Redesigning Heat Transfer Learning for Diverse Jakarta Elementary Students: A Cognitive Load–Optimized Virtual Lab (CLOVL)

Abstract

Elementary schools in Jakarta face gaps in science learning resources, resulting in uneven and often theoretical practical experiences for students. Virtual laboratories offer a pathway toward equity, yet poorly structured designs can trigger cognitive overload and undermine learning. This study examines how a virtual lab optimized with cognitive load theory—the Cognitive Load–Optimized Virtual Lab (CLOVL)—influences fifth graders’ conceptual understanding of heat transfer in six elementary schools (three central and three suburban) in Jakarta. Using an explanatory sequential mixed-methods design, we involved 120 students and 12 science teachers. The experimental group used CLOVL, which incorporated five-minute segments, pre-training of key vocabulary, narrative–visual modalities, and personalization through culturally Indonesian avatars. The control group used a conventional virtual lab. Quantitative data consisted of pre- and post-tests, an adapted cognitive load scale, and eye-tracking for a sub-sample (n = 40). Qualitative data were gathered from teacher interviews, classroom observations, and student focus groups. Results indicated higher conceptual understanding in the CLOVL group (normalized gain ⟨g⟩ = 0.52) compared to the control group (⟨g⟩ = 0.23), with a large standardized effect size (Hedges’ g ≈ 1.20), particularly in suburban schools, where gains were approximately 45% greater than in central schools. Integrative analyses showed shorter fixation durations and a higher proportion of gaze directed toward relevant areas. These findings support the application of cognitive load principles and underscore the role of cultural relevance in virtual lab design for urban Indonesian settings, while highlighting the theoretical, practical, and policy implications for further research.