This article is for use only of students enrolled in GPS 520: Advanced Social Psychology for purposes associated with this course and may not be retained or further disseminated.

Hergenhahn, B. R., & Olson, M. H. (1999).  An introduction to theories of personality (5^th Ed).  Upper Saddle River, NJ: Prentice Hall.

CHAPTER ELEVEN

Biographical Sketches

Albert Bandura

Albert Bandura was born on December 4, 1925, in Mundare, a small town in the province of Alberta, Canada. His parents were wheat farmers of Polish heritage. The high school he attended had only 20 students and 2 teachers. Following graduation from high school, Bandura spent the summer working on the Alaskan highway. Many of the men with whom he worked had fled to Alaska to escape "creditors, alimony, and probation officers." Working with such characters instilled in Bandura "a keen appreciation for the psychopathology of everyday life" (American Psychologist, 1981). Bandura entered the University of British Columbia in 1946 and obtained his BA in 1949, with a major in psychology. He then went to the University of Iowa where he obtained his MA in 1951 and his PhD in 1952. It was at the University of Iowa that Bandura met his future wife, Virginia (Ginny) Varns, who was teaching in the school of nursing there. Bandura and his wife eventually had two daughters, Mary and Carol. After a year's clinical internship at the Wichita, Kansas, Guidance Center, he moved to Stanford University, where he has been ever since.

After arriving at Stanford, Bandura began working on the familial causes of aggression with his first graduate student, Richard Walters (1918‑1967). It was during this work that Bandura became aware of the importance of modeling and observational learning for personality development. A section on observational learning appears later in this chapter. Bandura's first book (coauthored with Richard Walters) was Adolescent Aggression (1959). His second book (also written with Richard Walters) was Social Learning and Personality Development (1963). Subsequent books were Principles of Behavior Modification (1969), Aggression: A Social‑Leaming Analysis (1973), Social Learning Theory (1977), Social Foundations of Thought and Action: A Social Cognitive Theory (1986), Self‑Efficacy in Changing Societies (1995) of which he was editor; and Self‑Efficacy: The Exercise of Control (1997). In addition to his books, Bandura has written many influential articles.

Included among Bandura's many honors are: Fellow of the Center for Advanced Study in the Behavioral Sciences, 1969 to 1970; a Guggenheim fellowship, 1972; a Distinguished Scientist Award from Division 12 (Clinical Psychology) of the American Psychological Association (APA), 1972; a Distinguished Scientific Achievement Award from the California Psychological Association, 1973; president of the APA, 1974; awarded an endowed chair, the David Starr Jordan Professor of Social Science in Psychology, 1974; recipient of the James McKeen Cattell Award, 1977; recipient of an honorary Doctor of Science degree from the University of British Columbia, 1979; elected fellow of the American Academy of Arts and Sciences, 1980; recipient of a Distinguished Contribution Award from the International Society for Research on Aggression, 1980; served as president of the Western Psychological Association, 1980; recipient of the Distinguished Contribution Award from the APA, 1980; and elected a member of the Institute of Medicine of the National Academy of Sciences, 1989.

Walter Mischel

Walter Mischel, the second son of upper‑middle‑class parents, was born on February 22, 1930 in Vienna, Austria, within walking distance of Freud's house. In 1938, when Mischel was 9 years old, the Nazis invaded Austria, and his family moved to the United States. After living in various parts of the country, the family settled in Brooklyn, New York in 1940. There, Mischel attended primary and secondary school and earned a college scholarship, but because his father became ill he was forced to work. He worked as a stock boy, elevator operator, and assistant in a garment factory before he was finally able to attend New York University, where he pursued his interests in painting, sculpture, and psychology.

In college, Mischel was disenchanted by behavioristic psychology but was attracted to psychoanalysis. At the same time, he strengthened his humanistic inclinations by reading existential philosophy and poetry. After graduation from college Mischel entered the MA program in clinical psychology at City College of New York. While working on his degree Mischel became a social worker working primarily with juvenile delinquents in the lower east side slums of New York. It was this work that caused Mischel to doubt the usefulness of psychoanalytic theory.

Between 1953 and 1956 Mischel worked on his PhD at Ohio State University where he came under the influence of George Kelly and Julian Rotter. Rotter's work emphasized the importance of expectancies in human behavior, and Kelly stressed the importance of the formulation of mental concepts (personal constructs) in dealing with the world. Both Rotter and Kelly emphasized cognitive events in dealing with the current situations and deemphasized the importance of traits and early developmental experience. The influence of both men is seen in Mischel's work. George Kelly's theory is covered in chapter 13.

From 1956 to 1958, Mischel lived in a Trinidad village in the Caribbean studying religious cults that practiced spirit possession. It was at this time that Mischel observed that some people have the ability to reject small, immediate re­wards in favor of larger, but delayed, rewards. Mischel also observed that natives with this ability to delay gratification had higher needs for achievement and showed more social responsibility (1958, 1961a, 1961b). As we will see later in this chapter, studying delayed gratification became one of Mischel's lifelong pas­sions.

Mischel next taught for two years at the University of Colorado before joining the Department of Social Relations at Harvard. While at Harvard his interests in personality theory and assessment were furthered by discussions with Gordon All­port. It was also at Harvard that Mischel met his future wife, Harriet Nerlove, who was a graduate student in cognitive psychology. The Mischels eventually had three daughters and have collaborated on several studies. In 1962, Mischel moved to Stanford where he became a colleague of Bandura. In 1983, after more than 20 years at Stanford, Mischel returned to New York City, where he joined the faculty at Columbia University. There, Mischel continues to pursue his long-standing in­terests in delayed gratification, self‑control, and the cognitive processes utilized by individuals in their interactions with the world.

In 1978, Mischel received a Distinguished Scientist Award from the clinical division of the American Psychological Association, and its Distinguished Scien­tific Contribution Award in 1982. His books include Personality and Assessment (1968) and Introduction to Personality, first published in 1971 and revised in 1976, 1981, 1986, and 1993.

Consistency of Human Behavior

Through the years, most personality theorists have assumed that a person's behavior is fairly consistent over time and across similar situations. That is, it was assurned that how people act at one time in their lives will be more or less how they act at other times, and that they will tend to respond to similar situations in similar ways. For example, if a person was outgoing in one social situation, it was assurned he or she would be outgoing in other situations and will continue to re spond in that characteristic way throughout most of his or her life. It was also assumed that scores on various personality tests and questionnaires would correlate significantly with actual behavior. That is, if a person scored high on a scale intended to measure introversion, he or she would tend to be introverted in social situations. The question was not whether behavior was consistent; rather, theorists attempted to account for the consistency they assumed existed. Psychoanalytic theory attempted to account for it by postulating repressed experiences, complexes, fixations, or internalized values. Within psychoanalytic theory, however, the conclusion of consistency can sometimes only be reached by a trained psychoanalyst because it is assumed that sometimes extreme aggression really means passivity, love sometimes really means hate, and repulsion sometimes really means attraction, to give but a few examples. Trait theory postulated enduring traits to explain why, for example, a neat person tended to be neat in a wide variety of situations. Learning theory emphasized the role of reinforcement. That is, behavior that was reinforced tended to persist and to transfer to situations similar to the one in which the reinforcement had occurred.

Traditionally, personality theory has attempted to account for individual differences among people, but it was always assumed that although people differ from one another, the behavior of any given individual tended to be consistent across similar situations and over time. In the early 1960s, Mischel was given the task of predicting the success of Peace Corps volunteers. In this effort, he found that standardized personality tests designed to measure traits were weak predictors of behavior (Mischel, 1965). He found that people are better predictors of their own behavior than the best available personality tests. It was this work that led to Mischel's highly influential book Personality and Assessment (1968). In this book, Mischel reviewed many studies designed to measure consistency of behavior across situations or to measure the relationship between performance on personality questionnaires and actual behavior. He found the typical correlation to be about 0.30. Mischel called this weak correlation the personality coefficient, and he argued that the weakness of the coefficient was not due to problems in measuring such characteristics as traits or behavior but, rather, to the fact that human behavior is simply not very consistent.

