American Psychologist
© 1991 by the American Psychological Association Volume 46(2), February 1991, p 157–162
Human Agency: The Rhetoric and the Reality

Bandura, Albert1

1Stanford University.



Rottschaefer's (1991, this issue) thoughtful commentary on the philosophical aspects of human agency seeks clarification on the issues of ontological reductionism and plurality. Mental events are brain processes, but emergent brain activities are not ontologically reducible. To use Bunge's (1977) analogy, the emergent properties of water, such as fluidity, viscosity, and transparency are not simply the aggregate properties of its microcomponents. With regard to ontological plurality, certain brain structures are specialized for mentation. The higher psychoneural systems are involved in the regulation of other subsystems. Thus, an emergent interactive agency assumes ontological nonreductionism and plurality of subsystems.

This does not mean that psychological theory is reducible to neurobiological theory. Much of psychology is concerned with discovering laws about how to structure psychosocial influences outside the organism that create brain states and orchestrate brain processes that subserve given purposes. Most of the subject matter of psychological theory regarding psychosocial factors does not have counterparts in neurobiological theory. For example, knowledge of the brain circuits involved in learning does not tell one much about how best to devise conditions of learning in terms of level of abstractness, novelty, and challenge; how to provide incentives to attend to, process, and organize relevant information; in what modes to present information; and whether learning is better achieved independently, cooperatively, or competitively. The optimal conditions must be specified by psychological laws.

Powers (1991, this issue) takes umbrage at my portrayal of the model of cybernetic control. My brief comments about control theory did not assess the cybernetic approach per se, but concerned the scope of some of the cybernetic analogues to human self-regulation of motivation and action. In these conceptions, individuals are typically portrayed as locked in negative feedback loops as the ruling control system constantly striving to reduce negative discrepancies between their goals and perceived attainments. My main point was that self-regulation by negative discrepancy tells only half of the story, and not necessarily the more interesting half.

People are proactive, aspiring organisms. Their capacity for forethought enables them to organize and regulate their lives proactively. Human self-regulation relies on discrepancy production as well as discrepancy reduction. People motivate and guide their actions through proactive control by setting themselves valued goals that create a state of disequilibrium and then mobilizing their abilities and effort on the basis of anticipatory estimations of what is required to reach the goals. Reactive feedback control comes into play in subsequent adjustments of strategies and effort to attain desired results. After people attain the goal they have been pursuing, those with a strong sense of efficacy set higher goals for themselves. Adopting further challenges creates new motivating discrepancies to be mastered. Self-regulation of motivation and action thus involves a dual control process of disequilibrating discrepancy production (proactive control), followed by equilibrating discrepancy reduction (reactive control).

In a cybernetic analogue, proactive and self-evaluating functions can be delegated to an executive control system. In the article under discussion (Bandura, September 1989), I identified a number of self-evaluating and proactive functions that have been verified in psychosocial research as important for self-regulation. A cybernetic analogue for human functioning must encompass such properties. They include predictive anticipatory control of strategies and effort; affective self-evaluative reactions to one's perceived performance, as rooted in a value system; self-appraisal of efficacy for goal attainment; and self-reflective metacognitive activity concerning the adequacy of one's self-efficacy appraisals and the suitability of one's strategies for goal setting.

Powers (1991) misperceives a unidimensional view of motivation that does not represent my conception. One must distinguish two definitional levels of a construct, each designed to serve different functions. Psychological constructs are defined at a generic level to delineate the boundaries of a selected phenomenon and to organize the body of knowledge relevant to it. But it is not the generic version of the construct that guides inquiry into the determinants and mechanisms governing the phenomenon of interest. The generic construct usually includes different facets, each with distinctive determinants and regulatory mechanisms. It is the particularized version of the construct that guides the selection of variables and investigatory methods. Consider, for purposes of illustration, the construct aggression. At the generic level, aggression is usually conceptualized in terms of actions intended to injure or destroy. Although injury and destruction are the common unifying properties, there are many different ways of hurting people and destroying things. To advance knowledge, one does not study an abstract aggression but rather its major variants—physical assault, verbal derogation, homicide, rape, child abuse, terrorism, political oppression, institutional discrimination, and the like. These different manifestations of aggression, such as child abuse and terrorism, must be analyzed separately because their determinants, mechanisms, and effects differ in important respects.

