Implicit Theories Of Personality And Attributions Of .

Implicit Theories and Hostile Intent Attributionsschool students’ responses to personal experiencesof peer conflict. They showed that those with moreof an entity theory were more likely to attributeinstances of victimization directed at them to apeer’s personal qualities. This characterologicalattribution mediated an effect on a greater desire to“get back at,” “hurt,” or “punish” the transgressingpeer. Next, Yeager et al. (2011) used a brief experiment to change adolescents’ implicit theoriestoward more of an incremental view, and showedthat the incremental theory reduced the belief that ahypothetical peer who bullied them had negativepersonal qualities. This, in turn, reduced the desirefor revenge.Although instructive, this past research did notinvestigate the role of implicit theories in shapingthe interpretation of negative events in which theintent of an anonymous peer was ambiguous—which is the critical test of an attributional bias (cf.Dodge, 1980). The Yeager et al. (2011) scenariosinvolved peers who explicitly and unambiguouslywere bullying the participants repeatedly on purpose. Thus, it is currently unknown whether implicit theories of personality would predict the hostileattributional bias. Moreover, both the Erdley andDweck (1993) and the Yeager et al. (2011) researchrelied on hypothetical scenarios and did not measure attributions following actual experiences ofpeer conflict. Therefore, the present studies extendedpast research by testing whether implicit theoriesmight predict attributions of hostile intent followingambiguous provocations from unknown peers inhypothetical scenarios and in controlled behavioralexperiences of peer exclusion.The Origins of Implicit TheoriesAlthough the developmental origins of an entitytheory are not fully known, some past research supports the notion that entity theories can result fromnonhostile or even positive influences, and yetnonetheless predict negative attributions and reactions. In analogous past experiments on implicittheories of intelligence, positive, well-intentionedpraise for being “smart” could induce an entitytheory of intelligence (Mueller & Dweck, 1998).Moreover, Dweck et al. (1995) reported that a fixedor entity theory about intelligence is not simply theresidue of poor academic performance. Many highachieving students hold an entity theory about theiracademic ability, yet those with more of an entitytheory of academic ability, regardless of their actuallevels of ability, tend to respond more negatively inthe face of academic challenges, for example, by3blaming their failures on themselves, lying aboutlow grades, considering cheating, or giving up(Blackwell, Trzesniewski, & Dweck, 2007; Mueller& Dweck, 1998; for an overview, see Dweck, 2006).More directly relevant to the present research, Chiuet al. (1997, Study 5) changed participants’ implicittheories of personality with a brief scientific articlesummarizing longitudinal studies that eithershowed that people’s traits become set, like plaster(in an entity theory condition) or that they can bemolded, like clay, throughout life (in an incrementaltheory condition). This brief article led participantsto adopt entity versus incremental frameworkswhen judging others. Hence, past research is suggestive that implicit theories can arise even frompositive or nonhostile influences (such as scientificinformation) and are not only the product of hostileexperiences. Building on this past research, we propose that adolescents who have learned more ofan entity theory of personality, regardless of thefrequency of past negative events or their overallnegative schemas about people, will respond toambiguous peer conflicts with a different pattern ofattributions and behavioral desires than those whohave learned more of an incremental theory.The Present ResearchStudy 1 reports 11 original correlations that werecollected from eight independent samples of adolescents. Five of these samples included one test of therelation between implicit theories and the hostileattributional bias, and three included two testsemploying measures collected at different times.These 11 tests of our hypothesis employed variousquestions, stimuli (e.g., scenarios and behavioralprovocations), and lengths of time between measurements, and they were aggregated into a singleeffect size estimate using meta-analytic methods, tofacilitate more general conclusions. In addition, totest for whether implicit theories would predictattributions among adolescents from relatively hostile as well as relatively nonhostile contexts, Study1 included samples from both lower violence andhigher violence neighborhoods. Last, Study 1 metaanalytically summarized tests of whether attributions of hostile intent mediated the effect of implicittheories on aggressive desires.Study 2 was an experiment designed to test thecausal effect of implicit theories on hostile attributional biases. In it we changed implicit theoriestoward more of a malleable or incremental view ofpersonality and then measured short-term changesin attributions of hostile intent. As in Study 1, this

