Abstract:

Developing theoretically grounded and transferable approaches to computational thinking (CT) instruction in primary education remains a persistent challenge for educators worldwide. While educational robotics shows significant promise in supporting CT development by externalizing abstract programming concepts through tangible artifacts, the specific interactional mechanisms that drive conceptual growth in collaborative settings remain insufficiently theorized. This study addresses this gap by presenting a pedagogical model for cross-age peer-mediated robotics instruction, designed to enhance CT competencies such as problem decomposition, algorithmic reasoning, and debugging.

The research utilized a design-based research (DBR) methodology, involving two iterative design cycles to develop and refine the instructional model. The pedagogical framework was grounded in social interdependence theory, generative learning, and distributed cognition. Key refinements across the cycles included clarifying role scripts, introducing scaffolded explanation prompts, and implementing structured debugging routines. In the second cycle, fifth-grade students acted as tutors for fourth-grade tutees as they engaged in robotics challenges using the LEGO SPIKE Prime platform.

Quantitative and qualitative analyses were conducted to evaluate the model's effectiveness. Results indicated that both tutors and tutees demonstrated significant improvements in key CT skills, specifically sequencing, decomposition, and debugging. Notably, there were no significant differences in learning gains between the two roles after controlling for baseline performance, suggesting a symmetrical benefit to the peer-tutoring arrangement. Qualitative findings further revealed that the structured interdependence among peers, the requirement for generative explanations, and the artifact-mediated nature of robotics debugging were the primary drivers of these conceptual gains.

The study concludes by articulating four transferable design principles for structuring peer-mediated CT instruction in elementary classrooms. These principles offer a scalable solution for schools, particularly those lacking specialist computing teachers, by leveraging peer-to-peer cognitive support and distributed scaffolding. This research advances a theoretically integrated model that bridges the gap between collaborative learning theory and robotics-based computational thinking education.

Keywords:

Computational thinking, robotics, peer tutoring, design based research.