bandura_1982_self-efficacy mechanism in human agency

Upload: alexandre-baudet

Post on 03-Jun-2018

218 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    1/26

    Self-Efficacy Mechanism inHuman Agency

    ALBERTB A N D U R A Stanford University

    ABSTRACT:This articleaddresses th ecentrality of theself-efficacy mechanism in human agency. Self-per-cepts of efficacy influence thought patterns actionsand emotional arousal.In causal tests the higher thelevel of induced self-efficacy, the higher the perfor-mance accomplishments and the lowerthe emotionalarousal .Different lineso f research arereviewed show-ing that th e self-efficacy mechanism may have wideexplanatory power. Perceivedself-efficacy helps to ac-countforsuch diverse phenomenaaschangesincopingbehavior produced by different modes of influencelevelof physiological stressreactions self-regulation ofrefractory behavior resignation and despondency tofailure experiences self-debilitating effects of proxycontrol and illusory inefficaciousness, achievementstrivings growth of intrinsic interest and careerpur-suits. The influential role of perceived collective e f f icacy in socialchangeisanalyzed as are the socialcon-ditionsconducivetodevelopmentofc ol lect ive inefficacy.Psychological theorizingandresearch tendtocen-ter on issues concerning either acquisition ofknowledge or execution of response patterns. Asa resultthe processes governingthe interrelation-ship between knowledge and action have beenlargely neglected (Newell,1978).Some of the re-cent efforts tobridgethisgaphave been directedat thebiomechanicsproblemhowefferent com-mandsofaction plans guidethe productionof ap-propriateresponse patterns (Stelmach,1976,1978).Others have approached the matter in termsofalgorithmicknowledge,whichfurnishesguidesf orexecuting actionsequences(Greeno, 1973; Newell,1973).Knowledge, transformational operations, andcomponent skillsare necessarybut insufficient foraccomplished performances. Indeed,people oftendo notbehaveoptimally, even though they knowfull well what to do. This is because self-referentthought also mediates the relationship betweenknowledge and action. The issues addressed in thisline of inquiry are concerned with how peoplejudge their capabilitiesandhow, throughtheirself-percepts of efficacy, they affect their motivationand behavior.122 F E B R U A R Y1982 AM ERICAN PSYCHOLOG IST

    Recent years have witnessed a growing conver-genceoftheoryandresearchon the influentialroleof self-referent thought inpsychological function-ing (DeCharms, 1968; Garber &Seligman, 1980;Lefcourt , 1976; Perlmuter &Monty, 1979; Rotter,Chance, &Phares, 1972; White, 1959). Althoughth e research isconducted from anumber of dif-ferent perspectives -under avarietyofnames, thebasic phenomenonbeingaddressed centersonpeo-ple's senseofpersonal efficacy toproduce and toregulate events in their lives. ,Efficacy in dealing with one's environment isnot a fixed act orsimplyamatterofknowingwhatto do.Rather, it involves a generative capabilityinwhich component cognitive, social, and behav-ioral skills must be organized into integratedcourses of action to serve innumerable purposes.A capability isonlyasgood as itsexecution. O p-erative competence requires orchestration andcontinuous improvisation of multiplesubskills tomanage ever-changing circumstances. Initiationand regulation of transactions with the environ-ment are therefore partly governed by judgmentsof operative capabilities. Perceived self-efficacy isconcerned with judgments of how well one canexecute courses of action required to deal withprospective situations.

    Function and Diverse Effects of Self-Percepts of EfficacySelf-percepts of efficacy are not simply inertes-timates of future action. Self-appraisalsof opera-

    Thisarticle was presented as a DistinguishedScientificContri-bution Award address at the meeting of the American Psycho-logical Association, Los Angeles,August 1981.The research by the author reported in this article was sup-ported by Research GrantM-5162fromthe National Institutesof Health, U.S. Public Health Service.Requests for reprints should be sent to Albert Bandura, De-partmentofPsychology, Stanford University, Building 420, Jor-dan Hall, Stanford, California94305.Vol. 37, No.2,122-147

    Copyright1982by theAmerican PsychologicalAssociation,Inc.0003-066X/82/3702-Ol22$00.75

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    2/26

    tive capabilities function as one set of proximaldeterminantsof howpeople behave, their thoughtpatterns, and the emotional reactions they expe-rience in taxing situations. In their daily lives peo-ple continuously make decisions about what coursesofaction to pursue and how long to continue thosethey have undertaken. Becauseactingonmisjudg-ments of personal efficacy can produce adverseconsequences, accurate appraisal ofone'sown ca-pabilities ha sconsiderable functional value. Self-efficacy judgments, whether accurate or faul ty ,influence choice of activities an d environmentalsettings. People avoid activities that they believeexceed their coping capabilities, but they under-take andperform assuredly those that they judgethemselvescapableofmanaging (Bandura,1977a).*Judgments of self-efficacy also determine howmucheffort people will expendand howlong theywillpersist in the face ofobstacles oraversiveex-periences. When beset withdifficultiespeoplewhoentertain serious doubts about their capabilitiesslacken their effortsor give up altogether, whereasthose who have a strong sense of efficacy exertgreater effort to master the challenges (Bandura Schunk, 1981; Brown &Inouye, 1978; Schunk,1981; Weinberg, Gould, &Jackson,1979). Highperseverance usually produces high performanceattainments.

    High self-percepts of efficacy m ay affect pre-paratory and performance effort differently, inthat some self-doubt bestirs learning but hindersadept executionofacquired capabilities. Inapply-ingexisting skills strong self-efficaciousness inten-sifies and sustains th e effort needed fo r optimalperformance, which isdifficult torealize if one isbeleaguered byself-doubts. In approaching learn-ing tasks, however, those whoperceivethemselvesto be supremely self-efficaciousin the undertakingfeel little need to invest much preparatory effortin it. Salomon (in press) provides some evidencebearing on this issue. Hefoundthat high perceivedself-efficacy as a learner isassociated with heavyinvestment ofcognitive effort and superior learn-ingfrom instructional media that children considerdifficult, bu t with less investment of effort andpoor learning from media that they believeto beeasy.Thus some uncertainty has preparatory ben-efits.An aid to good performance is a strong senseof self-efficacy towithstand failures coupled withsome uncertainty (construed intermsof the chal-lengeof thetask, rather than fundamental doubtsabout one's capabilities) to spur preparatory ac-quisition of knowledge an dskills.

    People's judgments of their capabilities addi-

    tionallyinfluence their thought patterns and emo-tional reactions during anticipatory and actualtransactions with the environment. Those whojudge themselves inefficacious incoping with en-vironmental demands dwell on their personal de-ficiencies andimagine potentialdifficultiesasmoreformidablethan they really are (Beck, 1976; Laz-arus &Launier, 1978; Meichenbaum, 1977; Sar-ason, 1975). Such self-referent misgivings createstressand impair performancebydiverting atten-tion from howbest to proceed withth eundertak-ing toconcerns overfailings and mishaps.I ncon-trast,persons who have astrong sense of efficacydeploy their attention .and effort to the demandsof th e situation and are spurred togreater effortby obstacles.

    Microanalyt ic Research StrategyPsychologicaltheories postulate intervening mech-anisms through which external factors affect be -havior. Attemptstoverify atheory commonly seekevidence ofcovariation between behavior and theexternal factorsbelieved toinstatetheinterveningevents, without including independent probes ofthe postulated mediator. Demonstrationsofenvi-ronmental-action covariation increase confidencein a theory, but they do not establish firmly itsvalidity because the covariation can be mediatedthrough other mechanisms capable of producingsimilar effects. A postulated mediator is not di-rectly observable, nevertheless it should haveob-servable indicants other than the actions it pre-sumablygoverns. Hence themost stringent testofatheoryisprovidedbyanchoringthehypothesizedmediator in an independently measurable indicantand confirming that external factors are indeedlinked to anindicant of theinternal mediator andthat it, inturn, islinked toovert behavior.In testing propositions about the origins andfunctions of perceived self-efficacy, a microana-lytic methodology isemployed (Bandura,1977a).Individualsare presented with graduated self-ef-ficacy scales representing tasks varying in diffi-cul ty,complexity,stressfulness, orsome other di-mension, depending on the particular domainof

    1 In the case ofhabitual routines, people develop their self-knowledge throughrepeated experiences, to the point wherethey no longer need to judge their efficacy oneach occasionthatthey perform the same activity. They behave in accordancewithwhat theyknowtheycan orcannotdo,withoutgivingth ematter much further thought. Significant changes in taskde-mands orsituational circumstances, however, prompt self-ef-ficacyreappraisalsa sguidesfo raction under altered conditions.A M E R I C A N PSYCHOLOGIST F E B R U A R Y1982 123

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    3/26

    I N T R A S U B J E C TI I IMEDIUM HIGH LO W ME IUM HIGH

    L E V E L O F S E L F - E F F I C A C YFigure 1.Mean performance attainmentsas a function

    ofdifferential levelso fperceived self-efficacy. (The leftpanel showstheperformancesofgroupsofsubjects whoseself-percepts of efficacy were raised toeither low, m e-dium, orhigh levels; th e right panel showst heperfor-mances of the same subjects at different levels ofself-efficacy[Bandura, Reese, &Adams,in press].)functioning being explored. They designate thetasksthat they judge theycan do and their degreeofcertainty.Anadequate efficacyanalysis requiresdetailedassessmentof thelevel, strength,a nd gen-erality of perceived self-efficacy commensuratewith theprecision with which performanceismea-sured. This methodology permits microanalysisofthedegree ofcongruence between self-perceptsofefficacy andactionat thelevelo findividualtasks.2