The construct of motivation likewise involves a dual level of specification serving different purposes. At the generic level, motivation is usually characterized in terms of directive and activating properties. But at the particularized level, motivation is analyzed in social cognitive theory as a multidimensional phenomenon indexed in terms of selection of pursuits from competing alternatives, intensity of effort, and persistence of exertion. Attempts to explain the motivational sources of behavior, therefore, are directed not at the generic notion, but rather at clarifying the determinants and intervening mechanisms that govern the selection, activation, and sustained direction of behavior.

Powers's (1991) comment regarding effort confounds several issues: effort as an index of motivation, the modifiability of effort by different means, and the multidimensionality of the construct of motivation. According to Powers, control theory uniquely identifies four ways in which effort is increased, as by raising goals, by strengthening commitment, or by imposing obstacles. These types of determinants had been identified years ago and thoroughly researched by goal theorists (see Locke & Latham, 1990). Moreover, sociocognitive perspectives specify conditions under which the simple operations that Powers describes as effort enhancers can, in fact, diminish effort. For example, burdening people with increased obstacles is likely to lower their perceived self-efficacy for goal attainment and decrease effort rather than raise it. Strong commitment will reduce effort if it is to a low goal that deters adoption of a higher goal calling for more effort. Because effort can be heightened in different ways, we are told that control theory shows that motivation is not a unitary concept. In fact, the multidetermination of effort shows that the same phenomenon (i.e., effort) can be produced in different ways, not that different phenomena are merged into a catchall concept of motivation. I doubt that anyone holds the view that effort raised by threat involves identical processes to effort raised by personal goal setting, nor that one need not be concerned with the distinctive processes by which different determinants heighten effort.

Locke (in press) has argued that much of control theory involves translation of the principles and knowledge of goal theory into a stilted machine language without providing a new perspective or predictive benefits. He further showed that adherents of control theory have now grafted so many ideas from other theories on the negative feedback loop as to remedy its prediction problems that control theory has lost its distinctiveness. Powers's (1991) discussion of the effect of belief on effort gives some evidence for Locke's observations. Sociocognitive approaches distinguish between different types of beliefs such as beliefs about one's capabilities, the nature of human ability, contingency structures, controllability of the environment, and social roles and norms, to mention just a few. Research has advanced our understanding of the determinants of these different beliefs and the processes through which they affect human motivation, affect, and action. Powers's translation of beliefs and related processes into such common terms as “optimistic goals,” “imagined success,” and “perceived effectiveness” offers neither any definitional refinements of constructs nor any new implications.

Powers (1991) invests control theory with all kinds of conceptual and investigatory virtues. It allegedly removes natural language ambiguities, provides new ways of interpreting commonsense ideas, generates different implications, and refines experimental investigations. Control theory has been around for a long time. Where is the empirical evidence for its predictive and operative superiority? Powers's view that the current social sciences are not up to the task of providing reliable data to test a working model of control theory suggests that this theory need not meet the test of reality until the prescribed salvation of the social sciences is achieved.

Corcoran (1991, this issue) succeeds in misunderstanding the issues he raises about perceived self-efficacy. He claims that perceived self-efficacy and locus of control are essentially equivalent constructs, but manages to ignore the empirical evidence that disputes that very claim. These two constructs are clearly distinguishable both conceptually and empirically. Perceived self-efficacy is concerned with people's beliefs about their capabilities to organize and execute designated courses of action. Locus of control refers to people's beliefs that outcomes are dependent on their actions or are the result of chance, fate, or luck. Beliefs about whether one can produce certain performances cannot, by any stretch of the imagination, be considered the same as beliefs about whether actions affect outcomes.

