Abstract:

In recent years, there has been increasing demand for classroom practices that foster self-directed, interactive, and deep learning. Among these approaches, role-play–based learning has gained notable attention, as it allows learners to explore alternative viewpoints and problem-solving strategies by engaging in specific roles or value orientations. However, when there is a substantial mismatch between a learner’s personality traits and the role assigned to them, performing the role can become psychologically burdensome. This misalignment often discourages learners from constructing arguments from the role’s standpoint, thereby limiting the effectiveness of collaborative learning.

To address this issue, this study proposes a metaverse-based role-play–based learning environment that enables learners to immerse themselves in assigned roles while maintaining an appropriate psychological distance from their real identities. Within the proposed environment, learners communicate through avatars and engage in discussions on task themes while portraying diverse personas, such as highly proactive or assertive characters. During role assignment, the system automatically provides avatars whose visual attributes and role-related guidance correspond to the characteristics of the assigned role. By perceiving the avatar’s characteristics as extensions of themselves, learners are expected to exhibit behaviors and cognitions aligned with the role. Furthermore, interacting through a “virtual self” distinct from the real self is anticipated to reduce the psychological burden associated with speaking or expressing opinions.

To facilitate these role experiences, the system incorporates a mechanism that dynamically reconfigures the metaverse learning environment according to changes in assigned roles. After each discussion activity, a structured reflection phase allows learners to exchange feedback on their statements and discussion processes, thereby deepening their understanding of their own communication tendencies. This integrated design—from role-based discussion to reflection—encourages flexible engagement with diverse roles and fosters metacognitive awareness of communicative behavior.

We implemented a metaverse environment for role-playing. First, we created objects present in a real-world classroom using Blender, a 3D CG software. Next, we built a learning space using Unity (2021.3.12f1), a game engine serving as the environment for constructing the metaverse system, and set up avatar models capable of freely moving within that space. This avatar allows not only viewpoint movement and object selection via controllers, but also enables users to freely change models according to their role and check behavioral conditions and speech guidelines. Additionally, we built features to support efficient problem-solving, such as the ability to record speech content spatially and organize opinions through drawing on a board. Furthermore, to manage the entire system, we implemented Photon Unity Networking 2 (PUN2), a dedicated Unity network framework. PUN2's matchmaking and object synchronization capabilities enabled real-time communication between users and synchronized objects within the metaverse environment.

Overall, this research presents a new framework that leverages the affordances of the metaverse to mitigate psychological burdens inherent in traditional role-play–based learning, thus promoting more immersive and proactive communication learning.

Keywords:

Role-playing, metaverse, virtual self.