    O fcentral interest to self-efficacy theory isthedynamic interplay among self-referent thought,action, and affect. In this approach, self-referentthought isindexed intermsof particularized self-percepts of efficacy that canvary across activitiesand situational circumstances rather than as aglobal disposition assayedby anomnibus test. Mea-sures of self-percepts are tailored to the domainofpsychologicalfunctioningbeing explored.Aspe-cial merit of the microanalytic approach is thatparticularized indices of self-efficacy provide re-'fined predictions of human action and affectivereactivity.CausalAnalysis of Self-Perceptsof EfficacySome of theresearch conducted withinthe efficacyframework has sought to clarify th e causal linkbetweenself-perceptsof efficacy andaction (Ban-dura, Reese, &Adams, inpress). Fo r this purposedifferential levels of perceived efficacy were in -duced inphobic subjects, whereupon their coping

    behavior was measured. In one experiment thelevelofperceivedself-efficacy wasraised throughenactive mastery of progressively more threaten-in g activities. This was achieved through a se-quential procedure in which masteryofeach taskwasfollowedby aself-efficacy probe until subjectsachieved their preassigned low, moderate,orhighlevel of self-efficacy. Th e next phase of thestudyincluded successive modifications of self-efficacylevel within th esame subjects.Inspection ofFigure 1shows that performancevariesas a function ofperceived efficacy. Increas-ing levels of perceived self-efficacy both acrossgroups and within th e same subjects gave rise toprogressively higher performance accomplish-ments.Judgment of self-efficacy from enactive infor-mation is an inferential process inwhich the rel-ative contribution of personal andsituational fac-tors mustbeweighted and integrated. Fine-grainanalysisofenactive masteryand thegrowthofself-efficacyduringth ecourseoftreatment reveals thatself-percepts of efficacy may exceed, 'ma tch , orremain belowenactive attainments, depending onho w they are appraised.3That self-efficacy is not

    2The question arisesregarding whethermaking self-efficacyjudgments in itself can affect performance by creating publiccommitment and pressures for consistency (Rachman, 1978).In applying the microanalytic procedure, special precautionsare taken tominimizeanypossible motivationaleffects of theassessmentitself.Judgments ofself-efficacy aremadeprivately,rather than statedpublicly. Judgmentsofleveland strengthofefficacy aremadefor avariety ofactivitiesin different situa-tionsinadvanceofbehavior tests,ratherthan immediatelypriortoeach performance task. Research on the reactiveeffects ofefficacy assessment shows that performance and fear arousalare the same regardlessof whether peopledo or do not makepriorself-efficacyjudgments (Bandura,Adams,Hardy,&How-ells, 1980; Brown &Inouye, 1978). Nor are people's perfor-mancesaffected bywhether theymaketheirself-efficacy judg-ments publicly or privately (Gauthier & Ladouceur, 1981;Weinberg,Yukelson, &Jackson,1980).Contraryto the consis-tency demand notion,degree ofcongruence between self-ef-ficacy judgment and actionisunaffected orreducedwhen self-efficacy judgmentsarereportedpublicly,with knowledgethattheywillbeinspected,rather than if they are made privatelyunder conditions in which no one will eversee them (Telch,Bandura,Vinciguerra,Agras, &Stout,1981).When public in-spectionoftheirjudgments ismadesalient,peopleare inclinedto become conservative in their self-appraisals, which createsefficacy-actiondiscordances.Veridicalself-appraisalisthus bestachievedunder test conditionsthat reducesocial evaluativefac-tors.3Duringtheefficacy-induction phasethemasterytaskswerepresented in astandardhierarchical order, rather than variedinaccordancewith changes in subjects' perceived efficacy. Ifa small successinstilled a large increaseinperceived self-effi-cacy,topresentnext acorrespondinglyhighmasterytask wouldrisk raisingself-efficacy beyond the preassigned level. Thesetreatment processdatareveal the impact of eachincrementalmastery experience on subsequent self-percepts of efficacy.Aftersubjectsreached their preassigned levelofperceivedself-

    124 F E B R U A R Y 1982 A M E R I C A N PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    4/26

    10< 8 100K-0 80zi600UJ 400.t-g 20oQ:n 0

    - o_ 0--000--0-1

    ~v s *

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    5/26

    taking or suggests limitations to their mode of cop-ing, they register a decline inself-efficaciousnessdespite theirsuccessful performance. In such in-stances apparent successes leave them shakenrather than emboldened. As they gain increasingability topredictand tomanage potential threats,they develop a robust self-assurance that servesthem wellin mastering subsequent challenges.V I C A R I O U SINDUCTION OF DIFFERENTIAL LEVELSOFPERCEIVED SELF-EFFICACYAfurther experiment was designed to provide aneyen morestringenttest of the causal contributionof perceived self-efficacy to action by creatingdifferential levels of self-efficacy vicariously. Inthis mode of efficacy induction, persons observecoping strategies being modeled, but they them-selves do not execute any actions. Consequently,motoric mediators and their effects do not comeinto play. In vicarious influenceobservers have torelysolelyon whattheysee informing generalizedperceptions of their coping capabilities.

    The same causal paradigm was used in whichlevel of performance was examined as a conse-quence of induced differential levelsof self-effi-cacy. The model displays emphasized two as-pectspredictability andcontrollabilitythatareconduciveto the enhancement of self-perceptsofefficacy.In demonstrating predictability the modelrepeatedly exemplified how feared objects arelikely tobehave in each of many di fferent situa-tions. Predictability reduces stress and increasespreparedness in coping with threats (Averill,1973';M ille r, 1981). In modeling controllability themodel demonstrated highly effective techniquesfor handling threats in whatever situation mightarise.

    Self-efficacy probes were made at selectedpoints in the modeling of coping strategies untilsubjects' perceived self-efficacy was raised topreassigned low or medium levels. The" thirdl eve lmaximal self-efficacywas no t includedbecause some phobics would undoubtedly haverequired at least some performance mastery ex-periences to attain complete self-efficaciousness.AsshowninFigure4, thehigher levelof perceivedself-efficacy produced the higher performanceat-tainments.

    Thecombinedfindingslend validityto thethesisthat self-percepts of efficacy operate ascognitivemediators of action. The efficacy-action relation-shipisreplicated acrossd if ferent modesof efficacyinduction, across di fferent types of phobic dys-

    t o o F908070605040 INTER6ROUP

    IINTRASUBJECT

    IL O W MEDIUM LO W MEDIUMLEVEL OF SELF-EFFICACY

    Figure4 .M e anperformance a t tainments by differentgroups of subjects at different levels ofperceived self-efficacy (intergroup)and by thesame subjectsathigherlevels ofperceived self-efficacy (intrasubject) [Banduraetal, iii press].functions,andin both intergroup and intrasubjectexperimental designs. Microanalyses of efficacy-action congruences reveal a close fit of perfor-mance to self-percepts of efficacy on ,individualtasks. People successfully execute tasks that fal lwithin their enhanced range ofperceived self-ef-ficacy, but shun or failthose that exceed their per-ceived coping capabilities.PredictiveGenerality Across Modesof InfluenceIn th e social learning view, judgments of self-ef-ficacy, whether accurate or f au l ty , are based onfour principal sources of information. These in-clude performance attainments; vicarious experi-ences of observing the performances of others; ver-bal persuasion and allied types of social influencesthat onepossesses 'certain capabilities; and phys-iological states from which people partly judgetheircapability,strength,and vulnerability.

    Enactive attainments provide the most influ-ential sourceofefficacy informationbecauseit canbe based on authentic mastery experiences. Suc-cesses heighten perceived self-efficacy; repeatedfailures lower it , especially if failuresoccur earlyin the course of events and do not reflect lacko fef for t oradverse externalcircumstances..

    Peopledo notrelyonenactiveexperienceas thesolesource of. information about their capabilities.Efficacy appraisalsare partly influencedbyvicar-ious experiences. Seeing similar others performsuccessfully can raise efficacy expectations in ob-servers who then judge that they too possess the

    126 F E B R U A R Y1982 A M E R I C A N PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    6/26

    capabilities to master comparable activities. By thesame token, observing others who are perceivedto be ofsimilar competence faildespite high effortlowersobservers' judgmentsoftheir owncapabil-ities (Brown &Inouye,1978). Vicariously derivedinformationalters perceived self-efficacy throughwaysother than social comparison.Aspreviouslynoted, modeling displays convey information aboutth e nature and predictability of environmentalevents. Competent models also teach observers ef-fective strategies fo r dealing with challenging orthreatening situations.

    Verbal persuasioniswidely used to get peopleto believe they possesscapabilitiesthat will enablethem to achieve what they seek. Although socialpersuasion alone may be limited in its power tocreate enduring increases in self-efficacy, it cancontributetosuccessful performanceif theheight-ened appraisal iswithin realistic bounds. Persua-sive efficacyinfluences, therefore, have their great-est impact on people who have some reason tobelieve that they can produceeffectsthrough theiractions (Chambliss &M u rr ay , 1979a, 1979b).TOthe extent that persuasive boosts in self-efficacylead them to try hard enough to succeed, suchinfluences promote development of skills and asense of personal efficacy.

    People rely partly on information from theirphysiological state in judging their capabilities.They read their visceral arousal in stressful andtaxingsituationsas anominous signofvulnerabil-ity to dysfunction. Because high arousal usuallydebilitatesperformance,peoplearemore inclinedto expect success when they are not beset by aver-sivearousal than if they are tense and viscerallyagitated. Inactivities involving strengthandstam-ina, people readtheir fatigue, aches, and pains asindicants ofphysical inefficacy.Informationthatisrelevantfo rjudging personalcapabilitieswhetherconveyed enactively, vicar-iously,persuasively,or physiologicallyis not in-herently enlightening. Rather, it becomes instruc-tiveonly through cognitive appraisal. The cognitiveprocessing of efficacy information concerns thetypesofcues people have learned to use as indi-catorsofpersonal efficacy and the inference rulesthey employ fo r integrating efficacy informationfrom different sources (Bandura,1981).