4Yeager, Miu, Powers, and Dweckexperiment was conducted in two contexts withhighly different levels of neighborhood violence, todemonstrate the generality of this process. Finally,Study 3 was a longitudinal experiment that administered a brief (two class session) incremental theoryintervention and measured attributions of hostileintent and vengeful desires 8 months later.Study 1low socioeconomic status, and came from neighborhoods with violence rates above the national average, according to federal statistics (United StatesDepartment of Justice, Federal Bureau of Investigation, 2011). In the remaining two samples themajority of students came from middle- or highincome families and lived in neighborhoods withviolence rates below the national average. Responserates in all studies were over 75%. Parental consentand student assent were obtained for all ts were students in Grades 8, 9, and 10from eight independent samples coming from sixdifferent schools in the United States (samples collected from the same schools included different students surveyed in different years; N 1,128 uniqueparticipants provided a total of N 1,659 observations). See Table 1 for a summary of sample characteristics. Some of these studies were conductedexpressly to test the present hypotheses, and otherswere designed to test additional hypotheses thatwill be included in future research articles. However, critically, all the samples included here wereadministered measures of implicit theories andattributions of hostile intent, and none of the relations between these variables have been reported inanother research article.In the aggregate, 49% of participants werefemale. Twelve percent self-identified as Black orAfrican American, 17% as White, 26% as Asian,42% as Hispanic or Latino, and the rest as anotherrace or ethnicity. Schools were located in New YorkCity or the San Francisco Bay Area. In six of theeight samples, the majority of students at the schoolreceived free or reduced-price lunch, an indicator ofIn all studies, students completed activities andquestions as a class on computers in a school computer lab during school hours. The session wasadministered by a trained research assistant wholed students to work quietly and privately for theroughly 30 min required. Four of the studiesincluded an ambiguous provocation that studentsbelieved was a real interaction with another student(described below). In each case, students were thoroughly debriefed after the experience. At the end ofall of the studies, students were thanked for theirparticipation and were told how their responseswould be used to advance science and help futurestudents like themselves.MeasuresEntity TheoryParticipants in all studies answered five questions measuring an entity theory of personalitytraits relevant to peer conflicts in high school. Thesequestions were developed and used in past research(Yeager et al., 2011). Participants were asked toTable 1Samples Summarized Meta-Analytically in Study 1Sample12345678N391281792043630521126Grade n% female% Black% White% Asian% Hispanic or LatinoNew York, NYBay Area, CABay Area, CANew York, NYBay Area, CABay Area, CABay Area, CABay Area, 46610283858356158080Note. Participants could indicate as many races or ethnicities as they wished, or none at all, and therefore row percentages need notadd up to 100%. Samples 1, 2, 3, 4, 7, and 8 were collected from schools located in neighborhoods with higher levels of poverty andviolence, whereas samples 5 and 6 were located in upper-middle-class neighborhoods with lower levels of poverty and violence.

Implicit Theories and Hostile Intent Attributionsagree or disagree with statements such as “Bulliesand victims are types of people that really can’t bechanged” (1 strongly disagree, 6 strongly agree;below we report analyses showing that implicit theories are distinct from overall negative schemasabout people). These items were internally consistent across studies: as ranged from .66 to .83.Attributions of Hostile IntentIn each of the tests of our hypothesis, one of twodifferent ambiguous provocations was used toevoke attributions of hostile intent. One was ahypothetical scenario and one was a behavioralexperience of social exclusion. Following the designemployed in past demonstrations of an attributionalbias (e.g., Dodge, 1980), in each of these provocations the identity of the peer was unknown to theparticipant. Indeed, a crucial precondition for a testof an attributional bias is to examine person perceptions in the absence of any previous knowledgeabout the character of the provoking peer.Hypothetical scenario. Seven tests of our hypothesis employed an ambiguous scenario. This was aparadigmatic scenario used to illustrate the hostileattributional bias and was based on an exampledescribed by Dodge (2006, p. 791):Imagine that you were walking in a crowdedhallway at school and everybody was rushing toget to the next class so they wouldn’t be late.While you were looking the other way, you andanother student bumped into each other (prettyhard), so it hurt your shoulder and you droppedthe books that you were carrying. The other student paused briefly, looked at you quickly, andthen turned away and hurried to class.A pilot experiment showed that this stimuluswas, in fact, ambiguous with regard to hostileintent (see online Supporting Information).Cyberball exclusion. In four tests of our hypothesis,students experienced a standardized (but short-livedand harmless) instance of social exclusion duringan online game of catch (“Cyberball”) with two“peers” purportedly from their secondary schoolbut lo

entity theory would predict greater attributions of hostile intent following both imagined and experi-enced provocations. Study 1 also explored whether these increased hostile attributions might statisti-cally mediate the effect of an entity theory on the desire to respond aggressively. Study 2 addressed the causal role of implicit theories.