Comparative evaluation of medium-term retention effects by English speaking training based on QA and repeating tasks using robot-assisted language learning system

Abstract

To help learners of English develop oral skills, we have been developing a joining-in-type robot-assisted language learning (JIT-RALL) system in which two humanoid robots facilitate a learner to speak with them. The system trains the learner to use a focused form of English by asking questions after playing a model question-and-answer session using the robots. The system has asked learners to respond with their own answer basically, but has asked some of them to repeat the answer that a robot has made just before the learner’s turn if they cannot create their answer. To verify the learning and retention effects of the two training methods based on the question-and-answer task and repeating task, we conducted a comparative learning experiment that consists of a 5-day training period and three retention tests after 10 days, 30 days, and 100 days with 47 Japanese university students. The experimental results showed that the retention rate of the focused form was significantly higher for learners who received the question-and-answer-based training than for the learners who received the repeating-based training.