L&C - Extinction and Stimulus Control (8)

extinction

an operant procedure that we use in order to distinguish behaviour when behaviour of interest either decreases significantly or stops because it is no longer reinforced; if it stops for any other reason (e.g., punishment), it is not extinction

partial extinction

behaviour that is decreased because it is not reinforced

escape extinction

when a behaviour is reduced or stopped because it is no longer negatively reinforced

extinction burst

a secondary effect of extinction where a behaviour spikes in frequency or intensity following an extinction procedure — it gets worse before it gets better (ex. stop giving tantrums attention and they get worse before becoming extinct)

response variation

a secondary effect of extinction where new behaviours are attempted to receive reinforcement

spontaneous recovery

a secondary effect of extinction where a behaviour returns without reinforcement, however, if the behaviour continues to not be reinforced, this phenomenon goes away

resistance to extinction

how long the behaviour continues without reinforcement

high reinforcement

behaviour continues for a long time after reinforcement has stopped

low reinforcement

behaviour stops shortly after reinforcement stops

renewal

following successful extinction, the behaviour comes back in another context/new situation (ex. child loves to swear and we extinct this behaviour but he goes to grandma’s and the swearing comes back)

reinstatement

following successful extinction, behaviour might come back if we use the original reinforcer to either reinforce the new behaviour or as a non-contingent reinforcer (ex. a kid usually gets candy after whining and you stop giving him candy but dad comes back with lots of candy and the whining starts again)

resurgence

Extinction procedure was successful for behaviour A but while extinguishing A, we start reinforcing a more desirable (B) behaviour. If a some point B stops being reinforced, then A is likely to come back (ex. A = stop whining, B = ask politely… but one day you scream at the kid while he is asking politely and whining comes back)

discriminative stimulus (SD)

stimulus that let’s us know that if we engage in a behaviour, we will be reinforced for it

discriminative stimulus for extinction (S∆)

stimulus that let’s us know that if we engage in a behaviour, we will not be reinforced for it

differential responding

if our behaviour varies depending on the context/stimulus, this means that our behaviour has come under stimulus control (ex. only telling dad a joke when he is in a good mood — telling him a joke regardless of his mood means we are not under stimulus control)

generalization gradient

a graph where the x axis represents all the stimulus (conditions) the subject will be exposed to and the y axis represents the number of responses per stimulus; the strongest response should be to the SD if we are under stimulus control

flat (generalization gradient)

no discrimination, high generalization (we respond the same regardless of the stimulus/condition)

narrow (generalization gradient)

sharp point, flare base; high discrimination, low generalization (we only respond to the SD)

broad (generalization gradient)

sharp point, wide and gradual base; some discrimination/some generalization (we respond more to SD but still to other stimuli)

stimulus generalization

We have been successfully reinforced to engage in behaviour of interest in the presence of a specific SD. Afterwards, we start doing the behaviour in the presence of similar SDs without training or reinforcement (e.g. we only tell friends jokes when we can read that they’re in a good mood, not just with dad)

stimulus discrimination

we are reinforced when we engage in a desired behaviour in the presence one SD and only this one

absolute theory of stimulus control

we learn about the characteristics of two stimuli separately (e.g., pigeon learns to peck on a pink light to get food

relational theory of stimulus control

we learn about the characteristics of two stimuli in relation to each other (e.g., peck on the less bright light to get food)

transposition

evidence to support relational theory; we learn a rule in one context and we apply it to a different context

peak shift

evidence for relational theory; this happens with discrimination theory where the strongest association is not to the SD the subject is trained with but for the stimulus that is further away from S∆ to seek more reinforcement (ex. you usually wear loose dresses but when you wear a tight one, you get a lot of compliments so you switch to wearing more tight dresses)

behavioural contrast

a change in reinforcement to one SD causes an unintended change in behaviour towards the other SD despite the other SD not changing

positive behavioural contrast

let’s say we are on 2 separate and distinct schedules of reinforcement, A and B… something changes in A (it becomes thinner—less reinforcement) so response to B will increase despite it not changing (ex. your friend is fighting with her bf so she starts hanging out with you more)

negative behavioural contrast

let’s say we have two separate and distinct schedules of reinforcement, A and B… A becomes richer (more reinforcement) but B stays the same so B gets less attention even thought it remains the same (ex. your friend’s bf starts giving her more attention so she starts spending less time with you)

local (behavioural contrast)

behaviour in the present can be affected by our previous experiences

anticipatory (behavioural contrast)

behaviour in the present can be affected by future events