Journal of Personality and Social Psychology2001, Vol. 80, No. 1, 125-135

Copyright 2001 by the American Psychological Association, Inc.0022-3514/01/$5.00 DOI: 10.1O37//O022-3514.80.1.125

Sociocognitive Self-Regulatory Mechanisms GoverningTransgressive Behavior

Albert BanduraStanford University

Gian Vittorio Caprara, Claudio Barbaranelli,and Concetta Pastorelli

University of Rome, "La Sapienza"

Camillo RegaliaUniversita Cattolica del Sacro Cuore, Milan

This longitudinal research examined a structural model of the self-regulatory mechanisms governingtransgressive conduct. Perceived academic and self-regulatory efficacy concurrently and longitudinallydeterred transgressiveness both directly and by fostering prosocialness and adherence to moral self-sanctions for harmful conduct. The impact of perceived social self-efficacy was mediated throughprosocialness. Moral disengagement and prosocialness affected transgressiveness through the mediatinginfluence of irascible affectivity and hostile rumination. Ruminative affectivity, in turn, both concurrentlyand longitudinally affected transgressiveness. Moral disengagement also contributed independently tovariance in transgressiveness over time. This pattern of relations was obtained after controlling for priortransgressiveness. The structural model was replicated across gender and provided a better fit to the datathan did several alternative models.

Social cognitive theory analyzes human self-development, ad-aptation, and change from an agentic perspective (Bandura, 1999b,2001). The capacity to exercise some measure of control overone's thought processes, motivation, affect, and action operatesthrough mechanisms of personal agency. Diverse lines of researchhave documented the prominent role that self-regulatory mecha-nisms play in the development and pursuit of socially valued lifecourses (Bandura, 1995, 1997; Caprara & Cervone, 2000; Cervone& Shoda, 1999; Schunk & Zimmerman, 1994; Zimmerman, 1989).The longitudinal research presented in this article extends this lineof inquiry to self-regulatory mechanisms governing transgressiveforms of behavior.

Among the mechanisms of human agency, none is more focal orpervasive than beliefs of personal efficacy. In social cognitive

Albert Bandura, Department of Psychology, Stanford University; GianVittorio Caprara, Claudio Barbaranelli, and Concetta Pastorelli, Diparti-mento di Psicologia, University of Rome, "La Sapienza," Rome, Italy;Camillo Regalia, Department of Psychology, University Cattolica delSacro Cuore, Milan, Italy.

The research reported in this article was supported by grants from theGrant Foundation, the Spencer Foundation, and the Jacobs Foundation.This article was prepared while Claudio Barbaranelli was a visiting scholarat Stanford University on a grant from the Fulbright Exchange Program.

We thank Kay Bussey for her helpful comments on an earlier version ofthis article.

Correspondence concerning this article should be addressed to AlbertBandura, Department of Psychology, Stanford University, Stanford, Cali-fornia 94305-2130, or Gian Vittorio Caprara, Dipartimento di Psicologia,Universita Degli Studi di Roma, University of Rome, "La Sapienza," Viadei Marsi 78, 00185 Roma, Italy. Electronic mail may be sent tobandura@psych.stanford.edu or caprara@uniromal.it.

theory, the self-efficacy belief system is the foundation of humanmotivation, well-being, and personal accomplishments. Unlesspeople believe that they can bring about desired outcomes andforestall undesired ones by their actions, they have little incentiveto act or to persevere in the face of difficulties and adversities.Whatever other factors may operate as guides and motivators, theyare rooted in the core belief that one has the power to influenceone's own functioning and life circumstances.

Perceived self-efficacy plays a pivotal role in causal structuresbecause it affects courses of actions not only directly but alsothrough its impact on cognitive, motivational, and affective deter-minants. Such beliefs influence whether people think productively,self-debilitatingly, pessimistically, or optimistically; how well theymotivate themselves and persevere in the face of adversities; theirvulnerability to stress and depression; and the life choices theymake (Bandura, 1995, 1997; Maddux, 1995; Schwarzer, 1992).These diverse effects identify the multiple pathways throughwhich a strong sense of efficacy oriented toward positive self-development can affect transgressive behavior. It does so, in largepart, by promoting prosocialness, curtailing the propensity to dis-engage moral self-sanctions from socially alienating and harmfulconduct, and countering ruminative and vengeful affectivity (Ban-dura, Barbaranelli, Caprara, & Pastorelli, 1996a; Bandura, Pas-torelli, Barbaranelli, & Caprara, 1999). That a resilient sense ofpersonal efficacy is a vital personal resource has been amplydocumented by meta-analyses of findings from diverse spheres offunctioning by heterogeneous populations in a variety of environ-mental conditions (Holden, 1991; Holden, Moncher, Schinke, &Barker, 1990; Multon, Brown, & Lent, 1991; Stajkovic & Luthans,1998).

125

126 BANDURA ET AL.

A growing body of evidence shows that efficacy beliefs arelinked to domains of functioning rather than conforming to anundifferentiated trait (Bandura, 1997). In the present study, weexamined the role of three major domains of perceived personalefficacy that have been verified cross-culturally (Pastorelli et al., inpress) and shown to be predictive of developmental outcomes(Bandura, 1997). These domains include academic, social, andself-regulatory efficacy to resist peer pressure for transgressiveactivities. However, social cognitive theory identifies several pro-cesses that can produce some covariation even across distinctactivity domains. One such process involves metacognitive self-regulation. Proficient performance is partly guided by higher orderself-regulatory skills, which include generic skills for evaluatingtask demands, constructing alternative options, setting proximalgoals to guide one's efforts, and creating self-incentives to sustainengagement in taxing activities and to manage stress and debili-tating intrusive thinking. Such generalizable self-regulatory skillsenable people to improve their performance in a variety of activ-ities (Meichenbaum & Asarnow, 1979; Zimmerman, 1989). Ge-neric metastrategies learned in one realm of activity tend to beused in other activity domains (Bandura, Jeffery, & Gajdos, 1975).

A second process that can promote cross-domain covariationsconcerns beliefs in one's learning efficacy. Perceived learningcapability affects how people approach the mastery of new chal-lenges. Having achieved success in a particular activity domaincan create a more general sense of efficacy to learn in other lifesituations. To the extent that people consider their self-regulatorycapabilities and learning efficacy in their self-appraisals, they willexhibit at least some generality in their sense of personal efficacyacross different activities.

Codevelopment across diverse domains can also give rise tocovariation when the development of competencies in differentspheres are promoted together. For example, students are likely todevelop comparably high perceived self-efficacy in language andmathematics in superior schools but relatively low perceived self-efficacy in these diverse subjects in ineffective schools, which donot promote much in the way of academic competencies in anysubject matter. Adoption of particular lifestyles also clusters ac-tivity domains for codevelopment. Thus, students who are aca-demically oriented are likely to pursue a supportive constellationof extracurricular activities, cultivate prosocial relationships, andshun involvement in delinquent pursuits (Donovan & Jessor, 1985;Elliott, 1993). Thus, although self-efficacy beliefs are multifacetedrather than undifferentiated, some interdomain covariance isexpected.

The self-regulatory mechanisms through which moral agency isexercised are of special relevance to the self-management oftransgressive behavior. In social cognitive theory (Bandura,1999a), moral agency has dual aspects—inhibitive and proactive.The inhibitive form of morality is expressed in the power to refrainfrom behaving inhumanely. The proactive form of morality isexpressed in the power to behave humanely. After individualsadopt personal standards, their negative self-sanctions for actionsthat violate their standards and their positive self-reactions forconduct faithful to their moral standards serve as the regulatoryinfluences (Bandura, 1991). These self-reactive influences serve asthe motivational and cognitive regulators of moral conduct.