Nor does the claim of equivalence survive empirical scrutiny. If two constructs are the same, then measures of them should be very highly correlated and should, in turn, correlate similarly with other variables. A number of studies have included measures of perceived self-efficacy and locus of control, assessed by either the Rotter scale or the more specific health locus-of-control scale. The specific scale presumably affords better predictions because it is tailored to the domain of study. That research shows that locus of control and perceived self-efficacy bear little or no relation to each other. The correlations are r = .11 (Smith, 1989), r = -.01 (Manning & Wright, 1983), and r = -.22 averaged across self-efficacy subscales (Taylor & Popma, 1990). With regard to the pattern of correlates, perceived self-efficacy predicts such diverse events as academic performance, proneness to anxiety, level of pain tolerance, and career decision making, but locus of control does not predict any of these events (Manning & Wright, 1983; Smith, 1989; Taylor & Popma, 1990). The more circumscribed health locus-of-control measure does not fare any better. Perceived self-efficacy is a uniformly good predictor of adoption of diverse forms of health behavior, whereas health locus of control does not predict any of these behaviors (Alagna & Reddy, 1984; Beck & Lund, 1981; Brod & Hall, 1984; Kaplan, Atkins, & Reinsch, 1984; Walker & Franzini, 1983). So much for the alleged equivalence.

Self-efficacy theory was initially examined in terms of perceived capabilities to exercise control over performances to fulfill coping task demands. The theory has since been extended to perceived capabilities to exercise control over one's own thought processes and affective states, to the self-regulation of goal-directed pursuits and impulsive and addictive behavior, to the exercise of control over social environments, and to collective efficacy in which the group's belief about its capabilities as a whole is the unit of analysis. Contrary to Corcoran's (1991) belief, the definition and measurement of perceived self-efficacy in terms of perceived capability has never changed. Rather, the contributory role of self-efficacy beliefs has been extended to diverse facets of psychosocial functioning. Obviously, people act on their self-efficacy beliefs to exercise some control over events that affect their lives. Corcoran has apparently misread this to infer a definitional change that has no foundation in fact. Nor does including outcome expectations in a model of human behavior in any way change the definition of the construct of self-efficacy.

Human adaptation requires recognizing conditional relations between environmental events and between actions and outcomes. In a program of research with Robert Wood on managing a simulated environment, I demonstrated that experimentally enhanced perceived self-efficacy was accompanied by efficient use of analytic strategies to discover the rules to manage that social environment. In contrast, diminished perceived self-efficacy leads to erratic analytic thinking. Corcoran (1991) misconstrues reference to predictive rules for altering organizational outcomes as a further indication of some shift in the definition of self-efficacy. On the basis of an erroneous equation of particularized predictive rules with generalized expectations, Corcoran launches into amusing diagnostic misattributions that I failed to link the two because generalized expectations have “behavioral overtones” (p. 156). There exists a simpler and less subterranean explanation: They are unlinked because specific rules are not generalized expectations.

Corcoran's (1991) puzzling claim that my view of motivation relies solely on a self-efficacy source is also unfounded. It disregards the text of the article, which explicitly specified multiple sources of motivation, including outcome expectations. People do not go around exercising their perceived self-efficacy in activities devoid of any outcomes. Corcoran asks what makes something fearful or depressing? The answer is hardly a huge mystery. Events possess inherent and endowed properties of a positive or negative sort. People are inclined to become anxious when they perceive themselves as unable to manage aversive events. They are likely to become depressed when they perceive themselves as unable to prevent an important loss or to gain what they value highly. Because losses of what one values highly often produce aversive outcomes as well, perceived self-efficacy is usually both distressing and depressing (Bandura, 1986).