Thus, according to Mischel, although traits and various other inner states have been used to describe behavior for a long time, they are of limited use in actually predicting behavior. The use of traits to describe and explain behavior, says Mischel, often indicates what the theorist thinks should be the case rather than what actually is the case. Mischel (1990) uses the term consistency paradox to describe the fact that both laypersons and professional psychologists persist in believing that people's behavior is consistent when evidence indicates that it is not (e.g., Carlson & Mulaik, 1993; Hayden & Mischel, 1976). Chapman and Chapman (1969) coined the term illusory correlation to describe the belief or perception that variables are correlated when, in fact, they are not. According to Mischel, the belief that traits and behavior are highly correlated is an example. Mischel ended Personality and Assessment (1968) with the plea to abandon the trait approach to understanding personality, "This conceptualization of man, besides being philosophically unappetizing, is contradicted by massive experimental data" (p. 301).

            Since 1968, Mischel and many others have performed research that confirms the inconsistency of human behavior. For example, in a much cited study, Mischel and Peake (1982) examined the consistency of college students on two "traits"conscientiousness and friendliness. While common sense, and many psychologists, would predict that students would behave consistently across situations according to the degree to which they possess the traits in question, the prediction was not confirmed. There was essentially no intersituational consistency (r = .13).

            In the years since 1968, Mischel has moderated his attack on trait theory (1979, 1984, 1990). He no longer denies the value of describing people in terms of such traits as intelligence, friendliness, and aggressiveness. He does say, however, that it is practically impossible to predict how these, and other, traits will manifest themselves in a person's behavior in any given situation. Thus, Mischel no longer wishes to purge psychology of the concept of trait. Rather, he seeks to clarify how the traits that an individual may possess actually contribute to that person's behavior. For example, he wants to be able to make conditional statements such as "person A, in situation x, tends to do y" (Wright & Mischel, 1987). In other words, Mischel remains opposed to making generalized predictions about people based on the traits that they supposedly possess. Whatever consistency a person may demonstrate can only be discovered by observing how he or she responds to specific situations. Cantor and Mischel (1979) say, "Consistency of personality may be demonstrable only when theorists abandon their search for distinctive nornothetic traits and instead look for more idiographic patterns of person‑environment interactions" (p. 43).

                Although Mischel's observation that human behavior is not as consistent as was traditionally believed caused quite a stir among personality theorists, he believes their concern was misguided. Mischel believes that the degree of consistency previously thought to exist may actually be maladaptive.

Personologists have long searched for behavioral consistencies from situation to situation as if such consistency were the essence of personality. Perhaps that form of consistency will prove to be more characteristic of rigid, maladaptive, incompetent social functioning than of the integrated individual. (1984, p. 360)

            Mischel's major criticism of traditional personality theories is that they emphasize person variables and deemphasize situation variables. Person variables are those aspects of a person, like traits, habits, and repressed experiences, that are assumed to cause the person to act consistently in a variety of similar situations.

According to those theories emphasizing person variables, behavior is consistent because internal variables are enduring and thus continue to generate the same behavior patterns. Situation variables consist of the environmental circumstances in which the person finds himself or herself.

Although Mischel believes situation variables have been deemphasized by most personality theories, he believes they have been overemphasized by certain behavioristic theories-for instance, Skinner's. According to Mischel, Skinner's attempt to explain behavior entirely in terms of environmental conditions overlooks the significant contributions made by the person to his or her own behavior. What is needed, says Mischel, is a theory that considers the contributions of both the person and the situation.

Reciprocal Determinism

The position taken by the social cognitive theorist is called reciprocal determinism that means person variables, situation variables, and behavior continuously interact with one another. Situation variables provide the setting in which a person behaves, person variables determine how a situation is analyzed and which behaviors are chosen, and behavior both provides information concerning the person's analysis of the situation and modifies the environment. Bandura (1986, p. 24) diagrams reciprocal determinism as follows, where P is the person, E is the environment, and B is the person's behavior:

One implication of reciprocal determinism is that people's beliefs about themselves and the world will influence both how they behave and the environments in which they place themselves. In turn, feedback from behavior and environmental experience will confirm or disconfirm people's beliefs. For example, a gregarious person believing that people are generally friendly will tend to respond warmly to strangers. Such positive responses tend to create an environment that encourages positive human interactions. Thus the person's gregarious tendency is confirmed and strengthened. In contrast, a shy, withdrawn person will dread interpersonal interactions and therefore avoid or minimize them. Such reactions create social environments that are negative and encourage the person to remain shy and withdrawn. However, beliefs are not the only aspect of a person that influence his or her environment. Person variables also include a person's gender, social position, size, and physical attractiveness. Bandura (1978) says, "People activate different environmental reactions, apart from their behavior, by their physical characteristics (e.g., size, physiognomy, race, sex, attractiveness) and socially conferred attributes, roles, and status" (p. 346).

Social cognitive theory, then, does not exclude person variables. In fact, they are considered extremely important. The person variables postulated, however, are not the traditional types such as traits, habits, and repressed experiences. Rather, they are beliefs, values, and information-processing strategies. Furthermore, they are important only insofar as they manifest themselves in the present.

Cognitive Social Person Variables

Social cognitive theory views the interaction between the person and the environment as highly complex and individualistic. Each individual brings to each situation the remnants of previous experience, which are used to deal with the present situation. The outcome of the negotiations with the present situations, in turn, influence how similar situations are dealt with in the future. How a given individual will interact with a situation is determined by what Mischel calls cognitive social person variables (1993, p. 403). It is these variables that determine which of the myriad stimuli confronted by a person are perceived, selected, interpreted, and used. Unlike the traditional person variables, the person variables in social cognitive theory are active cognitive processes that operate in the present. These variables provide dynamic, reciprocal relationships among the person, the person's behavior, and the environment.

Mischel (1993, pp. 403-411) describes five cognitive social learning person variables.

1. Encoding strategies: How we see things. People not only select different aspects of the environment to attend to, but they also assign different meanings to the stimuli selected. Mischel (1993) says:

People differ greatly in how they encode (represent, symbolize) and group information from stimulus inputs. The same "hot weather" that upsets one person may be a joy for another who views it as a chance to go to the beach. The same stranger in the elevator who is perceived as "dangerous" by one person may be seen as "attractive" by another. (p. 404)

To understand a person it is necessary to know the categories with which he or she embraces the world. Here we see the influence of George Kelly (see chapter 13) on Mischel. Kelly suggested that any given event can be construed (interpreted) in any number of ways and that people are free to choose the constructs (concepts, categories, symbols, or words) with which they interpret their experiences. This means that in the same physical situation, people will encode it, construe it, and respond to it differently. Furthermore, because person variables are dynamic, no reason exists to assume that even for a given person the same situation will be construed the same way twice. It is a person's encoding strategies that provide some consistency in behavior, but the fact that those strategies can be changed by the individual at any time at least partially accounts for the fact that cross-situational behavior is not very consistent. Encoding strategies determine what aspects of the world are attended to and how they are interpreted.

2. Expectancies: What we think will happen. The person variable of encoding strategies determines how one categorizes experience. At some point, however, people must actually act on the environment. The most important variable for actual performance is a person's expectations. For example, in a given situation, the person hypothesizes, If I act in this way, it will have the following result. This is called a behavior-outcome expectancy. In the absence of any information about a specific situation, one tends to create expectancies based on past experience with similar situations. If specific information is available, however, one's expectations change accordingly. In preparation for a job interview, for example, if one hears that the job interviewer is especially impressed by assertiveness, one might expect that acting assertively would increase the probability of getting the job, and then would act on this newly formed expectancy.

Mischel refers to a second type of hypothesis we formulate in dealing with our experience as a stimulus-outcome expectancy. We learn that if event 1 occurs, event 2 is likely to follow. For example, if we hear a siren, we expect a speeding emergency vehicle will soon be seen; or, noting that it is six O'clock, we expect that dinner will soon be available.