Personal standards do not function as invariant internal regula-tors of conduct, however. Self-sanctions do not impinge on con-

duct unless they are activated, and there are numerous sociocog-nitive maneuvers by which moral self-reactions can be selectivelydisengaged from detrimental conduct. In the conception of moralagency, social cognitive theory specifies eight mechanisms ofmoral disengagement that operate at different points in the controlof behavior by moral self-sanctions (Bandura, 1991, 1999a). Theycenter on the construal of injurious conduct itself, the sense ofpersonal agency for the actions taken, the representation of theinjurious effects that flow from actions, and the characterization ofthe recipients of maltreatment.

One set of disengagement practices operates on the cognitiveconstrual of the conduct itself. Through moral justification, detri-mental conduct is made personally and socially acceptable byportraying it as serving socially worthy or moral purposes (Kelman& Hamilton, 1989; Rapoport & Alexander, 1982; Reich, 1990;Sanford & Comstock, 1971). Language shapes the thoughts onwhich actions are based. Sanitizing euphemisms and convolutedlanguage is widely used to make harmful conduct look respectableor benign (Bollinger, 1982; Diener, Dineen, Endresen, Beaman, &Fraser, 1975; Lutz, 1987). How behavior is viewed is also coloredby what it is compared with. Through advantageous exonerativecomparison, detrimental conduct can lose its repugnancy or evenappear benevolent by contrasting it with more flagrant inhumani-ties (Bandura, 1991).

Moral control operates most strongly when people regard them-selves as contributors to harmful outcomes. The second set ofdisengagement mechanisms operates by obscuring, minimizing, ordisclaiming the agentive role in the harm that one causes. Thisdisengagement is achieved by displacement and diffusion of re-sponsibility (Bandura, Underwood, & Fromson, 1975; Diener,1977; Milgram, 1974; Zimbardo, 1995). Further ways of weaken-ing moral control operate by misrepresenting the harm caused byone's conduct (Klass, 1978). As long as the harmful effects areignored, minimized, distorted, or disbelieved, there is little reasonfor self-sanctions to be activated.

The final set of disengagement mechanisms operates on therecipients of detrimental acts. Ascription of blame to the victimsfor their plight or to compelling circumstances can serve self-exonerative purposes (Crick & Dodge, 1994; Darley, Klosson, &Zanna, 1978; Ferguson & Rule, 1983; Weiner, 1986). The strengthof moral self-sanctions also depends partly on how perpetratorsview those whom they maltreat. To perceive another person ashuman activates, through perceived similarity, empathic reactionsthat counteract cruelty (Bandura, 1992). However, self-censure forcruel conduct can be disengaged by dehumanization that stripspeople of human qualities or invests them with demonic or bestialqualities (Bandura, Underwood, et al., 1975; Haritos-Fatouros,1988; Keen, 1986).

A substantial body of evidence has demonstrated the disinhibi-tory power of moral disengagement. The moral disengagement isshown in the perpetration of large-scale inhumanities (Andrus,1969; Bandura, 1990; Kelman & Hamilton, 1989; Rapoport &Alexander, 1982; Reich, 1990) and social punitiveness in labora-tory conditions conducive to disengagement of moral self-sanctions (Bandura, Underwood, et al., 1975; Diener, 1977; Mil-gram, 1974; Tilker, 1970; Zimbardo, 1969). Research in whichproneness to moral disengagement is assessed verifies some of theprocesses through which it is presumed to operate. To justify anddisown responsibility for the harm done to others and to dehuman-

SELF-REGULATORY MECHANISMS IN TRANSGRESSIVENESS 127

ize and blame them for their maltreatment are not conducive toprosocial relationships. Effective moral disengagement also freesone from the restraints of self-censure experienced as anticipativeguilt for detrimental conduct. Self-exoneration for wrongdoingfosters a self-righteousness that not only justifies one's conduct butalso breeds inimical rumination. Indeed, high moral disengagersexperience low guilt over injurious conduct, are less prosocial, andare more prone to vengeful rumination (Bandura, Barbaranelli,Caprara, & Pastorelli, 1996a).

Prosocialness operates as another influential factor in the pos-ited causal structure. In recent years, a conceptual shift has beenwitnessed from the prevailing focus on the impact of negative riskfactors on developmental trajectories toward the influential role ofpositive enablement factors in shaping the directions that livestake. Prosocialness, as reflected in cooperativeness, helpfulness,sharing, and empathicness, is one such factor that helps to promoteadvantageous self-development (Bandura et al., 1996b; Caprara,Barbaranelli, Pastorelli, Bandura, & Zimbardo, 2000; Eisenberg &Fabes, 1998; Patterson, Reid, & Dishion, 1992; Wentzel, 1991). Aprosocial orientation also deters aggressive conduct by fosteringsocial networks conducive to harmonious relationships and em-pathic and vicarious emotional arousal over the suffering of others(Bandura, 1992; Feshbach & Feshbach, 1986; Miller & Eisenberg,1988).

An agentic perspective assigns a prominent role to self-generated emotional arousal and affect regulation in the self-management of socioemotional life (Bandura, 1997, 1999b). Situ-ationally induced anger arousal has been shown to increase thelikelihood and intensity of injurious behavior (Berkowitz, 1993;Zillmann, 1983). People live extensively in a psychic environmentthat is largely of their own making. The instigators to detrimentalconduct often involve problems of thought. Anger arousal dissi-pates with time, but it can be repeatedly regenerated by cognitiveself-arousal (Bandura, 1973). Many people are quick to rousethemselves into angered states by ruminating over perceived in-sults, inequities, and indignities. Anger arousal often fuels retali-

ative schemes. Hostile ruminative affectivity not only distortsthinking but also predisposes people to untoward conduct, espe-cially if combined with infirm capability for affect self-regulation.The behavioral effects of ruminative affectivity have been corrob-orated experimentally in simulated conditions in which partici-pants can inflict shocks of varying intensity on provocateurs.Individuals who have a low threshold for anger arousal and areprone to hostile rumination behave more punitively than those whoare slower to anger and disinclined to dwell on grievances andpossible retaliations (Caprara, Coluzzi, Mazzotti, Renzi, & Zelli,1985; Caprara, Renzi, Alcini, D'Imperio, & Travaglia, 1983; Ca-prara, Renzi, Amolini, D'Imperio, & Travaglia, 1984; Caprara etal., 1986).

The conceptual analysis presented thus far has examined howparticular sociocognitive determinants separately foster or deterdetrimental conduct and the supportive empirical evidence fortheir contributory role. In the present study, we examined howthese various self-regulatory determinants operate in concertwithin an integrative causal structure in governing transgressivebehavior. Transgressiveness encompassed varied detrimental con-duct, including interpersonal breaches by lying, cheating, andstealing; destructiveness; verbal and physical assaults; and sub-stance abuse. We proximally and longitudinally analyzed the pathsof influence.

Figure 1 summarizes schematically the direct and mediatedlinks in the posited structural model. Perceived self-efficacy toregulate one's academic activities and to ward off peer pressure toengage in detrimental activities both directly and mediationallyaffects transgressive behavior by supporting prosocialness, adher-ence to moral self-sanctions, and low proneness to vindictiverumination. In addition, perceived social efficacy affects transgres-sive conduct through the mediated effects of prosocialness.

We posited the following structural relations among the medi-ating factors. Children who exhibited high prosocialness, as man-ifested in helping, sharing, and empathicness, would refrain fromruminative vengefulness toward others. Those who strongly ad-

Figure I. Posited causal structure through which perceived self-efficacy and moral disengagement operate inconcert with other sociocognitive factors to concurrently and longitudinally affect transgressive conduct.

128 BANDURA ET AL.

hered to moral self-sanctions against hurting others would be morelikely than high moral disengagers to act prosocially in theirinterpersonal relationships and to exhibit a low propensity forirascible affectivity and hostile rumination. Low ruminative affec-tivity, in turn, would reduce the likelihood of transgressive behav-ior. Prosocialness and moral disengagement would concurrentlyand longitudinally affect transgressive behavior both directly aswell as mediationally through ruminative affectivity.

