Nature of learning
Learning is any fairly lasting change in behaviour or knowledge that happens because of practice or experience.
- It always involves experience; changes due to maturing, tiredness, illness or drugs are not learning.
- It is inferred from performance; we cannot see it directly.
- It is a process that goes through steps and it can transfer to new situations.
Main paradigms: classical conditioning, operant conditioning, observational learning, cognitive learning, verbal learning, concept learning and skill learning.
Classical conditioning
Ivan Pavlov studied dogs. Food in the mouth makes a dog salivate without any learning.
- US (unconditioned stimulus): food. UR (unconditioned response): salivation to food.
- CS (conditioned stimulus): a bell, at first neutral.
- After many bell + food pairings, the bell alone brings saliva: the CR (conditioned response).
Determinants of classical conditioning
- Time relation between CS and US: delayed (CS starts first and stays till US comes) works best; simultaneous, trace (CS ends before US) and backward (US first) work less or poorly.
- Type of US: appetitive (food, water) or aversive (shock, loud noise). Aversive ones are learned fast.
- Intensity of the CS: stronger, clearer signals are learned faster.
Operant conditioning
B.F. Skinner used a box with a lever. A hungry rat moving around presses the lever by chance and food drops. It then presses more often. An operant is a voluntary action that works on the surroundings.
Determinants of operant conditioning
- Type of reinforcer: positive (something pleasant added, like praise or food) or negative (something unpleasant removed, like stopping an alarm). Both raise behaviour. Primary reinforcers meet body needs; secondary ones (money, grades) are learned.
- Number/amount of reinforcement: more, bigger rewards speed learning up to a point.
- Schedule: continuous (every time) builds learning fast; partial (sometimes) makes it last longer and resist extinction.
- Delay: the sooner the reward, the better the learning.
Punishment is a bad result that lowers a behaviour. It is not the same as negative reinforcement. Avoidance and escape learning also come from unpleasant events.
Key learning processes
- Reinforcement: anything that strengthens a response.
- Acquisition: the response builds up with pairings or rewards.
- Extinction: the response fades when reinforcement stops.
- Spontaneous recovery: after a rest, an extinguished response comes back a little.
- Generalisation: a response to one stimulus also comes to similar stimuli (a child afraid of one dog fears all dogs).
- Discrimination: responding only to the right stimulus and not to similar ones.
- Shaping and chaining: rewarding small steps to build a complex behaviour.
Key formulas and definitions
- Learning: lasting change from practice or experience
- Classical: CS + US (paired) → later CS alone → CR
- Operant: behaviour → consequence → more or less behaviour
- Positive reinforcement: add pleasant → behaviour ↑
- Negative reinforcement: remove unpleasant → behaviour ↑
- Punishment: unpleasant result → behaviour ↓
- Partial reinforcement → slower extinction
Worked examples
1. Every time Aman's phone buzzes, a friend's funny message follows. Now he feels happy at the buzz alone. Name US, CS and CR.
US = funny message (makes him happy naturally). CS = the buzz. CR = feeling happy at the buzz alone.
2. A driver fastens the seat belt to stop the beeping. Which process?
Negative reinforcement: removing an unpleasant sound makes seat-belt use more likely.
3. A child afraid of one barking dog is now afraid of all dogs. Later, after good experiences, she fears only that one dog. Name both processes.
First generalisation (fear spreads to similar stimuli), then discrimination (fear only to the specific one).
4. Why do people keep playing lottery games even though they rarely win?
Wins come on a partial (variable) schedule of reinforcement, which makes behaviour very hard to extinguish.
Common mistakes
- Thinking negative reinforcement means punishment. Negative reinforcement removes something bad and raises behaviour.
- Calling growth changes learning. Growing taller or a baby starting to walk because of maturation is not learning.
- Believing extinction erases learning. Spontaneous recovery shows it was only suppressed.
- Mixing classical and operant: in classical the stimulus comes first; in operant the consequence comes after an action.