Lecture 5 - Mechanisms of CC Flashcards
Main conclusions of Pavlovs theory of temporal contiguity
Main conclusions drawn from 7000 conditioning experiments
CS and US must come into temporal contiguity
Do so repeatedly
Fear conditioning ratio: strong conditioning is a ratio close to 0.
Eyeblink conditioning experiment in rabbits
Trial block increase, conditioned response increases
Will level off
Learn to anticipate air puffs
Two kinds of conditioning associations in Pavlovian conditioning
Stimulus -> Stimulus
Stimulus -> Response
AutoShaping in pigeons
Appetitive. Increase learning trials leads increase pecking rate.
Pairing key light and food.
Peck key light in anticipating food. Not Physical limitations
Change response-outcome contingency
What factors determine the rate of learning?
Events associated paired in time so that they occur together
Hume 1700s
Hebb 1949 - Cells that fire together, wire together
Outline second-order conditioning and sensory preconditioning
Introduce food after initial stage. Normally, second event needs be something important for learning to show. Do not need awareness or conscious processing. Only later becomes fully formed association
Outline Pavlov’s 1927 theory of Temporal Contiguity
Conditioning occur with delays several mins, provided CS continued act until onset US - delay conditioning
CS comes end before onset US
(trace conditioning) much harder form CRs. Trace intervals 2-3 mins
Outline temporal contiguity generally improves learning
Something interesting going on, distract you from less salient stimulus (issues simultaneous conditioning)
Outline Bernstein’s 1978 experiment on taste aversion learning
Ppts undergoing chemotherapy (lose appetite, result of food being associated with illness), ice-cream 1 hour before session.
Reduced preference for ice-cream on test day if experience chemo
Taste aversion on Pavlov’s theory of temporal contiguity
Trace conditioning temporal contiguity between CS and US important
Taste aversion show temporal contiguity not necessary
Outline Mackintosh 1976 effects of stimulus intensity: Overshadowing
Lip suppression - animal motivated to drink - licking. Then get stimulus associated foot shock.
Ratio gets closer 0 result higher conditioning after several presentations
Conditioning depend if other cues around same time competing attention.
Outline latent inhibition
Exposure animal cue many times over and nothing happens.
Find later group no pre-exposure signal weak.
Past experience and exposures familiarity reduce conditioning cue
CI = previously trained positive association. Informative outcome expect isn’t going to happen
Latent inhibition apply both excitatory and inhibitory.
Outline CI
Conditioned inhibitory present the CS is not reinforced
Outline the effects of an excitatory with an inhibitor or another excitator
If certain level conditioned responding, then bring in inhibitor, move conditioning opposite direction.
Bring in another excitor we can increase conditioning response
Outline hypothetical effects of pairing an inhibitor with a US
Inhibition occur through extra stimulus but also through schedule where relationship between signal and outcome degraded.
CS isn’t predicting US.
Outline Rescorla’s experiments: CS-US contingency
CS isn’t predicting US
0 contingency = equally get CR after signal as well as randomly
Negative Conditioning = signal ends up being a sign we are not getting food any time soon
Outline Kamin’s blocking experiment
Give animals a predictor
Then give them another cue, it doesn’t tell them anything they didn’t already know. Blocks info.
Second stage introduce light - prior learning blocks association learning of that outcome
Inbuilt control for overshadowing
Outline Kamin’s unblocking experiment
Inbuilt control for overshadowing, which in theory may also reduce conditioning to light.
When change the outcome from phase 1 to phase 2. Introduce light into phase 2 light no longer redundant
Design and results of a relative validity experiment
Correlated group has better predictor of X
Uncorrelated group X no better predicting than other predictors
Attention maintained to X as its consequences are unknown.
Uncorrelated group better learning of X as if attention maintained due to uncertainty of schedule
Functional significance of learning
Restrict learning to events likely indicate causes of important outcomes
Timing: temporal contiguity necessary but not sufficient
Some potential signals don’t stand out from background = overshadowing
Pre-exposure can render stimulus irrelevant = latent inhibition
Significance some events may be qualified by other stimuli e.g. inhibitors, other exictors Some cues redundant = blocking
Some stimuli poorly correlated with outcomes = relative validity
