7.1.1 Learning: conditioning, observational learning and transfer (exam summary)

Explained

1. Classical conditioning: the term table.

Timing of CS and UCS (best to worst):

• Delayed (forward): CS starts first and overlaps the UCS; most effective.
• Trace: CS starts and ends before the UCS; a gap the animal must bridge in memory.
• Simultaneous: CS and UCS start together; weak.
• Backward: UCS comes before CS; least effective.

Human examples: fears and phobias (Little Albert), taste aversion after one bout of food poisoning, anxiety at the sight of a hospital, advertisements pairing a product with a pleasant image. Therapy uses it in reverse: systematic desensitisation (Wolpe) and aversion therapy.

2. Operant conditioning: the term table.

• Primary reinforcers meet biological needs (food, water); secondary (conditioned) reinforcers gain value through pairing (praise, money, grades, tokens).
• What makes reinforcement work: immediate rather than delayed reinforcement; amount and quality of the reinforcer; the schedule (continuous is fastest for acquisition, partial is most resistant to extinction: the partial reinforcement effect).
• Extinction, generalisation, discrimination and spontaneous recovery occur in operant learning too. A discriminative stimulus signals that a response will now be reinforced (the "open" sign on a shop).
• Punishment's limits: it shows what not to do, not what to do; it suppresses rather than removes a behaviour; it can breed fear, anger and avoidance of the punisher. It works best when immediate, consistent and paired with reinforcement of a correct alternative.
• Learned helplessness explained: in Seligman and Maier's study, dogs that had been able to stop shocks by pressing a panel later learnt to jump a barrier to escape; dogs given the same shocks with no control lay down and endured them. Lack of control, not the shock itself, produced the passivity. Human parallels: repeated failure in school, long-term unemployment, prisoners and institution inmates who stop trying. Remedy: experiences of success and control.

3. Classical vs operant conditioning.

4. Observational learning: what to remember.

• Attention (to a model who is attractive, similar, powerful or successful) → retention (in images and words) → reproduction (having the skill to copy) → motivation (a reason to perform).
• Vicarious reinforcement / punishment: seeing the model rewarded or punished changes how likely the observer is to copy.
• Learning vs performance (Bandura, 1965): children who saw a model punished for aggression copied less, but when later offered a reward for copying, the gap between the groups almost vanished: they reproduced the aggression about as well as the others. So they had learnt it; consequences decided only whether they performed it.
• Uses: teaching skills by demonstration, role models, peer tutoring; concern over violence in films and TV; modelling therapy for fears.

5. Insight learning: the features. Sudden solution after a pause; depends on seeing the whole problem; once found, it is repeated without error, retained and transferred to similar problems. Compare trial and error (Thorndike): gradual, random, error falls slowly.

6. Transfer of training in one table.

• Theories in one line each: formal discipline: hard subjects train the mind's "faculties" like muscles (rejected by James and Thorndike); identical elements: transfer only as far as the tasks share elements; generalisation: understanding a principle transfers (Judd's darts at an underwater target); transposition: transfer of a grasped relationship (Gestalt); ideals: values transfer when taught as ideals.
• Learning to learn (learning set): Harry Harlow (1949): monkeys given hundreds of different discrimination problems got faster at each new one until they solved it in a single trial. Practice in a type of problem transfers.
• Teaching for transfer: teach principles, not only facts; use varied examples; link lessons to real life; point out the similarities between old and new tasks; practise to mastery.

7. Learning and motivation.

• Motivation starts, directs and keeps up learning (motive cycle and Maslow are in EDU-LR-05). Without some drive or interest, practice is mechanical and little is learnt.
• Intrinsic vs extrinsic: learning for its own sake (curiosity, mastery) lasts longer than learning for marks or prizes.
• Overjustification effect: Lepper, Greene and Nisbett (1973): nursery children who enjoyed drawing were promised a reward for it; later they drew less in free time than children who were not promised one. An expected external reward can weaken intrinsic interest.
• Praise and reproof: Elizabeth Hurlock (1925): in arithmetic classes, the praised group improved most, the reproved group next, and the ignored group least. Any attention beats none, and praise works best.
• Knowledge of results (feedback) improves learning, especially when immediate (the basis of programmed learning; EDU-LR-03).
• Level of aspiration, achievement motivation, competition and co-operation also shape effort (EDU-LR-05).
• Too much arousal hurts: exam anxiety lowers performance on hard papers (Yerkes-Dodson).