    The aimofa comprehensive theory is to providea unifyingconceptual framework that can encom-pass diverse modes of influence known to alterbehavior. In any given activity skillsand self-be-liefsthat ensure optimal use ofcapabilitiesare re-quired fo r successful functioning. If self-efficacy

    is lacking, people tend to behave ineffectually,eventhough theyknowwhat to do. Social learningtheorypostulatesacommon mechanismofbehav-ioral changedifferent modes of influence altercoping behavior partly by creating and strength-eningself-percepts of efficacy.The.explanatory and predictive power of thistheory was tested in a series of experiments inwhich severe snake phobies received treatmentsrelying on enactive, vicarious, emotive,and cog-nitive modes of influence (Bandura & Adams,1977; Bandura, Adams, & Beyer, 1977; Bandura,Adams, Hardy, &Howells, 1980). This type ofdisorder permits the mostprecisetests of mecha-nisms of change because participants rarely, ifever, have contact with reptiles while the treat-ment is in progress. Consequently, the changesaccompanying treatment are not confounded byuncontrolledexperiences arisingfromcontact withthe threats between sessions. In each study in thisseries, the1level, strength,andgeneralityofcopingself-efficacy for avarietyof threatening tasksw asmeasured prior to and after treatment.Inthe treatment employing enactive mastery asthe principal vehicle of change, phobies are as-sisted by performance induction aids in dealingwithwhat they fear. Astreatment progresses th eprovisional aids are withdrawn,and self-directedmasteryexperiences are then arranged to authen-ticate and generalize personal efficacy. In thevi-carious mode of treatment, phobies merely observethe model perform progressively more threateningactivities without any adverse effects. In the thirdtreatment tested,which draws heavily on a cog-nitive modality (Kazdin, 1973),phobies generatecognitive scenarios in which multiple models ofdiffering characteristics cope with and masterthreatening activities.As a further testof the gen-eralityofefficacy theory,anemotive-oriented pro-cedure was also examined. In this desensitizationtreatmentpeoplevisualize threatening scenes whiledeeply relaxed until they no longer experience anyanxiety arousal. Imaginal conquestoffear and ac-quisition of a self-relaxation coping skill can boostperceived self-efficacy.Results of these studies confirm that differentmodes of influence al l raise an d strengthen self-percepts of efficacy. Moreover, behavior corre-sponds closelytolevelof self-efficacy change,re-gardless of the method by which self-efficacy isenhanced (Figure 5). The higher the level of per-ceived self-efficacy, the greater the performanceaccomplishments. Strengthofefficacyalso predictsbehavior change. The stronger the perceived ef-A M E R I C A N PSYCHOLOGIST F E B R U A R Y 1982 127

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    7/26

    oI-

    Q

    98765

    4321

    ENACT IVE V I CAR IOUS CONTROL

    -oSELF-EFFICACYPERFORMANCE

    PRETEST POSTTEST PRETEST POSTTEST PRETEST POSTTESTFigure 5.Levelo fperceived self-efficacy andcoping behavior displayedby subjectstoward

    threats afterreceiving treatments relyingoneither enactive, vicarious; emotive,o rcognitivemodes of influence. (In the posttest phase, level of self-efficacy wa s measured prior to andafterthetestofcoping behavior.Thescores representthemean performance attainmentswi thsimilara ndgeneralization threats [Bandura &Adams, 1977; Bandura,Adams, & Beyer, 1977;Bandura, Adams, Hardy, &Howelis,1980].)

    ficacy, themore likelyarepeopletopersistintheirefforts until theysucceed.Consistent with self-ef-ficacytheory, enactive mastery producesthehigh-est, strongest, and most generalized increases incopingefficacy The latter finding iscorroboratedby other comparative studies demonstrating thatenactive mastery surpasses persuasive (Biran &W ilson,1981),emotive (Katz,Stout, Taylor, Home, Agras, Note 3), and vicarious (Feltz, Landers, Raeder,1979) influences increating strongself-percepts of efficacy;

    Self-efficacy theory explainsrateofchange dur-ing the course of treatment as well (Bandura&Adams, 1977). Self-percepts of efficacy formedthrough partial mastery experiences at differentpoints in treatment predict,at a high level of ac-curacy, subsequent performance of threateningtasksthat subjectshad never done before.Thedegreeof relationship between self-percepts

    .of efficacy and action can bequantified inseveralways.Correlations'pan be computed between ag-gregate scoresof perceived self-efficacy an d per-formance attainments. At a more particularized.level of analysis, degree of congruence betweenself-perceptsandactioncan begauged by record-ing whether persons judge themselves capableofperformingeachof thevarious tasks usinga cutoffstrength value and computing the percentageofaccurate correspondence between self-efficacyjudgment and actual performance on individualtasks. Dichotomizing self-efficacy judgments onth e basis of a minimal strength value inevitablyloses somepredictive information. The most pre-cise microanalysis of congruence is provided by

    computing th e probability of successful perfor-mance as a function ofstrength ofperceived self-efficacy. Allthree indexes reveala close relation-ship between self-percepts of efficacy and actionregardless of whether efficacy is instated by enractive mastery, vicarious experience, cognitivecoping,oreliminationofanxiety arousal (Bandura,1977a; Bandura etal, 1980).Influences that operate through nonperfor-mance modes are of particular interest becausethey providenobehavioralinformation fo rjudgingchangesinone'sself-efficacy. Persons havetoinfertheir capabilities from vicarious and symbolicsources of efficacy information. Even in the caseof enactiyely instated self-efficacy, performanceisnot thegenesisof thecausal chain. Performanceincludes among its determinants self-percepts ofefficacy. W e know from th e research of Salomon(inpress),forexample, that self-perceived learningefficacy affects how much effort is invested ingiven activitiesan dwhat levelsofperformance areattained. Thus, judgments of one's capabilitiespartly determine choice of activities and rateofskillacquisition,andperformance mastery,inturn,canboost perceived self-efficacy in amutuallyen -hancing process. It is not asthough self-rperceptsof efficacy affect fu tureperformances but play norole whatsoever in earlier performance attain^ments.Questions about causal ordering of factorsarise in enactively based influences when inter-active processes are treated as linear sequentialones and causally prior self-efficacy determinantsof past performance accomplishments gounmea-sured.

    128 F E B R U A R Y1982 A M E R I C A NP SYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    8/26

    PredictiveGenerality Across Domainsof FunctioningThe preceding experiments examinedtheexplan-atoryandpredictive generalityofself-efficacy the-oryacross different modesofinfluenceapplied toth esame type of dysfunction. Tests of the gener-ality ofthis theory have been extended todiverseareas of functioning. O ne study designed fo r thispurpose included severe agoraphobics, whose liveswere markedly constricted by profound copinginefficacy that makes common activities seemfilled with danger (Banduraet al., 1980).

    The treatment included group sessions in whichthe participants were taught how to identify sit-uational arid ideational elicitorsofanxiety,how tomanage anxiety arousal through thoughta nd self-relaxation, and how to use proximal goal settingingaining coping skill. But the critical ingredientof treatment involved field mastery experiences.Therapists, who accompanied the agoraphobicsinto community settings, drew on whatever per-formance induction aids were required to enabletheir clients to cope successfully with what theydreaded. As treatment progressed therapists re-duced their guided participation andassigned theclients progressively more challenging taskstoper-form o n their own. ,

    Assessment of self-efficacy and performanceac-complishmentsinpreviously dreaded situationstraveling by automobile,using elevators and es-calators, climbing stairs to high levels, dining inrestaurants, shopping in supermarkets, an d ven^turing forthalone into public placesrevealssub-stantial increases in perceived coping efficacy(Figure 6) .Inm icroanalysesconducted both priprto and at thecompletion oftreatment, behavioralchange corresponded closely to level of self-effi-cacy change.

    A variety of studies applying different modesof influence to diverse domains of functioningspeak furtherto theissueofperceived self-efficacyasa common mechanism mediating psychologicalchanges. Perceived self-efficacypredicts degreeofchangeindiverse typesofsocial behavior (Kazdin,1979; Barrios, Note 4); varieties of phobic dys-functions (Biran &Wilson,1981; Bburque & La-douceur, 1980); stress reactions and physiologicalarousal (Banduraet al., inpress); physical stamina(Weinberg et al., 1979; Weinberg, Yukelson,Jackson, 1980); self-regulationofaddictive behav-ior (Gondiotte&Liechtenstein,1981; DiClemente,1981);achievement strivings (Bandura &Schunk,1981; Collins, 1982; Schunk, 1981); and career

    choice and development (Betz &Hackett, 1981;Hackett &Betz, 1981; Hackett, Note 5). In thesediverse lines of research, predictive success isachieved across time, settings,performance vari-ants, expressive modalities, and vastly differentdomains of psychological functioning. Moreover,measures of self-percepts of efficacy using themicroanalyticapproach predict variationsinlevelofchanges produced bydifferent modes of influ-ence,variations among persons receivingthesamemodeof influence, andeven variations within in -dividuals inregardto the particulartasks theyarelikely tomaster or fail (Bandura,1977a;Banduraet al., 1980). Some of these areas of research arediscussedmorefullybecause theyclarify differentaspectsof the mediating self-efficacy mechanism.Although self-efficacy judgmentsarefunction-allyrelated toaction,anumber offactors can af-fect th estrengthof the relationship.Discrepanciesm ay arise because of faulty self-knowledge, mis-judgment of task requirements, unforeseen situa-tional constraints on action, disincentives to act onone's self-percepts of efficacy, ill-defined globalmeasures of perceived self-efficacy or inadequateassessmentso fperformance, and newexperiencesthat prpmpt reappraisalso fself-efficacy in thetimeelapsing between probes of self-efficacy and ac-tion. These and other sources of discordance arediscussed fully elsewhere(Bandura, in press) andwillnot bereviewed here.Perceived Self-Regulatory EfficacyExercise of influence over one's ow n behavior isnotachieved by afeat ofwillpower. Self-regula-tory capabilities require toolsof personal agencyand the self-assurance to use them effectively(Bandura, in press). People who are skeptical oftheir ability to exerciseadequatecontrol over theiractions tend to undermine theireffortsin situationsthat ta x capabilities. Relapses in self-regulationofrefractory consummatory behavior provide a fa-miliar example.M ar la t t and Gordon (1980) have postulated acommon relapse process in heroin addiction, al-coholism, and smoking in which perceived self-regulatory efficacy operatesasacontributing fac-tor. People who have the skillsand assurance intheir coping efficacy mobilize th e effort neededtosucceed inhigh-risk situations. Masteryofpro'b-lem situations further strengthens self-regulatoryefficacy.Incontrast,whencopingskillsareunder-developed andpoorly used because ofdisbelief inone's efficacy, arelapse will occur. Faultless seuv

    A M E R I C A N PSYCHOLOGIST F E B R U A R Y1982 129

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    9/26

    controlis not easy tocomeby for pliant activities,let alone for addictive substances. Nevertheless,thosewho perceive themselves to be inefficaciousare more prone to attribute a slip to pervasive self-regulatory inefficacy. Further cpping efforts arethen abandoned, resultingin atotal breakdowninself-control.

