Virtual Educational Agents in Immersive Virtual Reality Learning Environments: A Scoping Review of Terminology, Conceptualizations, and Taxonomy Requirements

The growth of immersive virtual reality learning environments (IVRLEs) coincides with the emergence of numerous virtual educational agents that offer support to their users. Researchers suggested various terms to describe these agents; yet, this terminology fragmentation creates conceptual ambiguity, rendering them hard to compare. The purpose of this scoping review was to map out the terms used to describe educational agents in IVRLEs and look at how they have been conceptualized, the characteristics that have been used to differentiate them, and the educational roles and associated learning functions they have. In addition, the goal was to identify the requirements for the development of a relevant taxonomy. Arksey’s and O’Malley’s methodological framework and the PRISMA Extension for Scoping Reviews were applied. Out of the 1867 articles found in the Scopus, Eric, and LearnTechLib databases, published between 2015 and 2026, 129 met the eligibility criteria. The findings revealed substantial conceptual fragmentation, with similar educational entities being described using different terminology while identical terms were frequently applied to conceptually distinct systems. Characteristics such as intelligence, embodiment, interaction modality, adaptivity, affective and social characteristics, and context awareness distinguished these agents rather than terminology alone. Advances in artificial intelligence have further blurred the boundaries of traditional categories. Overall, this review provides a theoretical foundation for future research aimed at developing more coherent conceptual frameworks and standardized approaches to the design, classification, and evaluation of virtual educational agents.