Use of an expert system to predict language learning success

Abstract

The major purpose of this study was to examine a number of language learning (LL) background variables in relationship to rate of progress through an intensive English program (IEP). In the first stage of the study, a knowledge bank of 40 background variables for 201 subjects was refined to those variables which most clearly showed “maximum interclass difference” when comparing three categories of success. In the second stage of the study, the extent to which success category could be accurately predicted by an expert system was measured, basing prediction on (a) all 40 variables, (b) learner-reported language learning background variables, and (c) IEP entry proficiency measurements. Results indicated that entry proficiency variables were most effective when predicting high and low success category, but language learning background variables were most effective when predicting medium success. A brief introduction to knowledge base expert systems is provided as part of the background to the study.