Investigating the Impact of Learning-by-Teaching a Social Robot on Students with Different Prior Knowledge Levels.

Social robots are increasingly recognised as promising assistive technologies for enhancing engagement and learning in educational settings. This study investigates the use of a learning-by-teaching (LbT) paradigm, in which primary school children tutor a small social robot during a vocabulary learning task. Students from the same year group, but with varying levels of prior knowledge, were randomly assigned to either an LbT condition or a self-practice condition using a tablet. Results indicate that children with lower baseline knowledge spent more time engaging with the material and demonstrated higher post-test and retention gains in the LbT condition. While these differences did not reach statistical significance, they suggest that the LbT approach may foster deeper, self-reflective learning, particularly among students who typically struggle with independent practice. These findings support the potential of social robots as inclusive learning companions, and motivate future work to refine interactive teaching strategies and assess their long-term impact on students with learning difficulties.