Abstract:

Young children are known to learn less effectively from video than from real-world interactions, a phenomenon referred to as the video deficit. Co-viewing with a social partner has been shown to mitigate this effect, and robots have been explored as potential co-learners. However, it remains unclear whether the physical presence of a robot enhances children's engagement in foreign language video learning, or whether on-screen presentation of a robot model may be equally or more effective. This study examined whether a physically present robot could serve as a learning rival that increases preschoolers' engagement in foreign language learning from video, as measured by their vocal responses to target vocabulary, compared with an on-screen robot model.

Thirty Japanese children (mean age: 3.54 years) were randomly assigned to either a robot-observation group (n = 15) or a video-viewing group (n = 15). Both groups watched a Chinese-language learning video twice. In a co-learner modeling paradigm inspired by the Model/Rival method, a social robot demonstrated target Chinese vocabulary by repeating each word from the video. In the robot-observation group, the robot was physically present beside the child and its voice was produced from the robot itself. In the video-viewing group, the same robot demonstration was presented on screen, with both the instructional video and the robot's voice output from the same monitor. Changes in children's Chinese word repetition attempts between the first and second video viewings were compared across groups as an index of learning engagement.

The video-viewing group showed a significantly greater increase in Chinese word repetition attempts than the robot-observation group (z = -2.43, p = 0.008). No significant group differences were found for total utterance counts or combined Chinese and Japanese word production. Notably, the majority of children (23 out of 30) produced no utterances during the first viewing, indicating that the task was challenging for this age group.

These results suggest that the physical presence of a robot co-learner may not enhance preschoolers' engagement in foreign language learning from video under the present conditions. The robot-observation group showed lower increases in vocal responses relative to on-screen presentation, and several factors may have contributed to this pattern. The brief familiarization period (only a few minutes of one-way introduction by the robot) may not have been sufficient for children to become comfortable with the robot, possibly limiting its effectiveness as a co-learner. Additionally, in the robot-observation group, the instructional video and the robot's voice came from spatially separate sources, potentially creating a split-attention effect that made it harder for children to associate the model utterances with the learning content. Under these conditions, the robot did not appear to function as a learning rival that stimulated children's engagement. These findings point to the importance of considering not only the presence of a robot but also the degree of children's familiarization with the robot and the spatial arrangement of audio-visual cues when designing robot-assisted language learning environments for young children.

Keywords:

Social robots, Foreign language learning, Video-based learning, Preschool children.