Immersive TOEFL Preparation in the Metaverse: Usability and Navigability of a Roblox-Based Game Developed via GDLC

Abstract

Despite the growing adoption of game-based learning in language education, three-dimensional metaverse platforms for TOEFL preparation remain critically underexplored, leaving learners reliant on drill-based 2D media that lack immersion and sustained engagement. This study addresses that gap by developing a Roblox-based TOEFL learning game using the Game Development Life Cycle (GDLC) and evaluating user understanding of its game flow through graduated formative evaluation. An R&D design was employed, implementing six GDLC stages alongside Tessmer's formative evaluation: one-on-one (n=3), small group (n=5), and field testing (n=40). Data were gathered via observation, interview, and Likert-scale questionnaires, and analyzed descriptively. The game was realized as an area-based environment with three thematic zones and a multi-level navigation system. Evaluations showed progressive quality improvement: one-on-one (M=3.25), small group (M=4.00), and field testing (M=3.84, 96.3% positive response), indicating satisfactory usability and navigability. However, awareness of the Challenge Room and return-route comprehension remain areas requiring refinement. Theoretically, this study demonstrates that GDLC paired with formative evaluation provides a structured and iterative framework for validating metaverse-based educational games. Practically, the findings offer actionable design principles for educators and developers building immersive, game-based language learning environments within social 3D platforms