Self-Efficacy: Toward A Unifying Theory of Behavioral Change

Self-efficacy: Toward a Unifying Theory of Behavioral Change
Albert Bandura
Stanford University
The present article presents an integrative theoretical framework to explain

and to predict psychological changes achieved by different modes of treatment.
This theory states that psychological procedures, whatever their form, alter the
level and strength of eel J-e ficac y. It is hypothesized that expectations of per-
sonal efficacy determine whether coping behavior will be initiated, how much
effort will be expended, and how long it will be sustained in the face of ob-
stacles and aversive experiences. Persistence in activities that are subjectively
threatening but in fact relatively safe produces, through experiences of mastery,
further enhancement of self-efficacy and corresponding reductions in defensive
behavior. In the proposed model, expectations of personal efficacy are derived
from four principal sources of information : performance accomplishments,
vicarious experience, verbal persuasion, and physiological states. The more de-
pendable the experiential sources, the greater are the changes in perceived self-
efficacy. A number of factors are identified as influencing the cognitive
processing of efficacy information arising from enactive, vicarious, exhortative,
and emotive sources. The differentia! power of diverse therapeutic procedures is
analyzed in terms of the postulated cognitive mechanism of operation. Findings
are reported from microanalyses of enactive, vicarious, and emotive modes of
treatment that support the hypothesized relationship between perceived self-
efficacy and behavioral changes. Possible directions for further res‘earch are
discussed.
Current developments in the field of be- apparent divergence of theory and practice
havioral change reflect two major divergent can be reconciled by postulating that cognitive
trends. The difference is especially evident in processes mediate change but that cognitive
the treatment of dysfunctional inhibitions events are induced and altered most readily by
and defensive behavior. On the one hand, the experience of mastery arising from effective
mechanisms by which human behavior is performance. The distinction between process
acquired and regulated are increasingly for- and means is underscored, because it is often
mulated in terms of cognitive processes. On assumed that a cognitive mode of operation
the other hand, it is performance-based pro- requires a symbolic means of induction.
cedures that are proving to be most powerful Psychological changes can be produced through
for effecting psychological changes. As a con- other means than performance
sequence, successful performance is replacing accomplishments. Therefore, the explanatory
symbolically based experiences as the prin- mechanism developed in this article is designed
ciple vehicle of change. to account for changes in behavior resulting
The present article presents the view that from diverse modes of treatment.
changes achieved by different methods derive
from a common cognitive mechanism. The
Cog›tifiue Locus oJ 0 Aeration
Psychological treatments based on learning
The research by the author reported in this principles were originally conceptualized to
article was supported by Research Grant M-5162
from the National Institutes of Health, United operate through peripheral mechanisms. New
States Public Health Service. behavior was presumably shaped automat-
Requests for reprints should be sent to Albert ically by its effects. Contingency learning
Bandura, Department of Psychology, Stanford through paired stimulation was construed in
Uni- versity, Stanford, California 94305.
192 ALBERT BANDURA
connectionist terms as a process in which re- structed from observing the effects of one's
sponses were linked directly to stimuli. Al- actions rather than from the examples pro-
tering the rate of preexisting behavior by re- vided by others.
inforcement was portrayed as a process Changes in behavior produced by stimuli
wherein responses were regulated by their that either signify events to come or indicate
immediate consequences without requiring probable response consequences also have
any conscious involvement of the responders. been shown to rely heavily on cognitive rep-
Growing evidence from several lines of re- resentations of contingencies. People are not
search altered theoretical perspectives on how much affected by paired stimulation unless
behavior is acquired and regulated. Theo- they recognize that the events are correlated
retical formulations emphasizing peripheral (Dawson & Furedy, 1976 ; Grings, 1973) .
mechanisms began to give way to cognitively Stimuli influence the likelihood of a be-
oriented theories that explained behavior in havior’s being performed by virtue of their
terms of central processing of direct, vicari- predictive function, not because the stimuli
ous, and symbolic sources of information. are automatically connected to responses by
Detailed analysis of the empirical and con- their having occurred together. Reinterpreta-
ceptual issues (see Bandura, 1977) falls tion of antecedent determinants as predictive
beyond the scope of the present article. To cues, rather than as controlling stimuli, has
summarize briefly, however, it has now been Shifted the locus of the regulation of be-
amply documented that cognitive processes havior from the stimulus to the individual.
play a prominent role in the acquisition and The issue of the locus at which behavioral
retention of new behavior patterns. Transitory determinants operate applies to reinforce-
experiences leave lasting effects by being ment influences as well as to antecedent en-
coded and retained in symbols for memory vironmental stimuli. Contrary to the common
representation. Because acquisition of review that behavior is controlled by its im-
sponse information is a major aspect of learn- mediate consequences, behavior is related to ing,
much human behavior is developed its outcomes at the level of aggregate con-
through modeling. From observing others, one sequences rather than momentary effects
forms a conception of how new behavior pat- (Baum, 1973) . People process and synthesize
terns are performed, and on later occasions feedback information from sequences of
the symbolic construction serves as a guide events over long intervals about the situa- for
action ( Bandura, 1971 ) . The initial ap- tional circumstances and the patterns and
proximations of response patterns learned rates of actions that are necessary to produce
observationally are further refined through given outcomes. Since consequences affect be-
self-corrective adjustments based on informa- havior through the influence of thought, be- tive
feedback from performance. fiefs about schedules of reinforcement can
Learning from response consequences is exert greater influence on behavior than the
also conceived of largely as a cognitive reinforcement itself (Baron, Kaufman, &
process. Consequences serve as an unarticu- Stauber, 1969 ; Kaufman, Baron, & Kopp,
lated way of informing performers what they 966) . Incidence of behavior that has been
must do to gain beneficial outcomes and to positively reinforced does not increase if in-
avoid punishing ones. By observing the dif- dividuals believe, based on other information,
ferential effects of their own actions, in- that the same actions will not be rewarded
dividuals discern which responses are ap- OR future occasions (Estes, 1972 ) ; and the
propriate in which settings and behave ac- same consequences can increase, reduce, or
cordingly ( Dulany, 1968) . Viewed from the have no effect on incidence of behavior de-
cognitive framework, learning from differ- pending on whether individuals are led to
ential outcomes becomes a special case of believe that the consequences signify correct
observational learning. In this mode of con- responses, incorrect responses, or occur non-
veying response information, the conception contingently (Dulany, 1968) .
of the appropriate behavior is gradually con- The discussion thus far has examined the
SELP-EPPICACY THEORY
193
role of cognition in the acquisition and PERSON BEHAVIOR OUTCOW

regulation of behavior. Motivation, which is
primarily concerned with activation and per-
sistence of behavior, is also partly rooted in
cognitive activities. The capacity to repre- Pigure J. Diagrammatic representation of the difler-
sent future consequences in thought provides ence between efficacy expectations and outcome ex-
one cognitively based source of motivation.
Through cognitive representation of future
outcomes individuals can generate current and in integrating the results accompanying
mo- tivators of behavior. Seen from this per- diverse modes of treatment. The present
spective, reinforcement operations asect be- article outlines a theoretical framework, in
havior largely by creating expectations that which the concept of sefJ-eQcacy is assigned
behaving in a certain way will produce antici- a central role, for analyzing changes achieved
pated benefits or avert future difficulties in fearful and avoidant behavior. The ex-
(Bolles, 1972b). In the enhancement of planatory value of this conceptual system
previously learned behavior, reinforcement is is then evaluated by its ability to predict
conceived of mainly as a motivational de- behavioral changes produced through dif-
vice rather than as an automatic response ferent methods of treatment.
strengthener.
A second cognitively based source of mo-
tivation operates through the intervening in-
8uences of goal setting and self-evaluative
reactions (Bandura, 1976b, 1977). Self-
motivation involves standards against which The present theory is based on the prin-
to evaluate performance. By making self- cipal asssumption that psychological pro-
rewarding reactions conditional on attaining a cedures, whatever their form, serve as means
certain level of behavior, individuals create of creating and strengthening expectations
self-inducements to persist in their efforts of personal eEicacy. Within this analysis,
until their performances match self- e&cacy expectations are distinguished from
prescribed standards. Perceived negative respon utcome expectancies. The di8er-
discrepancies between performance and ence is presented schematically in Figure 1.
standards create dissatisfactions that motivate An outcome expectancy is defined as a
corrective changes in behavior. Both the person’s estimate that a given behavior will
anticipated sa’tislactions of desired lead to certain outcomes. An eRcacy ex-
accomplishments and the negative appraisals pectation is the conviction that one can
of insufficient performance thus provide successfully execute the behavior required to
incentives for action. Having accomplished a produce the outcomes. Outcome and efficacy
given level of performance, individuals expectations are differentiated, because in-
often are no longer satisfied with it and make dividuals can believe that a particular course
further self-reward contingent on higher of action will produce certain outcomes, but
attainments. if they entertain serious doubts about whether
The reconceptualization of human learn- they can perform the necessary activities
ing and motivation in terms of cognitive such information does not influence their
processes has major implications for the behavior.
mechanisms through which therapeutic pro- In this conceptual system, expectations of
cedures alter behavioral functioning. Al- personal mastery aifect both initiation and
though the advances in cognitive psychology persistence of coping behavior. The strength
are a subject of increasing interest in specu- of people’s convictions in their own
lations about behavioral change processes, effectiveness is likely to aifect whether they
will even try to cope with given situations. At
few new theories of psychotherapy have been
this initial level, perceived self-e&cacy
proposed that might prove useful in stimu- influences
lating research on explanatory mechanisms
19 ALBERT
choice of behavioral settings. People fear and