A third type of hypothesis we use in dealing with the world is called a self-efficacy expectancy. It is one thing to know what behavior would be effective in a given situation and another to be able to perform that behavior. One's ability to perform the behaviors required in a particular situation is called self-efficacy. Because what one can actually do often differs from what one thinks he or she is capable of doing, social cognitive theory places great stress on perceived self-efficacy which is what a person thinks he or she is capable of doing in various situations. We return to a discussion of self-efficacy later in the chapter.

In general, the person variable of expectancies answers the questions, What should I expect if I act in a certain way? If I see one thing, what should I expect to see next? and Am I capable of doing what I think needs to be done?

3. Subjective values: What is worth having or doing? Even if a person has a strong behavior outcome expectancy and a strong self-efficacy expectancy, he or she may decide not to translate those expectancies into behavior because what would be gained simply is not seen as worth the effort. For example, a student may know exactly what needs to be done to write an outstanding term paper, and may believe he or she has the ability to do what is necessary but may decide that getting an A on a term paper is not worth the time and effort. Another student in the same situation may value an A more, and thus expend the time and energy and engage in the behaviors necessary to obtain it. Likewise, a given student may, at one time, under certain conditions, decide an A is worth pursuing and at another time, under different circumstances, decide that it is not. One's values will to a large extent determine whether or not the other person variables will be translated into performance. The subjective values of a person determine what is worth having or doing.

4. Self-regulatory systems and plans: How do we attain our goals? According to social cognitive theory, human behavior is largely self-regulated. Performance standards are established, and when actual performance meets or exceeds those standards, one feels good; when it does not, one feels bad. Behavior, then, is influenced more by intrinsic (internal) reinforcement and punishment than it is by extrinsic (external) reinforcement and punishment.

Humans also set future goals and then plan the events in their lives so that they are compatible with those goals. Typically, a major future goal is approached by first reaching a sequence of lesser short-term goals. For example, the goal of obtaining a college degree is met by doing well in high school, graduating from high school, applying for admission into a college, surviving the first term in college, and so on. Thus social cognitive theory views much human behavior as teleological, that is, purposive. Self-regulatory systems and plans make it possible to attain those goals deemed important. We say more about self-regulated behavior later in this chapter.

5. Competencies: What we are capable of doing. Through observational learning, which is discussed shortly, the person acquires information about the physical and social worlds and his or her relationship to them. The person develops skills, concepts, and problem-solving strategies that are actively used in dealing with the environment. Mischel stresses the fact that these competencies are not static memories that are mechanically activated by environmental stimuli. Rather, they are active processes that can be used by the person to generate a wide variety of creative constructions or responses to any given situation. Competencies, then, are seen as tools available to the individual in doing commerce with the environment. Like any tools, what can be done with them is limited only by the imagination of those using them. Competencies refer to what a person knows and what he or she is capable of doing.

It is not only one's competencies that are learned from observational learning but also one's encoding strategies, expectancies, values, and self-regulatory systems and plans. It is to the important topic of observational learning that we turn next.

Observational Learning

At the heart of social cognitive theory is the notion of observational learning. The most important fact about observational learning is that it requires no reinforcement. According to Bandura and Mischel, humans learn what they attend to, and therefore, for them, learning is a perceptual process. Thus social cognitive theory contrasts sharply with the theories of Skinner and Dollard and Miller that rely heavily on the concept of direct reinforcement.

Although within social cognitive theory, reinforcement is not believed to influence learning, it is believed to be importantly related to performance and to perceptual processes. In other words, reinforcement is thought to influence what is attended to, and thus learned and, more importantly, it is thought to determine which aspect of what has been learned is translated into behavior.

The distinction between learning and performance is illustrated in a famous study performed by Bandura (1965). In this experiment, a five-minute film was shown of a model aggressing toward a large Bobo doll. In social cognitive theory, a model is anything that conveys information-for instance, a person, television, a book, a film, a demonstration, or instructions. In this experiment, the model was an adult human.

The film began with a scene in which a model walked up to an adult-sized plastic Bobo doll and ordered him to clear the way.  After glaring for a moment at the noncompliant antagonist the model exhibited four novel aggressive responses each accompanied by a distinctive verbalization.

First, the model laid the Bobo doll on its side, sat on it, and punched it in the nose while remarking, "Pow, right on the nose, boom, boom." The model then raised the doll and pommelled it on the head with a mallet. Each response was followed by the verbalization, "Sockeroo ... stay down." Following the mallet aggression, the model kicked the doll about the room, and these responses were interspersed with the comment, "Fly away." Finally, the model threw rubber balls at the Bobo doll, each strike punctuated with "Bang." This sequence of physically and verbally aggressive behavior was repeated twice. (Bandura, 1965, pp.590-591)

Following presentation of the film, one group of children were shown a second film in which a second adult was shown praising the model for his aggressive behavior and rewarding him with pop and candy. A second group of children saw a film in which a second adult punished the model for his aggressive behavior. In this group, the model was called a coward and a bully, spanked with a rolled-up newspaper, and threatened with a beating if he was caught being aggressive again. A third group of children, the control group, saw only the first five-minute film. The children were then exposed to a doll similar to that in the film, and their aggressiveness toward it was measured. Results indicated that the group who saw the model being reinforced for aggressiveness was most aggressive; the group who saw the model punished for aggressiveness was least aggressive; and the control group was between the other two groups in aggressiveness.

This portion of Bandura's study demonstrated that children performed on the basis of what they had seen happen to someone else. That is, rather than performing on the basis of direct reinforcement and punishment, they were responding to vicarious reinforcement and vicarious punishment. To put this in terms of the person variables described earlier, it was as if the children had formulated behavior-outcome expectancies on the basis of what they had seen in the film. The children who saw the model's aggressiveness reinforced appeared to develop the hypothesis that if they acted aggressively, they, too, could expect reinforcement. Those children who saw the model punished for aggressiveness expected similar behavior on their part would yield similar results.

The first phase of Bandura's study seemed to indicate that reinforcement was still important for learning, but that it could be vicariously experienced and need not be contingent on one's own behavior. This finding alone was in contradiction to the theories of Skinner and Dollard and Miller which state that for reinforcement to be effective, it must be contingent on one's own behavior. In their view, seeing someone else's behavior being reinforced or punished should have no effect on one's own behavior.

The results of the second phase of Bandura's study were even more surprising. In this phase, all the children were offered attractive incentives (stickers and juice) to act as the model had acted in the film. Remember that in the first film the model was shown aggressing toward the doll, but the consequences of the model's aggression varied according to what group the children were in. When offered the incentive to replicate the model's behavior, the children in all three groups did so. This means that all the children had learned what they had observed-that is, the model being aggressive-but they had translated what had been learned into the type of behavior they expected would be either reinforced or at least not punished. Bandura's experiment shows that indeed what is observed is learned, and that how what is learned is translated into performance depends on the learner's behavior-outcome expectancies. The behavior- outcome expectancies in this case came vicariously from observing the consequences of a model's behavior. Bandura indicates that the research just discussed made the Bobo doll famous in psychology, "When I'm introduced at invited lectures at other universities the students place a Bobo doll by the lectern. From time to time I have been asked to autograph one. The Bobo doll has achieved stardom in psychological circles" (quoted in Evans, 1989, p. 23).

Research has shown that emotional responses can also be learned vicariously (for example, Bandura & Rosenthal, 1966; Berger, 1962; Craig & Weinstein, 1965). Berger (1962) found that when a model was shown reacting with pain, presumably in response to extreme electric shock, following the sound of a tone, the observer, too, would develop an emotional response to the tone. Physiological measures indicated that after observing a model experience what appeared to be a tone-shock relationship, the observer responded emotionally when the tone sounded even when the model was no longer present. Thus a conditioned emotional response was developed on the basis of what appeared to be happening to another person.

