ベイジアンネットワークモデルを用いた英語の比較変化の習得に関する検討

概要

論文の詳細を見る
The purpose of this study is to examine the states of the acquisition of comparison with using Bayesian network (BN) model. The model represents causal relationships among variables as a directed acyclic graph. We applied the model to the data on the quiz of comparison conducted for 8th grade Japanese students in the English class. The two BN models were examined; one is hypothesized that the students acquires both regular and irregular comparison simultaneously, the other is that the acquisition of regular comparison is preceded irregular one. The result indicated the latter model that represented the teacher's treatment fit the data better. The applicability of the BN model to the data on periodic examinations conducted to evaluate achievement is discussed. Furthermore, we emphasize the result implies that the data of the examinations can be utilized to assess learners' knowledge states diagnostically.
早稲田大学の論文
2003-09-30