sonal mastery. Moreover, expectations are
tend to avoid threatening situations they
usually assessed globally only at a single
believe exceed their coping skills, whereas
point in a change process as though they
they get involved in activities and behave
represent a static, unidimensional factor.
assuredly when they judge themselves cap-
Participants in experiments of this type are
able of handling situations that would other-
simply asked to judge how much they expect
wise be intimidating.
to benefit from a given procedure. When
Not only can perceived self-efficacy have asked to make such estimates, participants
directive influence on choice of activities and assume, more often than not, that the benefits
settings, but, through expectations of eventual will be produced by the external ministra-
success, it can affect coping efforts once they tions rather than gained through the de-
are initiated. Efficacy expectations determine velopment of self-efficacy. Such global mea-
how much effort people will expend and how sures reflect a mixture of, among other things,
long they will persist in the face of obstacles hope, wishful thinking, belief in the potency
and aversive e.xperiences. The stronger the of the procedures, and faith in the therapist. It
perceived self-efficacy, the more active the therefore comes as no surprise that outcome
efforts. Those who persist in subjectively expectations of this type have little relation to
threatening activities that are in fact magnitude of behavioral change (Davison &
relatively safe will gain corrective experiences Wilson, 1973, Lick & Bootzin, 1975).
that reinforce their sense of efficacy, thereby
Efficacy expectations vary on several di-
eventually eliminating their defensive behavior.
mensions that have important performance
Those who cease their coping efforts
implications. They differ in magnitude. Thus
prematurely will retain their self- debilitating
when tasks are ordered in level of difficulty,
expectations and fears for a long time.
the efficacy expectations of different individ-
The preceding analysis of how perceived uals may be limited to the simpler tasks, ex-
self-efficacy influences performance is not tend to moderately difficult ones, or include
meant to imply that expectation is the sole even the most taxing performances. Efficacy
determinant of behavior. Expectation alone expectations also differ in generalit y. Some
will not produce desired performance if the experiences create circumscribed mastery ex-
component capabilities are lacking. Moreover, pectations. Others instill a more generalized
there are many thinp•s that people can do sense of efficacy that extends well beyond the
with certainty of success that they do not specific treatment situation. In addition,
perform because they have no incentives expectancies vary in strength. Weak expecta-
to do so. Civen appropriate skills and ade- tions are easily extinguishable by discon-
quate incentives, however, efficacy expecta- firming• experiences, whereas individuals who
tions are a major determinant of people’s possess strong expectations of mastery will
choice of activities, how much effort they will persevere in their coping efforts despite dis-
expend, and of how long they will sustain confirming experiences.
effort in dealing with stressful situations.
An adequate expectancy analysis, there-
fore, requires detailed assessment of the
Etinicnsions oJ E fficac y Es per.Nations magnitude, generality, and strength of ef-
ficacy expectations commensurate with the
precision with which behavioral processes are
Empirical tests of the relationship between measured. Both efficacy expectations and per-
expectancy and performance of threatening formance should be assessed at significant
activities have been hampered by inadequacy junctures in the change process to clarify
of the expectancy analysis. In most studies their reciprocal effects on each other. Mastery
the measures of expectations are mainly con- expectations influence performance and are,
cerned with people’s hopes for favorable out- in turn, altered by the cumulative effects of
comes rather than with their sense of per- one's efforts.
SELF-EFFICACY 19
Sources oJ E ficac y Expectations

because it is based on personal mastery ex-
In this social learning analysis, expecta- periences. Successes raise mastery expecta-
tions of personal efficacy are based on four tions; repeated failures lower them, par-
major sources of information: performance ticularly if the mishaps occur early in the
accomplishments, vicarious experience, course of events. After strong efficacy ex-
verbal persuasion, and physiological states. pectations are developed through repeated
success, the negative impact of occasional
Figure 2 presents the diverse influence failuresIS likely to be reduced. Indeed, oc-
procedures commonly used to reduce
casional failures that are later overcome by
defensive behavior and presents the principal determined effort can strengthen self-moti-
source through which each treatment operates
vated persistence if one finds through ex-
to create expectations of mastery. Any given
method, depending on how it is applied, may perience that even the most difficult ob-
of course draw to a lesser extent on one or more stacles can be mastered by sustained effort.
other sources of efficacy information. For ex- The effects of failure on personal efficacy
ample, as we shall see shortly, performance- therefore partly depend on the timing and
based treatments not only promote be- the total pattern of experiences in which the
havioral accomplishments but also failures occur.
fear arousal, thus authenticating self-efficacy Once established, enhanced self-efficacy tends
through enactive and arousal sources of in- to generalize to other situations in which
formation. Other methods, however, provide performance was self-debilitated by pre-
fewer ways of acquiring information about occupation with personal inadequacies (Ban-
one’s capability for coping with threatening dura, Adams, & Beyer, in press; Bandura,
situations. By postulating a common mecha- Jeffery, 4t Gajdos, 1975). As a result, im-
nism of operation, this analysis provides a provements in behavioral functioning transfer
conceptual framework within which to study not only to similar situations but to activities
behavioral changes achieved by diBerent that are substantially different from those on
modes of treatment. which the treatment was focused. Thus, for
Performance accomplishments. This example, increased self-efficacy gained through
source of efficacy information is especially rapid mastery of a specific animal phobia can
influential
EFFICACY EXPECTATIONS
SOURCE
MOOE OF INDUCTION
_-PARTICIPANT MOOELING
-- --PERFORMANCE DESENSITIZAFION
PERFORMANCE •¿ ”PERFQRMANCE EXPOSURE
”SELF- INSTRUCTED PERFQWAtiCE
-LIVE MODELING
- -- -SYMBOLIC MODELING
VICARIOUS EXPERIENCE
• -EXHORTATION
¿- - -SELF- INSTRUCTION
VERBAL PERSUASION ” “INTERPRETIVE TREAT MENTS
_ -ATTRIBUTION
----RELAXATlOti, BIOFEEDBACK
EMOTIONAL AROUSAL • - SYMBOLIC DESENSITIZATION
“”SYMBOLIC EXPOSURE
higure 2. Major sources of elficacy inf.ormation and the principal sources through which diderent
modes of treatment operate.
19 ALBERT
increase coping efforts in social situations Sherman, 1972 ; Strahley, 1966) . Physiolog-
as well as reduce fears of other animals. ical measures yield similar results. Symbolic
However, the generalization effects occur desensitization reduces autonomic responses
most predictably to the activities that are to imagined but not to actual threats, whereas
most similar to those in which self-efficacy performance desensitization eliminates auto-
was restored by treatment (Bandura, Blanch- nomic responses to both imagined and actual
ard, & Ritter, 1969 ) . threats ( Barlow, Leitenberg, Agras, & Wincze,
Methods of change that operate on the 1969) . The substantial benefits of successful
basis of performance accomplishments convey performance are typically achieved in less
efficacy information in more ways than simply time than is required to extinguish arousal to
through the evidence of performance im- symbolic representations of threats.
provements. In the course of treatments em- More recently, avoidance behavior has
ploying modeling with guided performance, been treated by procedures involving massive
participants acquire a generalizable skill for exposure to aversive events. In this approach,
dealing successfully with stressful situations, intense anxiety is elicited by prolonged ex-
a skill that they use to overcome a variety of posure to the most threatening situations
dysfunctional fears and inhibitions in their ev- and sustained at high levels, without relief,
eryday life ( Bandura et a1., in press; Bandura et until emotional reactions are extinguished.
al., 1975) . Having a serviceable coping skill Several investigators have compared the
at one’s disposal undoubtedly contributes to relative success of prolonged exposure to
one's sense of personal efficacy. Behavioral aversive situations in imagery and actual en-
capabilities can also be enhanced through counters with them in ameliorating chronic
modeling alone (Bandura, 1971 ; Flanders, agoraphobias. Real encounters with threats
1968). However, participant modeling pro- produce results decidely superior to imagined
vides additional opportunities for translating exposure, which has weak, variable effects
behavioral conceptions to appropriate actions (Emmelkamp & Wessels, 1975 ; Stern & Marks,
and for making corrective refinements toward 1973 ; Watson, Mullett, & Pillay, 1973) .
the perfection of skills. Prolonged encounters that ensure behavioral
Most of the treatment procedures de- improvements are more effective than
veloped in recent years to eliminate fearful distributed brief encounters that are likely to
and defensive behavior have been imple- end before successful performance of the
mented either through performance or by activity is achieved ( Rabavilas, Boulougouris,
symbolic procedures. Regardless of the & Stefanis, 1976).
methods involved, results of comparative The participant modeling approach to the
studies attest to the superiority of perform- elimination of defensive behavior utilizes
ance-based treatments. In the desensitization successful performance as the primary vehicle
approach devised by Wolpe ( 1974) , clients of psychological change. People displaying
receive graduated exposure to aversive events intractable fears and inhibitions are not
in conjunction with anxiety reducing ac- about to do what they dread. In implement-
tivities, usually in the form of muscular ing participant modeling, therapists therefore
relaxation. A number of experiments have structure the environment so that clients can
been reported in which relaxation is paired perform successfully despite their incapaci-
with scenes in which phobics visualize ties. This is achieved by enlisting a variety of
themselves engaging in progressively more response induction aids, including preliminary
threatening activities or with enactment of modeling of threatening activities, graduated
the same hierarchy of activities with the tasks, enactment over graduated temporal
actual threats. Findings based on different intervals, joint performance with the thera-
types of phobias consistently reveal that pist, protective aids to . reduce the likelihood
performance desensitization produces sub- of feared consequences, and variation in the
stantially greater behavioral change than does severity of the threat itself ( Bandura, Jeffery,
symbolic desensitization (LoPicollo, 1970 ; & Wright, 1974) . As treatment progresses,
SELF-EFFICACY 19
the supplementary aids are withdrawn so