Bandura (1986) notes that the ability to learn by observing the consequences of other people's behavior not only enhances survival but also makes life less tedious:

If human behavior depended solely on personally experienced consequences, most people would not survive the hazards of early development. Of those who managed to outlive their mistakes, each would have to rediscover, through tiresome trial and error, what works and what fails in everyday transactions with their environment. Fortunately, people are spared many hazards and much tedium by their capacity to benefit from the experiences of others. (p. 283)

Elsewhere, Bandura says, "One would not teach military recruits how to lob grenades through trial and error. This mode of learning would create a situation of, 'Look, sergeant, no arms.' It would give new meaning to the wartime movie, 'A Farewell to Arms.' (quoted in Evans, 1989, p. 8).

We do, of course, learn by observing the consequences of our own behavior but, according to Bandura, whatever can be learned by direct experience can also be learned by vicarious experience. Bandura (1986) says, "Virtually all learning phenomena resulting from direct experience, can occur vicariously by observing other people's behavior and its consequences for them" (p. 19).

News and Entertainment Media as Models. Because, according to social cognitive theory, we learn from what we observe, it follows that newspapers, television, and films can act as influential models. Bandura (1986) makes the following observations about fictional television:

Analyses of televised programs reveal that violent conduct is portrayed, for the most part, as permissible, successful, and relatively clean.... Witnesses to the violence in the dramatic presentations are more likely to approve of such behavior or to join in the assaults rather than to seek alternative solutions. Violence not only is shown to pay off but is readily used by superheroes, who dispose of their adversaries in a quick, perfunctory way as though slaying human beings was of no great concern. (p. 292)

Since Bandura's early work on modeled aggression, numerous studies have examined the relationship between television violence and aggression and other forms of antisocial behavior. Typically these studies confirm a positive relationship between the two (e.g., Geen & Thomas, 1986; Huesmann & Malamuth, 1986; Turner, Hesse, & Peterson-Lewis, 1986). Recent research has indicated that despite considerable current societal concern, 58% of television programs still contain violence and that 78% of that violence is depicted as at least unremorseful, if not actually glamorized. Furthermore, evidence indicates that the use of warning labels has not prevented children from watching violent programs (Seppa, 1997). It has also been found that violent rock videos can create, in some men, violent attitudes toward women (Peterson & Pfost, 1989).

Clearly, not everyone exposed to violence becomes violent nor does everyone exposed to sexually explicit literature or movies become sexually deviant. For some, however, such modeling provides learning experiences that promote antisocial behavior. Nor are the news and entertainment media the only places in society where violence is modeled, for example, parents who employ corporal punishment in their child rearing practices are modeling aggression. Bandura (1973) observes that children whose parents employ considerable corporal punishment often become highly aggressive. Such children also tend to be less obedient (Power & Chapleski, 1986). Bandura supports neither the extensive use of corporal punishment nor "unconditional love." He believes that unconditional love is selfdefeating because it eliminates the informative relationship between performance and reward. In the absence of this relationship children are directionless.

Some child-rearing authorities have popularized the view that healthy personality development is built on "unconditional love." If this principle were, in fact, unfailingly applied, parents would respond affectionately regardless of how their children behaved-whether or not they mistreated others, stole whatever they wanted, disregarded the wishes and rights of others, or demanded instant gratification. Unconditional love, were it possible, would make children directionless and quite unlovable. Most readers are undoubtedly acquainted with families where parents who attempted to approximate this condition succeeded in producing "self-actualized" tyrants. (Bandura, 1977, p. 102)

In any case, news and entertainment media are among society's influential models and "humane living requires reducing social influences that promote cruelty and destructiveness [and] ... society has the right to regulate obscene materials that can cause harm" (Bandura, 1986, p. 296). However, Bandura recognizes that controlling such material is complex. First, there is disagreement over what is harmful. Second, the suppression of material considered harmful may conflict with the basic right of free expression.

Although it is accurate to say that, according to social cognitive theory, what is observed is learned, certain processes influence what is attended to, what is retained, how what is learned is translated into behavior, and why it is translated into behavior. Bandura (1986) describes four such processes.

Attentional Processes

These processes include aspects of the environment that influence attention, such as the complexity, distinctiveness, and prevalence of the stimulation. For example, screeching brakes attract almost everyone's attention. Certain characteristics of models determine the extent to which they are observed. For instance, it has been found that models are attended to more if they are seen as similar to oneself, are respected, are considered powerful, or are attractive. Related to the concept of power, it has been found that adults who control resources, that is, have the ability to dispense rewards, are more powerful models for children than adults who do not control resources (Bandura, Ross, & Ross, 1963; Grusec & Mischel, 1966).

Attentional processes also include observer characteristics such as sensory capacities. Blind and deaf people, for example, do not respond to the same stimuli as people with normal sight and hearing.

Also, the consequences of past behavior can create a perceptual set in the observer. For example, if attending to certain stimuli in the past resulted in positive consequences, a tendency will exist to attend to similar stimuli in similar situations.

Retentional Processes

What is learned by observation is of no value unless it is retained. We saw earlier that how information is encoded and construed varies from person to person but, in general, Bandura (1986) says that experiences are stored either imaginally or verbally. That is, we either retain an actual cognitive picture of what was experienced or we retain the words that describe the experience. These memories make delayed modeling possible. Delayed modeling refers to the fact that often information gained by observational learning is first translated into behavior long after the time that it had been learned.

Motor Reproduction Processes

To translate learning into performance, one needs to have the necessary motor apparatus. Also, even if one has the necessary motor apparatus, one can be temporarily prevented from performing because of injury, fatigue, or illness. Even with the necessary motor system available and functioning well, complex skills cannot simply be observed and immediately translated into performance. First, with complex skills, many observations may be required before all of the relevant information can be attended to and remembered. Second, if all the relevant information is learned, many rehearsals that attempt to match performance with what had been learned and retained may be necessary. According to Bandura (1986), it is the observer's conceptualization of modeled behavior that provides the frame of reference necessary for skill improvement:

The conceptual representation provides the internal model for response production and the standard for response correction. Behavioral production primarily involves a conception-matching process in which the incoming sensory feedback from enactments is compared to the conception. The behavior is then modified on the basis of the comparative information to achieve progressively closer correspondence between conception and action. (p. 64)

Motivational Processes

No matter how much one has learned, and no matter what one's capabilities are, learning will not be translated into performance unless there is an incentive to do so.  Observational learning may create potentially effective behavior-outcome expectations, but unless the person believes that behavior will yield something valued, no behavior will occur. According to social cognitive theory, reinforcement, either direct or vicarious, provides the information necessary for the development of effective behavior-outcome expectancies. Even with direct reinforcement, however, it is assumed that learning is observational and not an automatic, unconscious strengthening of response tendencies, as Skinner and Dollard and Miller had assumed. According to social cognitive theory, a person learns from observing the consequences of either his or her own behavior (direct reinforcement) or from observing the consequences of other people's behavior (vicarious reinforcement). In either case, information is provided about what behaviors lead to what consequences.

In addition to providing information concerning what behaviors lead to what consequences, reinforcement provides an incentive for action. Within social cognitive theory, reinforcement is equated with something of value to the person under existing circumstances. People, then, are motivated to act in ways that provide things they value and that allow them to avoid things considered aversive. Bandura's views on observational learning and the translation of that learning into behavior can be summarized as follows: A person must observe something; remember what was observed; be able to perform the behaviors necessary to reproduce what was observed; and want to reproduce those behaviors.

Self-Regulated Behavior

Social cognitive theory maintains that most human behavior is self-regulated. Through cumulative direct and vicarious experience, people develop performance standards that they use to evaluate their own behavior. Almost constantly, the person compares what he or she does in a situation with some performance standard. If performance meets or exceeds the standard, the person experiences intrinsic reinforcement. If a performance falls short of a standard, the person experiences intrinsic punishment.

Modeling has been found to influence the formulation of one's performance standards. Bandura and Kupers (1964), for example, found that children exposed to models who set high performance standards reinforced themselves only for superior performance, whereas children exposed to models accepting minimal performance standards reinforced themselves for minimal performance. It would be expected, then, that relevant people in a child's life-for instance, parents, siblings, and peers-would have a profound influence on the development of a child's performance standards.

