The purpose of this study is to explore how university students perceive the possibilities and risks of metaverse schools in terms of learning opportunities for school non-attending children and countermeasures against cyberbullying. University students discussed the current situation of school non-attendance and existing cases of metaverse use in education, using the argument-mapping tool Personary to create discussion maps and submitting short written reports. The data were examined through qualitative content analysis, supplemented by co-occurrence network analysis with KH Coder 3 based on the Jaccard coefficient. The results indicate that students regard metaverse schools as a promising option for expanding learning opportunities and enhancing psychological safety for school non-attending children, while at the same time being highly aware of risks such as inequalities arising from devices, network access, and paid items, as well as the increase of new forms of cyberbullying. The findings also suggest that implementing public, tuition-free metaverse schools requires careful institutional design, including clear legal positioning, the allocation of teachers and support staff, and rules for logging and monitoring.

View full abstract

This study analyzed the effects of the spatial designs of metaverse campus in the virtual space on learning motivation for students. The results showed that the impression of the learning environment elements influenced the motivation to learn, the ease of continuing and the difficulty of starting. It was also evident that these impressions could lead to both positive and negative attitudes.

View full abstract

This study explores the potential of information and communication technology (ICT) and artificial intelligence (AI) to support “thinking and discussion-based” moral education in Japan by developing and implementing a metaverse–generative AI–integrated learning environment. We designed a virtual reality (VR) scenario in which an avatar portrays a child struggling with friendship issues, and elementary school students engage in spoken dialogue with the avatar. The system was implemented on Meta Quest devices; learners’ utterances were captured via the built-in microphone, transcribed using an automatic speech recognition model, and responded to by a generative AI model. The AI-generated replies were converted into speech and synchronized with the avatar’s lip movements and facial expressions, thereby enabling an immersive, quasi face-to-face conversational experience. The prototype was piloted in a local event in City A with participating elementary school students. From the collected interaction data, five dialogue logs were selected for qualitative analysis and coded into emergent categories. The results indicated that many students spontaneously verbalized concrete relational repair strategies such as “playing together” or “talking more” with peers. In some dialogues, nascent forms of reflection on self-trust, future orientation, and cause exploration were also observed, occasionally coupled with explicit emotional support for the avatar. At the same time, most utterances remained at the level of behavioral advice, suggesting the need for carefully designed prompts and teacher-led reflection to elicit value-laden reasoning and justification. These findings suggest that metaverse–AI environments can provide an experiential and dialogic space that fosters students’ engagement and empathic understanding, while potentially reducing teachers' preparation burden and reliance on instructional skills. However, the results also highlight the importance of ongoing scenario refinement and post-VR reflective activities in moral lessons.

View full abstract

This paper reexamines the ethical challenges of education in the age of AI and the metaverse from a civilizational perspective. It proposes a theoretical framework for reconstructing education as a “system for the regeneration of will,” grounded in the three principles of will, ownership, and justice. In contemporary society, technology increasingly substitutes for human judgment, leading to a progressive loss of autonomous will. In particular, transactions lacking genuine human intent—as typified by unauthorized access incidents—and the spread of AI-based automated decision-making have undermined the very structure of human responsibility. This situation is not merely a technical issue but a crisis that shakes the ethical foundations of modern civilization. Accordingly, this study asks how education should respond to this civilizational pathology of “the absence of will.” In conclusion, it argues that metaverse-based education must be redesigned not merely as a virtual environment but as an ethical and institutional space for restoring human free will.

View full abstract

The rapid diffusion of generative AI has increased the importance of developing the ability to use AI to solve social problems. However, there is a significant gap in higher education between introductory, concept-oriented AI materials and advanced, programming-based curricula. This makes it challenging for humanities students to grasp and apply AI in meaningful contexts. This study aims to develop and implement an AI literacy education program for humanities undergraduates based on Goal-Based Scenario (GBS) theory to examine its educational effectiveness. The program is grounded in the GBS principle of goal-oriented task performance and consists of two components: a　generative AI–based idea-support exercise for first-year students and a text-mining analysis exercise for second-year students. Questionnaire surveys using a five-point scale and paired t-tests revealed significant improvements in perceived cognitive support, concentration during AI use, and motivation to enroll in AI-related courses. Effect sizes (Cohen’s d) indicated small to moderate educational effects. Qualitative analysis of open-ended responses showed that students developed a deeper understanding of the usefulness and limitations of AI and an enhanced ability to critically interpret AI outputs. These findings suggest that instructional design based on GBS theory is an　effective approach for fostering AI literacy among humanities students.

View full abstract

With the growing importance of keyboard typing in today’s information society and the spread of tablets in schools, accurate Romanized input has become indispensable. In the 2025 revision of the Course of Study, programming skills in “Information I” will be assessed in the Common Test for University Admissions, underscoring the need for precise input as a foundation for learning. This study examined how Romanization patterns, input methods, and accuracy influence typing speed among 272 second-year high school students in Saitama Prefecture. Using input analysis software, students’ performances were recorded and analyzed. Results revealed variation in input methods, particularly in voiced sounds, contracted sounds, and Romanization forms. A significant positive correlation was observed between accuracy and typing speed, and regression analysis confirmed that accuracy strongly affects speed. These findings suggest that mastering accurate Romanized input enhances typing efficiency and contributes to programming literacy. Moreover, classifying learners by input characteristics provides implications for differentiated instruction, emphasizing the need to address diverse input strategies to strengthen the foundation of digital education.

View full abstract

This qualitative case study examined a remote physical education lesson conducted through a shared metaverse environment connecting two geographically distant schools. The purpose of this study was to explore how an inclusive learning environment was constructed and how students’ participation was sustained in a school-to school context mediated by immersive digital technology. Data were collected through lesson video recordings, field notes, students’ verbal statements, and teacher interviews. These data were analyzed qualitatively to examine instruc tional design, patterns of participation, and students’ engagement within the virtual environment. The analysis fo cused on observable actions and recorded statements rather than inferred internal states. The findings indicate that a clear division of instructional roles among teachers supported smooth lesson progression while enabling diverse learners to participate without being separated from the group. One teacher primarily supported students’ physical engagement in the classroom, while the other monitored the virtual space and provided guidance based on avatar representations. In addition, avatar-based representation enabled students from different schools to engage in　physical activity as a single learning community. Physical differences among students were not visually emphasized, and wheelchair users participated in the same activity as their peers, with engagement observed through both behavior and verbal expressions. These results suggest that immersive digital technologies can function as mediational environments that restructure participation in physical education rather than merely compensating for　physical differences. The study highlights the potential of shared metaverse environments to support inclusive and embodied learning experiences across schools.

View full abstract