Are You Paying Attention? The Effect of Embodied Interaction with an Adaptive Robot Tutor on User Engagement and Learning Performance.

Robots are becoming increasingly popular as a teaching aid in language learning. For language learning, which relies on inter-personal interactions and references to the physical world, an agent’s embodiment and ability to adapt to the student are both important factors. In this study, adaptive behavior and embodiment were combined in robot-assisted language learning. An online brain-computer interface (BCI) was used to monitor student’s brain activity and prompt adaptive responses from the robot whenever a lapse of attention was detected. The response involved additional repetition of the latest word and an iconic gesture to illustrate its meaning. To isolate the effect of embodied interaction in such a system, participants completed learning tasks in two conditions: one where the adaptive robot was physically present, and another where videos of the robot appeared on a screen. Despite no changes in robot’s behavior, participants reported higher engagement and more positive impressions of the robot, and also showed increased learning outcomes during the embodied interaction. This study confirms the importance of embodied interaction during adaptive learning and highlights the effectiveness of BCI systems in the design of future pedagogical robots.