
ceptions of a perpetrator’s intent might not explain how adoles-
cents respond when it is clear that a transgressor has victimized
them on purpose. That is, given clear hostile intent, why do some
adolescents seek more revenge after conflicts?
The present research has suggested that implicit theories play an
important role in responses to unambiguously intentional conflicts.
To address this question directly, we asked participants to rate how
much they thought the acquaintance in the recalled conflict had
acted negatively toward them on purpose (1 ⫽completely by
accident,7⫽completely on purpose) as part of Study 1. We found
that this did not matter. In other words, when we controlled for the
degree to which adolescents believed that the transgression had
been committed on purpose, the effect of an entity theory on the
desire for vengeance was not changed in either of the samples.
Therefore, it seems that implicit theories predict responses to
conflict regardless of perceived intent.
However, an entity theory may well be related to attributions of
hostile intent when intent is ambiguous. An adolescent with more
of an entity theory may see ambiguous actions as driven by the
traits of transgressors and, therefore, as arising from relatively
clear intent. Yet, an adolescent with an incremental theory, with an
appreciation of situational or developmental influences on behav-
1103
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

ior, might not reach such a conclusion. We are currently examining
these possibilities.
Limitations
The present research is not without its limitations. The most
important caveat is that Study 3 was a demonstration, not an
intervention. It is important to determine in future research
whether an incremental theory of personality can be taught to
adolescents in a lasting way and how to teach it appropriately.
Adolescents who learn an incremental theory might incorrectly
believe they are supposed to change a bully by themselves, poten-
tially leading to guilt or danger if their efforts are not effective. An
intervention should make it clear that adolescents ought not always
believe that it is their responsibility or within their power to change
a transgressor. It should also teach that change is neither easy nor
certain and may only happen over time—but it is usually possible.
Additionally, an incremental intervention, while teaching victims
that bullies are not inherently bad, also needs to teach the victims
that they themselves are not to blame. A task for future research,
then, is to determine how best to teach adolescents to flexibly
apply malleable theories about transgressors and about themselves,
an issue that must be addressed in the design and implementation
of real-world interventions.
A potential limitation in our research is that the measurement
and manipulation of implicit theories combined both implicit the-
ories about the self and theories about others. This was done
intentionally to maximize ecological validity. In pilot investiga-
tions and focus group interviews with adolescents, many adoles-
cents had global mind-sets that viewed people’s traits as generally
either fixed or malleable. We also believed it was more ethical in
Study 3 to manipulate both implicit theories for the reasons men-
tioned earlier. However, it is possible for these implicit theories to
be separated and to have separate effects. That is, some adoles-
cents may have an entity theory of others but not of the self or vice
versa. In the former case, they would be predicted to make dispo-
sitional (bad person) attributions about others but not to make
negative, dispositional attributions about themselves. As such,
some victims of bullying might think that they can improve over
time, even though the bullies will continue to be bad people. Thus,
future research might assess the potentially separate effects of
implicit theories about the self and about others.
Next, the present research focused on the desire for vengeance
(Studies 1–3) and on vengeful intentions (Study 3) and not on
actual vengeful behavior. We think it is likely that implicit theories
would relate to behavioral responses to peer conflict, because
previous studies have found a relation between an entity theory
and increased aggressive behavior (measured by self-reports and
parent reports; Rudolph, 2010) and between an incremental theory
and increased prosocial confronting behavior (Rattan & Dweck,
2010). Nevertheless, an important next step is to investigate
whether teaching an incremental theory reduces actual vengeful
behavior following peer conflicts.
It is also important to note that, as in much of the literature,
many of our measures, such as the measures of attributions and
emotions, consisted of single items. Single-item measures typi-
cally have lower reliability, and measurement error in mediators
often leads to overestimation of the direct effects of independent
variables (Bullock, Green, & Ha, 2010). This bias, however, would
work against the significance of the indirect effects, a fact that
highlights the potential importance of the mediators we identified
in our studies. Nevertheless, future studies should use more exten-
sive and more fully validated measures (e.g., Tangney et al.’s,
1996, measure of shame). Along similar lines, future research
might investigate additional emotions that could arise from im-
plicit theories. For example, in line with Tangney et al.’s (1996)
formulation, incremental theorists might feel more guilt rather than
shame for their part in a conflict, which could lead to constructive
actions rather than aggressive ones.
