operant conditioning
BF Skinner
learning in which an individual performs a behaviour and the following consequence (reward or punishment) increases or decreases the chance that the behaviour will be performed again
we operate based on a reward or fear of punishment
the law of effect, Edward Thorndike’s cats
behaviours with satisfying outcomes are more likely to happen again
unsatisfying outcomes are less likely to happen again
positive and negative reinforcement
skinner’s contributions
refinement of the concept of operant conditioning and the law of effect
schedules of reinforcement
the shaping of new behaviours through reinforcement
he believed that we don’t have a lot of agency/free will and that we are always being driven by outside forces
reinforcement
increases the chance that a behaviour will occur again
positive reinforcement
giving a pleasant stimulus that increases the likelihood of a behaviour occurring again
negative reinforcement
removal of an unpleasant stimulus that increases the likelihood of the behaviour occurring again
punishment
decreases that chance that a behaviour will occur again
positive punishment
presenting an unpleasant stimulus to decrease the chance the behaviour will occur again
negative punishment
the removal of a pleasant stimulus to decrease the chance the behaviour will occur again
the skinner box
- looking at rats and pigeons in a box, a lever was in the box and each time the rats pressed the lever they would get food (positive reinforcement)
- the information from the animals generalises to humans
- skinner used shaping to teach his animals
- pigeons: food was given at regular intervals regardless of the pigeon’s behaviour
- overtime they associated whatever random action they were doing when they got the food with receiving the food
- this made them repeat the behaviour, believing this action was what made the food appear
shaping
gradually reinforcing smaller behaviours towards an overall desired behaviour, giving small positive reinforcements along the way
skinners box limitations
animals and rats were used
there are limited options available to the rat, unlike in the real world
done in a lab setting
animals were tested one at a time
the validity of the skinner box to humans and in the real world is therefore limited
overly simplistic view of behaviour, ignores cognition and freewill
skinners box strengths
scientific rigor, ability to produce reliable, quantifiable data on learned behaviours
useful for applications like animal training, basic behavioural therapy
wide applicability
identification of reinforcement principles
examples in human behaviour
drug addiction - dopamine pathways positively reinforce drug use whilst withdrawing symptoms negatively reinforce use
gambling - functional magnetic resonance imagining was used to find there was activation of the neural structures relating to the rewards system when anticipating a money reward
social media - neural structures were investigated during the use of social media and found activation in the brain’s reward circuits when participants received likes
biological brain reward system
- positive reinforcement tends to work better long term as it creates a stronger neural pathway
- the reward systems of the brain are a group of structures that activate whenever we experience something rewarding
- structures that are considered to be a part of this reward system are found along the primary dopamine pathways of the brain
the reward pathway
- perception of a rewarding stimulus (hippocampus and amygdala)
- release of dopamine
- dopamine causes feelings of pleasure
- assessment of experience connects the feeling of pleasure with the stimulus (prefrontal cortex)
- this pathway operates on a feedback loop
operant learning strengths
effective for behaviours modification, creates new behaviours through shaping, practical applications e.g therapies, offers a framework for understanding consequences
operant learning weaknesses
potential for longer term negative effects of punishment
risk of eroding intrinsic motivation
ignores cognitive and social factors
reliance on external validation
generalisability from animal studies