    Studiesofbehavior that isamenableto changebut difficult to sustain over an extended periodconfirm that perceived inefficacy increasesvul-nerability to relapse. In this research, investigatorsmeasured the self-judged efficacy of cigarette

    smokersto resist smoking under various social andstressful inducementsafter theyhadquit smokingthrough various means (DiClemente, 1981;Mc-Intyre etal Note 2). Although allparticipantsachievedthesame terminal behavior, theydid notexhibit the same level of self-regulatory efficacy.Compared to abstainers, relapsers expressed lowerself-efficacy at the end of treatment about theirability to resist smoking under subsequent insti-gating conditions. The higher the perceived self-regulatoryefficacy,themore successfully smokingwaschecked during the follow-up period. Incon-

    10090

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    10/26

    trast, neither demographic factors nor smokinghistory and degree of physicaldependenceon nic-otine differentiated relapsers from abstainers.Ina microanalysis of the relation between self-perceptsofefficacy and smoking,Condiotte andLichtenstein(1981) assessed, at the completion oftreatment,subjects'perceived capability to resistthe urge to smoke in a variety of situations. Per-ceived self-regulatory efficacy predicted monthslater which participants would relapse, how soonthey would relapse, and even the specific situationsin which they experienced their first slip. More-over, perceived self-efficacy at the end of treat-ment predicted how participants were likely torespond to a subsequent relapse, shouldit occur.Thehighly self-efficacious subjectsreinstatedcon-trol following a slip, whereas th e less self-effica-cious ones displayed a marked decrease in per-ceived self-efficacy and relapsed completely.Evidence that changes in self-percepts of efficacypredict coping and self-regulatory behavior sug-geststhat self-efficacy probes during th ecourseoftreatment can provide helpful guides for imple-menting a program of personal change.

    Interactive Perceived Efficacy andPostcoronary RehabilitationSocialenvironmentsm ayplace constraintsonwhatpeople do or may aid them to behave optimally.Whether their endeavorsare socially impeded orsupported willdepend, inpart,on how efficaciousthey are perceived to be. The impetus for inter-personal judgmentsof efficacy isstrongestincloserelationships involving interdependent conse-quences. Thisisbecause actionsof apartner basedon faulty self-percepts of efficacy ca n producedetrimental consequences for, allconcerned. Sinceriskyactionsare alsot hemeanso fsecuring valuedbenefits, veridical mutual judgments of efficacyprovide a reliable basis to promote advantageousendeavors and to dissuade foolhardy ones. Fullunderstanding of how perceptions of efficacy af -fect courses of action under close social interde-pendencies requires analysisofinteractive efficacydeterminants .Recovery from a heart attack presents an im-portantproblemi nwhichtostudyboththeimpactofinteractive efficacy and the contribution of self-percepts ofefficacy tohealth-promoting habits. Inrecovering from aheartattack, the restoration ofperceived physical efficacy is an essential ingre-dient in the process. The heart heals rapidly, but

    psychological recovery is slow fo r patients whobelieve they lack the physical efficacy to resumetheir customary activities. They avoid physicalex-ertion and recreational activities that they previ-ously enjoyed, they areslowtoresume vocationaland social life in the belief that they will over-burdentheirdebilitatedcardiaccapacity, andtheyfearthat sexual activities will do them in. The re-habilitative task is to restore a sense of physicalefficacy sothat postcoronary patientscanlead full,productive lives.

    Physicians typicallyuse one ormoreof the fourprincipal sources of efficacy information to raiseand strengthen perceptionsofcardiacrobustnessinpostcoronary patients. Enactive efficacy infor-mation iscompellingly conveyed through stren-uous treadmill exercises. Vicariousefficacy infor-mation isprovided byenlisting the aid offormerpatients who exemplify active lives. Persuasiveefficacy information is furnished by informingpatients about what they are capable ofdoing.Aheart attack is apt to give rise to overattentivenesstocardiacactivity and misattribution offatigue toan impaired heart. Themeaning ofphysiologicalefficacy information is explained to ensure thatpatients do not misread their physiology, for ex-ample, by interpreting cardiac acceleration as por-tending areinfarction.

    Asa firststep towardclarifyingsome aspectsofthe efficacy restoration process, a research projectbeing conducted in collaboration with Ewart, Tay-lor, DeBusk, and Reese is examining the impactofenactivea nd persuasive efficacy informationonresuming physical activities. Several weeks afterpatients have experiencedamyocardial infarction,their self-percepts of physical efficacy are mea-sured for physical exertion, cardiac capability,emotional stress,and sexual>a ctivities.Psychological recovery from aheart attack is asocial, rather than an individual, matter. Becauseone spouse's notions about theother's physicalca-pabilities can aid or retard the recovery process,the spouse's judgments of the patient's physicalefficacy are measured under three levels of in-volvementin the treadmill activity. All of the pa-tients being studied are men, so the wives' judg-ments ofhusbands' efficacy are tested:when sheisuninvolvedin thetreadmill exercises; whensheis present to observe the husband's stamina as heperformson thetreadmill under increasing work-loads; or when she performs the strenuous tread-millexercises, toexperiencepersonally thephysicaldemands of the task, whereupon she observes herhusband do the same. In the informative consul-

    A M E R I C A N PSYCHOLOGIST F E B R U A R Y1982 131

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    11/26

    vCJ

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    12/26

    CultivatingIntrinsic Interest ThroughDevelopment of Self-EfficacyM ost of thethingspeopleenjoydoingfortheiro wnsake originally had little or no interest for them.But under appropriate learning experiences, al-mostany activity, however silly it mayappear tomanyobservers,canbecomeimbued with consum-ing significance.Theprocessbywhichpeoplede-velop interest in activities in which they initiallylack skill, interest, and self-efficacy is an issueofsome importance. Positive incentives are widelyusedto promote such changes. Some writers (Deci,1975;Lepper&Greene, 1978) have questioned thewisdom ofsuch anapproach, on thegrounds thatrewarding people fo r engaging in an activity ismore likely to reduce than to increase subsequentinterest in it.Extrinsic incentives presumably de-crease interest by weakening competency drivesor by shifting causal attributionsfor performancefrom internal motivators toexternal rewards.The effects ofextrinsic incentives have receivedextensive study.Resultsshow that rewardscan in-crease interest in activities,reduceinterest, or haveno effect (Bates, 1979; Kruglanski, 1975; Lepper,1980; Ross,1976).In evaluating the role of incen-tives inhuman functioning, it isimportant todis-tinguish between whether incentives are used tomanage performance or to cultivate personal ef-ficacy.

    TASK-CONTINGENT INCENTIVES

    Extrinsicrewards are most likely to reduce interestwhen they aregiven merely fo rperforming overand over again an activity that isalready ofhighinterest (Condry, 1977; Lepper &Greene, 1978).Insuch situations rewardsa regainedregardlessofthelevelorqualityofperformance. However, evenunder the limiting conditions wherein rewards arebelieved to produce reductive effects, incentivessometimes enhance interest (Arnold, 1976; Dav-idson & Bucher, 1978), boost low interest but di-minish or do not affect high interest (Calder &Staw, 1975; Loveland & Olley, 1979; McLoyd,1979),orreducelowinterestbut do notaffec t highinterest (Greene, Srernberg, &Lepper,1976).Ap-parently a wide array of other factorslevel ofpreexisting interest and ability, magnitude and sa-lience of rewards, type of activity,degree of re-ward contingency, accompanying social mes-sagescan radically alter or override the effectsof rewards given simply fo r undertakingatask.

    COMPETENCE-CONTINGENT INCENTIVES

    The controversy over the effects of performance-irrelevant reward on high interest has led to ne-glect of the important issueofwhether incentivesfo r performanceattainments cultivate interest andself-perceptsofefficacy. Rewardsfo rtask mastery,which reflect on personal efficacy, should bedis-tinguished from performance-contingent rewardsgainedb yperforming routine activities.Agarmentworkerpaid on apiece-rate basisforsewing shirtsday in and day ou t isunlikelytodevelopagrowingfondness for sewing, even though rewards arehighlycontingent onperformance.Conceptual analysesof intrinsic interest withinth e framework ofsocial learning theory (Bandura,in press) and the theory of intrinsic motivation(Deci, 1975; Lepper & Greene, 1978) assign per-ceived competence a mediating role. The alter-nativetheoreticalapproaches, however, postulatesomewhat different underlying mechanisms. Incognitive evaluation theory (Deci,1975), interestisanexpressionof aninborn drive fo r competenceandself-determination;inattribution theory(Bern,1972; Lepper &Greene, 1978), interestis aproductof retrospective judgmentsof the causes of one'sperformances; insocial learning theory (Bandura,1981, in press), interest grows from satisfactionsderivedfrom fulfillinginternal standardsand fromperceived self-efficacy gained from performanceaccomplishments and other sourcesof efficacy in-formation.There are several ways inwhich incentives fo rtask mastery ca n contribute to the growth of in-terest an d self-efficacy. Positive incentives fosterperformance accomplishments. Gaining knowl-edge and skills that enable one to fulfill personalstandards of merit tend to heighten interest anda firm sense of personal efficacy. Success in at-taining desired outcomes through challenging per-formances can fur ther verify existing competen-cies.Thisisbecause people usuallydo notperformmaximal ly , though they possess th e constituentskills. It is under incentives that test upper limitsthatpeoplefind out what they are able to do. Bymobilizing higheffort , incentives canhelpto sub-stantiate talents,eventhough no newskillsare ac-quired in the process.Rewardsalso assumeefficacy informative valuewhen competencies are difficult to gauge fromperformance alone, which is often the case. Tocomplicate fur therthecompetencevalidation pro-cess, most activities involve diverse facetssothatperceived adequacy may vary widely,depending

    A M E R I C A N PSYCHO LOGIST F E B R U A R Y 1985J 133

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    13/26

    on how the differing aspects are subjectivelyweighted. Because of these ambiguities level ofreward imparts social informationon the qualityof performance. Inthis processcompetentperfor-mancesareperceivedas thereasonfortherewards,ratherthan the rewards being viewed as the causeofcompetent performance (Karniol& Ross,1977).