Method

Participants

The participants in this longitudinal study were 564 children testedinitially at 11 years of age with staggered annual starts for two separatecohorts. There were 304 boys and 260 girls. We selected the participantsfrom two schools in a residential community located near Rome, Italy. Foreach cohort, the sociocognitive factors and transgressiveness assessed inthe sixth grade served as predictors of level of involvement in transgressiveactivities in the eighth grade.

Adaptation in the adolescent phase was selected for study because it isan especially important period in the life course when adolescents have toconcurrently manage major biological, educational, and social role transi-tions (Eccles & Midgley, 1989; Furstenberg, Eccles, Elder, Cook, &Sameroff, 1999; Graber, Brooks-Gunn, & Petersen, 1996). It is also aperiod of experimentation with risky activities and substance use (Elliott,1993; Jessor, 1986).

The community from which the participants were selected represents amicrocosm of the larger society, containing families of skilled workers,farmers, professionals, and local merchants and their service staffs. Fifteenpercent were in professional or managerial ranks, 38% were merchants oroperators of other types of businesses, 16% were skilled workers, 30%were unskilled workers, and 1% were retired. The socioeconomic diversityof the sample adds to the generalizability of the findings.

This community adheres to a stringent consent procedure for the conductof research in the schools. Each research proposal must be approved by aschool council composed of parent and teacher representatives as well asstudent representatives at the junior high and high school levels. In addi-tion, parents must give consent, and children are free to decline to take partif they so choose. The study was structured to the parents and the childrenas a project designed to gain a better understanding of child development.Informed consent was obtained from 100% of the families, with 91 % of thesample reassessed in the follow-up phase of the study. The low attritionwas due chiefly to relocation from the area or absence from school at thetime of the assessment. The latter children did not differ significantly fromtheir counterparts on any of the variables in the initial assessment. The dataregarding the variables of theoretical interest were collected in the chil-dren's classrooms by two female experimenters. The various sociocogni-tive measures were administered over a period of several days at eachphase of the study.

Perceived Self-Efficacy

Children's beliefs in their efficacy were measured by 37 items repre-senting seven domains of functioning (Bandura, 1990; Bandura et al.,1996b). For each item, children used a 5-point response format to rate thestrength of their belief in their capability to execute the designatedactivities.

Perceived academic self-efficacy measured the children's belief in theircapabilities to master different areas of course work. The subjects includedmathematics, science, reading and writing language skills, and socialstudies. A second set of scales measured children's perceived self-efficacyfor regulating their own learning activities (Zimmerman, Bandura, &Martinez-Pons, 1992). These scales assessed children's efficacy to arrange

environments conducive to learning, to plan and organize their academicactivities, to use cognitive strategies to enhance understanding and memoryof the material being taught, to seek pertinent information and get teachersand peers to help them with academic problems when needed, to motivatethemselves to do their schoolwork, to get themselves to complete scholasticassignments within set deadlines, and to pursue academic activities whenthere are other interesting things to do. The item "How well can you getteachers to help you when you get stuck on schoolwork?" measuredperceived self-efficacy to enlist enabling social resources. The item "Howwell can you study when there are other interesting things to do?" mea-sured children's perceived efficacy to motivate themselves for academicpursuits in the face of competing attractions.

A third set of scales assessed efficacy for leisure and extracurricularactivities involving mainly group extracurricular activities. A fourth set ofscales assessed children's self-regulatory efficacy to resist peer pressure toengage in high-risk activities involving the use of alcohol and drugs andtransgressive behavior that could get them into trouble. For example, thefollowing item assessed perceived self-regulatory efficacy to rebuff pres-sures exerted by peers to drink alcoholic beverages: "How well can youresist peer pressure to drink beer, wine, or liquor?"

Perceived social self-efficacy measured children's beliefs in their capa-bilities to form and maintain social relationships, work cooperatively withothers, and manage different types of interpersonal conflicts. Self-assertiveefficacy measured children's beliefs in their capabilities to voice theiropinions when they differ from those held by others, to stand up tomistreatment or harassment, and to refuse unreasonable requests. "Howwell can you express your opinions when other classmates disagree withyou?" is a sample item assessing perceived self-assertive efficacy. Per-ceived self-efficacy to meet others' expectations assessed children's beliefsin their capabilities to fulfill what their parents, teachers, and peers expectof them and to live up to what they expect of themselves. "How well canyou live up to what your parents expect of you?" typifies items in theperceived efficacy domain to fulfill social expectations.

A principal-components factor analysis revealed a three-factor structure.The first factor, Perceived Academic Self-Efficacy, included high loadingson 19 items measuring perceived capability to manage one's own learning;to master academic subjects; and to fulfill personal, parental, and teachers'academic expectations. Perceived Social Self-Efficacy constituted the sec-ond factor. The 13 items loading on this factor included perceived capa-bility for peer relationships, for self-assertiveness, and for leisure-timesocial activities. The third factor, Perceived Self-Regulatory Efficacy, wasrepresented by 5 items measuring perceived capability to resist peer pres-sure to engage in high-risk activities. These three factors consti-tuted 16.2%, 8.2%, and 5.4% of the variance, respectively.

The triadic factor structure of these multidimensional efficacy scales hasbeen replicated cross-nationally with Italian, Hungarian, and Polish chil-dren (Pastorelli et al., in press). The predictive validity of these differentforms of perceived self-efficacy is supported by findings of prior studies(Bandura et al., 1996b; Bandura, Barbaranelli, Caprara, & Pastorelli, inpress; Zimmerman et al., 1992).

The reliability of the factors of perceived self-efficacy was assessed bythe squared multiple correlations of factor scores. Coefficients of .70 orhigher are indicators of stable factors (Tabachnik & Fidell, 1989). Theestimated reliabilities were .90 for Perceived Academic Self-Efficacy, .80for Perceived Social Self-Efficacy, and .83 for Perceived Self-RegulatoryEfficacy.

Prosocial Behavior

Children rated their prosocialness on a 10-item scale that assessed theirdegree of helpfulness, sharing, consoling, kindness, and cooperativeness(Caprara & Pastorelli, 1993). "I try to help others" and "I try to make sadpeople happier" are sample items. To avoid a possible response bias, wealso included several control items in this scale. The internal consistency

SELF-REGULATORY MECHANISMS IN TRANSGRESSIVENESS 129

reliability was .77. The factor structure and the concurrent validity of thismeasure have been corroborated in studies relating children's self-ratingsto level of prosocialness as rated by parents', teachers', and peers' socio-metric nominations (Caprara & Pastorelli, 1993).

Ruminative Affectivity

The affective contributors to transgressive behavior were measured by alatent variable labeled ruminative affectivity comprising two facets—rumi-nation self-arousal and irascibility. Rumination self-arousal assessedwith 15 items the level of preoccupation with personal grievances andhostile and retaliative ideation. For example, the items "When I amoutraged, the more I think about it the angrier I feel" and "Sometimes Ican't sleep because of a wrong done to me" assessed ruminative self-arousal of anger. The irascibility measure, comprising 14 items, measuredproneness to negative affectivity in terms of ready anger arousal in socialtransactions and high testiness even to slight provocations. "It takes verylittle for things to bug me" is a sample item. This latent variable, previouslycalled "aggression proneness," was labeled ruminative affectivity so as tobe more descriptive of its dual affective components. The alpha reliabilitycoefficients were .86 for hostile rumination and .83 for irascible affectivity.As we noted earlier, the predictive validity of these measures has beenverified experimentally. The high factor loadings of .76 and .63 for theboys and the girls, respectively, showed it to be a well-defined latentconstruct.

