Principles of Behavior Modeling

Behavior modeling or imitation learning was a virtually unresearched and unknown topic prior to the 1941 publication of Social Learning and Imitation by Miller and Dollard. Their studies lead them to view imitation learning as a special form of the behavioral conditioning process.

You can download excellent powerpoint slides on HR management, business strategy and personal development HERE.

Essentially, the trainer must provide a sample of the behavior, the learner must respond in a way that matches the sample, and the imitation must be positively reinforced. In Bollard’s and Miller’s view, the “model” simply informs the learner where to go or how to behave for reinforcement. The learner does not acquire new, previously unexhibited behavior from the model. Though much of Miller’s and Bollard’s interpretation of results and theorizing has been questioned recently, they deserve credit for prim-ing the pump, for beginning to re¬search the question, “How and what do people learn simply by watching others?”

Groundbreaking and impressive modeling research has been conducted by Br. Albert Bandura of Stanford University. In a typical experiment, Bandura, Ross and Ross showed nursery-school children a motion picture of an adult displaying “aggressive behavior” toward a large, inflatable rubber clown— the kind that bounces back for more because of a bottom full if sand. Boys and girls who watched the film behaved aggressively toward the clown themselves, closely mimicking the adult in the film. They lifted and threw, kicked and hit, and beat the clown with a hammer exactly as the adult had. Bandura has repeatedly found that most children will, with little or no prodding, thus imitate the novel behavior of a model.

In addition to establishing that modeling does occur, Bandura has investigated the ways in which it can be increased or decreased. He has found, or example, that children are more likely to imitate the behavior of a Model they see rewarded than one hey see punished. In one experiment, three groups of children watched films if a model who yelled at and punched a Bobo doll. In the film one group saw, the model was punished by an authority for punching the doll; in the film a second group saw, the model was praised for his aggressive behavior; in the film the third group saw, the model received neither praise nor punishment.

When put into a situation similar to that of the model, the children who had seen him praised were much more likely to imitate him than those who had seen him punished. Bandura hypothesized that the children identified with the model and experienced reward or punishment vicariously as they watched the film. Therefore, when given a chance to act as the model had, they tended to behave as if they themselves (instead of the model) had earlier been praised or punished for hitting the doll.

Bandura’s approach to understanding the modeling phenomenon is more complex and cognitively oriented than Bollard’s and Miller’s. To Bandura, models influence learning in an information-transmitting fashion. The observer acquires a symbolic representation of the modeled activities rather than a group of stimulus-response associations. Because there are cognitive or mental, as well as behavioral, components to modeling, Bandura suggests that there are four processes governing the modeling phenomenon: attention, retention, re-production, and reinforcement/motivation.

You can download excellent powerpoint slides on HR management and HRD strategy HERE.

The attention process

The principle here is that a person cannot learn much by observation if he or she does not attend to, or recognize, the essential features of the model’s behavior. The learner must be sen¬sitized to look for the things we want them to learn or they will learn some¬thing else — or nothing at all. Research suggests there are the follow¬ing natural factors which shape the attention patterns of learners:

• Similarity of model and learner. The people with whom one regularly associates influence the types of behavior one will repeatedly observe and learn thoroughly. Opportunities for learning aggressive behavior are most prevalent for children of the urban poor than for offspring of the rural Quaker.

• Functional value. Behavior which has an intuitive value for the learner in the learner’s environment is more likely to be attended to. Subway-map reading doesn’t have a high functional value to a Los Angelino.

• Interpersonal attraction. Models who possess interesting and winsome qualities are sought out, whereas those without such characteristics tend to be ignored or rejected, even though they may excel in other ways.

• Media attractiveness. Models presented in televised form are so effective in capturing attention that viewers learn the depicted behavior regardless of the presence or absence of incentives for learning.

You can download excellent powerpoint slides on HR management, business strategy and personal development HERE.

Retention processes

If one is to reproduce a model’s behavior when the latter is no longer present to serve as a guide, the response patterns must be represented in memory in symbolic form.