Goals and plans extend self-directed behavior over long periods of time. Once a future goal has been established, one organizes his or her experiences so they increase the probability of goal attainment. Bandura (1991) says, "People form beliefs about what they can do, they anticipate the likely consequences of prospective actions, they set goals for themselves, and they otherwise plan courses of action that are likely to produce desired outcomes" (p. 248). Experiences that keep the person on track toward the goal result in self-reinforcement; those that are incompatible with the goal cause self-punishment. As mentioned earlier, major goals are seldom accomplished all at once. Rather, a series of subgoals are established, and as they are attained, one approximates the major goal. Personal frustration and depression can occur if goals are either too distant or too difficult. Goals must be related to one's capabilities and must be attainable through the achievement of reasonable subgoals that are of moderate difficulty relative to a person's competencies. Performance standards, too, must be realistic. If they are too lenient, they will be too easily met, and little, if any, self-reinforcement will result from performing in accordance with them. If they are too stringent, one will experience frustration or worse. Bandura (1986) says, "In its more extreme forms, harsh standards for self-evaluation give rise to depressive reactions, chronic discouragement, feelings of worthlessness, and lack of purposefulness" (p. 358). Bandura (1991) finds that people prone to depression generally judge themselves more harshly than people not prone to depression, "Compared to nondepressed persons, those who are prone to depression react less self-rewardingly for similar successes but more self-critically for similar failures" (pp. 274-275).

The situation is best if one's performance standards are modified as a function of one's accomplishments and failures. Bandura (1986) notes:

One's previous behavior is continuously used as a reference against which ongoing performance is judged. In this referential process, self-comparison supplies the measure of adequacy. Past attainments affect self-appraisal mainly through their effects on standard setting.... After a given level of performance has been attained, it is no longer challenging, and people seek new self-satisfactions by means of progressive improvement. Hence, people tend to raise their performance standards after success and to lower them to more realistic levels after repeated failure. (pp. 347-348)

Self-Efficacy

Bandura (1989) observes, "Among the mechanisms of personal agency, none is more central or pervasive than people's beliefs about their capabilities to exercise control over events that affect their lives" (p. 1175). Self-efficacy refers to what a person is actually capable of doing. More important, however, is what a person thinks he or she is capable of doing, or perceived self-efficacy. Perceived self-efficacy is influenced by several factors-for example, personal accomplishments and failures, seeing models perceived as similar to oneself succeed or fail at various tasks, and verbal persuasion. Through verbal persuasion, persons can often be encouraged to try to achieve goals they would otherwise avoid; an example is when a coach "fires up" his team before a game. The effects of verbal persuasion are short-lived, however, if the resulting perceived self-efficacy is not manifested in real performance. It has also been found that people tend to judge their self-efficacy in a potentially threatening situation in terms of their emotional arousal. Because high arousal typically inhibits performance, strong emotion is usually correlated with feelings of perceived self-inefficacy. Conversely, being relaxed in such a situation facilitates perceived self-efficacy.

Research performed by Bandura and others (for a summary of this research see Bandura, 1986, 1989, 1995) has demonstrated that compared to individuals with low perceived self-efficacy, individuals with high perceived self-efficacy have the following characteristics:

they set more challenging goals and performance standards

they persist longer in the pursuit of goals

they are more venturesome in their behavior

 they recover more quickly from setbacks and frustrations

they experience less fear, anxiety, stress, and depression

As with goals and performance standards, the situation is best when one's perceived self-efficacy is in line with one's true capabilities. Thinking one can do more than one can actually do results in frustration. Thinking that one is not capable of doing something that one is actually capable of doing inhibits personal growth. Both distortions of one's self-efficacy result in dysfunctional (erroneous) self-expectancies which, if severe enough, can cause a person to seek psychotherapy.

Bandura's concept of perceived self-efficacy has become one of the most popular research topics in the realm of personality theory. What follows represents only a sample of that research. Perceived self-efficacy has been found to be related to: the effectiveness of weight loss programs (Mitchell & Stuart, 1984; Weinberg, Hughes, Critelli, England, & Jackson, 1984); the effectiveness of programs designed to help people quit smoking (Baer, Holt, & Lichtenstein, 1986; Garcia, Schmitz, & Doerfler, 1990; Wojcik, 1988); the effectiveness of alcohol treatment programs (Annis, 1990); AIDS prevention (Bandura, 1990a; O'Leary, 1992); academic performance (Multon, Brown, & Lent, 1991; Tuckman, 1990); depression (Bandura, 1989; Bandura, Adams, Hardy, & Howells, 1980; Davis-Berman, 1990); job burnout (Meier, 1983); maternal competence (Teti & Gelfand, 1991); occupational choice (Bonett, 1994; Bores-Rangel, Church, Szendre, & Reeves, 1990; Betz & Hackett, 1981); athletic performance (Gould, Hodge, Peterson, & Giannini, 1989); managerial decision making (Bandura & Jourden, 1991; Bandura & Wood, 1989; Wood & Bandura, 1989; Wood, Bandura, & Bailey, 1990); gender differences (Lent, Brown, & Larkin, 1986; Nevill & Schlecker, 1988; Poole & Evans, 1989); developmental experiences (Skinner, Chapman, & Baltes, 1988; Holden, Moncher, Schinke, & Barker, 1990); physiological reactions (Bandura, Reese, & Adams, 1982; Bandura, Taylor, Williams, Mefford, & Barchas, 1985); immune system functioning (Bandura, Cioffi, Taylor, Brouillard, 1988; Bandura, O'Leary, Taylor, Gauthier, & Gossard, 1987; Wiedenfeld, O'Leary, Bandura, Brown, Levine, & Raska, 1990); coping ability (Ozer & Bandura, 1990); job satisfaction and performance (Saks, 1995).

Later in this chapter we will also see that the concept of perceived selfefficacy is central to Bandura's approach to, and understanding of, psychotherapy.

Moral Conduct

Standards of right and wrong are also highly personal and are also derived from one's direct and vicarious experience. As with performance standards, moral principles are usually modeled by a child's parents and are eventually internalized. Once internalized, these moral principles determine which behaviors and thoughts are self-sanctioned and which result in self-contempt. Thus moral behavior comes to be self-regulated and is maintained independently of, and in many cases despite, environmental consequences. Bandura (1977) says, "The anticipation of self-reproach for conduct that violates one's standards provides a source of motivation to keep behavior in line with standards in the face of opposing inducements. There is no more devastating punishment than self-contempt" (p. 154).

Certain cognitive mechanisms, however, allow a person to act contrary to his or her moral principles without experiencing self-contempt. These self-exonerating mechanisms (Bandura, 1986, pp. 375-385) are summarized here:

1. Moral justification. One's otherwise reprehensible behavior becomes a means to a higher purpose and therefore is justifiable. "I committed the crime so I could provide food for my family." Bandura (1986) says, "Over the years, much destructive and reprehensible conduct has been perpetrated by decent people in the name of religious principles, righteous ideologies, and nationalistic imperatives" (p. 377).

2. Euphemistic labeling. By calling an otherwise reprehensible act something other than what it really is, one can engage in an act without self-contempt. For example, nonaggressive individuals are far more likely to aggress toward another person when doing so is called a game.

Through the power of hygienic words, even killing a human being loses much of its repugnancy. Soldiers "waste" people rather than kill them, CIA operatives "terminate (them) with extreme prejudice." When mercenaries speak of "fulfilling a contract," murder is transformed by admirable words into the honorable discharge of duty. (Bandura, 1986, p. 378)

3. Advantageous comparison. By comparing one's self-deplored acts with even more heinous acts, it makes one's own reprehensible acts look trifling by comparison. "Sure I did that, but look at what he did."

Promoters of the Vietnamese war and their supporters, for example, minimized the slaying of countless people as a way of checking massive communist enslavement. Given the trifling comparison, perpetrators of warfare remained unperturbed by the fact that the intended beneficiaries were being killed at an alarm  ing rate. Domestic protesters, on the other hand, characterized their own violence against educational and political institutions as trifling, or even laudable, by comparing it with the carnage perpetrated by their country's military forces in foreign lands. (Bandura, 1986, p. 379)

4. Displacement of responsibility. Some people can readily depart from their moral principles if they believe a recognized authority sanctions their behavior and takes responsibility for it. "I did it because I was ordered to do so."

