Rescorla & Wagner’s Model

Question 1

what did the Rescorla and Wagner Model (RWM) find?

Answer 1

• found that RWM could acquire (λ = 1) and lose (λ = 0) associative strength.
• In each case learning realistically levelled off at λ.
• Our computations realistically captured learning curves for acquisition and extinction of associative learning.

Question 2

what is rescorla and Wagner’s model?

Answer 2

This is the Rescorla-Wagner equation. It specifies that the amount of learning (the change ∆ in the predictive value of a stimulus V) depends on the amount of surprise (the dif- ference between what actually happens, λ, and what you expect, ΣV).

Question 3

Rescorla and Wagner equation

Answer 3

∆V = αβ(λ − ΣV)

Question 4

what is ∆V?

Answer 4

the change in associative strength on the trial you’re computing

Question 5

what is α?

Answer 5

A learning-rate parameter ‘alpha’

Question 6

what is β?

Answer 6

a learning rate parameter ‘beta’

Question 7

what is λ ?

Answer 7

the asymptote, lambda

Question 8

what is ΣV?

Answer 8

the total associative strength from previous trials, ‘sigma V’. it’s a running total of the associative strength and gets bigger during conditioning until it matches λ

Question 9

what can RMW explain?

Answer 9

• blocking
• overshadowing
• CS-UCS contingency effects

Question 10

what does RMW fail to explain?

Answer 10

• downshift unblocking
  one-trial overshadowing
• latent inhibition