    Several lines of research confirm that positiveincentives promote interest when they enhance orauthenticate personal efficacy. Both children andadults maintain orincreasetheir interest in activ-itieswhen rewarded forperformance attainments,whereas their interest declines when they are re-warded fo r undertaking activities irrespectiveo fhowwell they perform (Boggiano&Ruble, 1979;Ross, 1976). The larger the extrinsic reward forperformances signifying competence, the greaterthe increase in interest in the activity (Enzle &Ross, 1978). Even incentives for undertaking atask, rather than fo r performance mastery, canraiseinterestifengagementin theactivity providesinformation about personal competence(Arnold,1976). Whenmaterial reward for each task com-pletion is accompanied by self-verbalization ofcompetence,fchildren sustain high interest in theactivity (Sagotsky&Lewis, Note 6).P R O X I M A LSELF-MOTIVATIONContingent incentives are not necessarilythe bestvehicle fo renlisting the typeofsustained involve-ment in activities that builds interest and self-ef-ficacywhere they are lacking. In social learningtheory an important cognitively based sourceo fmotivation operates through the intervening pro-cesses ofgoal setting and self-evaluative reactions(Bandura,1977b, in press). This form of self-mo-tivation, which involves internal comparison pro-cesses, requires personal standards against whichtoevaluate performance. Bymaking self-satisfac-tion conditional on acertainlevel of performancemastery, persons create self-incentives for theirefforts.Self-motivation isbestsummoned and sustainedby adopting attainablesubgoals that leadto largefu ture ones. Whereas proximal subgoals provideimmediate incentivesand guidesforaction, distalgoals are too far removed in time to effectivelymobilize effort or todirect what one does in the \hereand now. Proximal goals can also serve as animportant vehicle in the development ofself-per-ceptsofefficacy. Without standards against whichto measure their performance, people have littlebasisforjudginghowtheya redoingorforgauging

    their capabilities. Subgoal attainments provideclear markerso fprogress along the way to verifyagrowing senseof self-efficacy.

    Thereare at least two waysinwhich proximalgoalsmight contribute toenhancement ofinterestin activities. When people a im for, and master,desired levelsof performance, they experience asense of satisfaction (Locke, Cartledge, & Knerr ,1970). The satisfactions derived from subgoal at-tainments can build intrinsic interest. When per-formances aregauged against distal goals, similaraccomplishments may prove disappointing be-cause ofwide disparities between current perfor-mance and lofty future standards. As a result in -terest fails todevelop, even thoughskillsa rebeingacquiredin the process.Asalready noted,asenseof personal efficacy in mastering tasks ismore ap tto spark interest in them than is self-perceivedinefficacy inperforming competently.

    That proximal self-motivationcanbuild intrinsicinterest in disvalued activities receives supportfrom astudyinwhich childrenwhoexhibited grossdeficits and disinterest inmathematical taskspur-sued a program of self-directed learning underconditions involvingeitherproximal subgoals, dis-ta l goals, or no reference to goals (Bandura &Schunk,1981). Under proximal subgoals childrenprogressed rapidly, in self-directed learning,achievedsubstantial masteryofmathematical op -erations, and developed astrong sense of self-ef-ficacy in solving arithmetic problems (Figure8).Distal goals had no demonstrable effects. In ad-dition to itsother benefits, goal proximity fostersveridical self-knowledge of capabilities, as re-flected in high congruence between judgmentsofmathematical self-efficacy and subsequent math-ematical performance.Asshown in Figure 9, it was mainly childrenin the proximally self-motivatedcondition,all ofwhom felt highly efficacious, who displayed th enotable level of intrinsic interest. Children in theother conditions generally expressed self-doubtsconcerning their capabilities and showed littlespontaneous interest in solving arithmetic prob-lems. Regardlesso ftreatment conditions, self-per-ceptsofmoderate tohigh strength were positivelyrelated to interest.Th e relationshipof the growthfunctions of self-efficacy and interest warrants systematic investi-gation. There may exist some temporal lag be-tween newly acquired self-efficacy and corre-sponding growth of interest in activities that aredisvalued or even disliked. In the temporal lagpattern, self-efficacy fosters mastery experiences

    134 F E B R U A R Y1982 A M E R I C A N PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    14/26

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    15/26

    cupationsareinappropriate forthembecausetheylack thecapabilitiestomaster requisiteskills.Efficacy analyses of career decision making,(Betz,1981) revealthat males perceive themselvesjtobeequallyefficaciousfortraditionallymale andfemalevocations. Incontrast, females judgethem-selveshighlyefficacious fo rthetypeo f occupationstraditionallyheldbywomen, but inefficacious inmastering the educational requirements and jobfunctions of vocations dominated by men. Thesedifferential perceptions of personal efficacy areespecially strikingbecausethegroupsdo not differin their actual verbal and quantitative abilityonstandardized tests. It is not the subskills that se-lected college students possess,but how they per-ceive and use them that makes the difference.Regardless ofsex, level of perceivedself-efficacycorrelates positively with range of career optionsseriously considered and the degree of interestshown in them.Hackett (Note5) has devoted special attentionto perceived mathematical self-efficacy becausemodern technologies have made quantitative skillsincreasingly important to a wide range of careeroptions and professional advancement. Using apath analysis, Hackett found that sex,sexroleso-cialization,andhigh school preparation affec t per-ceived self-efficacy in quantitative capabilities.Perceived self-inefficaciousness in dealing withnumbersi nturnaffectsmathematical anxietyandmathrelatedness of college major.Thecausally prior contributionofperceived ef-ficacy to socialization practices and educationalpreparation remains an important problem of fu-ture research to determine through longitudinalanalysis. It follows from th e present model of ca-reer development that parental career-related ef-ficacy willinfluencetherangeofvocational optionsthey considerviable fortheiroffspring. Students'differential self-percepts of efficacy fo r masteringoccupational entry requirements are likely to in-fluencewhat typesofcoursesthey choosetopursueduring their secondary educational preparation.Societal practices require ofwomenarobust senseof self-efficacy notonlytoentercareersdominatedby men, but to fulfil th e heavy demands arisingfrom dual workloadsofcareer and household.

    Self-Efficacy Conception ofFearArousalPerceptions of self-efficacy affect emotional re -actions aswell asbehavior. This isespecially true

    of anxiety an d stress reactions to unfamiliar orpotentially aversive events. Self-efficacy theorysuggests an alternative way of looking at humananxiety. Psychodynarnic theories generally attrib-ute anxiety to intrapsychic conflicts over theexpression oftabooed impulses.Theexternal objectof anxiety is considered to be of limited signifi-cancebecausethethreatposed by theimpulsecanbe projected onto any number of things. In thisapproach anxiety isrooted in the prohibited im -pulse.

    Conditioningtheory assumes that formerlyneu-t ral events acquire fear-provoking properties byassociation with painfulexperiences. This theoryexternalizes the cause in the s t imulusIt is thestimulusthatissaid to1become aversive.If apersondevelopsaphobiaofmountain driving asaresultof running into a stately roadside redwood, it isno t the road that ischanged by the aversive ex-perience. Rather, it is perceived competence indriving and anticipatory thought patterns tha tundergo change.Fromthesocial learning perspective,it ismainlyperceived inefficacy in coping with potentiallyaversiveevents that makes them fearsome.To theextent that one can prevent, terminate, or lessentheseverityofaversive events, there islittle reasontofearthem. Hence experiences that increase cop-ing efficacy candiminishfeararousaland increasecommercewi thwhatwaspreviously dreaded andavoided.Asenseofcontrollability can beachieved eitherbehaviorally orcognitively (Averill,1973; Lazarus,1980; Miller, 1979). Inbehavioral control individ-ualstake actions that forestall or modify aversiveevents.Incognitive controlpeoplebelievetheycanmanage environmental threats, should they arise.These tw o forms of controllability are distin-guished because the relationship between actualandself-perceived copingefficacy is far from per-fect. Indeed, there are many competent peoplewho areplaguedby asenseofinefficacy,andmanyless competent ones who remain unperturbed byimpending threats because theyareself-assuredoftheircoping capabilities.BEH VIOR L ONTROL

    The effects ofbehavioral controlonfear reductionandstress responses have been amply documentedwith both children and adults. Ability toexercisebehavioral control over potentially aversive eventseliminates or decreases autonomic reactions tothem (G unnar-vonG nechten, 1978; Miller, 1979).