Nazi prison commandants and their staffs felt little personal responsibility for their unprecedented inhumanities. They were simply carrying out orders. Impersonal obedience to horrific orders was similarly evident in military atrocities, such as the My Lai massacre. (Bandura, 1986, p. 379)

Bandura (1990b) observes that terrorists who have taken hostages often warn officials of targeted nations that if they take retaliatory action they will be held accountable for the lives of the hostages" (p. 175).

5. Diffusion of responsibility. A decision to act in an otherwise reprehensible manner that is made by a group is easier to live with than an individual decision. Where everyone is responsible, no single individual feels responsible. "I couldn't be the only one saying no."

6. Disregard or distortion of consequences. Here people ignore or distort the harm caused by their conduct and therefore there is no need to experience self-contempt. The farther people remove themselves from the ill effects of their immoral behavior the less pressure there is to censure it. "I just let the bombs go and they disappeared in the clouds."

7. Dehumanization. If some individuals are looked upon as subhuman, they can be treated inhumanly without experiencing self-contempt. Once a person or a group has been dehumanized, they no longer possess feelings, hopes, and concerns, and they can be mistreated without risking self-condemnation. "Why not take their land, they are nothing but savages without souls."

Once dehumanized, [individuals] are no longer viewed as persons with feelings, hopes, and concerns but as subhuman objects demeaningly stereotyped as 11 gooks," "fags," or "niggers." Subhumans are presumably insensitive to maltreatment and influenceable only through more primitive methods. If dispossessing disfavored people of their humanness does not blunt self-reproof, it can be fully eliminated by attributing bestial qualities to them. They become "degenerates," 11 pigs," and other bestial creatures. Over the years slaves, women, manual laborers, and religious and racial minorities have been treated as chattel or as subhuman objects. (Bandura, 1986, p. 382)

8. Attribution of blame. One can always choose something that a victim said or did and claim that it caused one to act in a reprehensible way.

Rapists and males who acknowledge a proclivity to rape subscribe to myths about rape embodying the various mechanisms by which moral self-censure can be disengaged.... These beliefs hold rape victims responsible for their own victimization because they have supposedly invited rape by sexually provocative appearance and behavior and by resisting a sexual assault weakly. (Bandura, 1986,pp.384-385)

The self-exonerating mechanisms provide another reason why human behavior tends to be inconsistent. Even if a person's moral principles were known with certainty, employing one or more of these mechanisms would make the person's moral behavior unpredictable to the outside observer. Such behavior would undoubtedly seem consistent and logical to the person himself or herself, however.

Delay of Gratification

In a series of experiments that began soon after he received his PhD from Ohio State University in 1956, Mischel and his colleagues have extensively explored the variables related to the ability to delay gratification. In one early study, Mischel and Metzner (1962) found that the ability to delay gratification increased with age, intelligence, and with shorter intervals of delay. Later, Mischel and his colleagues began their influential research on 4- and 5-year-old children attending the Bing nursery school at Stanford University (Mischel & Ebbesen, 1970; Mischel, Ebbesen, & Zeiss, 1972). In these studies, children were given a choice between a small reward given immediately (e.g., one marshmallow or pretzel) or a larger reward given after a delay (e.g., two marshmallows or pretzels). There were four experimental conditions: (1) both the immediate and delayed rewards were visible; (2) neither the immediate nor the delayed rewards were visible; (3) only the immediate reward was visible; and (4) only the delayed reward was visible. The researchers predicted that being able to see the rewards would enhance the ability to delay. The opposite was true, children waited the longest when no rewards were visible. It turned out that when rewards were visible it caused the children to think about them, and thinking about them reduced their ability to delay gratification. It was also found that children who could delay the longest employed a number of self-distraction strategies in order to make the aversive delay period more pleasant. These children talked to themselves, sang, invented games to be played with their hands and feet, and even tried to fall asleep (one child was successful). Mischel and Moore (1973) found that instructing children to think of pictures of rewards instead of actual rewards increased their ability to delay gratification. Other studies showed that instructing children to think of otherwise desirable rewards in undesirable ways (e.g., to think of pretzels as little brown logs or of marshmallows as round white clouds or as balls of cotton) increased their ability to delay (Mischel & Baker, 1975; Mischel & Moore, 1980; Moore, Mischel, & Zeiss, 1976). From the experiments just described, Mischel (1993) concludes, "What is in the children's heads-not what is physically in front of them-determines their ability to delay [and] if you can make the waiting easier for yourself, you are more likely to wait successfully" (pp. 457, 458). Numerous studies have confirmed the contention that imposed or self-created distractions facilitate the ability to delay gratification (Mischel, Shoda, & Rodriguez, 1989).

In the more than 40 years of research on delay of gratification, one of the most dramatic findings is the relationship between the ability to delay gratification as a preschool child and a number of positive adolescent personality characteristics. Mischel and his colleagues followed up their research on the Stanford preschool children 10 years after the original research (Mischel, Shoda, & Peake, 1988; Mischel, Shoda, & Rodriguez, 1989; Shoda, Mischel, & Peake, 1990). Of special interest were the children who were able to delay even when rewards were visible and who were not given strategies to facilitate delay. It was assumed that such children were capable of generating their own self-distracting strategies, and that this capability would continue to exist later in life. These assumptions were clearly supported. Preschool children who could delay gratification with rewards visible and without imposed distractions were rated positively by their parents on a variety of social and academic skills when they were high school students. Table 11-1 summarizes the attributes these parents saw in these adolescents.

In addition to positive parental ratings, it was found that both verbal and quantitative scholastic aptitude test (SAT) scores were significantly related to the preschool ability to delay gratification in the condition in which rewards were visible and no strategies for distraction were supplied to the child. In contrast, preschool delay behavior when rewards were not visible or when distraction strategies were provided did not reliably predict either positive parental ratings or SAT scores. Mischel (1993, p. 460) notes evidence that people with the ability to self-distract can also deal more effectively with pain, stress, and severe life crises.

What determines the extent to which one possesses such an important personality attribute? As with other factors influencing self-regulated behavior, it has been found that the willingness to delay gratification can be learned by observing models (Bandura & Mischel, 1965; Mischel & Liebert, 1966; Stumphauzer, 1972). The fact that the ability to delay gratification can be taught to young children, and that the ability is clearly advantageous later in life, suggests that such teaching be incorporated into early education and in child rearing practices. In any case, it is vital that we understand the kind of self-control necessary to delay gratification because, according to Mischel (1993), without it humans are as impulsive as nonhuman animals, and the goal-oriented behavior necessary for civilized living is impossible.

Table 11-1  As adolescents, children who delayed gratification longer in preschool were described by parents as:

Less likely to be sidetracked by minor setbacks.

More likely to exhibit self-control in frustrating situations.

Copes well with important problems.

Capable of doing well academically when motivated.

Less likely to yield to temptations.

When faced with a choice, is less likely to settle for an immediate but less desirable choice.

More able to pursue goals when motivated.

 Intelligent.

When motivated, is capable of exhibiting self-control in tempting situations.

 Skilled in maintaining friendships and getting along with peers.

Not distractible when trying to concentrate.

Capable of exhibiting self-control when frustrated.

Can effectively pursue goals.

Can divert attention from the frustration of having to postpone a desired gratification while continuing to pursue it.

__________________________

From Mischel, 1993, p. 459

Dysfunctional Expectancies and Psychotherapy

According to social cognitive theory, psychological problems result from dysfunctional expectancies (that is, erroneous, nonfunctional, or faulty expectancies), and any type of therapy that corrects them-that is, brings them in line with reality will be, by definition, effective.

If, for example, one believes that developing a close relationship with a member of the opposite sex will bring pain and frustration, one will avoid such relationships. Such expectancies are usually based on real experiences, but they are overgeneralized and, when they are, prevent the person from having the types of experiences that would disconfirm them. The defensive behavior based on a dysfunctional expectancy is therefore often difficult to remedy.

