Modelling the Category-Order Effect with an Oscillator-Based Connectionist Network

Abstract

Numerous factors interact to affect a participant's ability to encode and recall information. One example of this interaction is known as the category-order effect (COE; Brooks and Watkins, 1990). The present study models earlier work performed by the authors (Schoenherr & Thompson, 2008) with an oscillator-based model of memory (Brown et al., 2000). The OSCillator-based Associative Recall (OSCAR) network developed by Brown et al.(2000) was adapted to examine the role that attention plays in the COE. A series of simulations demonstrate that both the differential allocation of attention to items, as well as the strength of items stored in memory, independently contribute to the COE. Further lines of experimental inquiry are also discussed.