persuade themselves that if others can do it,
that clients cope effectively unassisted. Self-
they should be able to achieve at least some
directed mastery experiences are then ar- improvement in performance (Bandura &
ranged to reinforce a sense of personal ef-
Barab, 1973). Vicarious experience, relying
ficacy. Through this form of treatment in- as it does on inferences from social com-
capacitated people rapidly lose their fears,
parison, is a less dependable soap ée of in-
they are able to engage in activities they formation about one’s capabilities than is
formerly inhibited, and they display general-
direct evidence of personal accomplishments.
ized reductions of fears toward threats Consequently, the efficacy exjiectations in-
beyond the specifically treated conditions
duced by modeling alone are likely to be
(Bandura, 1976a) . weaker and more vulnerable to change.
Participant modeling has been compared
A number of modeling variables that are
with various symbolically based treatments.
apt to affect expectations of personal efficacy
These studies corroborate the superiority of
have been shown to enhance the disinlubiting
successful performance facilitated by model-
influence of modeling procedures. Phobics
ing as compared to vicarious experience
benefit more from seeing models overcome
alone (Bandura et a1., 1969; Blanchard,
their difficulties by determined effort than
1970b; Lewis, 1974 ; Ritter, 1969 ; Riiper,
from observing facile performances by adept
Rachman, & Marks, 1975), to symbolic
models (Kazdin, 1973 ; Meichenbaum, 1971).
desensitization (Bandura et al., 1969;
Showing the gains achieved by effortful
Litvak, 1969), and to imaginal modeling in
coping behavior not only minimizes for ob-
which clients visualize themselves or others
servers the negative impact of temporary
coping successfully with threats (Thase &
distress but demonstrates that even the most
Moss, 1976). When participant modeling is
anxious can eventually succeed through per-
subsequently administered to those who
severance. Similarity to the model in other
benefit only partially from the symbolic
characteristics, which increases the personal
procedures, avoidance behavior is
relevance of vicariously derived information,
thoroughly eliminated within a brief period.
can likewise enhance the effectiveness of
The findings summarized above are con- symbolic modeling (Kazdin, 1974b).
sistent with self-efficacy theory, but they do Modeled behavior with clear outcomes
not shed much light on the mechanism by conveys more efficacy information than if the
which specific mastery experiénces produce effects of the modeled actions remain am-
generalized and enduring changes in biguous. In investigations of vicarious pro-
behavior. Verification of the operative cesses, observing one perform activities that
mechanism requires experimental evidence meet with success does, indeed, produce
that experienced mastery does in fact alter the greater behavioral improvements than wit-
level and strength of self-efficacy and that nessing the same performances modeled
self-efficacy is, in turn, linked to behavior. without any evident consequences (Razdin,
We shall re- turn later to research that 1974c, 1975). Diversified modeling, in which
addresses itself specifically to the linkages the activities observers regard as hazardous
between treatment procedures, perceived are repeatedly shown to be safe by a variety
self-efficacy, and behavior. of models, is superior to exposure to the same
Vicarious experience. People do not rely on performances by a single model (Bandura 4t
experienced mastery as the sole source of in- Menlove, 1968 ; Kazdin, 1974a, 1975, 1976).
, formation concerning their level of self- If people of widely differing characteristics
efficacy. Many expectations are derived can succeed, then observers have a reason-
from vicarious experience. Seeing others able basis for increasing their own sense of
perform. threatening activities without self-efficacy.
adverse consequences can generate The pattern of results reported above of-
expectations in observers that they too will fers at least suggestive support for the view
improve if they intensify and persist in their that exemplitications of success through sus-
esorts. They
19 ALBERT
tained effort with substantiating comparative Linder, 1971 ) . As in the “placebo” studies, it
information can enhance observers’ percep- is difficult to make conclusive interpretations
tions of their own performance capabilities. because the outcome expectations induced
Research will be presented below that bears suggestively are not measured prior to the
more directly on the proposition that model- assessment of behavior changes, if at all.
ing procedures alter avoidance behavior Simply informing participants that they will
through the intervening influence of efficacy or will not benefit from treatment does not
expectations. mean that they necessarily believe what they
Verbal persuasion. In attempts to influence are told, especially when it contradicts their
human behavior, verbal persuasion is widely other personal experiences. Moreover, in the
used because of its ease and ready avail- studies just cited the verbal influence is
a1›ili ty. People are led, through sugye.stion, aimed mainly at raising outcome expectations
into believing they can cope successfully with rather than at enhancing self-efficacy. It is
what has over whelmed them in the past. changes on the latter dimension that are most
Efficacy expectations induced in this manner relevant to the theory under discussion.
are also likely to be z'eaker than those arising Although social persuasion alone may have
from one’s own accomplishments because definite limitations as a means of creating an
they do not provide an authentic experiential enduring sense of personal efficacy, it can
base for them. In the face of distressing contribute to the successes achieved through
threats and a long history of failure in coping corrective performance. That is, people who
with them, whatever mastery expectations are socially persuaded that they possess the
are induced by suggestion can be readily capabilities to master difficult situations and are
extinguished by disconfirminp• experiences. provided with provisional aids for effective
Results of several lines of research attest action are likely to mobilize greater eflort than
to the limitation of 1•rocerlures that attempt those who receive only the performance aids.
to instill outcome expectations in people However, to raise by persuasion expectations
simply by telling them what to expect. In of personal competence without arrangin*
laboratory studies, “placebo” conditions de- conditions to facilitate effective performance
signed suguestivi•1y to raise expectations of will most likely lead to failures that discredit the
improvement produce little change in re- persuaders and further undermine the
f ractor y behavior (Lick & Bootzin, 1975 ; recipients’ perceived seli-efficacy. It is therefore
ñloore, 1965 ; Paul, 1966 ) . \Vhether this is the interactive, as well as the independent,
due to the low credibility of the suggestions effects of social persuasion on self-efficacy that
or to the weakness of the induced expecta- merit experimental consideration.
tions cannot be determined from these stud- Emotional arousal. Stressful and taxing
ies, because the expectations were not situations generally elicit emotional arousal that,
measured. depending on the circumstances, might have
Numerous experiments have been con- informative value concerning personal
ducted in which phobics receive desensitiza- competency. Therefore, emotional arousal is
tion treatment without any expectancy in- another constituent source of information
formation or with suggestions that it is either that can affect perceived self-efficacy in cop-
highly efficacious or ineffective. The differing with threatening situations. People rely
ential outcome expectations are verbally in- partly on their state of physiological arousal
duced pric›r to, during, or immediately after in judging their anxiety and vulnerability to
treatment in the various studies. The findings stress. Because high arousal usually debili-
generally show that desensitization reduces tates performance, individuals are more likely to
phobic behavior, but the outcome expectancy expect success when they are not beset by
manipulations have either no effect or weak, aversive arousal than if they are tense and
inconsistent ones (Howlett & Nawas, 1971 ;
viscerally agitated. Fear reactions gen-
McGlynn & Mapp, 1970 ; McGlynn, 3lealiea,
& Nawas, 1969 ; McGlynn, Reynolds, &
SELF-EFFICACY 19
era.te further fear of impending stressful
and the reinforcing function of arousal re-
situations through anticipatory self-arousal.
duction. Social learning theory, on the other
By conjuring up fear-provoking thoughts
hand, emphasizes the informative function
about their ineptitude, individuals can rouse
of physiological arousal. Simply acknowledg-
themselves to elevated levels of anxiety that
ing that arousal is both informative and mo-
far exceed the fear experienced during the
tivating by no means resolves the issue in
actual threatening situation.
dispute, because these are not necessarily two
As will be recalled from the earlier dis- separate effects that somehow jointly produce
cussion, desensitization and massive exposure behavior. Rather, the cognitive appraisal of
treatments aimed at extinguishing anxiety arousal to a large extent determines the level and
arousal produce some reductions in avoidance direction of motivational inducements to
behavior. Anxiety arousal to threats is like- action. Certain cognitive appraisals of one’s
wise diminished by modeling, and is even physiological state might be energising, whereas
more thoroughly eliminated by experienced other appraisals of the same state might not
mastery achieved through participant model- (Weiner, 1972 ) . Moreover, many forms of
ing (Bandura & Barab, 1973 ; Bandura et physiological arousal are generated cognitively
al., 1969 ; Blanchard, 1970a). Modeling ap- by arousing trains of thought. When motivation
proaches have other advantages for enhancing is conceptualized in terms of cognitive processes
self-efficacy and thereby removing dysfunc- (Bandura, 1977 ; Weiner, 1972), the
tional fears. In addition to diminishing prone- informational and motivational effects of
ness to aversive arousal, such approaches also arousal are treated as interdependent rather than
teach effective coping skills by demonstrating as separate events. We shall re- turn to this
proficient ways of handling threatening situa- issue later when we consider the differential
tions. The latter contribution is especially im- predictions made from social learning theory
portant when fear arousal partly results from and from the dual-process theory of avoidance
behavioral deficits. It is often the case that behavior concerning the behavioral effects of
fears and deficits are interdependent. Avoid- extinguishing anxiety arousal.
ance of stressful activities impedes develop-
Researchers working within the attribu-
ment of coping skills, and the resulting lack
tional framework have attempted to modify
of competency provides a realistic basis for
avoidance behavior by directly manipulating the
fear. Acquiring behavioral means for con-
trolling potential threats attenuates or elim- cognitive labeling of emotional arousal (Valins
inates fear arousal (Averill, 1973 ; Notter- & Nisbett, 1971) . The presumption is that if
man, Schoenfeld, & Bersh, 1952 ; Szpiler & phobics are led to befiere that the things they
Epstein, 1976) . Behavioral control not only have previously feared no longer affect them
allows one to manage the aversive aspects of internally, the cognitive reevalua- tion alone
an environment. It also affects how the en- will reduce avoidance behavior. In treatment
vironment is likely to be perceived. Poten- analogues of this approach, phobics receive
tially stressful situations that can be con- false physiological feedback sug- gesting that
trolled are construed as less threatening, they are no longer emotionally upset by
and such cognitive appraisals further re- threatening events. Results of this procedure are
duce anticipatory emotional arousal (Averill, essentially negative. Early claims that
1973). erroneous arousal feedback reduces avoidance
Diminishing emotional arousal can reduce behavior (Valins & Ray, 1967) are disputed by
avoidance behavior, but different theories methodologically superior studies showing that
' posit different explanatory mechanisms for false feedback of physiological tranquility in
the observed effects. In the theory from the presence of threats has either no
appreciable effect on subsequent fearful
which the emotive treatments are derived,
behavior (Gaupp, Stern, & Galbraith, 1972 ;
emotional arousal is conceived of as a drive
Howlett & Nawas, 1971 ; Kent, Wilson, &
that activates avoidance behavior. This view Nelson, 1972 ; Rosen, Rosen, & Reid, 1972 ;
stresses the energising function of arousal Sushinsky R Bootzin,
20 ALBERT
1970), or produces minor changes under Moreover, mislabeling arousal or attributing