Some dogs do bite, some airplanes do crash, some intimate relationships do result in pain and frustration, some members of minorities do commit crimes, and some men and some women are insensitive. To generalize on the basis of a few cases to all possible experiences, however, does not accurately represent reality. People forming strong expectancies on the basis of limited experiences need to have further experiences with the same types of objects, events, or people without those experiences being negative. Bandura (1986) observes, "What such individuals need in order to relinquish their erroneous beliefs are powerful disconfirming experiences, which verbal assurances alone do not provide" (p. 190). Providing powerful disconfirming experiences is a key ingredient of the kind of therapy promoted by social cognitive theory.

Another type of dysfunctional expectancy concerns perceived self-efficacy. As we have seen, if people believe they are incapable of doing something, they will not try to do it. If one believes he or she cannot touch dogs, cats, snakes, children, or members of the opposite sex, that person will not do so regardless of his or her true capabilities. Likewise, a person who believes he or she is incapable of handling success will avoid success. Thus, according to social cognitive theory, a major goal of psychotherapy is to change the client's perceived self-efficacy. The assumption is that if a person's perceived self-efficacy becomes more realistic, behavior will become more adaptive. Bandura summarizes how he believes perceived self-efficacy is most effectively changed during psychotherapy as follows: "Treatments that are most effective are built on an empowerment model. If you really want to help people, you provide them with. . . competencies, build [in them] a strong self-belief, and create opportunities for them to exercise these competencies" (quoted in Evans, 1989, p. 16).

The assumption that perceived self-efficacy relates directly to behavior was tested by Bandura, Adams, and Beyer (1977). Seven men and 26 women who suffered a chronic snake phobia were recruited through newspaper advertisements. Before treatment, each person was given a Behavioral Avoidance Test that consisted of 29 performance tasks requiring increasingly close interactions with a redtailed boa constrictor. Self-efficacy expectancies were also measured before treatment. Each person was given a list describing various interactions with a snake and was asked to indicate which ones he or she would be able to perform. Subjects were also asked to indicate the certainty with which they believed they could or could not engage in various interactions with the snake.

After pretesting, Bandura, Adams, and Beyer (1977) randomly assigned subjects to one of three treatment groups. The first was the participant modeling condition in which first a live model handled a boa constrictor in a series of interactions ranging from mildly threatening to extremely threatening. Then subjects were asked to perform these same interactions. The time required for these interactions ranged from 40 minutes to 7 hours, with the average time being 1 1/2 hours. In some cases, the fear of a subject was so intense that a baby boa had to be used initially. The second group was called the modeling condition where subjects observed a model engage in a series of interactions with the snake, but the observer did not actually come into contact with the snake himself or herself. In the control condition, the pretreatment measures were made but no treatment was given.

After the various treatments, subjects were again given the Behavior Avoidance Test, the final step of which was to let the boa constrictor crawl on the subject's lap while his or her hands were held passively to the side. Results indicated that both the participant modeling and the modeling conditions significantly reduced avoidance behavior, but the participant modeling condition was the more effective of the two treatments.

Efficacy expectations were measured again after conclusion of treatment. It was found that the participant modeling condition and the modeling condition significantly changed the self-efficacy expectations of being able to handle a snake. Again, although both techniques were effective, the participant modeling technique was most effective in changing self-efficacy expectations. Most important, however, was the finding that self-efficacy expectations were accurate predictors of behavior. Those subjects who indicated they now believed they could handle a snake actually did so. The relationship between perceived self-efficacy and actual performance existed for subjects in all groups. In almost every case, the type of interactions a person believed he or she could engage in were those in which he or she actually engaged. Maximal efficacy expectations, however, did not always result from maximal interactions with the snake; that is, some of the individuals in the participant modeling group who were able to perform the most threatening interactions with the snake during treatment did not form the expectation that they would be able to interact that way with a snake on subsequent occasions, and they were correct. The formation of efficacy expectations, then, is based on more than one's experience with successful performance, although successful performance has a powerful influence on the formation of efficacy expectations.

Further measures indicated that the reduction of avoidance caused by the treatments generalized to other snakes, that the results of the experiment persisted over time, and that the results generalized positively into other aspects of the subjects' lives. For instance, in some cases, improvement was reported in dealing effectively with other animal and social threats. Bandura, Reese, and Adams (1982) performed a similar experiment on people with spider phobias with essentially the same results as those just described..

In another study, Bandura, Blanchard, and Ritter (1969) tested the effectiveness of various techniques in dealing with a snake phobia. In this experiment, adults and adolescents were divided into four groups: symbolic modeling in which subjects were shown a film showing models interacting with a snake; live modeling with participation in which subjects interacted with a snake along with the model; systematic desensitization, in which subjects were asked to imagine interactions with a snake ranging from low-anxiety-producing interactions to those producing high anxiety. Subjects, starting with imagining low-anxiety scenes, continued to imagine each scene until it no longer caused anxiety, and continued ill this manner until the scenes that previously caused the greatest amount of anxiety no longer did so. The last group was the control condition in which subjects received no treatment. The ability to interact with a snake was measured for all subjects before and after the experiments. Results are shown in Figure 11-1.

It can be seen that again live modeling with participation was most effective, followed by symbolic modeling and then desensitization. Subjects in the control condition showed little or no improvement in their ability to interact with a snake, Follow-up research again indicated that the effects of the experiment endured and generalized positively into other areas that had produced fear prior to the experiment.

Many other experiments have shown the effectiveness of modeling in deal. ing with a wide variety of dysfunctional expectancies. In all of this research, the emphasis is on the person's current perceptions and expectancies. There is no mention of traits, reinforcement histories, or internal conflicts. In fact, Bandura (1986) believes those therapists looking for such things tell us more about themselves than they tell us about the source of their client's problems:

Advocates of differing theoretical orientations repeatedly discover their chosen motivators at work but rarely find evidence for the motivators emphasized by the proponents of competing viewpoints.... Hence, one can predict better the types of insights and unconscious motivators that persons are apt to discover in themselves in the course of psychodynamic analysis from knowledge of the therapists' conceptual belief system than from the clients' actual psychological status. (p. 5)

Social Cognitive Theory View of Human Nature

Freedom vs. Determinism

Social cognitive theory portrays human nature as complex and rational but is that the same as saying humans possess a free will? Bandura rejects the notion that humans are autonomous, that is, free to act independently of the environmental and personal influences impinging on them. He also rejects the notion that humans respond mechanistically to those influences. Rather, as we have seen, Bandura accepts reciprocal determinism in which people can influence both their behavior and their environment. People can even produce unique thoughts and actions "through their capacity to manipulate symbols and to engage in reflective thought, people can generate novel ideas and innovative actions that transcend their past experiences" (Bandura, 1989, p. 1182). However, says Bandura, "self-generated influences operate deterministically on behavior the same way as external sources of influence do" (1989, p. 1182). Thus, according to Bandura, humans are rational but they do not possess an autonomous free will.

William James (1884/1956) distinguished between hard and soft determinism. Hard determinism is the belief that the causes of human behavior function in an automatic, mechanistic manner, thus rendering the idea of personal responsibility meaningless. Hergenhahn (1997) describes soft determinism.

With soft determinism.... cognitive processes such as intentions, motives, beliefs, and values intervene between experience and behavior. The soft determinist sees human behavior as resulting from thoughtful deliberation of the options available in a given situation. Because rational processes manifest themselves prior to actions, the person bears responsibility for those actions. Although soft determinism is still determinism, it is a version that allows uniquely human cognitive processes into the configuration of the causes of human behavior. Soft determinism, then, offers a compromise between hard determinism and free will-a compromise that allows for human responsibility. (p. 13)

Bandura is a soft determinist. This conception allows human behavior to be viewed as teleological, or goal-oriented: "By representing foreseeable outcomes symbolically, people can convert future consequences into current motivators and regulators of foresightful behavior" (Bandura, 1986, p. 19). It also allows people to be viewed as at least partially responsible for their own behavior, "[People can] serve as a causal contributor to their own life course by selecting, influencing, and constructing their own circumstances" (Bandura, 1986, p. 38). Elsewhere, Bandura (1989) says, "the capacity to exercise control over one's own thought processes motivation, and action is a distinctively human characteristic. Because judgment and actions are partly self-determined, people can effect change in themselves ail their situations through their own efforts" (p. 1175). And finally, demonstrating that soft determinism is still determinism, Bandura (1989) says, "self-generated influences operate deterministically on behavior the same way as external sources of influence do" (p. 1182).