    136 FEBRUARY1982 AMERICAN PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    16/26

    Control over events makes them predictable, thusreducing uncer tainty, which in itself can be ame-liorative. I t might, therefore, be argued that i t ispredictability,rather than behavioralm astery,thatis stress reducing. However, behavioral controldecreases arousal overand above any benefitsde-rived from the ability to predict theyoccurrenceof stressors. If anything, having foreknowledgeofwh en aversive events will occur w ithou t being ableto do anything about them increases anticipatorystress reac tions, (G un na r, 1980; M iller,1981). Butsince predictability signalssafety aswella sdanger(Seligman & Binik, 1977), it can have oppositeeffects at different points in timeraising antici-patory arousal just prior to stressful events whi lereducing arousal during safe interim periods.Being able to manage what one fears can di-minish arousal because the capability is used toreduce or toprevent pain.Butthere ismoreto theprocess of stress reduction by behavioral controlthan si 'mply curtailing painful stimuli . In someforms of behavioral mastery, previously frighten-in g events occur undiminished,but they becomenonthreatening when activated personally (G un-nar-vonG nechten, 1978). H ere it is the personalagency of causali ty,not cur tai lment of the eventsthemselves, that reduces fe ar. And in si tuations inwhich th e opportunity to wield control existsbu tis unexercised, it is the self-knowledge of copingefficacy, rather than it s application, tha t reducesanxiety arousal (G lass, Re im , & Singer, 1971).COGNITIVE CONTROL

    Apainfuleventhastwoarousalcomponentstoitdiscomfort produced by the aversive stimulationand the thought produced arousal.It is the thoughtcomponentthe arousal generated by repetitiveperturbing ideationthat accounts fo r m u c h ofhu m an distress.Asnotedearlier, peoplewhojudgethemselves1inefficacious dwell on their coping de -ficiencies and view try ing si tuations asfraughtwi thperil. They not only magnify the severity of pos-siblethreatsbut worry aboutperilsthatrarely,ifever, happen. As a result-they experience a .,highlevel of cognitively generated distress. Elevatedarousal,in turn, heightens preoccupation with per-sonal inefficacy and potential calamities.Anticipatory thought that does not exceed re-alistic bounds has functional value in that i t mo-tivates development of competencies and plans fo rdealing with foreseeable threats.But to thosewhodoubt their coping self-efficacy, the anxiousantic-ipation canbecomea preoccupation thatoftenfar

    exceedstheobjectivehazards.In an intensive anal-ysis of.acuteanxiety reactions, Beck,Laude, an dBohnert (1974) found that almost without excep-tion, frightful cognitions occur just prior to theonset of anx iety attack s. The ideation often centersaround profound coping inefficacy, which resultsin drea dful physical and social catastfophies.Because stress-inducing thought plays a para-mou nt role in hum an arousal ,self-perceptsof cop-ing efficacy can reduce the level of arousal before,during, andaf ter a trying experience. In laboratorystudies of perceived control, people who believethattheycanexercisesome influence overaversiveevents display less autonomic arousal an d impair-men tinperformanc e than thosewho believe theylackany personal control , even thou gh both groupsare subjected to th e same aversive stimulation(Averill, 1973; M iller, 1979, 1980). M ere belief incoping efficacy similarly increases abil ity to w ith-stand pain (Neufeld & Thomas, 1977).SELF-EFFICACY AS A M E D I A TI N G M E C H A N I S MThat perceived self-efficacy operatesas aco gnitivemechanism by which controllability reduces feararousal receives support in the previously citedresearch designed to enhance coping efficacy insevere phobics (Bandura & Adam s, 1977; Ba nduraet al., 1977; Banduraetal., 1980). In these studies,after completing th e various forms of t r ea tment ,phobics designated the strength oftheir perceivedefficacy in performing different tasks varying inthreat value. During later behavioral tests theyreported the intensity of fear arousal that they ex-perienced in anticipation ofperformingeachtaskand, again, while they were performing theactivity.InFigure10 the intensityoffear arousalisplot-ted as afunction ofself-efficacy strength enhancedthrough four different modes of influence.Peopleexperience high anticipatoryandperformance dis-tress on tasksin which they perceive themselvesto be inefficacious,but as the strength of their self-judged efficacy increases, their fear arousal de-clines. At high strengths of self-efficacy, threat-ening tasks are pe rforme d w ith virtually no ap-prehensiveness.Studies in which perceived self-efficacy is in-duced to differential levels (Bandura et al., inpress) shed further empirical l ight on the notionthat fear arousal arises from perceived copinginefficacy.Herethedataofinterestare theamoun tof distress phobics at different levels ofperceivedself-efficacy experience while performingt he same

    A M E R I C A N PSYCHOLOGIST F E B R U A R Y 1982 137

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    17/26

    common task (Figure 11).The relationship be-tween perceived inefficacy and subjective distressis replicated, regardless of whether self-perceptsof efficacy are instated enactiyely or vicariouslyor whether the analysis involves anticipatory orperformance fearbasedonintergrouporintrasub-ject changes. The less efficacious subjects judgethemselves to be, themore fear theyexperience.

    The generality of the relationship betweenper-ceived inefficacy and stress reactions is furthercorroborated in a study using physiological indicesof arousal (Bandura et al., in press). Elevation inblood pressure and cardiac acceleration were mea-sured in severe spider phobics during anticipation

    andperformanceofintimidating tasks correspond-ing tostrong,medium, andweak strengthofper-ceivedself-efficaciousness. In thenext phaseof thestudy,self-perceptso fefficacywere raisedtomax-imalstrength, whereupon autonomic reactions tothe same tasks were again measured. Figure 12showsthemean change from the baseline levelinheartrateandbloodpressureas afunctionof dif-ferentialstrength of self^percepts of efficacy.Subjects were viscerally unperturbed by tasksthat they regarded with utmost self-efficacious-ness. Ontasks about which they were moderatelyinsecure concerning their coping efficacy, how-ever, their heart rate accelerated and their blood

    J

    1OctUJ~J

    o7654321o

    ANTICIPATORYo--o PERFORMANCE^

    \o\\\ENACTIVE N\i i i i v

    0

    - o

    VICARIOUS

    10-20 30-40 50-6070-8090-100 10-20 30-40 50-6070-8090-100EFFICACY STRENGTH

    ct 5

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    18/26

    1 0 F - A N T I C I P A T O R Y ^ P E R F O R M A N C E

    LOW MEDIUM HIGH LOW MEDIUM HIGHI N T E R G R O U P LEVELS I N T R A S U B J E C T L E V E L S

    LOW MEDIUM LOW MEDIUMI N T E R G R O U P L E V E L S INTRASUBJECT LEVELS

    LEVEL OF PERCEIVED SELF-EFFICACYFigure 11.Mean intensity of anticipatory and performance fear arousal experienced by

    different groupso f subjectsa t different levelso fperceived self-efficacy (intergroup)and bythesame subjectsatsuccessively higher levelsofperceived self-efficacy (intrasubject). (Self-perceptsof efficacy were raisedthrougheriactivemasteryin the two leftpanels and throughmodeling in the tworight panels [Banduraet al., in press].)

    pressure rose during anticipation and performanceof the activities. After self-percepts of efficacywere fully strengthened, these same task demandswere managed unperturbedly.

    When presented with tasks in the weak self-ef-ficacy range, most subjects promptly dismissedthem as too far beyond their coping capabilitiesto even attempt. Indeed, onlya few subjectswereableto do any ofthem. Althoughtoo few instanceswere available fo r ameaningful analysis of per-formancearousal, data fromthe anticipatory phaseshed somelight on howvisceral reactions changewhen people preclude transactions with threatsthat they judge will overwhelm their coping ca-

    pabilities. Cardiac reactivity subsided, but bloodpressure continuedtoclimb.After self-perceptsofefficacy were strengthened to themaximal level,everyone performed these previously intimidatingtaskswithoutany visceral agitation.

    Heart rate is likely to be affected more quicklythan blood pressure by personal restructuringofstressful demands, which m ay explain th e differ-ential patternofphysiologicalreactivity atextremeself-inefflcaciousness. There exists some evidencethat catecholamines are released in different tem-poral patternsin response toexternalevents(Mef-fordet al.,1981).Heart rateisespecially sensitiveto momentary changes in hormonal patterns, with

    8 s

    W Bj S ( S ) S ( M ) S ( W )11 - 1

    SYSTOLIC in 6

    843D 3Zuj2 91'ozOi

    DIASTOLIC

    B, S M WB2 S ( S ) S ( M ) S ( W ) B, S M WB2 S(S )S(M)S(W)

    STRENGTH OF S E L F - P E R C E P T S O F E F F I C A C YFigure 12. Mean change from the baseline level in heart rate and blood pressure during

    anticipatory an d performance periods, as a function ofdifferential strength of self-perceptsof efficacy. (B refers tobaseline, and S, M , and W signifystrong, medium,an d weak strengths.of perceived self-efficacy, respectively. Fo reach physiological measure the figure on the leftin the panel showsthe autonomic reactions related to self-percepts that differ in strength[performancearousalatweakself-efficacy isbasedo nonlya fewsubjectswhoexhibited partialperformances];the figure on the right in the same panel shows the autonomic reactions to thesameset oftasks after self-perceptsofefficacywerestrengthenedtomaximal level[Banduraetal.,in press].)

    A M E R I C A N PSYCHOLOGIST F E B R U A R Y1982 139

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    19/26

    epinephrine, which is rapidly discharged,havingamore pronounced effect oncardiac activity thanon arterial pressure. Understanding of the phys-iological mechanisms by which self-percepts ofefficacy give rise tostressreactions can be carriedon estep further bylinking strength of perceivedself-efficacy tohormonal releases.

    Perceived self-efficacy and emotional arousalundoubtedly involve,interactive (though asym-metr ical ) effects, with coping efficacy exercisingth emuch greater sway.That is ,perceived ineffi-caciousnessin coping with potential threats leadspeople to approach such situations anxiously, andexperiencing disruptive arousal may further lowertheir sense of efficacy that they will be able toperform skillfully. However, self-percepts of effi-cacy predict avoidance behavior, whereas auto-nomic arousal bears no uniform relationship to it(Bandura, 1978a; Bolles, 1972; Herrnstein, 1969;Leitenberg,Agras,Butz,&Wincze,1971). Peopleare thus much more likely to act on their self-per-ceptsofefficacy thanonvisceral cues. This shouldcome as no surprise, since information derivedfrom past accomplishments and comparative ap-praisals is considerably more indicative of capa-bleness than are theindefinite stirrings of the vis-cera. For example, accomplished actors interprettheir brief nervousnessbeforea play as a normativesituational reaction, rather than as an indicant ofpersonal.incapability,and are in no way dissuadedby their viscera fromgoing on stage and perform-ing well what they assuredly know they can doonce theyget started.Perceived Self-Inefficacy, Futilityand DespondencyInabil i ty to influence events and social conditionsthat significantly affec t one's life can give rise tofeelings of futility and despondency aswell as toanxiety.Self-efficacy theory distinguishes betweentw ojudgmental sourcesoffutil i ty.Peoplecangiveup trying because they seriously doubt that theycan dowhat isrequired. O rthey may beassuredoftheir capabilitiesb utgiveup t ry ingbecause theyexpect their efforts to produce no results due tothe unresponsiveness, negative bias, or punitivenessof the environment. These tw oseparate sourcesoffutility have quite d if ferent causes and remedialimplications.Tochange efficacy-basedfut i l i ty re -quires development of constituent competenciesand strong percepts ofself-efficacy. Incontrast, tochange outcome-based futility necessitates chang-ingthesocial environmentsothat peoplecangain

