The learning approach: imitation, conditioning and reinforcement

What the learning approach claims

Behaviour is acquired from the environment through three mechanisms: classical conditioning (an association between two stimuli), operant conditioning (consequences change how often a behaviour occurs) and social learning (observing a model and imitating). The approach is strong on explaining how behaviour is learned and how it can be unlearned; it is weak on biological predispositions and on thought processes.

Bandura, Ross and Ross (1961): transmission of aggression

Method: a laboratory experiment with 36 boys and 36 girls aged 37–69 months from the Stanford University nursery. Children were matched on pre-rated aggression and allocated to an aggressive model, a non-aggressive model, or a control group, with model gender also varied. After watching, every child was deliberately frustrated — shown attractive toys then told they were reserved for other children — before entering a room containing a Bobo doll, a mallet and other toys. Behaviour was observed through a one-way mirror in 5-second intervals for 20 minutes.

Results: children who saw the aggressive model imitated specific acts closely, including the mallet strike and verbalisations such as 'Pow'. Boys were more physically aggressive than girls, especially after a male model; girls showed more verbal aggression after a female model. Conclusion: aggression can be learned by observation and imitation, and imitation is stronger when the model is the same sex as the child.

Saavedra and Silverman (2002): button phobia

A case study of a nine-year-old boy with an intense fear of buttons that stopped him dressing for school. The fear began when he spilled a bowl of buttons at nursery and the teacher reacted sharply — classical conditioning. It was maintained because avoiding buttons removed the anxiety, which is negative reinforcement — operant conditioning. Treatment ran in two phases. Phase 1 was standard exposure therapy with rewards, working up a hierarchy of buttons; fear ratings dropped but then rose again. Phase 2 added imagery exposure in which the boy pictured buttons falling on him and discussed his feelings of disgust. Fear ratings fell far further and stayed down at a six-month follow-up. Conclusion: a phobia maintained by disgust as well as fear responds better when the disgust itself is addressed.

Fagen et al. (2014): training captive elephants

Five Asian elephants in Nepal were trained to accept a trunk wash — needed to test for tuberculosis — using secondary positive reinforcement. A whistle was paired with a food reward so that the whistle itself became a reward marker (a bridge), and the target behaviour was built up by shaping and successive approximation: trunk here, then trunk up, then hold, then blow. Training took 35 sessions of 5–10 minutes. Four of the five elephants completed the full trunk wash. Conclusion: operant techniques with positive reinforcement can replace coercive handling for a difficult veterinary procedure.