Freedom as Options. Within the context of social cognitive theory, freedom is defined as "the number of options available to people and their right to exercise them. The more behavioral alternatives and prerogatives people have, the greater is their freedom of action" (Bandura, 1986, p. 42). Bandura (1989) elaborates:

Given the same environmental conditions, persons who have developed skills for accomplishing many options and are adept at regulating their own motivation and behavior are more successful in their pursuits than those who have limited means of personal agency. It is because self-influence operates deterministically on action that some measure of self- directedness and freedom is possible, (p. 1182)

Anything that reduces a person's options, then, limits his or her freedom. According to Bandura (1986, pp. 42-43) the factors that can limit personal free. dom include the following:

deficiencies in knowledge and skills

perceptions of self-inefficacy

internal standards that are too stringent

social sanctions that limit a person's opportunities because of his or her skin color, sexual orientation, gender, religion, ethnic background, or social class

Chance Encounters and Life Paths. Bandura (1982) defines a chance encounter as an "unintended meeting of persons unfamiliar to each other" (p. 748). Such en- counters can also be called fortuitous because of their unplanned, accidental nature. Bandura (1982) observes that chance encounters can significantly impact one's life and provide another reason for the unpredictability of human behavior.

Some chance encounters touch people only lightly, others leave more lasting effects, and still others branch people into new trajectories of life, Psychology cannot foretell the occurrence of fortuitous encounters, however sophisticated its knowledge of human behavior becomes. This factor introduces a measure of Un-, predictability about the direction of the flow of human lives. Patterns of change can therefore vary widely in different individuals. They may display basic continuities throughout their life spans, continuities in some aspects of life but discontinuities in others, or a vase discontinuity at any point along the way affecting virtually all domains of functioning. (p. 749)

As an example of the latter, Bandura describes how Nancy Davis met her future husband, Ronald Reagan.

While pursuing her acting career, she began to receive in the mail announcements of communist meetings intended for another person bearing the same name, who appeared on a Hollywood list of communist sympathizers. Fearing that her career might be jeopardized by mistaken identity, she voiced concern to her film director, who arranged a meeting with Ronald Reagan, then president of the Screen Actors Guild. Before long they were wed. In this instance a coincidental likeness of names and a postal mix-up altered the course of lives. (1982, p . 749)

Reciprocal determinism is not violated by noting the importance of fortuitous experience. Such experience indicates an additional way in which the environment can influence people and their subsequent behavior.

Mind-Body Relationship

Although social cognitive theory gives cognitive events a prominent role as causative agents, it does not accept psychophysiological. dualism. Bandura (1989) states, "Thoughts are higher brain processes rather than psychic entities that exist separately from brain activities. Ideational and neural terminology are simply different ways of representing the same cerebral processes" (p. 1181). However, at least at the present time, psychological laws cannot be derived from neurophysiological laws:

One must distinguish between biological laws governing the mechanics of cerebral systems and psychological laws of how cerebral systems can be orchestrated to serve different purposes. Psychological knowledge of how best to structure influences to create belief systems and personal competencies is not derivable from knowledge of the neurophysical mechanisms that subserve such changes. Thus, understanding the brain circuits involved in learning does not tell one much about how best to present and organize instructional contents, how to code them for memory representation, and how to motivate learners to attend to, cognitively process, and rehearse what they are learning. Nor does understanding of how the brain works furnish rules on how to create social conditions that cultivate the skills needed to become a successful parent, teacher, or executive. The optimal conditions must be specified by psychological principles. (Bandura, 1989, p. 1182)

Thus while it is true psychological laws cannot violate the laws of the neurophysiological systems that subserve them, attempting to reduce psychology to biology serves no useful purpose: 

Were one to embark on the road to reductionism, psychology would be reduced to biology, biology to chemistry, and chemistry to physics, with the final stop in atomic particles. Neither atomic particles, chemistry, nor biology will provide the psychological laws of human behavior. (Bandura, 1989, p. 1182)

Evaluation

Empirical Research

Like the theories of Cattell, Eysenck, Skinner, and Dollard and Miller, the theory of Bandura and Mischel is well grounded in empirical research. Throughout this chapter we have sampled numerous experimental studies that Bandura and Mischel and their colleagues have run to test their concepts. Social cognitive theory continues to generate an enormous amount of research. Also, Mischel's book, Personality and Assessment (1968), triggered a debate among personality theorists concerning the consistency of human behavior that has lasted to the present time. This debate has stimulated research programs designed to answer the question, Is human behavior consistent? If so, to what extent, and in what areas? (For an extensive review of this research, see Zuckerman, 1991). It now appears that whether behavior consistency is found or not depends on what is meant by consistency and how the research is performed. For example, if average behavior is measured, it pools those behaviors that are consistent with those that are not. The result tends to support the notion that behavior is inconsistent. Bowers (1973) and Bem and Allen (1974) argue that research on consistency should be correlational rather than experimental because the correlational method preserves individual differences. Bern and Allen tested the hypothesis that "individuals who identify themselves as consistent on a particular trait dimension will in fact be more consistent cross-situationally than those who identify themselves as highly variable' (p. 512). The hypothesis was supported, indicating that some people are consistent on some traits and others are not; further, people themselves know whether or not they are consistent.

After reviewing the literature and the arguments on consistency versus inconsistency, Pertain (1984) reached the following conclusion:

Perhaps the conclusion that makes most sense is that most people are consistent in their behavior some of the time and variable in their behavior the rest of the time.... In other words, each person can be expected to be consistent in ways that are salient or meaningful for him or her. The areas of consistency differ with different individuals. Thus, for example, for some individuals it may be important to be always honest while for others it may be important to be always dominant. Each individual may have many or few such areas in which there exists cross-situational consistency. Sampling of a large group of individuals on any one characteristic might suggest little cross-situational consistency since it is Unlikely that more than a small proportion of the total sample will be consistent in that particular area. The consistency of a subgroup in the population can be masked by the situational specificity of the larger group: this is true for every personality characteristic that is assessed. The conclusion might then be that behavior is situationally specific rather than that behavior is stable in some ways for some individuals and in other ways for other individuals. What appears to be needed is a more differentiated picture of consistency and of the relationship between person and situational determinants. (p. 17)

The information just summarized refutes the position of neither the social cognitive theorist nor the trait theorist. Although the position that Mischel took in 1968 tended toward situationalism, by claiming that the probability of reward or punishment in any situation is the most powerful determinant of behavior, he never said that behavior was entirely inconsistent; if it were, he said, memory would have no value (1977, p. 333). In any case, regardless of Mischel's early position, he is, as we have seen, now an interactionist who stresses the importance of both person and situation variables. just as showing consistency of behavior is no longer a valid argument against social cognitive theory, finding inconsistency in behavior is not necessarily an argument against trait theory. We saw in chapter 7 that Allport believed the possession of personality traits did not cause a person's behavior to be fixed and stable. Rather, Allport viewed traits as representing ranges of possible behavior. He believed differing situations determined which behavior, in the possible range of behaviors, occurred. Thus Allport's theory was in fact similar to the interactionism of social cognitive theory. The major differences are in Allport's use of the term trait and his belief that a person's behavior must occur within a specified range of possibilities. Also, we saw in chapters 8 and 9 that Cattell and Eysenck believed traits interact with situational variables to produce behavior in any given situation. Thus neither Allport, Cattell, nor Eysenck believed behavior was entirely consistent or that it could be predicted on the basis of personality traits alone.