    OUTCOMEJUDGMENT

    UIoQ

    UU

    u.u i(0

    S O C I A L A C T I V I S MP R O T E S TG R I E V A N C EMILIEU C H A N G E

    R E S IG N A T IO NAPATHY

    A S S U R E D , O P P O R T U N EA C T I O N

    S E L F - D E V A L U A T I O ND E S P O N D E N C Y

    Figure 13.Interactive effects ofself-percepts of effi-cacy and response outcomeexpectationson behavior andaffective reactions.the benefits of the competencies they already pos-sess.In any given instance behavior would be bestpredicted by considering both self-efficacy andoutcomebeliefs.4As can beseeninFigure13,dif-ferentpatternsofoutcomeand efficacy beliefs arelikely to produce different psychological effects.Ahigh senseofpersonal efficacy and aresponsiveenvironment that rewards performance attain-ments fostersassured, activeresponsiveness. Con-sider next the pattern combining high self-efficacy

    4The typesof outcomespeople expect depend largely ontheir judgments of how well they willbe able toperform ingiven situations. For example, drivers who judge themselvesinefficaciousin navigating winding mountain roads will conjureupoutcomes of wreckageandbodily injury,whereas thosewhoare fully confident oftheir driving capabilities willanticipatesweeping vistas rather than tangled wreckage. Similarly thesocial reactions people anticipate forasserting themselves de-pend on their judgments of how adroitly they can do it. Insocial, intellectual, andathleticpursuits,those whojudge them-selveshighly efficaciouswillanticipatesuccessfuloutcomes andself-doubters willexpectmediocreperformancesofthemselvesand, thus, less favorable outcomes. Foractivities in which out-comes are either inherent to the actions or are tightly linkedby socialcodes,expected outcomes cannotbedisembodied fromthevery performance judgmentsonwhich theyare conditional.Outcome expectations are dissociable from self-efficacy judg-ments when extrinsic outcomes are loosely linked to level orquality of performance. Such structural arrangements permitsocial biases to come intoplay, so the same performance at-tainments may produce variable, and often inequitable, out-comes.Expectedoutcomesarealso partially separable from self-efficacyjudgments whenextrinsic outcomesarefixedto amin-imal level of performance, as when a designated level of workproductivity produces a fixed pay but higher performancebrings noadditional monetary benefits.140 F E B R U A R Y1982 A M E R I C A N PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    20/26

    with low environmental responsiveness. Effica-cious persons who cannot achieve positive out-comes by their actions will no t necessarily ceasebehaving. Those of low efficacy will give upreadily,should theirefforts failtoproduceresults.But self-efficacious individuals will intensify theirefforts and,if necessary, try to change the envi-ronment .The pattern in which competency goes unre-warded or is punished underscores the need todifferentiate two levels of controlcontrol overoutcomes and control over th e social systems thatprescribe whatth eoutcomes willbe. In addressingthis issue G uri n (in press) and L acey (1979) giveconsiderable attention to the exerciseof influenceoversocial systems, which typically receives scantnotice in psychological analysesof controllability.Conditions combining high self-efficacy with en-vironmental unresponsiveness tend togenerate re -

    sentment, protest,andcollectiveeffortstochangeexisting practices (Bandura,1973;Short & Wol f -gang, 1972). Should change be difficult to achieve,given suitable alternatives people will desert en-vironments that are unresponsive to their effortsand pursuetheir activities elsewhere.Considering the joint influence of self-efficacyand outcome beliefs provides a basis fo r differ-entiating conditions conducive to apathy fromthose likely to induce despondency. W hen peoplehave a low sense of personal efficacy and noamount of effort by themselves or comparativeothers produces results,they become apathetic andresigned to a dreary life. The pattern in whichpeople perceive them selves as ineffectual but seesimilar others enjoying the benefits of successfuleffort is apt to give rise to self-disparagement anddepression. Evident successes of others make ithard to avoid self-criticism.In the original theory of learned helplessness(Seligman,1975),peoplebecomeinactive andde-pressed iftheir actions cannotaffect what happensto them. Because they come to expect future re -sponding to be futile, they no longertry, even insituationsinw hich they can achieve results throughtheir behavior. The reformulated theory (Abram-s o n , Seligman, &Teasdale,1978) shifts the causallocus of detr imenta l effects from belief that one'sperformances will go unre wa rded (response-out-come independence) tobelief thato ne cannot pro-duce the performances. It singles out three di-mensions in causal judgments of failure: In-ternal i tyAre failures ascribed to personal or toexternalfactors?Stabil i tyArethe ascribed causesenduring or transient?G enera li tyArethe causes

    believed to operate in many situations or only afew? Attributingone's failures to personal defi-cienciesofgeneralized an dendu ring nature, whichis postulated to be most debilitating an d depress-ing,constitutesaprofound senseofpersonal inef-ficacy. Biases toward ascribing poor performancesto basic personal deficiencies increase pronenessto depression (Seligman, Abram son, Sem m el, &von Baeyer, 1979).The adequacyof performance attainmentsde-pendson thepersonal standards against which the yare gauged. Acomprehensive theoryof depressionmust therefore be concerned no t only with th eperceived causality of fail ure but also with internalstandardsbywhichattainmentswillbe self-judgedas successesor asfailuresto begin with. Depressivereactions often arise from stringent standards ofself-evaluation, which make objective successespersonal failures. Individuals who are prone todepression impose on themselves high perfor-mance demands and devalue their accomplish-m ents because they fallshortoftheir exacting stan-dards (Kanfer & H a ge rma n , 1980;Re hm,1977;Simon, Note8) .A theory must specify when perceived ineffi-cacy willgive risetoanxietyordespondency. Thenat ure of the outcom es over which personal controlis sought is one differentiating factor. People ex -perience anxiety when they perceive themselvesillequippedtomanagepotentiallyinjuriousevents.Attenu ation or control ofaversive outcom esis cen-traltoanxiety. P eopleare saddened and depressedby their perceived inefficacy in gaining highlyvalued outcomes. I rreparable loss or failu re to gaindesired rewarding outcomes figuresprominentlyin despondency. In the extreme cases individualsbecome sochronically preoccupied wi th self-de-preciation and their sense of worthlessness tha t thepursuit of personal satisfactions becomes futile(Beck, 1973). There are certain situations, ofcourse, in which perceived inefficacy in gaininghighlyvalued outcomes can be anxiety provokingas well. W hen the valued outcomes one seeks alsoserve to forestall future aversive events, as whenfailure tosecurea jobjeopardizes one's livelihood,perceived inefficacy is both distressing and de-pressing. Because of the interdependence ofevents, both apprehension and despair often ac-company perceived personal inefficacy.Undermining Self -Ef f icacy byRelinquishing Personal ControlW hen personal con trol is easy to exercise and en-ables one to dealeffectively with everyday events,

    A M E R I C A N P S Y C H O L O G I S T F E B R U A R Y1982 141

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    21/26

    it ishighly desired. Indeed, in laboratory studiesinwhich aversive stimulican becontrolledbysim-ple responses requiring neither skills nor expen-diture of effort and entailing no risks, controlla-bility is decidedly preferred (M il ler , 1979). Butthere is anonerous side topersonal control that israrely, if ever, incorporated in most of the para-digms-designed tostudy personal control.Self-de-velopment of efficaciousness requires mastery ofknowledge and skills that can be attained onlythrough long hours of arduous work. This oftennecessitates sacrificingm a n y immediate rewards.M oreover, maintaining proficiency in given en-deavors, which constantly change with social andtechnological advances, demands continued heavyinvestment oftime, effort , and resources.Inaddition to the work of self-development,inmany situations th e exercise of personal controlcarries heavy responsibilities and risks. For ex-ample, presidents of corporations are granted con-siderable controlling power, but they must bearpersonal responsibility for the negative conse-quences of their decisions and actions, some ofwhich have widespread repercussions. These bur-densome aspects dull theappetitefo rpersonal con-trol. Attractive incentives, privileges, and headysocialrewardsare therefore needed to get peopleto seek control involving complicated skills, labo-riousresponsibilities, an d heavy risks.P R O X YCONTROLPeopleare notaversetorelinquishing control overevents that affect their livesinorder tofreethem-selves of the performance demands an d hazardsthat the exercise of control entails. Rather thanseeking personal control, they seek their securityin proxy controlwherein they can exert someinfluence over those who wield influence andpower. Partof thepriceofproxy controlisrestric-tionofone's ownefficacyand a vulnerable securitythat restson thecompetenciesan d favorsofothers.

    Perceived inefficacy fostersdependence onproxycontrol* which further reduces opportunities tobuild therequisite skillsfor efficaciousaction. Theinfluential role of comparative self-ability evalu-ation in proxy control is revealed in studies byMiller and her associates (Miller, 1980). Peoplewho are led to believe that they possess superiorcopingability handle potential threats themselves,whereas those who believe themselves to be lessskilled readily yield controltootherstocopewiththe aversive environment. The dependent onesenjoy the protective benefits without the perfor-mance demands and attendant stresses, and the

    controllersdo theworkand suffer th edistress overrisks offailure.U N D E R M I N E R SO F P E R S O N A L EFFICACYThepreceding discussion focusedonpersonal inef-ficacy-arising from th e costs an d demands of ef-ficaciousbehavior. M a nyfactors operate inevery-da y life to undermine efficacious use of theknowledgeand skillsthat people possess. In an in-formative program ofresearch onillusory incom-petence,Langer (1979) has given us abetter un-derstanding of the diverse conditions that impairtheexerciseofcapabilities:Situational factors thatoftenaccompany poor performance can inthem-selves instill a sense of incompetence that is un-warranted. The mere presence of a highly confi-dent individual undermineseffective use ofroutineskills. Attending to what isstrange in new tasks,rather than what is famil iar and clearly withinone's range of capability, m ay similarly hinderperformance. And when people are cast in sub-ordinate roles or are assigned inferior labels, im -plying limited competence, they perform activitiesatwhich theya reskilled less well than when theydo notbear thenegative labelsor thesubordinaterole designations.

