Operant conditioning

Instrumental learning

• Edward Thorndike (1898) - experiments investigating 'instrumental' behaviour (voluntary, active behaviour that affects the environment)
• Placed a hungry cat inside a puzzle box and small amount of food outside the door
• Measured the time taken to escape over a number of trials

Findings and conclusions

• Learning more gradual than might have been expected
• Not due to insight or understanding, but through formation of an association between stimulus and response
• Law of effect: behaviours that are followed by a 'satisfying state of affairs' tend to be repeated and those that produce an unpleasant state of affairs are less likely to be repeated

operant conditioning

• B.F. Skinner (1983, 1953)
• Developed Thorndike's research
• Operant behaviour: behaviour that an organism produces that has some impact on the environment, which in turn changes because of that impact
• Environmental changes either reinforce or punish the behaviours that produced them

Operant conditioning chamber (Skinner box)

• A controlled environment for studying the behaviour of small organisms 
• Behaviour such as pressing the lever is either reinforced or punished in the presence of a discriminative stimulus
• The consequence of behaviour is the presentation or removal or a desirable or aversive stimulus (food or shock)
• Discriminative stimulus > operant behaviour > consequence
• Frequency of responses measured over time

Reinforcement

Positive Reinforcement

• Presentation of a desirable stimulus following a response
• Lever press (operant behaviour) -> food (consequence)
• The consequence of food leads to an increase in lever pressing
• E.g. reward after handing in a lost item

Negative reinforcement

• Removal of an aversive stimulus following a response
• Lever press (operant behaviour) -> Escape to avoid shock (consequence)
• The consequence of escaping the shock leads to increase in lever pressing
• E.g. Pain relief/prevention after taking an aspirin

Punishment

• Positive punishment: Presentation of an aversive stimulus following a response
• Negative punishment: Removal of a desirable stimulus following a response

Primary reinforcers/punishers:

• Do not need to be learned, gain their functions from biological mechanisms

Secondary reinforcers/punishers:

• Are not innately effective and need to be learned. typically acquire their effectiveness through pairing with primary reinforcers/punishers

Reinforcement vs Punishment

• Thorndike observed that punishment was relatively ineffective, and believed reinforcement to be critical to learning
• Punishment might be effective in stopping a behaviour, but it doesn't specify what the desired behaviour
• Learning is thus usually more effective through reinforcement of the desired behaviour, rather than punishment of the unacceptable behaviour

Schedules of reinforcements

• Different schedules of reinforcement produce different rates of responding
• Continuous reinforcement: reinforce every time behaviour occurs
• rapid learning - rapid extinction
• Partial/Intermittent reinforcement
• Sometimes reinforce, sometimes don't 
• Slower learning; more resistant to extinction
• interval and ratio schedules

 

Shaping - involves rewarding successive approximations of the target behaviour

Project Pigeon - During WWII, Skinner used operant conditioning to train pigeons to guide missiles.

Animal behaviour enterprises

• used operant conditioning to train animals for commercial purposes
• opened the IQ zoo in 1955
• never used punishment, only positive reinforcement
• Later developed programs for people with intellectual and developmental disabilities