ABSTRACT Video experiments and virtual experiments have proven effective in supporting online scientific inquiry. This study investigated optimal strategies for combining these two approaches, focusing on two critical multi‐representational learning factors based on the design, functions, tasks (DeFTs) framework: the learning sequence and the representational integration scaffold (RIS). A two‐level factorial quasi‐experiment was conducted to examine the main effects and interaction effects of these two factors on scientific inquiry. The participants included 149 seventh‐grade students randomly assigned to four conditions: video‐virtual without a scaffold ( n = 38), virtual‐video without a scaffold ( n = 38), video‐virtual with a scaffold ( n = 35), and virtual‐video with a scaffold ( n = 38). The results revealed a significant main effect of learning sequence: Compared with the “virtual‐video” sequence, the “video‐virtual” sequence led to superior accuracy of experimentation, conceptual understanding, and knowledge transfer, along with lower perceived difficulty. Furthermore, a significant interaction was found between the learning sequence and RIS: With respect to the RIS, the “video‐virtual” group outperformed the “virtual‐video” group across the accuracy of experimentation, conceptual understanding, and knowledge transfer; without the RIS, no sequence differences emerged. Lag sequential analysis further revealed distinct experimental behavioral patterns between the two scaffolded conditions. These findings provide actionable insights for designing and implementing effective online scientific inquiry learning with video and virtual experiments.
Chen et al. (2026) studied this question.