本研究以Spatial元宇宙平台為教學介入工具，探討其應用於大學日語自主學習之教學實踐歷程與學習成效，並從自主學習(Self-Directed Learning, SDL)與自我調節學習(Self-Regulated Learning, SRL) 理論觀點加以分析。隨著AI與元宇宙技術快速發展，沉浸式虛擬情境逐漸被視為促進學習動機與參與度的關鍵環境，但在外語教育情境中，特別是日語自主學習的脈絡下，其具體運作機制與教學意涵仍有待釐清。
本研究於111至113學年度間，將Spatial平台融入大學基礎日語與會話課程的期末自主學習活動，設計「旅遊照片分享」「我喜歡的歌曲」「偏鄉教材製作」等任務導向主題，引導學生在3D虛擬空間中以日語進行內容建構、文化詮釋與成果展示。研究採教學實踐導向之混合方法設計，研究對象為84位日語學習者，量化資料包含課後問卷與小組合作自評表，質性資料則來自學習作品、同儕互評、學生反思與教師觀察紀錄。
量化結果顯示，多數學生對元宇宙任務抱持正向態度，學習興趣、參與意願與團隊合作等向度平均分數皆高於中間值；語言評量亦呈現日語表達與文化理解之中高水準表現，同時學生自評其數位應用能力有所提升。質性分析進一步揭示，學生在目標設定、資料蒐集、時間管理、歷程監控與自我反思等面向展現具體SDL與SRL行為，但亦提及平台操作複雜、素材整合耗時與認知負荷偏高等問題。
研究結果指出，Spatial平台對日語自主學習的正向影響，並非僅源自技術的新穎性，而是建構於沉浸式情境、任務設計與教學鷹架三者的交互作用之上。 因此，在適當控制任務頻率與工作量並提供明確操作與反思支援的前提下，元宇宙學習環境可作為整合語言能力、文化理解與科技賦能之有效教學途徑。
This study examines the integration of the metaverse platform Spatial into university Japanese courses, with the aim of exploring how a 3D virtual environment can support self-directed learning (SDL) and self-regulated learning (SRL) in autonomous Japanese language projects. While recent developments in AI and metaverse technologies have highlighted the motivational potential of immersive learning environments, empirical research on their use in foreign language education—particularly in Japanese self-directed learning contexts—remains limited. This study therefore seeks to clarify how Spatial mediates learners’ task engagement, learning processes, and perceived outcomes when used as a core space for language and culture-oriented projects.
Over three academic years (2022–2024), Spatial was embedded into end-of-semester autonomous learning activities in basic Japanese and first-year conversation courses at a private university, involving task types such as travel photo narration, favorite song introduction, and teaching-material development for rural education. In these tasks, students were required to construct and present content in Japanese, design 3D exhibition spaces, and interact with peers within Spatial’s virtual environment. The study adopted a practice-oriented mixed-methods design and collected data from 84 university learners of Japanese, including post-activity Likert-scale questionnaires, group self-evaluation sheets, student artifacts in Spatial, peer feedback, reflective comments, and teacher observation notes.
Quantitative results indicate that learners reported high levels of interest, willingness to participate in similar activities, and positive evaluations of collaboration, with mean scores above the scale midpoint for motivation and engagement items. Teacher assessment further suggested that students reached an upper-intermediate level in Japanese accuracy, fluency, content organization, and cultural explanation in their project outputs, while learners also perceived noticeable growth in their digital skills. Qualitative analyses revealed rich evidence of SDL and SRL processes: students described setting personal goals, searching and selecting resources, planning and adjusting timelines, monitoring progress, and revising their work based on feedback. At the same time, many participants reported challenges related to platform complexity, time-consuming space construction, and divided attention between linguistic content and technical design, pointing to increased cognitive and technical load.
Overall, the findings suggest that the educational value of Spatial for self-directed Japanese learning lies not merely in its novelty, but in the interplay among immersive affordances, task design, and structured pedagogical scaffolding. When technological demands and workload are carefully calibrated and learners are provided with clear operational guidance and reflective supports, metaverse platforms such as Spatial can serve as effective environments for integrating language proficiency, cultural understanding, and digital literacy in higher education.