NAB-Mechanisms of Classical Conditioning

Unconditioned Stimulus (UCS): elicit a response automatically

Conditioned Stimulus (CS): previously a neutral stimulus, elicits a conditioned response after being associated with a UCS

E.g. Eyeblink conditioning in rabbits->association between sound and puff of air

>Sound+puff of air=blinking

>Sound=blinking

But, what are the rabbits learning? S-R associations (sound=blinking)? S-S associations(Sound=puff of air)?

Most often S-S associations, but S-R associations are also possible.

The events to be associated need to be paired in time so that they occur together:

"Two ideas will be associated if they occur closely together in time." (Hume, 1700s)

"Cells that fire together, wire together." (Hebb, 1949)

Second-order conditioning:

>CS1=US

>CS2=CS1

>CS2 will elicit the CR

Sensory preconditioning:

>CS1=CS2

>CS2=US

>CS1 will elicit the CR

Pavlov's theory of temporal contiguity where he drew conclusions from over 7000 conditioning experiments:

• The CS and UCS must come into temporal contiguity
• They must do so repeatedly
• Time is fundamental in CC
• Best conditioning: when the CS is followed by the US

Different ways to present CS and US in time:

1. Delay Conditioning-US is presented after CS
2. Trace Conditioning-CS and US are separated by a gap
3. Simultaneous Conditioning-CS and US are presented at the same time
4. Backward conditioning-US is presented before CS

Temporal contiguity generally improves learning. But not always.

=>Berstein's Taste aversion study (1978):

• Group of patients where the chemotherapy they have to undergo causes nausea.
• A new taste of ice-cream, Mapletoff was given to the conditioning group of patients one hour before treatment.
• They rejected Mapletoff ice-cream later.

Studies of trace conditioning show that temporal contiguity between CS and US is generally important, but the study of taste aversion show that temporal contiguity is not necessary for successful associative learning.

So, does the role of temporal contiguity depend on the importance of other factors too?

Overshadowing: effects of stimulus intensity

-Overshadowing: Two CS are paired together with a US. The strength of conditioning to each CS individually is weak

-Conditioned suppression: Decrease in response for reinforcement after the presentation of  a CS which predicts something aversive. Less conditioning= high suppresion ratio

Latent Inhibition

Phase 1) Pre-exposure: CSa=nothing, control: CSb=nothing

Phase 2) Pre-exposure: CSa=US, control: CSa=US

~Stimulus pre-exposure in ohase 1 reduces later learning in ohase 2 when the stimulus is now relevant

~LI can apply to both exitatory and inhibitory stimuli

Conditioned Inhibition: Inhibitory Learning

-When the CS is associated with the absense of the US where phase 1: CS1=US, phase 2: CS1+CS2=nothing

-Excitor and inhibitor neutralizes out each other

-When the inhibitor is later matched with a US, learning is slower

Blocking

-Prior conditioning with a CS1 blocks conditioning with a CS2 when they are presented together

-Learning occurs when the CS provides new information about the US.

-If the second CS2 predict something new, learning should occur. In group 2, CS2 predicts a new strong shock, rats learn the association between CS2 and the new shock.

-Some cues are redundant.

Relative validity

-In correlated group, A and B are bettwe predictors of the US than X. 

-In uncorrelated group, A and B are bad predictors of the US, but when the US is present X was always present.

-More conditioning to X in the uncorrekated group. Subjects learn about the best predictor.

-Some stimuli could be poorly correlated with outcomes.