Edward Crothers

A comparison of paired-associate learning models having different acquisition and retention axioms ☆

Abstract Several alternative interpretations of all-or-none processes for paired-associate learning and concept formation are examined. These models, along with three linear models, are applied to data from eight paired-associate learning experiments. The principal analyses involve goodness-of-fit tests for observed response sequences and conditional probabilities. The results favor a three-process model that postulates a distinction between long-term and short-term retention and allows for forgetting between successive presentations of the same stimulus item.

Learning model solution to a problem in constrained optimization

Abstract Two modes of presentation, A and X , are assumed to be available for teaching a concept or association. Learning models are applied to answer the question: give that i trials have been allotted a priori to the first mode and j trials for the second mode, what sequencing of the A and X trials will yield the maximum proportion correct on a test after Trial i + j ? The schedules are rank-ordered on the basis of predicted proportion correct on the test, and these rankings sharply contrast the long-short model with the random-trial increment, two-operator linear, and one-element models.