Long-lasting changes in behaviour caused by environmental events
Behaviour analysts study learning by
Looking at how behaviour is affected by environment (everything that happens before and after and around the behaviour)
Psychologists study learning by looking at how behaviour is affected by environment
Working assumption: both human and animal behaviour is made up of same simple elements
Animal behaviour is probably not affected by us observing it
Thorndike's puzzle box experiment
1. Cat in puzzle box
2. Get fish from outside of puzzle box (motivator)
3. Measure latency to escape
4. Cat stands on lever by accident, gets out (learning curve - more likely to do again in future)
Thorndike's Law of Effect
If response is followed by favourable consequence, response will be more likely to occur again in future
Operant/instrumental behaviour
Behaviour subject to the Law of Effect (things done in order to produce consequence)
Operant conditioning
Process of animal learning an association between response and its consequences
Reinforcer is most effective if it immediately follows the response
Contingency
Response causes the reinforcer to occur
Shaping
Reinforce closer and closer approximations to the response we want to produce
Positive reinforcement
Presenting a desired favourable event
Negative reinforcement
Withdrawing an aversive event
Positive punishment
Presenting an aversive event
Negative punishment
Withdrawing an appetitive event
Primary reinforcers
Can reinforce behaviour because of innate biological significance (e.g. food, warmth, escape from pain)
Secondary reinforcers (Conditioned reinforcers)
Previous neutral stimuli that get reinforcing properties by being paired with primary reinforcers in organism's history
Continuous reinforcement
Every response is reinforced
Intermittent reinforcement
Only some responses are reinforced, according to a schedule
Responses maintained by intermittent reinforcement persist much longer during extinction than those maintained by continuous reinforcement
Fixed ratio schedule
Reinforcing every fifth response
Fixed interval schedule
A response is reinforced after a fixed amount of time has elapsed since the last reinforcer
Variable interval schedule
A response is reinforced after a variable amount of time has elapsed since the last reinforcer
Differential reinforcement of other behaviour (DRO)
Reinforcer is delivered when fixed amount of time has elapsed since last response (reinforces non-responding)
Fixed time schedule
A reinforcer is delivered when a fixed amount of time has elapsed since the last reinforcer (non-contingent or response-independent reinforcement)
Skinner found that fixed time schedules can produce idiosyncratic, superstitious responding in pigeons
Superstitions in humans may be explained by the same process of adventitious (by chance) reinforcement of whatever behaviour was occurring just before the non-contingent reinforcement
Stimulus control of operant behaviour
The extent to which stimuli that precede or accompany operant behaviour come to affect the rate or probability of that behaviour
The Law of Effect is a bit oversimplified
When a response is reinforced
It becomes more likely to occur again in the same environment, or same stimulus context, in which it was reinforced
If we change the stimulus context, we can change the probability of response
Likely repeat reinforced behaviours when in same situation
Stimulus control
The extent to which stimuli that precede or accompany operant behaviour come to affect the rate or probability of that behaviour
More stimulus
More discrimination, less generalisation
Generalisation gradient
How much does original trained behaviour generalise to new stimuli
S+ (positive discriminative stimulus)
Signals what the consequences for responding are going to be, and that the consequences will be reinforcement