Adaptive Immersive Virtual Reality, Cognitive Load, Cybersickness, and Diagnostic Performance in Automotive Vocational Education: A PLS-SEM Multi-Group Study

This study examined the relationships among Adaptive Immersive Virtual Reality (IVR) Experience, Cognitive Load, Cybersickness, Learning Engagement, and Diagnostic Performance in automotive vocational education. A quasi-experimental explanatory study with a comparative multi-group design was conducted with 134 undergraduate students enrolled in the Automotive Engineering Education Program at Semarang State University, Indonesia. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) and Multi-Group Analysis (MGA). The results showed negative associations between Adaptive IVR Experience and Cognitive Load (β = −0.647) and between Adaptive IVR Experience and Cybersickness (β = −0.604). Learning Engagement was positively associated with Diagnostic Performance, while Cognitive Load and Cybersickness showed negative associations with learning-related outcomes. The indirect-effect analysis indicated significant indirect associations involving Learning Engagement, whereas the MGA identified significant differences in several structural relationships between the adaptive and standard IVR conditions, including the Cognitive Load–Diagnostic Performance relationship (Δβ = 0.165, p = .031) and the Cybersickness–Diagnostic Performance relationship (Δβ = 0.150, p = .042). These findings indicate that cognitive, physiological, and engagement-related factors are associated with diagnostic performance in immersive vocational learning, while the strength of these relationships differs across instructional conditions. The findings provide practical considerations for designing adaptive immersive learning environments that address cognitive demands, physiological comfort, and learner engagement in vocational education.