    The intervening mechanism through which de-moralizingconditions undermine effectiveuse ofwell-established skills remainsto beclarified. Stud-ie sin which self-percepts are measured under in-duced illusoryself-efficacy suggest that perceivedinefficacy, with itsconcomitant effects on choicebehavior,effortexpenditure, persistence, and self-debilitating thought,may be the operative mech-anism. This evidence comes from experimentsdemonstrating that changesinphysical staminaincompetitive situationsarepartly mediated throughself-percepts of efficacy (Weinberg et al., 1979;Weinberg et al., 1980). The lower the illusorilyinstated self-percepts of physical efficacy, th eweakerthecompetitive endurancein newphysicalactivities. Even the mere sight of a formidablelooking oponent instills lower self-percepts of ef-ficacythan does one wholooks lessimpressive.Asmightbeexpected preexistingself-percepts of ef-ficacyhave greatest impact oninitial competitiveperformance, whereas socially induced self-per-cepts affect the subsequent course of competitiveendurance (Weinberg, Gould, Yukelson, &Jack-son, inpress).Thepowerofself-efficacy belief overbrawnisunderscored furtherbyevidence that self-percepts of physical efficacy illusorily boosted infemales and illusorily diminished in males oblit-

    142 F E B R U A R Y1982 A M E R I C A N PSYCHOLOGIST

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    22/26

    crateslargepreexisting sexdifferences inphysicalstrength (Weinbergetal 1979).Collective EfficacyThe discussion thus far has focused mainly on thepersonal effects ofperceived self-efficacy. Peopledo not live their lives associal isolates.Manyofthe challenges and difficulties they face reflectgroup problems requiring sustained collective ef-fortto produce any significant change. The strengthof groups, organizations, an d even nations liespartly in people'ssense ofcollectiveefficacy thatthey can solve their problems and improve theirlives throughconcertedeffort.Perceived collectiveefficacy will influence what people choose to doasa group, how much effort they put into it, andtheirstaying power when group efforts failtopro*duce results. It should be noted that knowledge ofpersonal efficacy is not unrelated to perceivedgroupefficacy.Aswillbeshown shortly, collectiveefficacy isrooted in self-efficacy. Inveterate self-doubters are not easily forged into a collectivelyefficacious force.COLLECTIVEEFFICACY ANDSOCIAL CHANGEThe taskofsocial changehasnever been aneasyone. Those who seek to alter social systemsandtheir practices encounter opposition from powerholders and influential vested interests. Shouldchallengers resort to forceful social protest, puni-tive sanctionscan bebroughttobear against them.Thenumerous obstaclesandcoercive threatsdeterattempts to alter social conditions that adverselyaffect human lives.

    It isoftensaid that hopelessness breeds militantsocial action. However, the evidence would seemtodispute this view. Consistent with self-efficacytheory, studiesof socialand political activism in-dicatethatdetrimentalconditions prompt forcefulaction, not in those who have losthope, but in themore able members whose efforts at social andeconomic betterment havemetwithatleast somesuccess(Bandura, 1973). Consequently, they havereasontobelieve that some changescan bebroughtabout through forcefulgroup action.Among th e members ofdissident groups, thosewho protest social inequities, compared to non-participants, are generally better educated, havegreater self-pride, have a stronger belief in theirability to influenceeventsi ntheir lives,a nd favorcoercive measures, ifnecessary, to improve theirliving conditions (Caplan, 1970; Crawford & Na-

    ditch, 1970).Inmany nations university students,rather thanthe severely underprivileged segmentsof thesociety,are thespearhead ofpolitical activ-ism (Lipset, 1966). They are the ones whoofteninitiate the protest movements that eventuallyforcesocialreformsand topplegovernments. Re-sults of comparative studies indicate that peoplewho are most disposed to social action generallycome from familial backgroundsi nwhich the ex-ercise of social influence ha s been modeled an drewarded (Keniston, 1968; Rosenhan,1970).Mod-eling influences,however, which servea s a majorvehicle of social diffusion, ca n substantially alterthe personal and social correlates of activism overtime. Those who initiate collective action usuallydiffer incharacteristics from later adopters.

    Research includingefficacy probes speaks moredirectly to the issueofwhether perceived efficacyserves as one mechanism through which social dis-content gives rise tosocial activism. M u ch ofthisresearchreliesonglobalindicesof efficacy, oftenblending mixed contents (Balch, 1974). Even so,therelationships obtainedarefairlyconsistent.Thehigher the perceived efficacy,the greater the pro-pensity to social activism (Forward &Williams,1970; Marsh, 1977; Muller, 1972, 1979).However,sharper empirical testso ftheorywill require par-ticularizedmultifacetedmeasuresofefficacy, tap-ping perceived capabilities fo r fashioning and ex-ecuting different types of strategies designed toinfluence the courseofsocial events. Since socialoutcomes are typically achieved in concert withothers, perceptions of group as well as personalefficacywarrant examination.U N D E R M I N E R SOF COLLECTIVE EFFICACYRapidly changing conditions, which impair thequality of social life and degrade the physical en-vironment, call for wide-reaching solutions to hu-man problems and greater commitment to sharedpurposes. Such changes can be achieved onlythrough th emutualeffort ofpeople whohave th eskills, th e sense ofcollective efficacy, and the in-centives to shape the direction of their future en-vironment.As theneedforefficaciousgroup actiongrows,sodoesthesenseofcollectivepowerlessness.

    One can point to a number of factorsthat serveto undermine the development ofcollective effi-cacy.Modern lifeisextensively regulated bycom-plexphysical technologies that most people neithercomprehend nor believe they can do much to in-fluence. Pervasive dependence on technologiesthat govern major aspects of life imposes depen-denceon specialized technicians. The social ma-

    A M E R I C A N PSYCHOLOGIST F E B R U A R Y 1982 143

  • 8/12/2019 Bandura_1982_Self-Efficacy Mechanism in Human Agency

    23/26

    chinery of asociety is noless challenging. Layersof bureaucratic structures thwart effective social"action.Event hemoreefficacious individuals,whoare no t easilydeterred, find their efforts bluntedby mazy organizational mechanisms that diffuseand obscure responsibility. Rather than developingthe means for shaping their future , most peoplegrudginglyrelinquish control to technical special-ists and topublic officials.Effectiveactionfo rsocial change requires merg-ing diverse self-interests in support of commongoals.Disagreements among different constituen-cies that have a personal stake in the mattersofconcern create additional obstacles to successfulgroup action. Recent years have witnessed growingsocial fragmentation irito narrow-interest constit-uencies. Pluralism istaking the form of militantfactionalism. As aconsequence it iseasier toenlistdiverse factions to block coursesofaction than tomerge them intoa unifiedforceforsocial change.

    Inaddition to thedifficulties in enlisting sharedpurposes andcollectiveeffort intheir service, theinstitutions that are th e objectsof change mounttheirownforcefulcountermeasures. Becauseof themany conflicting forces that come into play, at-tempts to produce socially significant changes donot bring quick successes. Long delays betweenaction and noticeable results discourage manyoftheadvocates along theway,even though changesof long-termsignificancemayeventually occur. I tis ;difficult to develop and sustain a sense ofcol-lective efficacy when th e effects of group effortare notreadily noticeable.

    To complicate matters fur ther , life in today'ssocieties isincreasingly af fected by transnationalinterdependencies (Keohane &Nye, 1977). Whathappens in one p&rt of the world can affect thewelfare ofvast populations elsewhere". There arenohandy direct mechanisms by whichpeoplecanexercise reciprocal influenceontransnational sys-te'msthat affec t their daily lives.P rofound globalchangesburgeoning populations, shrinking re-sources,deteriorating environmentsarecreatingnew realities requiring transnational remedies.The subjectofcollective efficacy callsfor broadandcomprehensive research effort. Advancementinthisfieldofstudyrequires developmentof suit-able toolsfo rgauging groups' perceptions oftheirefficacy toachieve varying levelsofresults. Great-est progress will be made in elucidating the de-velopment, decline, and restoration of collectiveefficacy arid how it affects groupfunctioning, ifmeasures of perceived group efficacy are tiedclosely toexplicit indices ofgroup performance.

    National surveys havebeenconducted periodi-

    cally ofpeople'sgeheralsenseofpolitical efficacy,their confidence in their social institutions, andhow they view th e competence of those theychoose to lead them.Though such omnibus mea-sures leave much to be desired, they do provideevidence ofgrowing erosionofperceived efficacyof the citizenry and its social institutions to solvehumanproblems (Guest, 1974; Lipset &Schneider,1982).FACTIONAL EFFICACY AND COLLECTIVE ENDEAVOR

    I nanalyzing impedimentstohuman endeavors,i tisall tooeasy tolose sight of the fact that humaninfluence, whether individual orcollective, oper-ates in reciprocal, rather than in unidirectional,ways (Bandura, 1978b; Cairns, 1979; Endler &M agnusson,1976; Pervin&Lewis, 1978). Althoughth e degree of reciprocality m ay vary from onedomain ofactivity toanother, social transactionsare rarely unilateral. The amountofimbalanceofsocial power partly dependson theextenttowhichpeopleexercise the influence that istheirs to com-mand. The less they bring their influence to bearonothers, the more control they relinquish to them.

    It is theinternal barriers createdbyperceptionsof collective inefficacy that are especially perni-ciousbecause they are more demoralizing and be-haviorally self-debilitating than are external im-pediments. People whohave asenseof collectiveefficacy will mobilize their efforts and resourcestoCopewith externalobstaclesto thechangestheyseek. Butthose convinced of their inefficacy willcease trying even though changes are attainablethrough concerted effort.Thesocial system is not a monolith.Rather, itcomprises numerousconstituencies, eachvying fo rpower and lobbying fo r its ow ninterests. In thiscontinual interplay one and the same faction istransmuted from achallenger of the systemto aninfluentialconfederatein thesystem opposing rivalfactions, depending on the issues at stake. Thus,for example, the tobacco constituency fights thesystem in federal efforts to curtail s