A New Approach To Vocabulary Assessment: A Matrix Model For Predicting Future Vocabulary Performance

Abstract

This paper introduces a new approach to vocabulary assessment. We normally test learners’ performance to show what they have learned, but what if we could test learners to make predictions of their future performance? Teachers might be able to provide early intervention or affirmation to the learners to aid their learning. First, the paper introduces a model (Meara, 1990) which makes such predictions possible and explains how it works. Second, the paper reports on two studies conducted to validate the model and the accuracy of predictions. One study deals with retaining vocabulary knowledge and the other with learning new vocabulary. The first study was conducted with 43 EFL learners in Japan. They had already learned 200 new words. They were tested to see how well they could retain the words. The learners’ performance was compared with the predictions. The second study was conducted with 27 EFL learners in Japan. They learned 200 new words over 2 months, 20 words per week for 10 weeks. Predictions were made on the process of learning the new words. The results of the studies indicated that the model made reasonably accurate predictions of learners’ future performance both for retention and for learning.