Prompt A: Aliens!
You have been captured by aliens for research and are the subject of a category learning experiment with stimuli (@, #, $, %) matched to categories such as @# = Farnut, $# = wahool, $% = karrd, and @% = taroom. You successfully learn all the categories. To avoid further experimentation, you decide to speak out and educate your abductors on theories of categorization. Evaluate which theories can, and cannot, explain your performance. The theories you should be evaluating are the Rescola-Wagner theory, feature theory, prototype theory, and exemplar theory.
Prompt B: Matching-To-Sample
Consider the table below. Pigeons are trained in a symbolic matching-to-sample task in which many different sample stimuli are mapped to two comparisons: given Red (R) or Yellow (Y), pick (→) Vertical (V); given Green (G) or Blue (B), pick Horizontal (H). Then one group of pigeons is trained with everything reversed: given Red or Yellow, pick Horizontal; given Green or Blue, pick Vertical. The other group is given a partial reversal: given Red or Green, pick Horizontal; given Yellow or Blue, pick Vertical. Given what you know about "common coding" or mediated generalization, which reversal should be learned more easily? Why?
|
Training
|
Reversal
|
|
All groups
|
|
|
R→V
Y→V
G→H
B→H
|
Whole
|
|
R→H
Y→H
G→V
B→V
|
|
Partial
|
|
R→H
Y→V
G→H
B→V
|