Our investigation focused primarily on why some adolescents
who experience social setbacks turn to vengeful fantasies as a
coping strategy. Yet, implicit theories may also play a role in the
actions of those who initiate bullying or aggression toward others.
For instance, aggressors may justify their actions with fixed beliefs
that their victims are inherently losers or otherwise personally
flawed. However, if they humanized the potential victims and saw
them as people with thoughts and feelings—and not as people who
are permanently stuck in low-status social categories—then their
motivation to bully might be reduced. Thus, it may be interesting
in future investigations to attempt to reduce bullying by teaching
aggressors an incremental theory about their potential victims.
Finally, in Study 1, we found that adolescents’ implicit theories
predicted their desire for vengeance after a broad range of recalled
conflicts with acquaintances, who are more likely than close
friends to perpetrate intense victimization (Olweus, 1993). How-
ever, research on the domain specificity of aggression suggests that
responses to conflicts with close friends or romantic partners may
have a different relation to social–cognitive variables than do
responses to conflicts with acquaintances or enemies (e.g., Peets et
al., 2007, 2008; Pettit et al., 2010). Thus, it is important to conduct
future studies extending our research to other groups of peers.
Conclusion
It has become increasingly important to understand the psycho-
logical dynamics of peer conflict as adolescents have gained
unprecedented access to weapons and as society has increasingly
grappled with violence and suicide in schools. The present re-
search attempts to contribute to the knowledge of the motivation
behind violent retaliation after peer conflicts in general or victim-
ization in particular. Our findings are consistent in suggesting that
adolescents who construe their social world in terms of good
people and bad people can, after conflict or victimization, enter a
cycle of hatred and shame that leads to a desire for vengeance,
particularly when they believe that thoughts of revenge will make
them feel better about themselves. However, our findings also
suggest that a more positive cycle of thoughts, feelings, and desires
can be fostered by teaching adolescents to view themselves and
their peers as works in progress rather than as finished products.
References
Arbuckle, J. L. (2006). Amos 7.0 programming reference guide. Spring
House, PA: Amos Development Corporation.
Bandura, A. (1999). Moral disengagement in the perpetration of inhuman-
ities. Personality and Social Psychology Review, 3, 193–209. doi:
10.1207/s15327957pspr0303_3
Baumeister, R. F., Smart, L., & Boden, J. (1996). Relation of threatened
1104 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

egotism to violence and aggression: The dark side of high self-esteem.
Psychological Review, 103, 5–33. doi:10.1037/0033-295X.103.1.5
Baumeister, R. F., Vohs, K. D., DeWall, C. N., & Zhang, L. (2007). How
emotion shapes behavior: Feedback, anticipation, and reflection, rather
than direct causation. Personality and Social Psychology Review, 11,
167–203. doi:10.1177/1088868307301033
Bee, H. L., Barnard, K. E., Eyres, S. J., Gray, C. A., Hammond, M. A.,
Spietz, A. L., . . . Clark, B. (1982). Prediction of IQ and language skill
from perinatal status, child performance, family characteristics, and
mother–infant interaction. Child Development, 53, 1134–1156. doi:
10.2307/1129003
Beer, J. S. (2002). Implicit self-theories of shyness. Journal of Personality
and Social Psychology, 83, 1009–1024. doi:10.1037/0022-3514.83
.4.1009
Berger, K. S. (2007). Update on bullying at school: Science forgotten?
Developmental Review, 27, 90–126. doi:10.1016/j.dr.2006.08.002
Blackwell, L. S., Trzesniewski, K., & Dweck, C. S. (2007). Implicit
theories of intelligence predict achievement across an adolescent tran-
sition: A longitudinal study and an intervention. Child Development, 78,
246–263. doi:10.1111/j.1467-8624.2007.00995.x
Borum, R., Cornell, D. G., Modzeleski, W., & Jimerson, S. R. (2010).
What can be done about school shootings? A review of the evidence.
Educational Researcher, 39, 27–37. doi:10.3102/0013189X09357620
Brent, D. A., Baugher, M., Bridge, J., Chen, T., & Chiappetta, L. (1999).
