The Rescorla-Wagner model
DVA = aAb (l – VA)
VA = magnitude of the response on trial
A
DVA = change
in the VA (amt. of learning)
l = maximum
response possible
(l – VA) = difference between maximum and current resp.
aAb =
constant, based on intensity of CS and UCS
note that VA for a trial
will be the VA for the previous trial, plus DVA
assume l = 100 and aAb = .5
|
Trial |
VA |
(l – VA) |
DVA |
|
1 |
0 |
100 |
50 |
|
2 |
50 |
50 |
25 |
|
3 |
75 |
25 |
12.5 |
|
4 |
87.5 |
12.5 |
6.25 |
|
… |
… |
… |
… |
|
N |
100 |
0 |
0 |