Mathematical Models for Verbal Learning

Abstract

Publisher Summary The use of language is perhaps the most distinctive feature of human behavior. The laboratory study of how verbal associations are formed, in addition to having practical implications, has played an important role in testing theoretical ideas about the nature of the learning process. It is this last matter that will chiefly concern this chapter, and it concentrates on a particular kind of verbal learning problem known as paired associate learning. A number of alternative models are proposed to account for the learning of simple verbal associations. In the linear model, which is considered first, it is assumed that learning of an association proceeds gradually in an incremental fashion. The all-or-none model presented next assumes that learning takes place on a single trial; prior to that trial, the subject responds at random. The third model presented (the TDF model) postulates all-or-none acquisition and interference-produced forgetting. The models are applied to an experiment in which subjects were required to learn lists of 9, 15, or 21 stimulus-response pairs. The effect of increasing the number of pairs is to increase the number of errors in a way that is well accounted for by the TDF model.