such limited conditions as to be of little it to erroneous sources is unlikely to be of
practical consequence (Borkovec, 1973). much help to the highly anxious. Severe
Misattribution of emotional arousal is acrophobics, for example, may be temporarily
another variant of the attributional ap- misled into believing that they no longer
proach to modification of fearful behavior. fear high elevations, but they will reexperi- The
strategy here is to lead fearful people ence unnerving internal feedback when con- into
believing that their emotional arousal fronted with dreaded heights. It should also is
caused by a nonemotional source. To the be noted that in attributional explanations extent
that they no longer label their Of the success of behavioral treatments the
agitated state as anxiety, they will behave heavy emphasis on physiological arousal
more boldly. It may be possible to reduce derives more from speculations about the
mild fears by this means (Ross, Rodin, & nature of emotion (Schachter, 1964) than
Zimbardo, 1969) , but the highly anxious are from evidence that arousal is a major deter-
not easily led into misattributing their anxiety minant of defensive behavior.
to irrelevant sources (Nisbett R Schachter,
1966) . When evaluated systematically, mrs- CO ffttitize Processing oJ E fficac y In formation
attribution treatments do not produce sig The discussion thus far has centered pri-
nificant changes in chronic anxiety condi- marily on the many sources of information—
tions (Singerinan, Borkovec, & Baron, 1976) , enactive, vicarious, exhortative, and emotive
and some of the benefits reported with other —that people use to judge their level of self-
dysfunctions cannot be replicated (Bootzin, efficacy. At this point a distinction must be
Herman, & Nicassio, 1976 ; Kellogg R Baron, drawn between information contained in
1975) . There is also some suggestive evidence environmental events and information as
that in laboratory studies the attenuation of processed and transformed by the individual.
fear may be due more to the veridicality of The impact of information on efficacy ex-
arousal information than to misattribution of pectations will depend on how it is cognitively
fear arousal to an innocuous source (Calvert- appraised. A number of contextual factors,
Boyanowsky & Leventhal, 197S ) . including the social, situational, and temporal
Any reduction in fear resulting from de- circumstances under which events occur, enter
ceptive feedback is apt to be short-lived into such appraisals. For this reason, even
because illusory assurances are not an espe- success experiences do not necessarily create
cially reliable way of creating durable self- strong generalized expectations of personal
expectations. However, more veritable ex- efficacy. Expectations that have served self-
periences that reduce the level of emotional protective functions for years are not quickly
arousal can set in motion a reciprocal process discarded. When experience contradicts firmly
of change. In the social learning view, p- established expectations of self-efficacy, they
tential threats activate fear largely through may undergo little change if the conditions
cognitive self-arousal (Bandura, 1969, 197 of performance are such as to lead one to
Perceived self-competence can therefore af- discount the import of the experience.
fect susceptibility to self-arousal. Individuals The corrective value of information derived
who come to believe that they are less vuIner- from successful performance can be at-
able than they previously assumed are less tenuated in several ways. The first involves
prone to generate frightening thoughts in discrimination processes. The consequences
threatening situations. Those whose fears individuals anticipate were they to perform are
relatively weak may reduce their self- feared activities differ in circumstances which doubts
aura debilitating self-arousal to the vary in safeguards. As a result, they may point where
they perform successfully. Per- behave boldly in situations signifying safety, formance successes,
in turn, strengthen self- but retain unchanged their self-doubts under
efficacy. Such changes can, of course, be less secure conditions. Such mitigative dis-
reliably achieved without resort to ruses. criminations can e.xtend to the treatments
SELP-EPPICACY THEORY 201
themselves, as well as to the situational cir- the chances that behavior will be ascribed
cumstances in which behavioral attainments to external factors (Bern, 19f2 ; Weiner,
occur. This is especially true of treatments 1972).
relying solely on symbolic and vicarious ex- Even under conditions of perceived self-
perience. Achieving reductions in fear to determination of outcomes, the impact of
threats pre’sented symbolically is unlikely to performance attainments on self-efficacy will
enhance perceived self-efficacy to any great vary depending on whether one’s accomplish-
extent in people who believe that success in ments are ascribed mainly to ability or to
imagery does not portend accomplishments in effort. Success with minimal effort fosters
reality. Information conveyed by facilely ability ascriptions that reinforce a strong
modeled performances might likewise be sense of self-efficacy. By contrast, analogous
minimized by anxious observers on the grounds successes achieved through high expenditure
that the models possess special expertise of effort connote a lesser ability and are thus
enabling them to prevent injurious likely to have a weaker effect on perceived self-
consequences that might otherwise befall the efficacy. Cognitive appraisals of the difficulty
unskilled. Because such discriminations, even level of the tasks will further afiect the impact
though objectively mistaken, impede change of performance accomplishments on perceived
in self-efficacy, observers will be reluctant to self-efficacy. To succeed at easy tasks provides
attempt feared activities and will be easily no new information for altering one’s sense
dissuaded by negative experience. of self-efficacy, whereas mastery of
Cognitive appraisals of the causes of one’s challenging tasks conveys salient evidence of
behavior, which have been examined exten- enhanced competence. The rate and pattern of
sively in investigations of self-attributional attainments furnish additional information for
processes (Bern, 1972) , can similarly delimit judging personal efficacy. Thus, people who
gains in self-efficacy from behavioral attain- experience setbacks but detect relative progress
ments. It was previously shown that at- will raise their perceived efficacy more than
tributions of affect and actions to illusory those who succeed but see their performances
competence have little, if any, effect on leveling oS compared to their prior rate of
refractory behavior. This does not, of course, improvement.
mean that causal appraisals are of limited Extrapolations from theories about attribu-
importance in the process of behavior change. tion and self-perception to the field of be-
Quite the contrary, performance attainment is havioral change often imply that people must
a prominent source of efficacy information, but labor unaided or under inconspicuously ar-
it is by no means unambiguous. As already ranged influences if they are to convince
mentioned briefly, people can gain com- themselves of their personal competence
petence through authentic means but, be- (Kopel & Arkowitz, 1975) . Such prescrip-
cause of faulty appraisals of the circumstances tions are open to question on both con-
under which they improve, will credit their ceptual and empirical grounds. Cognitive
achievements to external factors rather than to misappraisals that attenuate the impact of
their own capabilities. Here the problem is one disconfirming experiences can be minimized
of inaccurate ascription of personal competency without sacrificing the substantial benefits
to situational factors. Successes are more likely of powerful induction procedures. This is
to enhance self- efficacy if performances are achieved by providing opportunities for self-
perceived as resulting from skill than from directed accomplishments after the desired
fortuitous or behavior has been established. Any lingering
' special external aids. Conversely, failures doubts people might have, either about their
would be expected to produce greater reduc- capabilities or about probable response con-
tions in self-efficacy when attributed to sequences under unprotected conditions, are
ability rather than to unusual situatioñal dispelled easily in this manner (Bandura
circumstances. The more extensive the situa- et al., 1975) . The more varied the circum-
tional aids for performance, the greater are stances in which threats are mastered in-
20 ALBERT
dependently, the more likely are success ex- elements are the models’ characteristics (e.g.,
periences to authenticate personal efficacy adeptness, perseverance, ap•e, expertness) , the
and to impede formation of discriminations similarity between models and observers, the
that insulate self-perceptions from difficulty of the performance tasks, the situa-
disconfirming evidence. tional arrangements under which the modeled
Results of recent studies support the achievements occur, and the diversity of
thesis that generalized, lasting changes in self- modeled attainments.
efficacy and behavior can best be achieved by Just as the value of efficacy information
participant methods using powerful induction generated enactively and vicariously depends on
procedures initially to develop capabilities, then coq•nitive appraisal, so does the information
removing external aids to verify personal arising• from exhortative and emotive sources.
efficacy, then finally using self-directed mastery The impact of verbal persuasion on self-efficacy
to strengthen and generalize expectations may vary substantially depending on the
of personal efficacy ( Bandura et a1., 1975). perceived credibility of the persuaders, their
Independent performance can enhance prestige, trustworthiness, expertise, and
efficacy expectations in several ways: (a) It assuredness. The more believable the source of
creates additional exposure to former threats, the information, the more likely are efficacy
which provides participants with further expectations to change. The influence of
evidence that they are no longer aversively credibility on attitudinal change has, of course,
aroused by what they previously feared. received intensive study. But its effects on
Reduced emotional arousal confirms increased perceived self-efficacy remain to be
coping capabilities. investigated.
(b) Self-directed mastery provides opportuni-
People judge their physiological arousal
ties to perfect coping skills, which lessen
largely on the basis of their appraisal of the
personal vulnerability to stress. (c) Indepen-
instigating conditions. Thus, visceral arousal
dent performance, if well executed, produces
occurring in situations perceived to be
success experiences, which further reinforce
threatening is interpreted as fear, arousal in
expectations of self-competency.
thwarting situations is experienced as anger,
Extensive self-directed performance of
and that resulting from irretrievable loss of
formerly threatening activities under pro-
valued objects as sorrow tHunt, Cole, &
gressively challenging conditions at a time
Reis, 1958 ) . Even the same source of phys-
when treatments are usually terminated could
iological arousal may be interpreted diNerently
also serve to reduce susceptibility to relearn- ing
in ambiq•uous situations dependinq• on the
of defensive patterns of behavior. A few
emotional reactions of others in the same
negative encounters among many successful
setting (Mandler, 1975 ; Schachter & Singer,
experiences that have instilled a strong sense
1962 ) .
of self-efficacy will, at most, establish dis-
When tasks are performed in ambiguous or
criminative avoidance of realistic threats, an
complex situations in which there is a variety
effect that has adaptive value. In contrast, if
of evocative stimuli, the informational value
people have limited contact with previously
of the resultant arousal will depend on the
feared objects after treatment, whatever
meaning imposed upon it. People who per-
expectations of self-efficacy were instated ceive their arousal as stemming from personal
would be weaker and more vulnerable to inadequacies are more likely to lower their
change. Consequently, a few unfavorable efficacy expectations than those who attribute
experiences are likely to reestablish defensive their arousal to certain situational factors.
behavior that generalizes inappropriately. Given a proneness to ascribe arousal to per-
We have already examined how cognitive sonal deficiencies, the heightened attention to
processing of information conveyed by model- internal events can result in reciprocally
ing might influence the extent to which escalating arousal. Indeed, as Sarason ( 1976)
vicarious experience effects changes in self- has amply documented, individuals who are
efficacy. Among the especially informative especially susceptible to anxiety arousal
SELP-EFFICACY THEORY 203
readily become self-preoccupied with their havior, whereas an omnibus motive does
perceived inadequacies in the face of dif- not. People will approach, explore, and try
ficulties rather than with the task at hand. to deal with situations within their self-
perceived capabilities, but they will avoid
transactions with stressful aspects of their
Differing Perl pecti»es on Sel]-eQcac y
environment they perceive as exceeding their
The phenomena encompassed by the con- ability.
struct of self-efficacy have been the subject The alternative views also difler on the
of interest in other theories of human be- origins of efficacy. Within the framework of
havior. The theoretical perspectives differ, eflectance theory, the effectance drive de-
however, in how they view the nature and velops gradually through prolonged trans-
origins of personal efficacy and the interven- actions with one’s surroundings. This theory
ing processes by which perceived self-efficacy thus focuses almost exclusively on the effects
affects behavior. In seeking a motivational produced by one’s ohm actions. In the social
explanation of exploratory and manipulative learning theory, self-efficacy is conceptualized
behavior, White ( 1959) postulated an “effec- as arising from diverse sources of information
tance motive,” which is conceptualized as an conveyed by direct and mediated experience.
intrinsic drive for transactions with the en- These differences in theoretical approach have
vironment. Unlike instigators arising from significant implications for how one goes about
tissue deficits, effectance motivation is be- studying the role of perceived self-efficacy in
lieved to be aroused by novel stimulation motivational and behavioral processes.
and is sustained when the resultant inquisi- Expectations of personal efficacy do not
tive and exploratory actions produce further operate as dispositional determinants
elements of novelty in the stimulus field. independently of contextual factors. Some
The effectance motive presumably develops situations require greater skill and more
through cumulative acquisition of knowledge arduous performances and carry higher risk
and skills in dealing with the environment. of negative consequences than do others.
However, the process by which an effectance Expectations will vary accordingly. Thus, for
motive emerges from effective transactions example, the level and strength of perceived
with the environment is not spelled out in self-efficacy in public speaking will differ
White’s theory. Nor is the existence of the depending on the subject matter, the format
motive easy to verify, because effectance of the presentation, and the types of audi-
motivation is inferred from the exploratory ences that will be addressed. The social
behavior it supposedly causes. Without an learning approach is therefore based on a
independent measure of motive strength one microanalysis of perceived coping capabilities
cannot tell whether people explore and manip- rather than on global personality traits or
ulate things because of a competence motive motives of effectance. From this perspective,
to do so, or for any number of other reasons. it is no more informative to speak of self-
Athough the theory of effectance motivation has efficacy in general terms than to speak of
not been formulated in sufficient detail to nonspecific approach behavior. To elucidate
permit extensive theoretical comparisons, there how perceived self-efficacy affects behavior
are several issues on which the social learning requires a microanalysis of both factors.
and effectance theories clearly differ. In the Discrepancies between efficacy expecta-
social learning analysis, choice behavior and tions and performance are most likely to
effort expenditure are governed in part by arise under conditions in which situational
percepts or self-efficacy rather than by a drive and task factors are ambiguous. When per-
condition. Because efficacy expectations are formance requirements are ill-defined, people
defined and measured independently of who underestimate the situational demands
performance, they provide an explicit basis will display positive discrepancies between self-
for predicting the occurrence, generality, and efficacy and performance attainments; those
persistence of coping be- who overestimate the demands will
2 ALBERT
exhibit negative discrepancies. Therefore, in ency to perceive events as being either per-
testing predictions from the conceptual
sonally or externally determined.
scheme presented here it is important that
The notion of locus of rontrol is often
subjects understand what kind of behavior
treated in the literature as analogous to self-
will be required and the circumstances in
efficacy. However, Rotter’s ( 1966) conceptual
which they will be asked to perform them.
scheme is primarily concerned with causal
Moreover, performances and the correspond-
beliefs about action—outcome contingencies
ing efficacy expectations should be analyzed
rather than with personal efficacy. Perceived self-
into separate activities, and preferably
efficacy and beliefs about the locus of causality
ordered by level of difficulty. In this type
must be distinguished, because convictions that
of microanalysis both the efficacy expecta-
outcomes are determined by one’s own actions
tions and the corresponding behaviors are can have any number of effects on self-efficacy
measured in terms of explicit types of per- and behavior. People who regard outcomes as
formances rather than on the basis of global personally determined but who lack the requisite
indices. skills would experience low self-efficacy and
The social learning determinants of self- view activities with a sense of futility. Thus,
efficacy can be varied systematically and for example, a child who fails to grasp
their effects measured. Hence, propositions arithmetic concepts and expects course grades to
concerning the origins of self-efficacy are be dependent entirely on skill in the subject
verifiable with some precision. A slowly de- matter has every reason to be demoralized. While
veloping motive, however, does not easily causal beliefs and self-efficacy refer to different
lend itself to being tested experimentally. phenomena, as we have already noted, causal
Another dimension on which the alternative ascriptions of behavior to skill or to chance
theories mip•ht be judged is their power to can mediate the effects of performance at-
produce the phenomena they purport to tainments on self-efficacy.
explain. As we shall see later, there are more
The theoretical framework presented in the
diverse, expeditious, and powerful ways of
present article is generalizable beyond the
creating self-efficacy than by relying solely
psychotherapy domain to other psychological
on novel stimulation arising from exploratory
phenomena involving behavioral choices and
actions.
regulation of effort in activities that can have
\Vith the ascendency of cognitive views of adverse effects. For example, the theory of
behavior, the concept of expectancy is as- learned helplessness advanced by Maier and
suming an increasingly prominent place in Seligman ( 1976) assumes that as a result of
contemporary psychological thought ( Bolles, being subjected to uncontrollable aversive
1972b; Heneman, & Schwab, 1972 ; Irwin, events, organisms acquire expectancies that
1971 ) . However, virtually all of the theoriz- actions do not affect outcomes. Because they
ing and experimentation has focused on action come to expect future responding to be futile,
—outcome expectations. The ideas advanced they no long•er initiate behavior in situations
in some of the theories nevertheless bear some where outcomes are in fact con- trollable by
likeness to the notion of self- efficacy. responses. Although this theory posits an
According to the theory of personality expectancy mechanism of operation, it focuses
proposed by Rotter ( 1966) , behavior varies as exclusively on response—outcome
a Iunction of generalized expectancies that expectancies.
outcomes are determined by one’s actions or by
Theorizing• and experimentation on learned
external forces be3•ond one’s control. Such
helplessness might well consider the con-
expectations about the instru- mentality of
ceptual distinction between efficacy and out-
behavior are considered to be largely a prod uct
come expectations. People can give up trvinp•
of one’s history of reinforcement. Much of the
because they lack a sense of efficacy in
research within this tradi- tion is concerned with
achieving the required behavior, or they may
the behavioral cor- relates of individual
differences in the tend- be assured of their capabilities but give up
SELF-EFFICACY 20
trying because they expect their behavior to

