Experiencing Flow in Virtual Reality: An investigation of complex interaction structures of learning-related variables

Theoretical background: Learning in virtual realities (VR) has become increasingly important. In this context, VR appears to be particularly conducive to affective learning objectives, such as perspective-taking in VR actors. Previous research has often focused on investigating cause-effect relationships that focus on the direct effects of different VR visualization technologies on learning outcomes. Little research has been conducted on more complex constellations of learning-related variables. Therefore, the present study aims at elaborating a research design that can be used to investigate both the direct effects of VR visualization technologies as well as the mediating effects of learning process variables latent in VR. For this purpose, the research design will be experimentally tested by comparing head-mounted display (HMD)-based VR with laptop-based VR with respect to the intended learning objectives, and controlling for influences by underlying learning processes (here: the experience of flow within VR). Methods: 132 students of grades eight and nine were recruited. The subjects were assigned to experimental conditions (HMD vs. laptop). The VR content dealt with the exploration of the hiding place of Anne Frank at the time of World War II in Amsterdam. Questionnaires were used to collect several data including sociodemographic characteristics, knowledge gain, perspective-taking, subjective evaluation, and the learning process variable flow. Results: A significant main effect for the variable VR visualization technology was uncovered averaged across all learning indicators. However, a superiority of HMD-based VR was found for the two evaluative indicators and also for the affective learning indicator. For the cognitive learning indicator, the effect was reverse. More relevant than unidirectional relationships are the mediating effects. Mediating effects through the experience of flow were discovered several times. Thus, the experience of flow within VR can significantly explain the cause-effect relationships between VR visualization technology and learning outcomes, even if, for the most part, only effects for evaluative indicators could be determined. Conclusion: The present study was able to exemplify that the investigation of complex interaction structures of VR visualization technologies and learning process variables can make a large contribution to the understanding of learning in VR environments. Advantages of HMD-based VR over laptop-based VR with respect to some learning indicators were uncovered. The significant findings of the mediation analyses point to the fact that the direct effects of VR visualization technologies on parameters of learning can be significantly explained by learning process variables such asflow experience and are systematically overestimated if such learning processes are not taken into account.