Support for self-regulated learning in immersive virtual reality – An experimental design for science learning

Past studies have found that immersive virtual reality (IVR) may increase students’ cognitive load, distract them from learning tasks, and hinder their knowledge acquisition. To address these challenges, this study explored the integration of self-regulated learning (SRL) strategies into IVR, and examined their impact on students’ cognitive load. An experimental design was employed, involving 105 students randomly assigned to either an SRL group (n = 52) or a non-SRL (NSRL) group (n = 53). A two-way mixed-design ANOVA was conducted to identify differences in the science learning outcomes of the two groups. In addition, partial least squares structural equation modeling (PLS-SEM) was applied to examine the structural relationships among affective factors in IVR environments, cognitive load, and students’ learning outcomes within both the SRL and NSRL conditions. The findings indicate that regardless of whether they received SRL support, students must bear an appropriate (germane) level of cognitive load to achieve comparable learning outcomes. Moreover, the sense of presence, control and active learning, and the components of cognitive load each exert distinct and sometimes inconsistent predictive effects on learning outcomes. These results underscore the necessity of designing balanced IVR learning environments that incorporate SRL elements to optimize cognitive load and enhance educational effectiveness. The study highlights practical implications for educators and instructional designers, such as providing explicit instruction in metacognitive strategies or embedding adaptive scaffolds and real-time cues. Consequently, a deeper exploration into how factors such as presence, control and active learning, and the structure of cognitive load influence learning in virtual environments is warranted.