performance sources of information for their
have no effect on an unresponsive environ-
efficacy expectations. Enactive and vicarious
ment or to be consistently punished. These
procedures were selected for study to assess
two separable expectancy sources of futility
the predictive value of self-efficacy created
have quite different antecedents and remedial
by quite different modes of treatment.
implications. To alter efficacy-based futility
The level, strength, and generality of the
requires development of competencies and
subjects’ efficacy expectations were measured
expectations of personal effectiveness. By
at critical junctures in the change process.
contrast, to change outcome-based futility
Subjects privately designated, on a list of 18
necessitates changes in prevailing environ-
performance tasks ranked in order of in-
mental contingencies that restore the in-
creasing threat, those tasks they considered
strumental value of the competencies that
themselves capable of executing. They then
people already possess.
rated the strength of their expectations for
each of these tasks on a 100-point probability
Microanalysis oJ SelJ-e ficac y and Behavioral scale ranging, in 10-unit intervals, from great
Change uncertainty, through intermediate values of
certainty, to complete certainty. They rated
To test derivations from the social learn- their efficacy expectations for coping with
ing analysis o( the process of change, an snakes of the same variety used in treatment
experiment was conducted wherein severe as well as dissimilar snakes to measure the
phobics received treatments designed to create generality of their efficacy expectations.
differential levels of efficacy expectations, and These measures were obtained prior to treat-
then the relationship between self- efficacy and ment, following treatment but before the
behavioral change was analyzed in detail behavioral posttest, and after completing the
(Bandura et a1., in press) . The experiment posttest. Approach behavior was assessed in the
proceeded as follows. Adult snake phobics, posttest by a series of performance tasks
whose phobias affected their lives adversely, requiring increasingly more threatening inter-
were administered for equivalent periods either actions with a different type of boa constrictor
participant modeling, modeling alone, or no from the one used in treatment and with a
treatment. In participant modeling, which
corn snake of markedly different appearance
operates through direct mastery experiences,
but equivalent threat value. Different phobic
subjects were assisted, by whatever induction
objects were used to provide a test of the
aids were needed, to engage in progressively
generalized effects of changes in efficacy
more threatening in- teractions with a boa
expectations along a dimension of similarity
constrictor. After completing all the
to the threat used in treatment.
therapeutic tasks, which included holding the
Subjects assigned to the control condition
snake, placing open hands in front of its head
participated in the assessment procedures
as it moved about the room, holding the snake
without receiving any intervening treatment.
in front of their faces, and allowing it to crawl
Following completion of the posttest, the
freely in their laps, the subjects engaged in a
controls and those in the modeling condition
brief period of self-directed mastery. In the
who failed to achieve terminal performances
present experiment, the modeling aid was used
received the participant modeling treatment.
only briefly if needed to help initiate
performance in order to minimize overlap of Consistent with the social learning analysis
this element in the two modes of treatment. of the sources of self-efficacy, experiences
based on performance accomplishments pro-
Subjects receiving the modeling treatment
duced higher, more generalized, and stronger
merely observed the therapist perform the
efficacy expectations than did vicarious ex-
same activities for an equivalent period.
perience, which in turn exceeded those in
These subjects did not engage in any behavior
the control condition. Figure 3 summarizes the
themselves, and consequently they had no
level of efficacy expectations and performance
as a function of treatment condi-
2 ALBERT
PERCENT SUCCESSFUL
PRE -TEST POST-TEST PRE-TEST I-•OST-TEST

SIMILAR THREAT DISSIMILAR THREAT
Figure 3. Level of efficacy expectations and approach behavior displayed by subjects toward threats
after receiving vicarious or enactive treatments, or no treatment (Bandura et aI., in press) .
tions at different phases of the experiment.

