A DISTINCTIVENESS MODEL OF SERIAL-LEARNING

An associative model of serial learning is described. The model is based on the assumption that the effective stimulus for a serial-list item is generated by adaptation-level coding of the item's ordinal position. The psychophysical concept of distinctiveness is used to quantify the strength of intralist competing associations. Illustrations are given on how the model may be used to generate predictions concerning the fundamental aspects of serial learning data: difficulty as a function of list length, performance as a function of amount of training, serial-position effects at various stages of learning, and the shape of the gradient of remote intrusions as a function of acquistion level.