Age- and sex-related risk factors for adolescent suicide. Journal of the
American Academy of Child and Adolescent Psychiatry, 38, 1497–1505.
doi:10.1097/00004583-199912000-00010
Bullock, J. G., Green, D. P., & Ha, S. E. (2010). Yes, but what’s the
mechanism? (Don’t expect an easy answer). Journal of Personality and
Social Psychology, 98, 550–558. doi:10.1037/a0018933
Bushman, B. J., Baumeister, R. F., & Phillips, C. M. (2001). Do people
aggress to improve their mood? Catharsis beliefs, affect regulation
opportunity, and aggressive responding. Journal of Personality and
Social Psychology, 81, 17–32. doi:10.1037/0022-3514.81.1.17
Chiu, C., Dweck, C. S., Tong, J. Y., & Fu, J. H. (1997). Implicit theories
and conceptions of morality. Journal of Personality and Social Psychol-
ogy, 73, 923–940. doi:10.1037/0022-3514.73.5.923
Chiu, C. Y., Hong, Y., & Dweck, C. S. (1997). Lay dispositionism and
implicit theories of personality. Journal of Personality and Social Psy-
chology, 73, 19–30. doi:10.1037/0022-3514.73.1.19
Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social
information-processing mechanisms in children’s social adjustment.
Psychological Bulletin, 115, 74–101. doi:10.1037/0033-2909.115.1.74
Dodge, K. A. (2008). Framing public policy and prevention of chronic
violence in American youths. American Psychologist, 63, 573–590.
doi:10.1037/0003-066X.63.7.573
Dodge, K. A., Coie, J. D., & Lynam, D. (2006). Aggression and antisocial
behavior in youth. In W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.),
Handbook of child psychology: Vol. 3. Social, emotional, and person-
ality development (6th ed.). New York, NY: Wiley.
Dodge, K. A., Greenberg, M. T., Malone, P. S., & the Conduct Problems
Prevention Research Group. (2008). Testing an idealized dynamic cas-
cade model of the development of serious violence in adolescence. Child
Development, 79, 1907–1927. doi:10.1111/j.1467-8624.2008.01233.x
Dweck, C. S., Chiu, C., & Hong, Y. (1995). Implicit theories and their role
in judgments and reactions: A world from two perspectives. Psycholog-
ical Inquiry, 6, 267–285. doi:10.1207/s15327965pli0604_1
Erdley, C., Cain, K., Loomis, C., Dumas-Hines, F., & Dweck, C. S. (1997).
The relations among children’s social goals, implicit personality theories
and response to social failure. Developmental Psychology, 33, 263–272.
doi:10.1037/0012-1649.33.2.263
Erdley, C., & Dweck, C. S. (1993). Children’s implicit theories as predic-
tors of their social judgments. Child Development, 64, 863–878. doi:
10.2307/1131223
Fitness, J., & Fletcher, G. (1993). Love, hate, anger, and jealousy in close
relationships: A prototype and cognitive appraisal analysis. Journal of
Personality and Social Psychology, 65, 942–958. doi:10.1037/0022-
3514.65.5.942
Gervey, B., Chiu, C., Hong, Y., & Dweck, C. S. (1999). Differential use of
person information in decision-making about guilt vs. innocence: The
role of implicit theories. Personality and Social Psychology Bulletin, 25,
17–27. doi:10.1177/0146167299025001002
Gilbert, P., & Andrews, B. (1998). Shame: Interpersonal behavior, psy-
chopathology, and culture. New York, NY: Oxford University Press.
Giles, J. W. (2003). Children’s essentialist beliefs about aggression. De-
velopmental Review, 23, 413–443. doi:10.1016/S0273-2297(03)
00039-X
Graham, S., Bellmore, A. D., & Mize, J. (2006). Aggression, victimization,
and their co-occurrence in middle school. Journal of Abnormal Child
Psychology, 34, 363–378. doi:10.1007/s10802-006-9030-2
Graham, S., & Juvonen, J. (1998). Self-blame and peer victimization in
middle school: An attributional analysis. Developmental Psychology, 34,
587–599. doi:10.1037/0012-1649.34.3.587
Haselhuhn, M. P., Schweitzer, M. E., & Wood, A. M. (2010). How implicit
beliefs influence trust recovery. Psychological Science, 21, 645–648.