will also contribute to variance in perform-
As shown in the figure, performance change ance.
corresponds closely to the magnitude of ex- Correlation coefficients based on aggregate
pectancy change. The greater the increments measures do not fully reveal the degree of
in self-perceived efficacy, the greater the correspondence between self-efficacy and per-
chanp•es in behavior. Similar relationships be- formance on the specific behavioral tasks
tween level of self-efficacy and performance are from which the aggrep•ate scores are obtained.
obtained when the data are considered A subject can display an equivalent number
separately for the two snakes. In accordance of efficacy expectations and successful per-
with prediction, participant modeling pro- formances, but they mip•ht not correspond
duced the more generalized increases in effi- entirely to the same tasks. The most precise
cacy expectations and the more generalized index of the relationship is provided by a
behavioral changes. microanalysis of the conq•ruence between self-
Although the enactive and vicarious treat- efficacy and performance at the level of in-
ments differed in their power to enhance self- dividual tasks. This measure was obtained by
efficacy, the efficacy expectations were equally recordin* whether or not subjects considered
predictive of su bsequent performance irrespective theirselves caI £f ble of performinp• each of
of how they were instated. The higher the level the various tasks at the end of treatment
of perceived self-efficacy at the completion of and by computing the percentage of accurate
treatment the hiyher was the level of approach correspondence between efficacy judgment
behavior for efficacy expectations instated and actual per formance. Self-efficacy was a
enactively (r = .83) and vicariously (r .54) . It uniformly accurate predictor of performance
might be noted here that all subjects had at their on tasks varying in difficulty wi th different
disposal the component responses for producing• threats regardless of whether the changes in self-
the interactive patterns of behavior, anri they all efficacy were produced through I• er formance
had some incentive to overcome their phobic be- accomplishments ( 89(c congruence) or by
havior. L’nder conditions in which people vicarious experience alone (86% con-
stiffer substantially in component capabilities gruence ) . The dc*ree of congruence between
and motivation, skill and incentive factors |ierceived self-efficacy and subsequent behav-
ior is equally high for enactive ( 82 Jo ) and
SELF-EFFICACY 20
vicarious ( 79/o ) treatments when the micro- predicts accurately the magnitude and gen-
analysis is conducted only on the subset of erality of behavioral change for efficacy ex-
tasks that subjects had never performed in the pectations induced enactively and vicariously.
pretest assessment. h'Ioreover, it orders variations in level of
In the preceding analysis efficacy expecta- behavioral change occurring within the same
tions were considered without regard to treatment condition. Subjects who received
strength. A weak sense of self-efficacy thus participant modeling, either as the primary or
received the same weight as one reflecting as the supplementary treatment, successfully
complete certitude. However, the intensity performed all of the behaviors in treatment
and persistence of effort, and hence level of that were later assessed in the posttest toward
performance, should be higher with strong different threats. Although all had previously
than with weak self-efficacy. The likelihood achieved maximal performances, not all
that a task will be performed as a function expressed maximal efficacy expectations. One
of the strength of the corresponding efficacy can therefore compare the error rates of
expectation therefore provides a further re- predictions made from maximal past
finement in the analysis of the relationship performance and from maximal efficacy ex-
between self-efficacy and performance. The pectations. It would be predicted from the
probability of successful performance of any proposed theory that among these successful
given task as a function of strength of efficacy performers, those who acquire maximal effi-
expectations is plotted in Figure 4. Because cacy expectations should attain terminal per-
the control subjects performed few responses formances, whereas those holding lower ex-
and had correspondingly restricted efficacy pectations should not. If one predicts that those
expectations, their data were plotted after who performed maximally in treatment will
they had received the participant modeling likewise achieve terminal performances when
treatment. In all conditions, the stronger the assessed with similar tasks, the error rate is
efficacy expectations, the higher was the relatively low for the similar threat (28 o )
likelihood that a particular task would be but high for the dissimilar threat ( 52
successfully completed. The positive relation- % ) . If, on the other hand, one predicts that
ship between strength of self-efficacy and those who express maximal expectations will
probability of successful performance is vir- perform maximally, the error rate is comparably
tually identical for the similar and the dis- low for both the similar ( 2 1/c ) and the
similar threats. dissimilar ( 24% ) threats. The predictive
In brief, the theory systematizes a variety
superiority of efficacy expectations over past
of findings. As the preceding results show, it
performance is significant for total
SIMILAR THREAT
LOO F— DISSIMILAR THREAT
0.60
* 0.40
0.20
“ 10-20 30-40 50-60 70-80 90-IOO 10-20 30-40 50-60 70-80 90-100
STRENGTH OF EFFICACY EXPECTATIONS
Figure 4. Probability of successful performance of any given task as a function of stren•_th of
sell-efficacy. The figure on the left shows the relationship for vicarious and enactive
treatments ; the figure on the right shows the relationship between strength of self-efficacv and
successful approach responses toward similar and dissimilar threats combined across treatments
(Bandura et al., in press) .
2 ALBERT
approach behavior and for approach behavior based on the assumption that anxiety ac-
toward the dissimilar threat. These differ- tivates defensive behavior (Wolpe, 19?4).
ential findings indicate that experienced According to this view, association of neutral
mastery altered subjects’ sense of personal events with aversive stimulation creates an
efficacy rather than merely providing be- anxiety drive that motivates defensive be-
havioral cues for judgments or self-efficacy. havior; the defensive behavior, in turrl, is
The theory also accounts for variations in reinforced by reducing the anxiety aroused
behavioral change produced by modeling by conditioned aversive stimuli. Hence, to
alone. To equate for duration of treatment, eliminate defensive responding, it is con-
subjects in the modeling condition were sidered necessary to eradicate its underlying
yoked to matched counterparts in participant anxiety. Treatment strategies are therefore
modeling, who received treatment until they keyed to reduction of emotional arousal.
performed all the therapeutic tasks. The Aversive stimuli are presented at graduated
subjects in the participant modeling condi- levels in conjunction with relaxation until
tion varied in the time they required to com- anxiety reactions to the threats are elimi-
plete treatment, so some of the subjects in nated.
the modeling conditions had only brief ex-
Although desensitizatien produces be-
posure to successful performances, whereas
havioral changes, there is little evidence to
others had the benefit of observing feared
support the original rationale that defensive
activities modeled repeatedly without any
untoward consequences. The findings are behavior is diminished because anxiety is
consistent with hypothesized increases in self- eliminated either by reciprocal physiological
efficacy as a function of repeated observa- inhibition or by associative recoupling of
tion of successful modeling. Brief exposure threatening stimuli to relaxation. Desensitiza-
produced limited increases in the level (9Qo ) tion does not require graduated exposure, and
and strength (5% ) of efficacy expectations anxiety-reducing activities are at most facili-
and correspondingly little behavior change tory, not necessary, conditions for eliminating
(10% ). In contrast, repeated observation of defensive behavior (Bandura, 1969; Wilson
successful performances increased by a sub- & Davison, 1971 ).
stantial amount the level (44Jo ) and strength The principal assumption that defensive
(38$o ) of self-efficacy which, in turn, was behavior is controlled by anxiety arousal is
accompanied by similarly large increments also disputed by several lines of evidence.
in performance (35% ). Autonomic arousal, which constitutes the
principal index of anxiety, is not necessary
for defensive learning. Because autonomic re-
Comparison of SelJ-e ficac y and Dual-Process actions take much longer to activate than do
Theory avoidance responses, the latter cannot be
caused by the former. Studies in which auto-
As a further test of the generality of the nomic and avoidance responses are measured
theory under discussion, a microanalysis was concurrently indicate that these two modes
conducted of efficacy expectations instated of activity may be partially correlated in the
by desensitization procedures, which are acquisition phase but are not causally re-
aimed at reducing emotional arousal. Social lated (Black, 1965) . Avoidance behavior, for
learning theory and the dual-process theory• example, can persist long after autonomic
of anxiety, on which the desensitization ap- reactions to threats have been extinguished.
proach is based, posit different explanatory Surgical removal of autonomic feedback ca-
mechanisms for the changes accompanying pability in animals has little effect on the
this mode of treatment. The alternative views acquisition of avoidance responses (Rescorla
therefore give rise to differential predictions & Solomon, 1967). Maintenance of avoidance
that can be readily tested. behavior is even less dependent on autonomic
The standard desensitization approach is feedback. Once defensive behavior has been
learned, depriving animals of autonomic feed-
SELF-EFFICACY 20
back does not hasten the rate at which such environment, not the stimuli, that are
activities are extinguished. changed by correlated experience. Stimuli
Research casts doubt on the postulated having predictive significance signal the
reinforcement sources, as well as the activat- likelihood of painful consequences unless
ing sources, of defensive behavior. In the dual- protective measures are taken. Defensive be-
process theory, the anxiety reduction havior, in turn, is maintained by its success
occasioned by escape from the feared stim- in forestalling or reducing the occurrence of
ulus presumably reinforces the defensive be- aversive events. Once established, self-pro-
havior. The evidence, however, reveals that tective behavior is difficult to eliminate even
whether or not defensive behavior removes though the hazards no longer exist. This is
the feared stimulus has variable effects on because consistent avoidance prevents a per-
the maintenance of the behavior (Ilolles, son from learning that the real-life conditions
1972a) . Moreover, defensive behavior can be have changed. Hence, the nonoccurrence of
acquired and maintained by its success in anticipated hazards reinforces the expectation
diminishing the frequency of aversive stim- that the defensive maneuvers forestalled
ulation, even though there are no feared them.
stimuli to arouse anxiety and to provide the From the perspective of dual-process
source of decremental reinforcement (Herrn- theory, thorough extinction of anxiety should
stein, 1969) . The substantial negative evi- eliminate avoidance behavior. Tn the desensi-
dence concerning an anxiety mediational tization treatment, however, anxiety reactions
mechanism in avoidance behavior suggests are typically extinguished to visualized rep-
that the effects of desensitization treatment must resentations of feared situations. One would
result from some other mechanism of operation. expect some transfer loss of extinction effects
Social learning theory regards anxiety and from symbolic to real-life threats, as is in-
defensive behavior as coeffects rather than as deed the case (Agras, 1967 ;• Barlow et a1.,
causally linked (Bandura, 197 7) . Aversive 1969) . It is not uncommon for people to
experiences, either of a personal or vicarious fear and avoid real-life situations to which
sort, create expectations of injurious effects they have been desensitized in imagery.
that can activate both fear and defensive be- Therefore, according to this view, thorough
havior. Being coeffects, there is no fixed extinction of anxiety to visualized threats
relationship between autonomic arousal and should produce substantial, though less than
actions. Until effective coping behaviors are complete, reductions in defensive behavior.
achieved, perceived threats produce high However, dual-process theory provides no basis
emotional arousal and various defensive for predicting either the level of behavior
maneuvers. But after people become adept at change or the variability in behavior displayed
self-protective behaviors, they perform them in by subjects who have all been
potentially threatening situations without having equally desensitized.
to be frightened (N otterman et at., 1952) . In the social learning analysis presented
Should their habitual coping devices fail, they earlier, reducing physiological arousal im-
experience heightened arousal until new
proves performance by raising efficacy ex-
defensive learning reduces their vulnerability.
pectations rather than by eliminating a drive that
Perceived threats activate defensive be-
havior because of their predictive value rather instigates the defensive behavior. This
than their aversive quality. That is, when information-based view of the mediating
formerly neutral stimuli are associated with mechanism predicts that the higher and
painful experiences, it is not that the stimuli stronger the efficacy expectations instated by
have become aversive but that individuals desensitization procedures, the greater are
have learned to anticipate aversive con- the reductions in defensive behavior. Because
sequences. It is people's knowledge of their arousal is only one of several sources of
efficacy information, and not necessarily the
most dependable one, extinguishing anxiety
2 ALBERT
OESENSITIZ ATION
o---o EFFICACY EXP.
e—•-+ BEHAVIOR
W0-
60 —
50-
PRE-TE9T POST-TE5T PRE-TEST POST-TEST

STHLAR THREAT DISSIMILAR THREAT
Figure S. Level of efficacy expectations and approach behavior displayed by subjects toward
different threats after their emotional reactions to symbolic representations of feared activities
were eliminated through systematic desensitization. (Bandura & Adams, in press.)
arousal is rarely a sufficient condition for efficacy at the end of treatment, the more
eliminating defénsive behavior. approach behavior they subsequently per- To
test the theory that desensitization formed in the posttest assessment (r = .74).
changes behavior through its intervening Results of the microanalysis of congruence
effects on efficacy expectations, severe snake between self-efficacy at the end of treatment
phobics were administered the standard and performance on each of the tasks ad-
desensitization treatment until their emo- ministered in the posttest are consistent with
tional reactions were completely extinguished the findings obtained from enactive and vi- to
imaginal representations of the most carious treatment. Self-efficacy was an ac-
aversive scenes (Bandura & Adams, in press) . curate predictor of subsequent performance
The assessment procedures were identical to on 85% for all the tasks, and 83% for the those
used in the preceding experiment. Sub- subset of tasks that subjects were unable to
jects’ approach behavior was tested on the perform in the pretest assessment. Subjects
series of performance tasks before and after successfully executed tasks within the range
the desensitization treatment. The level, of their perceived self-efficacy produced by
strength, and generality of their efficacy the desensitization treatment, whereas they
expectations were similarly measured before failed at tasks they perceived to be beyond
treatment, upon completion of treatment but their capabilities.
prior to the posttest, and following the post-
test. Microanalysis oJ SelJ-e ficac y and
The findings show that phobics whose Performance During the Process oJ Change
anxiety reactions to visualized threats have
been thoroughly extinguished emerge from The preceding series of experiments ex-
the desensitization treatment with widely amined the predictive value of self-efficacy at
differing efficacy expectations. As depicted the completion of different modes of treat- graphically
in Figure 5, performance cor- ment. A further study investigated the process responds closely to
level of self-efficacy. The of efficacy and behavioral change during the higher the subjects’ level
of perceived self- course of treatment itself. Participant model-
SELF-EFFICACY 'THEORY
211
ing was selected for this purpose because the

