L12 - Conditioned Inhibition

Question 1

What does extinction as inhibition mean?

Answer 1

Extinction creates new inhibitory association that counteracts original association

Question 2

What is inhibitory learning?

Answer 2

This is when the presence of the US is more likely when the CS is absent

Probability of the US following the CS is less than the prob that the US will occur without the CS

Question 3

What is a -ve contingency?

Answer 3

Presenting the US only when the CS is absent

Question 4

Real world example of conditioned inhibition

Answer 4

Catch the bus every morning on the way to school, bully bullies you every time except when the sister is on the bus, she is a conditioned inhibitor

Question 5

what are two tests to see that a stimulus has inhibitory properties

Answer 5

summation and redardation test

Question 6

What is a summation test?

Answer 6

If you have a stimulus that has inhibitory properties, if you were to present it with an excitatory at the same time, it would lead to a reduction in conditioned responding in comparison to showing the excitatory on its own

• The sum reduces responding to another CS+

Question 7

What is a retardation test?

Answer 7

If a CS has inhibitory properties it should take longer to associate to a US compared to a normal CS

Learning is impaired compared to a novel CS

Question 8

How does the R-W model explain inhibition?

Answer 8

A = excitatory CS, X = conditioned inhibitor CS

A is presented with X and no US is delivered (Lambda = 0 as no outcome occurs)

V decreases from 1 for CS A (excitatory)
For V of CS X, the V value is less than 0 - This explains X -no US association

Association of A will not drop to 0

Question 9

Discuss R-W model predictions when a CS A is reinforced 100% and CS B has partial rft of 50%?

Answer 9

Learning occurs faster and reaches higher asymptote for CS A

Question 10

What happens when you present a new CS with a partially reinforced 50% CS B followed by no US?

Answer 10

Animals quickly learn to expect no US due to the conditioned inhibitor

Question 11

What is conditioned inhibition?

Answer 11

Learning that a CS prevents the reward; therefore there is not point responding to it.

Negative discrepancy - the stimulus becomes an inhibitory because the associative strength becomes negative.

Question 12

Two different types of associative connection:

Answer 12

Excitatory link primes memory of the US.

Inhibitory link suppresses memory of US.

Question 13

How do we work out the associative strength of the inhibitory stimulus?

Answer 13

Conceptually, should expect the conditioned inhibitor (CI) to have equal magnitude to CS but in -ve direction

CI brings the CR back to baseline so needs to have the same magnitude in opposite direction in order to cancel this out

Question 14

Discuss what happens when you use pair a conditioned inhibitor with an excitatory CS during extinction

Answer 14

1st conditioning group : Form B->US association and then form B extinction assocation

2nd group: Form B > US association and then B+X (where X is a conditioned inhibitor)

B in first condition extinguishes faster

lack of US is associated with X so B doesn’t extinguish in 2nd group

Question 15

According to R-W, what should happen when an inhibitory CS is repeatedly presented on its own without the US?

Answer 15

Due to (0–V) delta V is +ve, so V increases up towards 0

i.e take a CI (-V) and present it on its own without any US: R-W predicts extinction of inhibition - we know this DOES NOT HAPPEN and CI’s do not extinguish

Question 16

What does the R-W model predict should happen when an inhibitory CS (X) is repeatedly presented with a neutral CS (Y) and no US?

Answer 16

DeltaV is positive for both X and Y

On the 1st trial, lambda = 0, there will be -ve V
Delta V will be +ve due to prediction error for both X and Y

Will end up extinguishing X and Y has an increase in associative strength above 0 suggesting that you have learnt something about Y and the US that X is inhibiting

Even tho never presented Y with the US at all - this doesn’t make sense

This is another failure of the R-W model

Question 17

Example of superconditioning experiment with a novel CS

RW model

Answer 17

Y > US
Y + X > US

where X is a conditioned inhibitor

Y is more powerful than X so:

Predictive value of
(1 - - 1) for trial 1