doi:10.1177/0956797610367752
Hong, Y., Chiu, C., Dweck, C. S., Lin, D., & Wan, W. (1999). Implicit
theories, attributions, and coping: A meaning system approach. Journal
of Personality and Social Psychology, 77, 588–599. doi:10.1037/0022-
3514.77.3.588
Hudley, C., & Graham, S. (1993). An attributional intervention to reduce
peer-directed aggression among African-American boys. Child Devel-
opment, 64, 124–138. doi:10.2307/1131441
Janoff-Bulman, R. (1979). Characterological versus behavioral self-blame:
Inquiries into depression and rape. Journal of Personality and Social
Psychology, 37, 1798–1809. doi:10.1037/0022-3514.37.10.1798
Juvonen, J., Graham, S., & Schuster, M. (2003). Bullying among young
adolescents: The strong, the weak, and the troubled. Pediatrics, 112,
1231–1237. doi:10.1542/peds.112.6.1231
Kaltiala-Heino, R., Rimpela¨, M., Marttunen, M., Rimpela¨, A., & Rantanen,
P. (1999). Bullying, depression, and suicidal ideation in Finnish adoles-
cents: School survey. British Medical Journal, 319, 348–351.
Kammrath, L. K., & Dweck, C. S. (2006). Voicing conflict: Preferred
conflict strategies among incremental and entity theorists. Personality
and Social Psychology Bulletin, 32, 1497–1508. doi:10.1177/
0146167206291476
Kaufman, P., Chen, X., Choy, S. P., Ruddy, S. A., Miller, A. K., Chandler,
K. A., . . . Klaus, P. (1999). Indicators of school crime and safety (NCES
1999–057/NCJ-178906). Washington, DC: U.S. Department of Educa-
tion and U.S. Department of Justice.
Killen, M., Crystal, D. S., & Watanabe, H. (2002). Japanese and American
children’s evaluations of peer exclusion, tolerance of differences, and
prescriptions for conformity. Child Development, 73, 1788–1802. doi:
10.1111/1467-8624.t01–1-00506
Killen, M., Kelly, M. C., Richardson, C., & Jampol, N. S. (2010). Attri-
butions of intentions and fairness judgments regarding interracial peer
encounters. Developmental Psychology, 46, 1206–1213. doi:10.1037/
a0019660
Kray, L. J., & Haselhuhn, M. (2007). Implicit negotiation beliefs and
performance: Longitudinal and experimental evidence. Journal of Per-
sonality and Social Psychology, 93, 49–64. doi:10.1037/0022-
3514.93.1.49
Kunda, Z., & Nisbett, R. E. (1986). The psychometrics of everyday life.
Cognitive Psychology, 18, 195–224. doi:10.1016/0010-0285(86)
90012-5
Lahti, A., Ra¨sa¨nen, P., Karvonen, K., Sa¨ rkioja, T., Meyer-Rochow, V. B.,
& Hakko, H. (2006). Autumn peak in shooting suicides of children and
1105
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE

adolescents from northern Finland. Neuropsychobiology, 54, 140–146.
doi:10.1159/000098265
Leary, M., Kowalski, R., Smith, L., & Phillips, S. (2003). Teasing, rejec-
tion, and violence: Case studies of the school shootings. Aggressive
Behavior, 29, 202–214. doi:10.1002/ab.10061
Levy, S., & Dweck, C. S. (1999). Children’s static vs. dynamic person
conceptions as predictors of their stereotype formation. Child Develop-
ment, 70, 1163–1180. doi:10.1111/1467-8624.00085
Levy, S., Stroessner, S., & Dweck, C. S. (1998). Stereotype formation and
endorsement: The role of implicit theories. Journal of Personality and
Social Psychology, 74, 1421–1436. doi:10.1037/0022-3514.74.6.1421
Loeb, I., & Dweck, C. S. (1994, July). Beliefs about human nature as
predictors of reactions to victimization. Paper presented at the Sixth
Annual Convention of the American Psychological Society, Washing-
ton, DC.
MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D. (2002). On
the practice of dichotomization of quantitative variables. Psychological
Methods, 7, 19–40. doi:10.1037/1082-989X.7.1.19
McGlothlin, H., & Killen, M. (2006). Intergroup attitudes of European
American children attending ethnically homogeneous schools. Child
Development, 77, 1375–1386. doi:10.1111/j.1467-8624.2006.00941.x
McGlothlin, H., & Killen, M. (2010). How social experience is related to
children’s intergroup attitudes. European Journal of Social Psychology,
40, 625–634. doi:10.1002/ejsp.733
Miller, C. H., Burgoon, J. K., & Hall, J. R. (2007). The effects of implicit
theories of moral character on affective reactions to moral transgres-
sions. Social Cognition, 25, 819–832. doi:10.1521/soco.2007.25.6.819
Nishina, A., & Juvonen, J. (2005). Daily reports of witnessing and expe-
riencing peer harassment in middle school. Child Development, 76,
435–450. doi:10.1111/j.1467-8624.2005.00855.x
Ojanen, T., Grönroos, M., & Salmivalli, C. (2005). An interpersonal
circumplex model of children’s social goals: Links with peer-reported
behavior and sociometric status. Developmental Psychology, 41, 699–
710. doi:10.1037/0012-1649.41.5.699
Olweus, D. (1993). Bullying at school: What we know and what we can do.
Cambridge, MA: Blackwell.
Peets, K., Hodges, R., Kikas, E., & Salmivalli, C. (2007). Hostile attribu-
tions and behavioral strategies in children: Does relationship type mat-
ter? Developmental Psychology, 43, 889–900. doi:10.1037/0012-
1649.43.4.889
Peets, K., Hodges, R., & Salmivalli, C. (2008). Affect-congruent social–
cognitive evaluations and behavior. Child Development, 79, 170–185.
doi:10.1111/j.1467-8624.2007.01118.x
Peled, M., & Moretti, M. M. (2007). Rumination on anger and sadness in
adolescence: Fueling of fury and deepening of despair. Journal of
Clinical Child and Adolescent Psychology, 36, 66–75. doi:10.1207/
s15374424jccp3601_7
Pettit, G. S., Lansford, J. E., Malone, P. S., Dodge, K. A., & Bates, J. E.
(2010). Domain specificity in relationship history, social-information
processing, and violent behavior in early adulthood. Journal of Person-
ality and Social Psychology, 98, 190–200. doi:10.1037/a0017991
Plaks, J. E., Stroessner, S., Dweck, C. S., & Sherman, J. (2001). Person
theories and attention allocation: Preference for stereotypic vs. counter-
stereotypic information. Journal of Personality and Social Psychology,
80, 876–893. doi:10.1037/0022-3514.80.6.876
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling
strategies for assessing and comparing indirect effects in multiple me-
diator models. Behavior Research Methods, 40, 879–891. doi:10.3758/
BRM.40.3.879
Punama¨ki, R., Nokelainen, P., Marttunen, M., & Tirri, K. (in press).
Violent behavior and school killings: Societal, developmental and social
factors. Finnish Academy of Science and Letters.
Raine, A., Dodge, K. A., Loeber, R., Gatzke-Kopp, L., Lynam, D., Reyn-
olds, C., . . . Liu, J. (2006). The Reactive–Proactive Aggression Ques-
tionnaire: Differential correlates of reactive and proactive aggression in
adolescent boys. Aggressive Behavior, 32, 159–171. doi:10.1002/
ab.20115
Rattan, A., & Dweck, C. S. (2010). Who confronts prejudice? The role of
implicit theories in the motivation to confront prejudice. Psychological
Science, 21, 952–959. doi:10.1177/0956797610374740
Reijntjes, A., Thomaes, S., Bushman, B. J., Boelen, P. A., Orobio de
Castro, B., & Telch, M. J. (in press). The outcast lash-out effect in youth:
Alienation increases aggression following peer rejection. Psychological
Science.
Robins, R. W., & Pals, J. L. (2002). Implicit self-theories in the academic
domain: Implications for goal orientation, attributions, affect, and self-
esteem change. Self and Identity, 1, 313–336. doi:10.1080/
15298860290106805
Rudolph, K. D. (2010). Implicit theories of peer relationships. Social
Development, 19, 113–129. doi:10.1111/j.1467-9507.2008.00534.x
Salmivalli, C., Ojanen, T., Haanpa¨a¨, J., & Peets, K. (2005). “I’m OK but
you’re not” and other peer-relational schemas: Explaining individual
differences in children’s social goals. Developmental Psychology, 41,
363–375. doi:10.1037/0012-1649.41.2.363
Scheff, T. J. (1994). Bloody revenge: Emotion, nationalism and war.
Boulder, CO: Westview Press.
Scheff, T. J., & Retzinger, S. (1991). Violence and emotions. Lexington,
MA: Lexington Books.