subjects could not perform in pretest because
amount of treatment can be well-regulated and
they found them too threatening, and did not
it promotes rapid change.
perform in treatment because the activities
As in the previous studies, adults whose extended beyond the failed block. Expecta- tions
lives were adversely affected by severe snake of personal effectiveness formed through partial
phobias were tested for their efficacy expecta- mastery experiences during the course of
tions and approach behavior using the micro- treatment predicted, at a 84 o level of accuracy,
analytic methodology described earlier. The performance on highly threatening tasks that
various treatment activities were segmented subjects had never done before.
into natural blocks of tasks of increasing
difficulty and threat value. Items in the initial
block included looking at a snake from Concluding Remarks
progressively closer distances ; intermediate
blocks required subjects to touch and to hold The present theoretical formulation orders
the snake with gloved and bare hands for variations in the level of behavioral changes
increasing intervals ; the terminal block re- produced by different modes of treatment ; it
quired them to tolerate the snake crawling accounts for behavioral variations displayed
about freely in their laps for an extended by individuals receiving the same type of
period. Subjects received the participant treatment ; and it predicts performance suc-
modeling treatment only for the block of cesses at the level of individual tasks during•
items they failed in the hierarchy of assess- and after treatment. It is possible to generate
ment tasks. Treatment was continued until alternative explanations for particular sub-
they could perform the activities in the sets of data, but the mechanism proposed in the
failed block, whereupon they were tested for present theory appears to account equally well
their efficacy expectations and approach re- for the diff erent sets of findings. It might be
sponses on the succeeding tasks. Subjects argued, for example, that self-efficacy proved
who attained terminal performances received to be an accurate predictor of performance in
no further treatment. For those who achieved the enactive mode of treatment because subjects
only partial improvement, the sequence of were simply judging their future performance
treatment on the failed block followed by from their past behavior. However, an
assessments of self-efficacy and approach be- interpretation of this type has no explanatory
havior on succeeding blocks was repeated until value for the vicarious and emotive treatments,
they achieved terminal performances. in which perceived self- efficacy was an equally
Findings of the microanalysis lend further accurate predictor of performance although
support to the postulated cognitive mecha- subjects engaged in no overt behavior. Even in
nism of change. Subjects who mastered the the enactive treatment, perceived self-efficacy
same intermediate performances during the proved to be a better predictor of behavior
course of treatment varied considerably in toward unfamiliar threats than did past
their behavioral attainments when tested on performance. Afore- over, self-efficacy derived
succeeding blocks of tasks. Past performance from partial enactive mastery during the
was therefore of limited value in predicting course of treatment predicted performance on
what subjects would be able to do when con- stressful tasks that the individuals had never
fronted with more threatening tasks. However, done before.
efficacy judgments proved to be good predic- As an alternative explanation, one could
tors of degree of behavioral change resulting invoke a superordinate mediator that con-
from partial mastery experiences. Self-efficacy trols both efficacy expectations and behax'ior.
predicted subsequent performance as mea- Although such a possibility is not inconceiv-
sured at different points in treatment in 92/o able, the mediator would have to be an ex-
of the total assessment tasks. This relationship ceedingly complex one to account adequatelv
holds even when the measure of congruence is for the diverse sets of relationships. To cite
based only on the subset of activities that but a few examples, it would have to affect
differentially efficacy expectations and be-
2 ALBERT
havior resulting from maximal enactive mas-
of different perspectives demonstrating that
tery; somehow, it would have to produce some of these factors afiect attitudinal and
different levels of self-efficacy from Equiva- behavioral changes has suggestive value. But
lent reductions in emotional arousal; and it it is investigations that include assessment of
would have to generate some variation in the intervening self-efficacy link that can best
efficacy expectations from similar partial provide validity for the present theory.
mastery experiences. The theory presented
The operative process involved in the rela-
here posits a central processor of efficacy tionship between etficacy expectations and
information. That is, people process, weigh, action also requires further investigation. It
and integrate diverse sources of information will be recalled that efficacy expectations are
concerning their capability, and they regulate presumed to influence level of performance
their choice behavior and effort expenditure by enhancing intensity and persistence of
accordingly. eflort. In the preceding experiments, the
Evidence that people develop somewhat behavioral tasks were ordered in level of
different efficacy expectations from similar difficulty and subjects either persisted in
enactive mastery and fear extinction warrants their efforts until they completed all of the
comment. One possible explanation for the tasks or they quit at varying points along
variance is in terms of differential cognitive the way. The number of tasks successfully
processing of efficacy information. To the ex- completed reflects degree of perseverance. As
tent that individuals differ in how they cog- a further step toward elucidating the inter-
nitively appraise their arousal decrements and vening process, it would be of interest to
behavioral attainments, their percepts of self- measure the intensity and duration of effort
efficacy will vary to some degree. A second subjects exert in attempts to master arduous
possibility concerns the multiple determination or insoluble tasks as a function of the level
of self-efficacy. Because people have met with and strength of their efficacy expectations.
different types and amounts of efficacy-altering Further research on the processes postulated
experiences, providing one new source of in the present theoretical formulation should
efficacy information would not be expected increase our understanding of the relation-
to a8ect everyone uniformly. Thus, for ship between cognitive and behavioral change.
example, extinguishing arousal’ to threats will
enhance self-efficacy, but more so in References
individuals whose past coping attempts have
occasionally succeeded than in those who Agras, W. S. Transfer dieing systematic desensitiza-
have consistently failed. tion therapy. Bahauiour Jtesearc8 and ThTr' ap y,
1967, /, 193-199.
The research completed thus far has tested Averill, J. R. Personal control over aversive stimuli
the predictive power of the conceptual scheme and its mlationship to streis. Psycñologisal BuI-
for efficacy expectations developed through /ctiN, 1973, 80 28&-303.
enactive, vicarious, and emotive-based pro- Bandura, A. Pri pipits o] baMvior modi5cotion.
New York: Holt, Rinehart & Winston, 1969.
cedures. Additional tests of the generality of Bandura, A. (Ed.) . Psycholo iMl modeling: Con-
this approach need to be extended to efficacy flicting theories. Chicago: A1dine-Atherton, 1971.
expectations arising from verbal persuasion Bandura, A. Effecting change through participant
and from other types of treatments aimed at modeling. In J. D. Rrumboltc & C. E. Thoresen
reducing emotional arousal. (Eds.), Counseling methods. New York: Holt,
Rinehart b Winston, 1976. (a)
Cognitive processing of emcacy informa- Bandura, A. Self-reinforcement: Theoretical and
tion, which is an important component func- methodological considerations. Bahaviorism, 1976,
tion in the proposed theory, is an especially %1351.(b)
relevant area for research. A number of fac- Bandura, A. Secret learning theory. Engiewood
tors were identihed as influencing the cog- Cli8s, N.J.: Prentice-Hall, 1977.
Bandura, A., b Adams, N. E. Analysis of self-
nitive appraisal of efficacy information con- chimney theory of behavioral change. Cognitñ›e
veyed by each of the major sources of self- P8cropy and ncscorcâ, in press.
efficacy. Previous research from a number Bandura, A., Adams, N. S., b Beyer, J. Cognitive
SELF-EFFICACY 2
processes mediating behavioral changes. /ournnf oJ
Calvert-Boyanowsky, J., & Leventhal, H. The role
Personalit y and Social Ps ycliofog y, in press.
of information in attenuating behavioral responses
Bandura, A., & Barab, P. G. Processes governing
to stress: A reinterpretation of the misattribution
disinhibitory effects through symbolic modeling.
phenomenon. Journal o] Personality otid Social
J'ot‹mat of Abnormal Ps}c/so/ogy, ‘19Y3, BZ, I—9.
Ps ycholog y, 1975, 32, 214-2 21.
Bandura, A., Blanchard, E. B., ñ Ritter, B. The
Davison, G. C., & Wilson, G. T. Processes of fear-
relative efficacy of desensitization and modeling
reduction in systematic desensitization: Cognitive
approaches for inducing behavioral, affective,
and social reinforcement factors in humans. Bc-
and attitudinal changes. Source/ o/ Pers onolit y
houior Therapy, 1973, 4, 1-21.
end Sociof Ps ycholo g y 1969, 13, 173—199.
Dawson, M. E., fi Furedy, J. J. The role of aware-
Bandura, A., Jeffery, R. W., It Gajdos, E. General-
ness in human diflerential autonomic classical con-
izing change th rough participant modeling with self-
ditioning: The necessary-gate hypothesis. Psy-
directed mastery. Behauiour 2testorch and Th«rap y,
cho physiolog y, 1976, JS, 50—53.
1975, 13, 141-152.
Dulany, D. E. Awareness, rules, and propositional
Bandura, A., Jeffery, R. W., & Wright, C. L.
control: A confrontation with S-R behavior
Efficacy of participant modeling as a function
theory. In T. R. Dixon & D. L. Horton (Eds.) ,
of response induction aids. /ourttof o/ Abnormol
Verbol behavior and generoi bz Senior theory.
Ps yc holog y, 1974, 83, 5h—64.
Engle wood Cliffs, h'.J.: Prentice-Hall, 1968.
Bandura, A., & Menlove, F. L. Factors determining
Emmelkamp, P. M. G., & Wessels, H. Flooding in
vicarious extinction of avoidance behavior through
imagination vs. flooding iii vivo: A comparison
symbolic modeling. Journal o f f'eriottofiiy end
with agoraphobics. Behaviour R csearch and
Social Ps ycholo g y, 1968, 8, 99—108.
Pfieropy, 1975, 13, 7—13.
Barlow, D. H., Leitenberg, H., Agras, W. S., 6 Estes, W. K. Reinforcement in human behavior.
Wincze, J. P. The transfer gap in systematic American Scientist, 1972, 60, 12 $-129.
desensitization: .In analogue study. Be hostour Re - Flanders, J. P. A review of research on imitative
searc h and Ther a p y, 1969, 7, 191—196.
behavior. Psychological Bulletin, 1968, dfi, 316—3 $ 1.
Baron, A., Kaufman, A., & Stauber, K. A. Effects of
Gaupp, L. A., Stern, R. M., & Galbraith, G. G. False