Moral Disengagement

Each of the eight mechanisms of moral disengagement was measured byfour subsets of items (Bandura et al., 1996a). They tapped readiness toconstrue injurious conduct as serving righteous purposes, masqueradingcensurable activities by palliative language or rendering them benign byadvantageous comparison, disowning responsibility for harmful effects bydisplacement or diffusion of responsibility, minimizing the harmful effectsof one's detrimental conduct, devaluing those who are maltreated, andattributing blame to them. To cite some examples, "If people are carelesswhere they leave things it is their own fault if they get stolen" was one ofthe items measuring attribution of blame to the victims. The item "Kidscannot be blamed for misbehaving if their friends pressured them to do it"measured displacement of responsibility. "Some people deserve to betreated like animals" measured proclivity for dehumanization.

The scale items encompassed diverse forms of detrimental conduct in avariety of contextual conditions and in different types of social relation-ships. The detrimental activities involved physically injurious and destruc-tive conduct, verbal abuse, deception, and theft. The social contexts en-compassed educational, familial, community, and peer relations. For eachof the 32 items, children rated the strength of their endorsement or

repudiation of moral exonerations of detrimental conduct. Their responseswere scored on a 5-point scale ranging from 1 (strongly agree) to 5(strongly disagree) for the various disengagement practices. High scoressignify a high level of moral disengagement, and low scores signify a highlevel of moral engagement. Factor analysis of the items revealed a one-factor structure with all items loading on the principal factor. Therefore, wesummed responses to the items to form a composite measure of moraldisengagement. The alpha reliability coefficient for this measure was .86.

Transgressive Behavior

Transgressive behavior was measured initially and 2 years later by theDelinquency subscale of the Child Behavior Checklist developed byAchenbach and Edelbrock (1978). Both the reliability and the predictivevalidity of this measure of transgressive behavior are well established(Achenbach, McConaughy, & Howell, 1987). The subscale, comprising 22items for boys and 19 items for girls, covers a wide range of transgressivebehaviors, including physical and verbal aggression, theft, cheating, lying,destructiveness, truancy, and use of alcohol and drugs. The participantsrecorded whether they engaged in such antisocial activities and, if they did,whether they did so only occasionally or often. The reliability coefficientswere .73 for girls and .85 for boys at the initial assessment and .82 forfemale adolescents and .87 for male adolescents at the second assessment.

Results

Prior to conducting the analyses, we examined the data forunivariate and multivariate outlying cases by using the proceduredevised by Tabachnik and Fidell (1989). Two participants withextreme standard scores on transgressive behavior were detectedand eliminated from subsequent analyses. Table 1 presents themeans and standard deviations for the various sociocognitive fac-tors and the matrix of correlations between them and transgressivebehavior at both time periods.

One-way analyses of variance of gender variations revealedsignificant differences on all of the assessed variables exceptperceived social efficacy. Compared with girls, boys expressedlower academic self-efficacy, F(l, 562) = 9.58, p < .01, and lowerself-regulatory efficacy, F(l, 562) = 13.73, p < .001; were lessprosocial, F(l, 562) = 29.68, p < .0001; were more prone todisengage moral self-sanctions for harmful conduct, F(l, 562) =9.09, p < .01; were more prone to ruminative self-arousal, F(l,562) = 9.91, p < .01, and irascible affectivity, F(l, 562) = 7.35,p < .01; and engaged more heavily in transgressive activities inboth the initial phase, F(\, 562) = 69.50, p < .0001, and the

Table 1Means, Standard Deviations, and Correlation Matrix for the Various Sociocognitive Self-Regulatory Factors and Transgressive

Behavior Assessed Concurrently and Longitudinally

Variable M SD 1 7

1. Academic efficacy2. Social efficacy3. Self-regulatory efficacy4. Prosocial behavior5. Moral disengagement6. Irascibility7. Hostile rumination8. Transgressive behavior 19. Transgressive behavior 2

3.894.224.12

24.3493.3559.9664.16

6.776.22

0.640.570.823.60

20.5214.2418.025.695.67

— .43** .25***.22***

.41***

.31***

.12**

__ 23***- .03- .18**-.19***

—

-.18***-.07- .02-.17***

.32***

- . 1 1 *- .01-.05_14***

.31***49***

-.45***-.16***-.28***-.25***

.33***

.36***

.28***

- .31***- .06-.27***-.22***

.31***

.27***

.22***

.52***

*p<.Q5. ** />< .01 . ***/><.001.

130 BANDURA ET AL.

longitudinal phase of the study, F(l, 562) = 65.43, p < .0001. Ina Gender X Time multivariate analysis of transgressive behavior,neither the main effect of time nor the Gender X Time interactionwas significant.

Paths of Influence

We tested the posited causal structure on the covariance matri-ces by using the EQS program (Bentler, 1995). Level of engage-ment in transgressive behavior at the longitudinal time point wasthe distal outcome variable. We analyzed the structural model byusing the multiple-groups model approach, which estimated simul-taneously the same pattern of relationships among variables in thetwo samples of boys and girls. Equivalence is evaluated by con-straints that impose identical estimates for the model's parameters(Byrne, 1994; Scott-Lennox & Scott-Lennox, 1995). In EQS, theplausibility of these equality constraints is examined by the La-grange Multipliers test (Bentler, 1995).

Figure 2 presents the results of the path analysis using Time 1predictors of Time 2 transgressive conduct. In accord with theposited structural model, perceived academic self-efficacy wasconcurrently and distally linked to low engagement in transgres-sive behavior both directly and through the mediation of proso-cialness and strong adherence to moral self-sanctions. Similarly, ahigh sense of efficacy to withstand peer pressure for transgressiveactivities curbed engagement in transgressive behavior at the ini-tial and the longitudinal phase both directly and by supportingmoral engagement. As we further hypothesized, the desistive im-pact of perceived social self-efficacy on transgressive behaviorwas mediated through enhancement of prosocialness.

Most of the posited structural links among the mediating factorswere verified. Children with a strong prosocial orientation re-frained from moral disengagement. Low prosocialness and facilemoral disengagement were accompanied by hostile ruminative

affectivity, which, in turn, contributed both concurrently and lon-gitudinally to variance in transgressive behavior.

The main departure from the posited conceptual model was thatperceived academic and self-regulatory efficacy were mediation-ally rather than directly linked to ruminative affectivity throughprosocialness and moral disengagement. The effect of prosocial-ness on transgressive behavior was also mediated through rumi-native affectivity. Moral disengagement contributed to concurrenttransgressiveness through ruminative affectivity but operated in-dependently on distal transgressive behavior. The analysis re-vealed no omitted paths of influence from those specified in thestructural model. The posited causal structure was fully replicatedacross gender, although moral disengagement contributed moreheavily to aggression proneness in boys than in girls, and proso-cialness operated as the counteractor of ruminative affectivity ingirls.

The trimmed model provided an excellent fit to the empiricaldata as shown by all the different goodness-of-fit indexes consid-ered. These tests yielded a nonsignificant ^ (53 , N = 564) =55.36, p = .39; a nonnormed fit index of .99; a comparative fitindex of .99; and a root-mean-square error of approximation of.009 (.00, .029). The model accounted for 36% of the variance inTime 1 transgressiveness for both boys and girls, 27% of thevariance in Time 2 transgressiveness for boys, and 26% of thevariance in Time 2 transgressiveness for girls.

Alternative Models

A number of alternative causal models were also tested. In onemodel, prior transgressiveness was assigned causal primacy affect-ing subsequent transgressiveness directly and through its impacton perceived self-efficacy, prosocialness, moral disengagement,and ruminative affectivity. The second model conferred causalprimacy on ruminative affectivity, which contributed to distal

Transgressive \ 29> ( TransgressiveBehavior 1 / r.29) V Behavior 2

Figure 2. Path analysis of the patterns of influence through which perceived self-efficacy and moral disen-gagement affect transgressive conduct. The first path coefficient on each of the structural links is for boys; thesecond coefficient in parentheses is for girls. All of the path coefficients are significant at the p s .05 level,except that boys' prosocial behavior was unrelated to ruminative affectivity. This nonsignificant path coefficientis in italic type. The coefficients with an asterisk on the paths differ significantly between boys and girls.

