Abstract
Stochostic learning models for dichotomous response trials and a finite number of absorbing states. A probabilistic model for a learning process has been built up in the context of the state models for the learning mathematical theor y. Two transitory states (for the acquisition and forgetting processes) and a finite number of absorbing states are considered. The transition probabilities are assumed to be controlled by the "response strength" of the Luce's β model, derived from his choice axiom.