Sirin, S. R. (2005). Socioeconomic status and academic achievement: A
meta-analytic review of research. Review of Educational Research, 75,
417–453. doi:10.3102/00346543075003417
Sourander, A., Jensen, P., Rönning, J. A., Niemela¨, S., Helenius, H.,
Sillanma¨ki, L., . . . Almqvist, F. (2007). What is the early adulthood
outcome of boys who bully or are bullied in childhood? The Finnish
“From a Boy to a Man” study. Pediatrics, 120, 397–404. doi:10.1542/
peds.2006-2704
Statistics Finland. (2009). Statistical yearbook of Finland 2009. Helsinki,
Finland: Author.
Swearer, S. M., Espelage, D. L., Vaillancourt, T., & Hymel, S. (2010).
What can be done about school bullying? Linking research to educa-
tional practice. Educational Researcher, 39, 38–47. doi:10.3102/
0013189X09357622
Swearer, S. M., Song, S. Y., Cary, P. T., Eagle, J. W., & Mickelson, W. T.
(2001). Psychosocial correlates in bullying and victimization: The rela-
tionship between depression, anxiety, and bullying. Journal of Emo-
tional Abuse, 2, 95–121. doi:10.1300/J135v02n02_07
Tangney, J. P., Miller, R. S., Flicker, L., & Barlow, D. H. (1996). Are
shame, guilt and embarrassment distinct emotions? Journal of Person-
ality and Social Psychology, 70, 1256–1269. doi:10.1037/0022-
3514.70.6.1256
Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral emotions and
moral behavior. Annual Review of Psychology, 58, 345–372. doi:
10.1146/annurev.psych.56.091103.070145
Thomaes, S., Bushman, B. J., Stegge, H., & Olthof, T. (2008). Trumping
shame by blasts of noise: Narcissism, self-esteem, shame, and aggres-
sion in young adolescents. Child Development, 79, 1792–1801. doi:
10.1111/j.1467-8624.2008.01226.x
Trach, J., Hymel, S., Waterhouse, T., & Neale, K. (2010). Bystander
responses to school bullying: A cross-sectional investigation of grade
and sex differences. Canadian Journal of School Psychology, 25, 114–
130. doi:10.1177/0829573509357553
Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). If
you can’t join them, beat them: Effects of social exclusion on aggressive
behavior. Journal of Personality and Social Psychology, 81, 1058–1069.
doi:10.1037/0022-3514.81.6.1058
Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski, W.
(2002). The final report and findings of the Safe School Initiative:
Implications for the prevention of school attacks in the United States.
1106 YEAGER, TRZESNIEWSKI, TIRRI, NOKELAINEN, AND DWECK

Washington, DC: U.S. Secret Service and U.S. Department of Educa-
tion.
Walker, D., Greenwood, C., Hart, B., & Carta, J. (1994). Prediction of
school outcomes based on early language production and socioeconomic
factors. Child Development, 65, 606–621. doi:10.2307/1131404
Weiner, B. (1986). An attributional theory of motivation and emotion. New
York, NY: Springer-Verlag.
Weiner, B. (1995). Judgments of responsibility: A foundation for a theory
of social conduct. New York, NY: Guilford Press.
Williams, K. D. (2009). Ostracism: A temporal need–threat model. In M.
Zanna (Ed.), Advances in experimental social psychology (Vol. 41, pp.
275–314). New York, NY: Academic Press. doi:10.1016/S0065-
2601(08)00406-1
Received June 6, 2010
Revision received January 5, 2011
Accepted January 19, 2011 䡲
Online First Publication
APA-published journal articles are now available Online First in the PsycARTICLES database.
Electronic versions of journal articles will be accessible prior to the print publication, expediting
access to the latest peer-reviewed research.
All PsycARTICLES institutional customers, individual APA PsycNET威database package sub-
scribers, and individual journal subscribers may now search these records as an added benefit.
Online First Publication (OFP) records can be released within as little as 30 days of acceptance and
transfer into production, and are marked to indicate the posting status, allowing researchers to
quickly and easily discover the latest literature. OFP articles will be the version of record; the
articles have gone through the full production cycle except for assignment to an issue and
pagination. After a journal issue’s print publication, OFP records will be replaced with the final
published article to reflect the final status and bibliographic information.
1107
IMPLICIT THEORIES AND ADOLESCENT VENGEANCE