SELF-REGULATORY MECHANISMS IN TRANSGRESSIVENESS 131

transgressiveness directly and through its influence on the medi-ating factors described above. In a third model, moral disengage-ment was the causally primary factor operating on distal transgres-siveness independently and through the mediating factors.

In each case, the alternative causal models provided a poorer fitto the empirical data than did the posited structural model. All ofthe chi-square tests were significant, and these alternative modelsfared less well on the other indexes of goodness of fit. Further testsof the aforementioned alternative models, but with reversed direc-tion of causation in which prosocialness curbed moral disengage-ment, similarly provided a poorer fit to the data.

Discussion

The findings of the present study lend considerable support forthe posited paths of influence through which self-regulatory fac-tors contribute to level of transgressiveness. With a few excep-tions, not only were the links in the postulated structural modelempirically verified, but the model provided a better fit to theempirical data than did several plausible alternative models.

Male adolescents, who were more heavily engaged than femaleadolescents in transgressive activities, differed on self-regulatoryfactors assigned a prominent role in social cognitive theory in theself-management of transgressive conduct. The male adolescentshad lower perceived academic and self-regulatory efficacy, weremore prone to disengage moral self-sanctions from detrimentalconduct, were quicker to rouse themselves to anger through hostilerumination, and were less prosocially oriented. These differencesin self-regulatory influence add to the growing body of evidencethat sociocognitive factors account for a major share of the vari-ance in gender differences in detrimental conduct (Bettencourt &Kernahan, 1997; Bettencourt & Miller, 1996; Bussey & Bandura,1999; Eagly & Steffen, 1986; Hyde, 1984; Perry, Perry, & Bold-izar, 1990). Although male and female participants differed inlevel of self-regulatory determinants, the posited causal structurewas essentially replicated across gender.

Perceived academic and self-regulatory factors concurrently andlongitudinally curtailed transgressiveness both directly and byfostering prosocialness and adherence to moral self-sanctions forharmful conduct. As we hypothesized, the impact of perceivedsocial self-efficacy was mediated through prosocialness. Adoles-cents who related prosocially to others and who eschewed moraldisengagement were disinclined to engage in angering and retali-ative rumination. Ruminative affectivity was both concurrentlyand longitudinally linked to transgressive conduct. Disengagementof moral self-sanctions also contributed independently to subse-quent transgressiveness.

Prior research has shown that the various facets of perceivedself-efficacy and moral disengagement operate as influential con-tributors to academic achievement (Bandura et al., 1996b). Evi-dence that these factors also operate influentially in transgressiveconduct adds to the explanatory and predictive generality of so-ciocognitive theory.

Only two paths in the structural model differed by gender. Moraldisengagement was accompanied by ruminative affectivity forboth boys and girls, but the link was stronger for boys. Themagnitude of difference may reflect the lower level and lesservariance on both factors for the girls. Prosocialness was linked tolow ruminative affectivity for girls but not for boys. Differential

associational networks may account for this gender difference.Peer groups are not homogeneous. Adolescents who are irascibleand hostilely ruminative may behave prosocially with like-mindedaggressive peers to whom they gravitate. In contrast, for adoles-cents who repudiate harmful conduct, their prosocialness is likelyto create little occasion both cognitively and through amicableaffiliations for anger arousal and vindictive ideation.

A second possible interpretation is in terms of the affectivequality of peer relationships. Girls are more likely than boys toengage their peers in discussions of their negative feelings towardothers (Buhrmester, 1990; Sharabany, Gershoni, & Hofman, 1981;Tannen, 1990). Talking things out may weaken the intensity oftroublesome experiences, permit corrections of misconstruals, andsuggest socially reparative actions. Ruminative hostility is likely toflourish in the absence of ameliorative social feedback. Genderdifferences in the various facets of prosocialness in the presentstudy are in accord with the latter interpretation. Boys and girls donot differ in the sociability facets, but girls are substantially moreconsoling, sharing, helpful, and affectionately demonstrative.Compatible associativity and emotional supportiveness may, ofcourse, operate as complementary rather than competing factors ingender differences in how prosocialness is linked to ruminativeaffectivity. Prosocialness would thus have differential effects onruminative affectivity depending on the types of peers with whomone associates and their level of emotional disclosure.

Several of the hypothesized paths in the structural model werenot verified. Perceived academic and self-regulatory efficacy af-fected ruminative affectivity through their positive impact onprosocialness and curtailment of moral disengagement rather thandirectly. Apparently, the more proximally germane empathic andmoral cognitive processes had determinative priority. Prosocial-ness mediationally but not directly affected transgressiveness bylessening ruminative affectivity and readiness to disengage moralself-sanctions. The nature of the affiliative linkages discussedabove may account for the absence of a residual direct link totransgressiveness. Transgressive adolescents may be prosocialwith peers who are antisocially oriented (Cairns & Cairns, 1991;Claes & Simard, 1992; Elliott, Huizinga, & Ageton, 1985), as arenondeviant adolescents whose peer associates are disinclined toengage in serious wrongdoing. Seriously antisocial individualsmay not be especially empathic and caring toward anyone (Mar-cus, 1996), but adjudicated delinquents fall at the extreme ofunstable social connectedness.

The old adage that birds of a feather flock together professes thesignificance of differential peer bonding. Beliefs of personal effi-cacy affect social relationships. For example, adolescents with ahigh sense of self-regulatory efficacy communicate openly withtheir parents about predicaments they face outside the home andenlist their parents' guidance and support in managing peer pres-sures to engage in detrimental activities (Caprara et al., 1998). Therole of sociocognitive influences in the selection and structuring ofpeer relationships and whether those social ties promote deviancyor prosocialness clearly warrant study.

Moral disengagement affected transgressiveness through its in-fluence on ruminative affectivity in the short run, but directly inthe long run. This pattern of linkage suggests that as cognitivemodes of moral disengagement get routinized, they no longer needto operate through self-arousing affective states. A similar regula-tive shift over time is observed in other forms of adaptation, such

132 BANDURA ET AL.

as avoidant lifestyles, in which emotional arousal initially plays afacilitative role, but as the behavior becomes habitualized, it isperformed independently of emotional arousal (Bandura, 1986). Inthe perpetration of large-scale inhumanities through collectivemoral disengagement, pernicious conduct gets routinized to thepoint where it is carried out with unexcited efficiency (Andrus,1969; Bandura, 1991; Kelman, 1973).

The participants rated the extent to which they engaged intransgressive activities and used alcohol and drugs. In the transi-tion to junior high school, their socioeducational realities becomemore diverse. They have multiple teachers who are a less reliableevaluative source because their observations are largely limited tobehavior in the particular course that they teach. Moreover, theyhave little firsthand accessibility to transgressive activities andsubstance use, which occur mainly outside the school. The group-ing of classmates also varies across the different academic coursearrangements. However, data from the elementary school period,in which social arrangements are more uniform, revealed signifi-cant congruence of ratings of aggressive and transgressive conductacross sources and methods. The children's self-ratings correlated.51 with parental ratings, .24 with teachers' ratings, and .37 withpeers' sociometric ratings. The high correlation with ratings byparents, who have the highest accessibility to their children'sbehavior, speaks most strongly to the reliability of participants'self-ratings. The findings of other studies attest to the generaliz-ability of particular paths of influence in the structural model. In anational longitudinal study in Korea, children who had a low senseof academic and self-regulatory efficacy and were facile to moraldisengagement showed a high propensity for antisocial conduct(Kwak & Bandura, 1997). Elliott and Rhinehart (1995) found thatproclivity to moral disengagement predicted both felony and mis-demeanor assaults and thefts regardless of age, sex, race, religiousaffiliation, and social class.