• Mental images. There is an almost automatic mental imaging function, but for practical use these “mind pictures” should be associated with common symbols, such as words. As Plato suggested, a bowl may be only a poor example of the concept bowl, but the concept name bowl invariably pulls forth an image of a specific bowl one has had experience with.

• Verbal coding. Observers who code modeled behavior into words, concise labels, or vivid imagery learn and retain the behavior better than those who simply observe or are men¬tally preoccupied with other matters while watching the performance of others.1 Consciously .counting Mark Spitz’s strokes per lap will help a swimmer retain and retrieve a mental image of Spitz’s swimming style.

• Rehearsal. People who mentally rehearse or actually perform modeled forget them. For example, mumbling your “opening lines” over and over to yourself makes them automatic.

Motoric reproduction processes

To actually do the behavior, the learner must put together a given set of responses according to the modeled patterns. The amount of the model which can be exhibited depends on acquisition of component skills.

• Parts must equal whole. A learner could have all the parts but not succeed with the whole. A young child might watch and be able to exhibit all the components of driving a car but still run the family Hupmobile through the garage wall.

• Skill level. Even if all components are learned, can be identified and talked about, performance of the whole will be poor until all the components are practiced and fitted together in a whole. A would-be tennis player can learn a lot by watching others play, talking about it and banging balls against the house. But, until one goes on the court and hits with another person, honing and fine tuning to the level of the model can’t take place.


A person can acquire, retain and possess the capabilities for skillful execution of modeled behavior, but the learning rarely will be activated il negatively sanctioned.

• Positive incentive. When positive incentives are provided, observational learning, which has been previously unexpressed, is promptly translated into action. Reinforcement influences not only the demonstration of already learned behavior, but it can affect the level of observational learning by influencing attending, coding and rehearsal phases.

• Anticipation of reinforcement. H the learner sees the model reinforced positively or avoiding punishment, learning is enhanced.

• Familiarity with subskills. If the learner has all the subskills and also has names or a coding system for them, seemingly complex skills can be modeled verbally — a factor most technical-manual writers rely upon heavily.

• Multiple models. The more models the learner is exposed to, the less likely he or she is to learn only one way of behaving. Learners exposed to three or four models exhibit behavior that is an innovative mix of all of the models seen. This would seem especially important for learning interpersonal skills that are personality dependent such as selling or interviewing.

• Outcome for the model. How the model is treated influences both the
probability of the learner learning the behavior and the learner’s attitude toward the model and the behavior the model exhibited. This concept has been especially useful in helping people de-learn phobias.

Where modeling works

• During the last decade, there has been copious research on modeling applications. And at least some should interest the training and development professional.

• Problem-solving strategies— both linear and creative — seem to be enhanced by a combination of process rules and extensive modeling of the processes.

• Rule following and positive attitude toward procedures and rules seem to be enhanced through discussion and viewing of filmed vignettes of people successfully following rules and procedures.

• Interviewing skills can be quickly and effectively learned by watching a videotape of a model conducting an effective interview.

• When watching a model of supervisory skills, trainees also tend to learn and assimilate the leadership style of the model.

• Self-disclosure in a closed group increases when the facilitator models self-disclosure and makes a high number of self-disclosing statements.

• Assertive behavior can be effectively learned watching a videotape of a low-assertive person becoming assertive with the help of an assertiveness trainer.

• Complex motor skills can be learned by watching a film of a correctly performing model or a film of someone learning to perform the motor skill.

• Tolerance to pain can be learned from watching a model endure what the viewer perceives to be a painful experience. (Especially useful for those who must conduct or endure long meetings or seminars.)

• And, finally, one researcher has determined that we are what we watch. Specifically, he found that those who watch “All in the Family” most frequently are most tolerant of and in sympathy with Archie Bunker’s attitudes and views. The researcher does not suggest a causal relationship but if some Sunday morning you notice the “Gilligan’s Island” return you’re watching is one you’ve
seen three times before— and you’re still enjoying it’—it might be time to consider the effect behavior modeling is having on you.

Source : Training Magazine June 1978.