Corcoran (1991) compounds confusion about the role of self-efficacy beliefs and outcome expectations in the exercise of control. A control process involves two control belief systems operating in concert—namely, self-efficacy beliefs and outcome expectations (Bandura, 1977). As explained in the article under discussion (Bandura, 1989), self-efficacy theory distinguishes degrees of controllability by personal means, which affects the extent to which outcome expectations contribute incremental prediction. In activities with outcomes that are highly contingent on quality of performance, self-judged efficacy accounts for most of the variance in expected outcomes. When variations in perceived self-efficacy are partialed out, the outcomes expected for given performances are redundant and do not add a predictive increment. Self-efficacy beliefs subsume only part of the variance in expected outcomes when outcomes are not completely controlled by quality of performance. And finally, expected outcomes are independent of perceived self-efficacy if contingencies are restrictively structured so that no level of competence can produce desired outcomes. Perceived environmental constraints and opportunity structures alter how efficacy and outcome information are cognitively processed. I have explained elsewhere the way in which different patterns of self-efficacy beliefs and outcomes expectations produce distinctive motivational, affective, and behavioral effects (Bandura, 1982b). Corcoran cites only the case in which outcomes are under full personal control, and then he concludes, illogically, that I am mistaken in assuming that all outcomes are under full personal control. Self-efficacy theory does not rest on the absurd assumption that life outcomes are completely under personal control—a circumstance that would require divine agency of the highest order.

Corcoran (1991) is displeased that I (Bandura, 1989) did not provide a criterion for when persistence is “maladjusted.” I have no interest in psychopathologizing human persistence. The functional value of veridical self-appraisal depends on the nature of the endeavor. In activities in which the margins of error are narrow, and missteps can produce costly or injurious consequences, personal well-being is best served by highly accurate self-appraisal. It is a different matter if difficult accomplishments can produce substantial personal or social benefits and if the personal costs involve time, effort, and expendable resources. Individuals have to decide for themselves which creative abilities to cultivate, whether to invest their efforts and resources in endeavors that are difficult to fulfill, and how much hardship they are willing to endure for pursuits strewn with obstacles. Societies enjoy considerable benefits from the eventual accomplishments of its persisters.

Corcoran (1991) tells us that the causal contribution of self-efficacy beliefs is questioned by evidence that wives' beliefs about their husbands' cardiac capabilities predicted the husbands' recovery of cardiac function as well as did the husbands' beliefs. Let us examine the logic of this argument. Spouses are known to talk to each other from time to time. If a wife knows how her husband judges his efficacy, then she has a pretty good idea of the type of future her husband is likely to produce. Marital partners also interact with each other in ways that mutually shape their developmental trajectories. A wife who believes her husband has a robust heart is more likely to encourage him to resume an active life, thus improving cardiac functioning, than is a wife who believes her husband's heart is impaired and at risk for further damage if he becomes active. It would require a complete spousectomy for wives to stay ignorant of their husbands' sense of efficacy, and for wives' beliefs about their husbands' cardiac capabilities to have absolutely no effect on which activities they encourage their husbands to undertake.

The causal contribution of self-efficacy beliefs to human functioning has, in fact, been subjected to stringent experimental tests. In these experiments, using diverse paradigms, self-efficacy beliefs were systematically varied by (a) observational exposure to modeled strategies, (b) controlling level of ability but varying perceived self-efficacy within ability levels, (c) introducing trivial factors that bias self-efficacy judgment, but are devoid of information to affect skills, (d) altering self-efficacy beliefs through bogus social comparative or normative information unrelated to one's actual performance, (e) creating changes that can impair functioning but in ways that raise perceived self-efficacy, and (f) conducting panel studies of multivariate relations that reveal how much variation in the predicted variable is explained by perceived self-efficacy when the influence of other determinants is controlled (Bandura, 1989). These diverse tests spanned different modes of efficacy induction, varied populations, multiple response systems and domains of functioning, and intergroup and intraindividual experimental designs that examined microlevel and macrolevel relations. The evidence consistently shows that perceived self-efficacy significantly contributes to level of motivation and to performance accomplishments. Evidence that divergent procedures produce convergent results adds strength to the explanatory and predictive generality of the efficacy mediator.

