Show Less
Restricted access

Confirmation Learning Theory


Raymond E. Hartley and Carl D. Williams

This work presents a cognitive and deterministic theory of learning and a related theory of performance. The authors assume that the organism is a logical system whose behavior is governed by its anticipation of the future, that this anticipation, called the «expectancy,» is the basic learning event, that the expectancy arises within a quasi-linguistic representational system, that the organism's confidence in the expectancy changes during learning, and that the strength of the expectancy in real organisms is subjective probability. Carnap's lambda-system of inductive logic is used as an approximation of subjective probability. Formulas from this framework are applied to conditioning, partial reinforcement, and other experimental data.
Contents: A precise definition of the expectancy is given and a method of measuring its strength is described. These formulations achieve quantitative predictions of performance in conditioning, latent inhibition, latent extinction, sensory preconditioning, reversal learning, and partial reinforcement experiments.