Research on the effects of situational provocation has addedsubstantially to our understanding of the conditions in whichinduced anger arousal is likely to spur aggressive actions (Berkow-itz, 1993; Rule & Nesdale, 1976; Zillmann, 1988). Cognitionshave been shown to operate as significant mediators of situationalincitements (Baumgardner, 1990; Crick & Dodge, 1994; Ferguson& Rule, 1983). Their contribution to aggression enhancement hasbeen studied primarily in terms of the cognitive labeling of emo-tional states in theories of emotion and in terms of ascriptions ofcausality and blame for detrimental conduct in attribution theory.Social cognitive theory of aggression broadens the cognitive con-tribution by adding the self-activation function of cognition (Ban-dura, 1973, 1986).

The findings of the present study underscore the influential roleof cognitive self-arousal in fostering detrimental conduct. Prone-ness to angering and retaliative rumination was positively linked totransgressiveness both concurrently and over time. It also played amediating role in the impact of prosocialness and moral disengage-ment on transgressive behavior.

As we previously noted, situationally induced anger dissipatesover time. If provocation produced only transitory arousal, itwould be of little importance. What gives enduring significance topast angering experiences is that they create proneness for affec-tive rearousal on future occasions. Several rearousal mechanismshave been proposed as explanatory contenders. One explanation isthat, through associative coupling, situational cues alone become

conditioned activators (Berkowitz, 1993). Situational reminders ofmaddening experiences can certainly rouse one to anger. Whetherthe situationally cued linkage is initially cognitively mediated butthen becomes thoughtless through routinized rearousal, is alwayscognitively mediated, or is thoughtless from the outset and remainsso thereafter is open to speculation and empirical verification. Insocial cognitive theory (Bandura, 1999b), efficient functioningrequires a dual-process mix of routinized action patterns underlower sensorimotor control and mindful activity regulatedcognitively.

Affect arousal is often construed in terms of a state-trait dis-tinction reflecting situational and dispositional activation. Socialcognitive theory treats personal dispositions, such as proneness toruminative affectivity, as dynamic self-arousal processes (Ban-dura, 1999b). The nature and locus of the events that set offperturbing thought processes may differ currently and on futureoccasions. But the self-arousal mechanism is similarly involved atdifferent time points regardless of whether self-arousal is situation-ally prompted and experienced as a current state or later cogni-tively activated by remembrances of past indignities. Experimentalinvestigations and field studies of thought sampling of self-arousallinked to contextual circumstances, accompanying trains ofthought, and adaptive actions can add to our understanding of hownegative affectivity contributes to detrimental conduct.

Everyday life is strewn with experiences that generate negativeaffect. People do not act naively on anger and vengeful thoughts.If every angering experience or vindictive thought triggered ag-gressive acts, individuals would be endlessly embroiled in serioussocial and legal troubles. Adaptiveness requires self-managementof affective states through exercise of self-regulatory capabilities.There is a major difference between possessing self-regulatoryskills and unwaveringly applying them in the face of stiff peerpressure for transgressive activities and maddening provocation.The findings of the present study verify the influential role ofperceived self-regulatory efficacy in the management of peertransgressive pressures.

Self-efficacy theory is currently being extended to the self-regulation of affective states. The self-management of affect takesseveral forms (Bandura, 1997). People influence their emotionalstates by whether they construe events in affect-laden or morebenign ways (Crick & Dodge, 1994; Lazarus & Folkman, 1984).The second mode of self-arousal operates through the trains ofthoughts that people conjure up and dwell on (Bandura, 1973;Wegner & Pennebaker, 1993). People can also moderate by pal-liative means their affective states once they arise without alteringtheir determinants.

Self-management of affect does not rely solely on cognitivemeans, however. People exert control over their emotional well-being by the environments that they select and create through theirstyles of behavior (Bandura, 1999b). For example, aggressiveindividuals manage to create by their actions hostile environments,whereas amicable individuals generate harmonious social milieus(Raush, 1965). Another form of affect regulation by behavioralmeans is concerned with whether people give vent to their emo-tions or restrain their expression (Gross, 1998).

Self-efficacy to manage one's emotional life is concerned withperceived capability to regulate both positive and negative emo-tions through construal, cognitive, palliative, and behavioral means(Bandura, 1997). Preliminary research in other domains of func-

SELF-REGULATORY MECHANISMS IN TRANSGRESSIVENESS 133

tioning has revealed that emotional self-regulatory efficacy con-tributes to variance in depression and prosocial behavior indepen-dently of perceived social efficacy (Caprara et al., 1999). Becausewe have clarified the contributory role of several aspects of per-ceived personal efficacy in the present study, an appropriate nextphase for research is to evaluate the unique contribution of per-ceived self-efficacy for affect regulation in the causal structure oftransgressive behavior.

References

Achenbach, T. M., & Edelbrock, C. S. (1978). The classification of childpsychopathology: A review and analysis of empirical efforts. Psycho-logical Bulletin, 85, 1275-1301.

Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1987). Child/adolescent behavioral and emotional problems: Implications for cross-informant correlations for situational specificity. Psychological Bulletin,101, 213-232.

Andrus, B. C. (1969). The infamous of Nuremberg. London: Fravin.Bandura, A. (1973). Aggression: A social learning analysis. Englewood

Cliffs, NJ: Prentice Hall.Bandura, A. (1986). Social foundations of thought and action: A social

cognitive theory. Englewood Cliffs, NJ: Prentice Hall.Bandura, A. (1990). Mechanisms of moral disengagement. In W. Reich

(Ed.), Origins of terrorism: Psychologies, ideologies, theologies, statesof mind (pp. 161-191). Cambridge, England: Cambridge UniversityPress.

Bandura, A. (1991). Social cognitive theory of moral thought and action. InW. M. Kurtines & J. L. Gewirtz (Eds.), Handbook of moral behavior anddevelopment (Vol. 1, pp. 45-103). Hillsdale, NJ: Erlbaum.

Bandura, A. (1992). Social cognitive theory of social referencing. In S.Feinman (Ed.), Social referencing and the social construction of realityin infancy (pp. 175-208). New York: Plenum.

Bandura, A. (Ed.). (1995). Self-efficacy in changing societies. New York:Cambridge University Press.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York:Freeman.

Bandura, A. (1999a). Moral disengagement in the perpetration of inhu-manities. Personality and Social Psychology Review, 3, 193-209.

Bandura, A. (1999b). A social cognitive theory of personality. In L. Pervin& O. John (Eds.), Handbook of personality (2nd ed., pp. 154-196). NewYork: Guilford Press.

Bandura, A. (2001). Social cognitive theory: An agentic perspective.Annual review of psychology (Vol. 52, pp. 1-26). Palo Alto, CA: AnnualReviews.

Bandura, A., Barbaranelli, C , Caprara, G. V., & Pastorelli, C. (1996a).Mechanisms of moral disengagement in the exercise of moral agency.Journal of Personality and Social Psychology, 71, 364-374.

Bandura, A., Barbaranelli, C , Caprara, G. V., & Pastorelli, C. (1996b).Multifaceted impact of self-efficacy beliefs on academic functioning.Child Development, 67, 1206-1222.

Bandura, A., Barbaranelli, C , Caprara, G. V., & Pastorelli, C. (in press).Self-efficacy beliefs as shapers of children's aspirations and careertrajectories. Child Development.

Bandura, A., Jeffery, R. W., & Gajdos, E. (1975). Generalizing changethrough participant modeling with self-directed mastery. Behaviour Re-search and Therapy, 13, 141-152.

Bandura, A., Pastorelli, C , Barbaranelli, C, & Caprara, G. V. (1999).Self-efficacy pathways to childhood depression. Journal of Personalityand Social Psychology, 76, 258-269.

Bandura, A., Underwood, B., & Fromson, M. E. (1975). Disinhibition ofaggression through diffusion of responsibility and dehumanization ofvictims. Journal of Research in Personality, 9, 253—269.

Baumgardner, S. R. (1990). Attribution of cause, responsibility and blame

among violent and nonviolent individuals. In V. K. Kool (Ed.), Perspec-tives on nonviolence (pp. 54-64). New York: Springer-Verlag.

