Instinct, Learning & Motivation L11 - Instrumental Conditioning- description and basic principles PDF

Preview

Summary

Instrumental Conditioning- description and basic principles...

Description

Instinct, Learning and Motivation Lecture 11 - Instrumental Conditioning: description and basic principles

Edward Thorndike • •

1898 book - Animal Intelligence Thorndike - Puzzle box - Cat put inside box when they were hungry and places food outside the box, so they had some motivation to escape the box - Trial and error learning - Gradual improvement - not sudden insight - Took multiple associations before cat could escape the box

Thorndike – IC and the Law of Effect Animal is instrumental in gaining reward by its own actions How does this work? When a response is followed by a satisfying state of affairs, that response will increase in frequency In first trial, cat displayed several behaviours that weren’t successful in helping the cat get out of the box

Over a number of trials, the cat only showed the same behaviour pattern learned through association to help it get out of the box = Law of Effect B F Skinner 1938 book The Behaviour of Organisms Operant behaviour – Behaviour that an organism produces that has some impact on the environment, which in turn changes because of that impact Reinforcement / punishment (can be applied in two different ways: positively and negatively)

Reinforcement 1. Positive reinforcement Receives pleasant stimulus for correct response Functionally, reward • Has beneficial consequence • Increases fitness 2. Negative reinforcement Receives unpleasant stimulus if it fails to make specified response Functionally, • Has detrimental consequence • Decreases fitness

Punishment 1. (positive) punishment Individual receives unpleasant stimulus if it makes specified response Functionally, • Has detrimental consequence • Decreases fitness 2. Negative punishment A reinforcing stimulus is removed after an undesired behaviour

Functionally, • Has detrimental consequence • Decreases fitness Punishment vs reward • Like rewards, punishment is more effective if it follows soon after the response • Punishment does not indicate what correct response should be • Punishment can elicit fear, hostility, retaliation In the morning you press a button to turn off your loud alarm. This is an example of: - Negative reinforcement Which of the following is an example of negative punishment? - A child starts a fight in class and has her favourite toy taken away Reinforcement and punishment: Summary In all cases individual learns that a change in behaviour results in a more pleasant outcome Functionally this means an increase in fitness Basic principles of instrumental/operant conditioning • Primary and secondary reinforcement and punishment • Neutrality of reinforcers – premack principle • Limiting conditions – over justification effect • Basics – discrimination, generalisation, context • Extinction • Schedules of reinforcement • Shaping through successive approximations • Superstitious behaviour (Red = cover in reading) Extinction in classical conditioning is: - The gradual elimination of a learned response that occurs when the CS (bell) is no longer followed by the US (salivation) Extinction is much harder to occur in instrumental conditioning than classical conditioning due to the idea that you are being rewarded every time during classical conditioning Primary and secondary reinforcement and punishment • Primary – innate (e.g food, social contact) • Secondary - learned (e.g money, trophies) Similar process to CC: Money = Neutral

After pairings with primary reinforcer (food) Money alone elicits a conditioned emotional response Neutrality of reinforcers – Premack principle (Premack 1962) = Find the activity someone prefers, and then use that to reinforce a non-preferred one Neutrality of reinforcers? Need to think of the individual Reinforcers are relative depending on the specific individual Limiting conditions – over justification effect Sometimes the presentation of a reward reduces performance of the behaviour! Lepper & Greene 1978

Gave group B a trophy after they had drawn = Group B spent more time drawing than group A When reward was removed = Group B spent less time drawing than group A Schedules of reinforcement in IC

Interval = reinforcement based on time elapsed since reinforcement Ratio – reinforcement based on how many response have been made

Fixed interval (FI) = Reinforcement becomes AVAILABLE after a fixed time period has elapsed (before every reifnorcer we see an increase in behaviour, due to anticipation of reward so start producin desired behaviour) Variable Interval (VI) = Same as FI BUT length of waiting period varies from one reinforce to the next Fixed ratio (FR) = Reinforcement will be delivered after a specific number of responses have been made e.g. reinforce after every 4th response Variable ratio (VR) = Reinforcement delivered after a specified average number of responses All intermittent reinforcement RATIO – high and consistent rates of responding as rewards are directly related to responses All produce behaviour more resistant to extinction than continuous reinforcement Summary Instrumental (=Operant) Conditioning • Individual gains reward by its own actions • Individual learns that a change in behaviour results in more pleasant outcome • Functionally this means increased fitness...