instructions and reinforcement-feedback on hu-
heart-rate feedback and reciprocal inhibition by
man operant behavior maintained by fixed-interval
aversion relief in the treatment or snake avoidance
reinforcement. lournol o/ the £zpcritncntaf Anaf-
behavior Behavior FAera9y, 1972, 3,
ysis o f Behavior, 1969, 12, 701—712.
7—20.
Baum, W. M. The correlation-based law of effect.
Grings, W. W. The role of consciousness and cogni-
Journal’ oJ the E zperimtntal Ariel ysis oJ Behavior,
tion in autonomic behavior change. In F. J. Mc-
1973, 20, 137-133.
Guigan & R. A. Schoonover (Eds.) , The ps y-
Bern, D. J. Self-perception theory. In L. Berkowitz
cho ph ysiolog y o] thinking. New York: Academic
(Ed.) , Advances in ex fieriment al s aciol ps ycholo g y
Press, 1973.
(Vol. h) . New York: Academic Press, 1972.
Heneman, H. G., III, & Schwab, D. P. Evaluation of
Black, A. H. Cardiac conditioning in curarized
research on expectancy theory predictions of
dogs: The relationship between heart rate and
employee performance Ps ycholo gicol Bulletin,
skeletal behaviour. In W. F. P rokasy (Ed.) , 1972, 78, 1-9.
Classical conditioning: A s ym posium. h'ew York: Herrnstein, R. J. Method and theory in the study
Appleton-Century-Crofts, 1965. of avoidance. Ps ycliologicof 2teview, 1969, 7d, 49-
Blanchard, E. B. The generalization of vicarious
69.
extinction effects. Behaoiour R eszorch and Howlett, S. C., & Nawas, M. M. Exposure . to
r;• p , redo, z, 3z3-3ao. (a) aversive imagery and suggestion in systematic
Blanchard, E. B. Relative contributions of model- desensitization. In R. D. Rubin, A. A. Lazarus, H.
ing, informational influences, and physical con- Fensterheim, & C M. Franks (Eds.) , Adoances
tact in extinction of phobic behavior. Journal o/ in behavior I hera p y. New York: Academic Press,
Abnormal Ps ych ofo g y, 1970, 7d, 55—61. (b) 1971.
Bolles, R. C. The avoidance learning problem. In Hunt, J. McV., Cole, M. W., & Reis, E. E. S.
G. Bower (Ed.) , The ps ycholo g y oJ leortiing and Situational cues distinguishing anger, fear, and
motivation (Vol. 6) . New York: Academic P rcss, sorrow. American Journal oJ fiychofogy, 1958, 71,
1972. (R) 136—151.
Bolles, R. C. Reinforcement, expectancy, and learn- Irwin, F. W. Inf enliottal behavior end motivation:
ing. Ps ycholo gical 2tevietr, 1972, 79, 394—409. lb) A coguilts e view. Philadelphia: Lippincott, 1971.
Borkovec, T. D. The role of expectancy and ph ysi- Kaufman, A., Baron, A., & Kopp, E. Some effects
ological feedback in fear research: A review with of instructions on human operant behavior. Ps y -
special reference to subject characteristics. Be - c ha nomic Mono gropm Su pplements, 1966, 1, 243- 250.
havior Th«ra p y, 1973, 4, 491—S05. Kazdin, A. E. Covert modeling and the reduction of
Bootzin, R. R., Herman, C. P., 6 Nicassio, P. The avoidance behavior. Journal oJ Abnocmal Ps y-
power of suggestion: Another examination of cholog y, 191$ , 81, 87—95.
misattribution and insomnia. 7 control o/ ter- Kazdin, A. E. Comparative effect.s of some varia-
so emit y and Social Psychology, 1976, a4, 673—679.
2 ALBERT
tions of covert modeling. Journof oJ BzMvior
of Personotit y and Sci:rat Ps ycñolog y, I9R, 17,
298—307.
225-232. (a)
Moore, N. Behaviour therapy in bronchial asthma:
Kazdin, A. E. Covert modeling, model similarity,
A controlled study. Journal oJ Ps ychosamotis
arid reduction of avoidance behavior. &clovior
JZezco/'r5, 1965, 0, 257—276.
Z'5cropy, 1974, /, 325—340. (b)
Kazdin, A. E. Eflects of covert modeling and re- Nisbett, R. E., & Schachter, S. Cognitive manipula-
inforcement on assertive behavior. /our›uif o/ tion of pain. Journ& o f Aspcrimenfof Social Ps E-
Abnormal Ps ycholog y, 1974, 8J, 240-252. (c) choing y, 1966, 2, 2 27-236.
Kazdin, A. E. Covert modeling, imagery assessment, Notterman, J. M., Schoenfeld, W. N., fi Bersh,
and assertive behavior. Correct o/ Cowutting P. J. A comparison of tlune extinction procedures
end Clinical PsEchoing y, 1915, 43, 716-724. following heart rate conditioning. Journal oJ Ab-
Kaidin, A. E. Effects of covert modeling, multiple normal and Saciat PsEchoing y, 1952, 47, 674-677.
Paul, G. L. might vs. dzs znsitizotion ia ps ycho-
models, and model reinfo.rcement on assertive be-
tlero9y. Stanford, Calif.: Stanford University
havior. BeLarior Tli«rap y, 1976, 7, 211-222.
Press, 1966.
Kellogg, R., & Baron, R. S. Attribution theory,
Rabavilas, A. D., Boulougouris, J. C., & Stefanis, C.
insomnia, and the reverse .placebo efftct: A re-
Duration of flooding sessions in the treatment of
versal of Storms and Nisbett's findings. /Purnaf
obussive-compubive patients. dcl tiiour 2t*-
o] Personolit y and Social Ps ychofog y, 19)5, 32,
231-Z36.
Rescorla, R. A., & Solomon, R. L. Two-process
Kent, R. N., Wilson, G. T., & Nelson, R. Elects of learning theory: Relationshi.ps between Pavlovian
false heart-rate feedback on avoidance behavior: conditioning and instrumental learning. Piycho-
An investigation of “cognitive desensitization.”
logical Jtrviee, 1967, 74, IN1-182.
Bahauior Therap y, 1972, 3, 1—6.
Ritter, B. The use of contact desensitization, demon-
Kopel, S., fi Arkowitz, H. The cole of attribution
and self-perception in behavior change: Implica- stration-plus-participation, and demonstration
tions for behavior therapy, Genetic Ps ycholog y alone in the treatment of acrophobia. Bthaoiour
If oRogrA§As, 1975, P2, 175-212. Research and Therap y, 1969, 7, 157-164.
Lewis, S. A comparison of behavior therapy tech- Riiper, G., Rachman, S., fi Marks, I. Passive and
niques in the reduction of fearful avoidance be- participant modelling in exposure treatment of
havior. Behavior Thera p y, 1974, 5, 648-655. obsessive-compubive neuroti‹s. Bcloriotir Acfaorch
Lick, J., & Bootzin, R. Expectancy factors in the end Thump y, 1975, 13, 2 71-2 79.
treatment of fear: Methodological and theoretical Rosen, G. M., Rosen, E., & Reid, J. B. Cognitive
issues. Psychological Butlctin, 1975, 82, 917—931. desensitization and avoidance behavior: A re-
Litvak, S. B. A comparison of two brief group evaluation. /ooritAf o/ d tiaor of PiychoIog ,
1972, 80, 176—182.
behavior therapy techniques on the reduction of
avoidance behavior. The Psychological R acord, Ross, L., Rodin, J., fi Zimbardo, P. T. Toward an
1969, J9, 329-334. attribution therapy: The reduction of fear through
LoPiccolo, J. Effective components of systematic induced cognitive-emotional misattribution. Jour-
desensi’tization (Doctoral dissertation, Yale Uni- nal oJ Pzrsonolit y and Social Ps ycholog y, 1969,
versity, 1969) . D’usertotion Abstrects International, J2, 27W288.
1970, 31, 1543B. (University Microfilms No. 70- Rotter, J. B. Generalized expectancies for internal
16300) versus external control of reinforcement. Psy-
Maier, S. F., & Seligman, M. E. Learned helpless- chological dfoaogro#ls, 1966, 80(1, Whole No.
ness: Theory and evidence. /ournof oJ £s§cri- 609) .
mental Psychology, 1976, JOJ, 3—46. Sarason, I. G. Anxiety and self-preoccupation. In
Mandler, G. Mind and emotion. New York: Wiley, I. G. Sarason fi C. D. Spielberger (Eds. ), Stress
1975. and a›cxieiy (Vol. 2) . Washington, D.C.: Hem-
isphere, 1976.
McGlynn, F. D., & Mapp, R. H. Systematic desen-
sitization of snake-avoidance following three Schachter, S. The interaction of cognitive and
types of suggestion. Behaviour R zszorcli and physiological determinants of emotional state.
T£ero§y, 1970, 8, 197-201. In L. Berkowitz (Ed.) , Advanczs in espcrimtiiief
secret piyc8oIogy. New York: Academic Press,
McGlynn, F. D., Mealiea, W. L., fi Nawas, M. M.
Systematic desensitization of snake-avoidance 1964.
under two conditions of suggestion. Ps ychologicol Schachter, S., fi Singer, J. E. Cognitive, social, and
2td§orfs, 1969, 2J 220-2 22. physiological determinants of emotional state.
Psychological Review, 1962, 69, J79—399.
McGlynn, P. D., Reynolds, E. J., & Linder, L. H.
Systematic desensi’tration with pre-treatment and Sherman, A. R., Real-life exposure as a primary
intra-treatment therapeutic instructions. Behaviour therapeutic factor in the desensitization treat-
Research and Th ere p y, 1971, P, 57-63. ment of fear. Journal oJ Abnormal Ps ycholo$ y,
Meichenbaum, D. H. Examination of model char- 1972, 79, 19-28.
acteristics in reducing avoidance behavior. torrent Singerman, K. J., Borkovec, T. D., fi Baron, R. S.
Failure of a “misattribution therapy” manipula-
SELF-EFFICACY 2
tion w'th a clinically relevant target behavior.
iii tlr de elo pment end treotmcnt oJ emotioit&
Be havior Therapy, 1976, f, 30b—315.
disorders. Morristown, N.J.: General Learning
Stern, R., & Marks, I. Brief and prolonged flooding:
Press, 1971.
A comparison in agoraphobic patients. A rchints
Valins, S., & Ray, A. Effects of cognitive desensi-
oJ General Ps yctiatry, 191$, 28, 2 70-2 76.
tization on avoidance behaviour. Journal ol
Strahley, D. F. Systematic desensitization and
Personalit y end Social Ps ycholo g y, 1967, 7, 345—
counterphobic treatment of an irrational fear of
350.
snakes (Doctoral dissertation, University of Ten-
Watson, J. P., Mullett, G. E., ñ Pi8ay, H. The
nessee, 1965) . Dissertation Adetracts, 1966, 27,
effects of prolonged exposure to phobic situations
973B. (University Microfilms No. 66-5366)
upon agoraphobic patients treated in groups. Be-
Sushinsky, L. W., & Bootzin, R. R. Cognitive
honiour Research and Therap y, 19 1$, 11, 531—
desensitization as a rriodel of systematic desensi-
545.
tization. Behaviour Research and Thtrop y, 1970,
Weiner, B. f’fieoriei o/ motivoti0ti. Chicago:
8, 29-33.
Markham, 1972.
Szpiler, J. A., & Epstein, S. Availability of an
avoidance response as related to autonomic White, R. W. Motivation reconsidered: The con-
arousal. I ournal oJ Abnarmof Ps Echoing y, 1916, cept of competence. ft ycfioJogreat Aevietr, 195S,
85, 73-82. dd, 297-333.
Thase, b1. E., ñ Moss, M. K. The rela0ve efficacy of Wilson, G. T., & Davison, G. C. Processes of tear
.covert modeling procedures and guided participant reduction in systematic desensitization: Animal
modeling in the reduction of avoidance behavior. studies. Piyc hofogicaf Bulletin, 1971, 7d, 1—14.
Sourer oJ Bcharter Therop y and Ex:perimental Wolpe, J. 'The 9roctice o/ bedouror tficropy. New
Psychiatry, 1976, I, 7—12. York: Pergamon Press, 1974.
Valins, S., & Nisbett, R. E. Attribution firocesses
Received June 30, 1976 •

Self-Efficacy: Toward A Unifying Theory of Behavioral Change

Uploaded by

Copyright:

Available Formats

Self-Efficacy: Toward A Unifying Theory of Behavioral Change

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Self-Efficacy: Toward A Unifying Theory of Behavioral Change

Uploaded by

Copyright:

Available Formats

Self-efficacy: Toward a Unifying Theory of Behavioral Change

The present article presents an integrative theoretical framework to explain

role of cognition in the acquisition and PERSON BEHAVIOR OUTCOW

choice of behavioral settings. People fear and

Sources oJ E ficac y Expectations

the supplementary aids are withdrawn so

1970), or produces minor changes under Moreover, mislabeling arousal or attributing

trying because they expect their behavior to

PRE -TEST POST-TEST PRE-TEST I-•OST-TEST

tions at different phases of the experiment.

PRE-TE9T POST-TE5T PRE-TEST POST-TEST

ing was selected for this purpose because the

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.