Bentler, P. M. (1995). EQS structural equations program manual. Encino,CA: Multivariate Software.

Berkowitz, L. (1993). Towards a general theory of anger and emotionalaggression: Implications of the cognitive-neoassociationistic perspec-tive for the analysis of anger and other emotions. In R. S. Wyer, Jr., &T. K. Srull (Eds.), Perspectives on anger and emotion: Advances insocial cognition (Vol. 6, pp. 1-46). Hillsdale, NJ: Erlbaum.

Bettencourt, B. A., & Kernahan, C. (1997). A meta-analysis of aggressionin the presence of violent cues: Effects of gender differences andaversive provocation. Aggressive Behavior, 23, 447-456.

Bettencourt, B. A., & Miller, N. (1996). Gender differences in aggressionas a function of provocation: A meta-analysis. Psychological Bulletin,119, 422-447.

Bollinger, D. (1982). Language: The loaded weapon. London: Longman.Buhrmester, D. (1990). Intimacy of friendship, interpersonal competence,

and adjustment during preadolescence and adolescence. Child Develop-ment, 61, 1101-1111.

Bussey, K., & Bandura, A. (1999). Social cognitive theory of genderdevelopment and differentiation. Psychological Review, 106, 676-713.

Byrne, B. (1994). Testing the factorial validity, replication, and invarianceof a measuring instrument: A paradigmatic application based on theMaslach Burnout Inventory. Multivariate Behavioral Research, 29,289-311.

Cairns, R. B., & Cairns, B. D. (1991). Social cognition and social net-works: A development perspective. In D. J. Pepler & K. H. Rubin (Eds.),The development and treatment of childhood aggression (pp. 249-278).Hillsdale, NJ: Erlbaum.

Caprara, G. V., Barbaranelli, C , Pastorelli, C , Bandura, A., & Zimbardo,P. G. (2000). Social foundations of children's academic achievement.Psychological Science, 11, 306-310.

Caprara, G. V., & Cervone, D. (2000). Personality: Determinants, dynam-ics and potentials. New York: Cambridge University Press.

Caprara, G. V., Coluzzi, M., Mazzotti, E., Renzi, P., & Zelli, A. (1985).Effect of insult and dissipation-rumination on delayed aggression andhostility. Archivio di Psicologia, Neurologia e Psichiatria, 46, 130-139.

Caprara, G. V., & Pastorelli, C. (1993). Early emotional instability, proso-cial behaviour, and aggression: Some methodological aspects. EuropeanJournal of Personality, 7, 19-36.

Caprara, G. V., Renzi, P., Alcini, P., D'lmperio, G., & Travaglia, G.(1983). Instigation to aggress and escalation of aggression examinedfrom a personological perspective: Role of irritability and of emotionalsusceptibility. Aggressive Behavior, 9, 345—353.

Caprara, G. V., Renzi, P., Amolini, P., D'lmperio, G., & Travaglia, G.(1984). The eliciting cue value of aggressive slides reconsidered in apersonological perspective: The weapons effect and irritability. Euro-pean Journal of Social Psychology, 14, 312-322.

Caprara, G. V., Renzi, P., D'Augello, D., D'lmperio, G., Rielli, G., &Travaglia, G. (1986). Interpolating physical exercise between instigationto aggress and aggression: The role of irritability and emotional suscep-tibility. Aggressive Behavior, 12, 83-91.

Caprara, G. V., Scabini, E., Barbaranelli, C , Pastorelli, C , Regalia, C , &Bandura, A. (1998). Impact of adolescents' perceived self-regulatoryefficacy on familial communication and antisocial conduct. EuropeanPsychologist, 3, 125-132.

Caprara, G. V., Scabini, E., Barbaranelli, C , Pastorelli, C , Regalia, C , &Bandura, A. (1999). Autoefficacia percepita emotiva e interpersonale ebuon funzionamento sociale [Emotional and interpersonal perceivedself-efficacy and social well-being], Giornale Italiano di Psicologia, 26,769-789.

Cervone, D., & Shoda, Y. (Eds.). (1999). The coherence of personality.New York: Guilford Press.

Claes, M., & Simard, R. (1992). Friendship characteristics of delinquent

134 BANDURA ET AL.

adolescents. International Journal of Adolescence and Youth, 3, 287-• 301.

Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of socialinformation-processing mechanisms in children's social adjustment.Psychological Bulletin, 115, 74-101.

Darley, J. M, Klosson, E. C , & Zanna, M. P. (1978). Intentions and theircontexts in the moral judgments of children and adults. Child Develop-ment, 49, 66-74.

Diener, E. (1977). Deindividuation: Causes and consequences. SocialBehavior and Personality, 5, 145-156.

Diener, E., Dineen, J., Endresen, K.,Beaman, A. L., & Fraser, S. C. (1975).Effects of altered responsibility, cognitive set, and modeling on physicalaggression and deindividuation. Journal of Personality and Social Psy-chology, 31, 328-337.

Donovan, J. E., & Jessor, R. (1985). Structure of problem behavior inadolescence and young adulthood. Journal of Consulting and ClinicalPsychology, 53, 890-904.

Eagly, A. H., & Steffen, V. J. (1986). Gender and aggressive behavior: Ameta-analytic review of the social psychological literature. Psychologi-cal Bulletin, 100, 309-330.

Eccles, J. S., & Midgley, C. (1989). State-environment fit: Developmen-tally appropriate classrooms for young adolescents. In R. Ames & C.Ames (Eds.), Research on motivation in education: Vol. 3. Goals andcognitions (pp. 139-186). New York: Academic Press.

Eisenberg, N., & Fabes, R. A. (1998). Prosocial development. In W.Damon & N. Eisenberg (Eds.), Handbook of child psychology: Vol. 3.Social, emotional, and personality development (pp. 701-778). NewYork: Wiley.

Elliott, D. S. (1993). Health-enhancing and health-compromising lifestyles.In S. G. Millstein, A. C. Petersen, & E. O. Nightingale (Eds.), Promotingthe health of adolescents: New directions for the twenty-first century (pp.119-145). New York: Oxford University Press.

Elliott, D. S., Huizinga, D., & Ageton, S. S. (1985). Explaining delin-quency and drug use. Beverly Hills, CA: Sage.

Elliott, D. S., & Rhinehart, M. (1995). Moral disengagement, delinquentpeers and delinquent behavior. Unpublished manuscript. University ofColorado, Institute of Behavioral Science.

Ferguson, T. J., & Rule, B. G. (1983). An attributional perspective on angerand aggression. In R. G. Geen & E. I. Donnerstein (Eds.), Aggression:Theoretical and empirical reviews (Vol. 1, pp. 41-74). New York:Academic Press.

Feshbach, S., & Feshbach, N. D. (1986). Aggression and altruism: Apersonality perspective. In C. Zahn-Waxler, E. M. Cummings, & R.Iannotti (Eds.), Altruism and aggression: Biological and social origins(pp. 189-217). Cambridge, England: Cambridge University Press.

Furstenberg, F. F., Eccles, J., Elder, G. H., Ir., Cook, T., & Sameroff, A.(1999). Adolescent development in urban communities: How familiesmanage risk and opportunity. Chicago: University of Chicago Press.

Graber, J. A., Brooks-Gunn, J., & Petersen, A. C. (1996). Transitionsthrough adolescence: Interpersonal domains and context. Mahwah, NJ:Erlbaum.

Gross, J. J. (1998). The emerging field of emotion regulation: An integra-tive review. Review of General Psychology, 2, 271-299.

Haritos-Fatouros, M. (1988). The official torturer: A learning model forobedience to the authority of violence. Journal of Applied Social Psy-chology, 18, 1107-1120.

Holden, G. (1991). The relationship of self-efficacy appraisals to subse-quent health-related outcomes: A meta-analysis. Social Work in HealthCare, 16, 53-93.

