Teacher-led robot intervention in early primary school classrooms improves pupil and teacher outcomes

Programming is often taught through robots in early primary education to support young children's computational thinking (CT), but many teachers lack the confidence and training to use them effectively. This paper presents a school-based robot intervention for children aged 4–7 (n = 430) and their classroom teachers (n = 17), delivered under three conditions: Intervention (robot intervention only), Intervention+ (intervention plus teacher education), and Control (no intervention). The two intervention groups assessed whether teacher education, in addition to classroom robot experience, influenced pupils' prediction and debugging, transferable skills (programming transfer and picture sequencing), and teachers' beliefs (enjoyment, relevance, self-efficacy, anxiety). The intervention improved children's prediction and debugging scores significantly, but only Intervention+ significantly outperformed Control for both prediction and debugging. Performance on the programming transfer and picture sequencing tasks improved across all groups. Teachers in both intervention groups reported improved relevance beliefs, though only Intervention+ showed a significant difference from Control. Self-efficacy also improved significantly in Intervention+ only. These findings offer practical guidance for embedding programming with robots in primary education and underscore the importance of teacher education for significant impact.