This study examines the effects of six augmented‑reality (AR)-supported lessons in lower‑secondary natural science classes in Vietnam on students’ conceptual understanding, engagement, and motivation. We used a quasi‑experimental, non-equivalent control group design in two semi‑rural public schools (grades 6–8; eight weeks). Intervention classes completed weekly AR‑supported activities aligned with the 2018 curriculum; comparison classes received regular instruction. Pre-test–post-test assessments and ANCOVA estimated groupdifferences, and a structural model related AR design features to engagement and learning. Interviews and focus groups with students and teachers contextualized mechanisms of change. Intervention classes scored higher on post‑tests than comparison classes (p < 0.001), with a large effect size at the class level; paths from engagement and motivation to learning were moderate to strong, suggesting that AR contributes primarily by sustaining attention and pacing information. Qualitative themes converged on three design levers – signaling,segmenting, and guided manipulability – that helped students construct more elaborated explanations when teachers orchestrated brief pauses and role rotation. These inferences are bounded by the study context (two semi‑rural schools, limited devices, and short teacher preparation). We provide practical guidance for lesson design and professional development and specify conditions under which AR is likely to be educationally useful in comparable settings.
Nguyen et al. (Tue,) studied this question.