Holden, G., Moncher, M. S., Schinke, S. P., & Barker, K. M. (1990).Self-efficacy of children and adolescents: A meta-analysis. Psycholog-ical Reports, 66, 1044-1046.

Hyde, J. S. (1984). How large are gender differences in aggression? Adevelopmental meta-analysis. Developmental Psychology, 20, 722-736.

Jessor, R. (1986). Adolescent problem drinking: Psychosocial aspects anddevelopmental outcomes. In R. K. Silbereisen, K. Eyferth, & G.Rudinger (Eds.), Development as action in context (pp. 241-264). Ber-lin, Germany: Springer-Verlag.

Keen, S. (1986). Faces of the enemy. San Francisco: Harper & Row.Kelman, H. C. (1973). Violence without moral restraint: Reflections on the

dehumanization of victims and victimizers. Journal of Social Issues, 29,25-61.

Kelman, H. C , & Hamilton, V. L. (1989). Crimes of obedience: Toward asocial psychology of authority and responsibility. New Haven, CT: YaleUniversity Press.

Klass, E. T. (1978). Psychological effects of immoral actions: The exper-imental evidence. Psychological Bulletin, 85, 756-771.

Kwak, K., & Bandura, A. (1997). Role of perceived self-efficacy and moraldisengagement in antisocial conduct. Unpublished manuscript, OsanCollege, Seoul, Korea.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. NewYork: Springer.

Lutz, W. D. (1987). Language, appearance, and reality: Doublespeak in1984. Quarterly Review of Doublespeak, 14, 382-391.

Maddux, J. E. (Ed.). (1995). Self-efficacy, adaptation, and adjustment:Theory, research and application. New York: Plenum.

Marcus, R. F. (1996). The friendships of delinquents. Adolescence, 31,145-158.

Meichenbaum, D., & Asarnow, J. (1979). Cognitive-behavioral modifica-tion and metacognitive development: Implications for the classroom. InP. C. Kendall & S. D. Hollon (Eds.), Cognitive-behavioral interven-tions: Theory, research, and procedures (pp. 11-35). New York: Aca-demic Press.

Milgram, S. (1974). Obedience to authority: An experimental view. NewYork: Harper & Row.

Miller, P. A., & Eisenberg, N. (1988). The relation of empathy to aggres-sive and externalizing/antisocial behavior. Psychological Bulletin, 103,324-344.

Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to academic outcomes: A meta-analytic investigation.Journal of Counseling Psychology, 38, 30—38.

Pastorelli, C, Caprara, G. V., Barbaranelli, C, Rola, J., Rozsa, S., &Bandura, A. (in press). Structure of children's perceived self-efficacy: Across-national study. European Journal of Psychological Assessment.

Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys.Eugene, OR: Castalia.

Perry, D. G., Perry, L., & Boldizar, J. P. (1990). Learning of aggression. InM. Lewis & S. Miller (Eds.), Handbook of developmental psychopathol-ogy (pp. 135-146). New York: Plenum.

Rapoport, D. C , & Alexander, Y. (Eds.). (1982). The morality of terror-ism: Religious and secular justification. Elmsford, NY: Pergamon Press.

Raush, H. L. (1965). Interaction sequences. Journal of Personality andSocial Psychology, 2, 487-499.

Reich, W. (Ed.). (1990). Origins of terrorism: Psychologies, ideologies,theologies, states of mind. Cambridge, England: Cambridge UniversityPress.

Rule, B. G., & Nesdale, A. R. (1976). Moral judgments of aggressivebehavior. In R. G. Geen & E. C. O'Neal (Eds.), Perspectives onaggression (pp. 37-60). New York: Academic Press.

Sanford, N., & Comstock, C. (1971). Sanctions for evil. San Francisco:Jossey-Bass.

Schunk, D. H., & Zimmerman, B. J. (Eds.). (1994). Self-regulation oflearning and performance. Hillsdale, NJ: Erlbaum.

Schwarzer, R. (Ed.). (1992). Self-efficacy: Thought control of action.Washington, DC: Hemisphere.

Scott-Lennox, J. A., & Scott-Lennox, R. D. (1995). Sex-race differences insocial support and depression in older low-income adults. In R. H. Hoyle

SELF-REGULATORY MECHANISMS IN TRANSGRESSIVENESS 135

(Ed.), Structural equation modeling: Concepts, issues, ami applications(pp. 199-216). Thousand Oaks, CA: Sage.

Sharabany, R., Gershoni, R., & Hofraan, J. E. (1981). Girlfriend, boy-friend: Age and sex differences in intimate friendship. DevelopmentalPsychology, 17, 800-808.

Stajkovic, A. D., & Luthans, F. (199S). Self-efficacy and work-relatedperformance: A meta-analysis. Psychological Bulletin. 124, 240-261.

Tabachnik, B. G., & Fidcll, L. S. (1989). Using mulrivariate statistics. NewYork: Harper Collins.

Tannen, D. (1990). Gender differences in topical coherence; Creatinginvolvement in best friend's talk. Discourse Processes, 13, 73-90.

Tilker, H. A. (1970). Socially responsible behavior as a function of ob-server responsibility and victim feedback. Journal of Personality andSocial Psychology, 14, 95-100.

Wegner, D. M., & Pennebaker, J. W. (J993). Handbook of mental control.Englcwood Cliffs, NJ: Prentice Hall.

Weiner, B. (1986). An attributional theory of motivation and emotion. NewYork: Springer-Verlag.

Wentzel, K. R. (1991). Social competence at school: Relations betweensocial responsibility and academic achievement. Review of EducationalResearch, 61, 1-24.

Zillmann. D. (1983). Arousal and aggression. In R. G. Geen & E. I.

Donnerstein (Eds.), Aggression: Theoretical and empirical reviews (pp.75-101). New York: Academic Press.

Zillmann, D, (1988). Cognition-excitation interdependencies in aggressivebehavior. Aggressive Behavior, 14, 51-64.

Zimbardo, P. G. (1969). The human choice: Individuation, reason, andorder versus deindividuation, impulse, and chaos. In W. J.Arnold & D.Levine (Eds.), Nebraska Symposium on Motivation, 1969 (pp. 237-309).Lincoln: University of Nebraska Press.

Zimbardo, P. G. (1995). The psychology of evil: A situationist perspectiveon recruiting good people to engage in anti-social acts. Research inSocial Psychology, 11, 125-133.

Zimmerman, B. J. (1989). A social cognitive view of self-regulated aca-demic learning. Journal of Educational Psychology, 81. 329-339.

Zimmerman, B. .!., Bandura, A., & Martinez-Pons, M. (1992). Self-motivation for academic attainment: The role of self-efficacy beliefs andpersonal goal-setting. American Educational Research Journal, 29,663-676.

Received April 11, 2000Revision received July 31, 2000

Accepted August 2, 2000

ORDER FORMStart my 2001 subscription to Journal of Personalityand Social Psychology! ISSN: 0022-3514

$170.00, APA Member/AffiliateS340.00, Individual Non-Member$812.00, InstitutionIn DC add 5.75% sales lax /In MD add 5% sales lax

TOTAL AMOUNT ENCLOSED $

Subscription orders must be prepaid. (Subscriptions are ona calendar basis only.) Allow 4-6 weeks for delivery of thefirst issue. Call for international subscription rates.

SEND THIS ORDER FORM TO:American Psychological AssociationSubscriptions750 First Street, NEWashington, DC 20002-4242

Or call (800) 374-2721, fax (202) 336-5568.TDD/TTY (202)336-6123. Email: subscriptions@apa.org

Send me a Free Sample Issue •

Q Check Enclosed (make payable to APA)

Charge my: Q VISA Q MasterCard Q American ExpressCardholder NameCard No. Exp. date

Signature (Required for Charge)

Credit CardBilling AddressCity State ZipDaytime PhoneSHIP TO:NameAddressCity State . . Zip .APA Customer #

GAD01

PLEASE DO NOT REMOVE - A PHOTOCOPY MAY BE USED