In recent years there have been major changes in the conception of human skill and competence (Bandura, 1990; Sternberg & Kolligian, 1990). A skill is not a fixed property that one does or does not have within one's behavioral repertoire. Rather, skill involves a generative capability in which cognitive, social, and behavioral skills must be organized and effectively orchestrated to serve a host of purposes. There is a marked difference between possessing knowledge and skills and being able to use them well under difficult circumstances. People with the same skills may perform poorly, adequately, or outstandingly depending on fluctuations of nonability influences. Corcoran (1991) reports that my previous critique of skill as a preformed entity residing in a “behavioral repertoire” reflects a common mistake that skill development involves learning any new competencies. My previous remarks regarding the conception of skill were, in fact, primarily concerned with the variable orchestration of preexisting subskills, not with their development. Nevertheless, I was very surprised to find that Corcoran (p. 156) views skill development as a “metaphoric expression,” and that skill training does not “teach them a new behavior” but only “increase(s) the occurrence of a given behavior.” Pity those indefatigable violinists, pianists, composers, managers, boxers, and graduate students who learned nothing new from their extended labors, beyond boosting their rates of selected responses. From Corcoran's perspective, Bach cannot be credited with creating new musical masterpieces—he merely arranged preexisting notes. Because of space constraints, I am happy to rest my case on the credibility of Corcoran's own testimony that neither acquisition of new competencies nor behavioral creativity exist, only changes in response frequencies.

Corcoran (1991) assumes that if skills exist within one's behavioral repertoire, managing environmental demands reduces to a simple matter of choice, and self-efficacy beliefs become behavioral predictions. In this agentless view, self-beliefs reside in the host organism as forecasters of future performances that get realized mainly by the work of environmental forces. There is nothing like a trivial explanation to bring inquiry to a halt. Numerous experiments have been conducted to clarify the psychological processes through which self-efficacy beliefs affect the course of human functioning. The findings of these different lines of research reveal that self-beliefs of efficacy are not just inert predictors of future performances. Self-efficacy beliefs alter thought processes, the level and persistency of motivation, and affective states, all of which contribute importantly to the types of performances that get realized (Bandura, 1989). In short, performances do not just happen to us, we do a lot to bring them about.

Corcoran's (1991) view of self-efficacy beliefs as inert forecasts is grounded in a misleading dichotomy that coping with threats and exercise of control over addictive behavior are matters of choice not skill, whereas physical activities involving hitting targets from a distance are expressions of skill. This supposition fails to recognize that shooting basketballs through hoops and controlling urges to engage in addictive behavior both involve the exercise of skills, but the latter relies on self-regulatory skills, the former on motor skills. It is not as though merely deciding to cease addictive behavior permanently eliminates it without the individual repeatedly mobilizing self-regulative efforts and strategies to bring it about. Corcoran's imaginary study offering make-believe payments for hypothetical performances simply demonstrates that, even in pretend undertakings, it is tough to raise self-efficacy by social persuasion alone for tasks requiring exacting accuracy. This pretend exercise has no relevance to the issue of self-belief causality. Space constraints do not permit a reply to related issues raised by Corcoran that have been addressed elsewhere in detail (Bandura, 1982a, 1984, 1986).

Corcoran (1991) claims that self-efficacy beliefs do not operate in activities that involve “overcoming anxiety (or expected anxiety)” (p. 156). This is a complete misrepresentation of the evidence. In fact, self-efficacy beliefs are good predictors of avoidant behavior that is accompanied by anxiety arousal. Williams (1987) has analyzed by partial correlation numerous data sets from studies in which perceived self-efficacy, anticipated anxiety, and phobic behavior were all measured. As shown in Table 1, perceived self-efficacy accounts for a substantial amount of variance in phobic behavior when anticipated anxiety is partialed out, whereas the relationship between anticipated anxiety and phobic behavior essentially disappears when perceived self-efficacy is partialed out.

Table 1 Comparison of the Relation Between Perceived Self-Efficacy and Coping Behavior When Anticipated Anxiety Is Controlled, and the Relation Between Anticipated Anxiety and Coping Behavior When Perceived Self-Efficacy Is Controlled

Perceived self-efficacy in controlling ruminative thought is emerging as an important factor in anxiety and depression (Kavanagh & Wilson, 1989; Kent & Gibbons, 1987). Recent evidence reveals a dual path of regulation of anxiety and avoidant behavior by perceived self-efficacy (Ozer & Bandura, 1990). One path of influence on avoidant behavior is mediated through the effects of perceived coping self-efficacy on perceived vulnerability and risk discernment; the second path of influence operates on avoidant behavior and anxiety arousal through the impact of perceived cognitive control self-efficacy on intrusive aversive thoughts. The analyses Corcoran (1991) cites regarding the determinants of anxiety arousal are based on studies in which the facet of perceived self-efficacy most relevant to anxiety (i.e., perceived capability to control aversive thought) was never measured.

Corcoran (1991) would lead readers to believe that I am an advocate of a one-factor theory of human behavior in which perceived self-efficacy carries the entire explanatory burden. The article under discussion (Bandura, 1989) is concerned with social cognitive theory, which includes perceived self-efficacy as one of many factors governing human motivation and action. I am at a loss to understand Corcoran's failure to distinguish between social cognitive theory and the self-efficacy component, considering that other classes of determinants were explicitly discussed in some detail. As explained in the article, social cognitive theory posits a multifaceted causal structure. Knowledge structures serve as guides for the construction of complex modes of behavior. The transformational and generative operations by which cognitive models are translated into proficient action were discussed, as well as the changes that occur in multilevel regulation of skills as they are perfected. In addition to the regulative function of self-efficacy appraisal, the anticipative mechanism of forethought in the regulation of human motivation and action received attention. Predictive knowledge of conditional relations between environmental events fosters foresightful adaptations. The ability to envision the likely outcomes of prospective actions is another way in which anticipative mechanisms contribute to human motivation and action. These outcome expectancies may take the form of external, vicarious, or self-generated consequences, often operating in concert to influence the course of human action. Cognized goals and internal standards rooted in value systems create self-incentives and guides for action through self-regulatory mechanisms. In short, a comprehensive psychological theory must address the multidetermination of human motivation and action beyond the simple notion that if a behavior occurs there must be some kind of reinforcement for it somewhere. Significant progress has been made in clarifying how perceived self-efficacy operates in a causal structure in conjunction with other determinants emphasized by social cognitive theory.


Alagna, S. W., & Reddy, D. M. (1984). Predictors of proficient technique and successful lesion detection in breast self-examination. Health Psychology, 3, 113–127. [Context Link]

Arnow, B. A., Taylor, C. B., Agras, W. S., & Telch, M. J. (1985). Enhancing agoraphobia treatment by changing couple communication patterns. Behavior Therapy, 16, 452–467.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191–215. [Context Link]

Bandura, A. (1982a). The assessment and predictive generality of self-percepts of efficacy. Journal of Behavior Therapy and Experimental Psychiatry, 13, 195–199. [Context Link]

Bandura, A. (1982b). Self-efficacy mechanism in human agency. American Psychologist, 37, 122–147. [Context Link]

Bandura, A. (1984). Recycling misconceptions of perceived self-efficacy. Cognitive Therapy and Research, 8, 231–255. [Context Link]

Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory. Journal of Clinical and Social Psychology, 4, 359–373. [Context Link]

Bandura, A. (1989). Perceived self-efficacy in the exercise of personal agency. The Psychologist: Bulletin of the British Psychological Society, 2, 411–424. [Context Link]

Bandura, A. (1990). Reflections on nonability determinants of competence. In R. J. Sternberg & J. Kolligian, Jr. (Eds.), Competence considered (pp. 315–362). New Haven, CT: Yale University Press. [Context Link]

Beck, K. H., & Lund, A. K. (1981). The effects of health threat seriousness and personal efficacy upon intentions and behavior. Journal of Applied Social Psychology, 11, 401–415. [Context Link]

Brod, M. I., & Hall, S. M. (1984). Joiners and non-joiners in smoking treatment: A comparison of psychosocial variables. Addictive Behaviors, 9, 217–221. [Context Link]

Bunge, M. (1977). Emergence and the mind. Neuroscience, 2, 501–509. [Context Link]

Corcoran, K. J. (1991). Efficacy, “skills,” reinforcement, and choice behavior. American Psychologist, 46, 155–157. [Context Link]

Kaplan, R. M., Atkins, C. J., & Reinsch, S. (1984). Specific efficacy expectations mediate exercise compliance in patients with COPD. Health Psychology, 3, 223–242. [Context Link]

Kavanagh, D. J., & Wilson, P. H. (1989). Prediction of outcome with a group version of cognitive therapy for depression. Behaviour Research and Therapy, 27, 333–347. [Context Link]

Kent, G., & Gibbons, R. (1987). Self-efficacy and the control of anxious cognitions. Journal of Behavior Therapy & Experimental Psychiatry, 18, 33–40. [Context Link]

Kirsch, I., Tennen, H., Wickless, C., Saccone, A. J., & Cody, S. (1983). The role of expectancy in fear reduction. Behavior Therapy, 14, 520–533.

Locke, E. A. (in press). Goal theory vs. control theory: Constrasting approaches to understanding work motivation. Motivation and Emotion. [Context Link]

Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood Cliffs, NJ: Prentice-Hall. [Context Link]

Manning, M. M., & Wright, T. L. (1983). Self-efficacy expectancies, outcome expectancies, and the persistence of pain control in childbirth. Journal of Personality and Social Psychology, 45, 421–431. [Context Link]

Ozer, E., & Bandura, A. (1990). Mechanisms governing empowerment effects: A self-efficacy analysis. Journal of Personality and Social Psychology, 58, 472–486. [Context Link]

Powers, W. T. (1991). Comment on Bandura's “human agency.” American Psychologist, 46, 151–153. [Context Link]

Rottschaefer, W. (1991). Some philosophical implications of Bandura's social cognitive theory of human agency. American Psychologist, 46, 153–155. [Context Link]

Smith, R. E. (1989). Effects of coping skills training on generalized self-efficacy and locus of control. Journal of Personality and Social Psychology, 56, 228–233. [Context Link]

Sternberg, R. J., & Kolligian, J., Jr. (Eds.). (1990). Competence considered. New Haven, CT: Yale University Press. [Context Link]

Taylor, K. M., & Popma, J. (1990). An examination of the relationships among career decision-making self-efficacy, career salience, locus of control, and vocational indecision. Journal of Vocational Behavior, 37, 17–31. [Context Link]

Telch, M. J., Agras, W. S., Taylor, C. B., Roth, W. T., & Gallen, C. C. (1985). Combined pharmacological and behavioral treatment of agoraphobia. Behaviour Research and Therapy, 23, 325–335.

Walker, W. B., & Franzini, L. R. (1983, April). Self-efficacy and low-risk aversive group treatments for smoking cessation. Paper presented at the annual convention of the Western Psychological Association, San Francisco. [Context Link]

Williams, S. L. (1987). On anxiety and phobia. Journal of Anxiety Disorders, 1, 161–180. [Context Link]

Williams, S. L., Dooseman, G., & Kleifield, E. (1984). Comparative power of guided mastery and exposure treatments for intractable phobias. Journal of Consulting and Clinical Psychology, 52, 505–518.

Williams, S. L., Kinney, P. J., & Falbo, J. (1989). Generalization of therapeutic changes in agoraphobia: The role of perceived self-efficacy. Journal of Consulting and Clinical Psychology, 57, 436–442.

Williams, S. L., & Rappoport, A. (1983). Cognitive treatment in the natural environment for agoraphobics. Behavior Therapy, 14, 299–313.

Williams, S. L., Turner, S. M., & Peer, D. F. (1985). Guided mastery and performance desensitization treatments for severe acrophobia. Journal of Consulting and Clinical Psychology, 53, 237–247.

Accession Number: 00000487-199102000-00014