In the realm of digital education, traditional virtual reality setups often confine students to basic interactions through devices like a mouse and handheld controllers, limiting immersive learning opportunities, particularly in disciplines requiring multi- modal knowledge such as hospitality and teacher training. Multi-modal knowledge encompasses physical actions, gestures, facial expressions, and other non-verbal communication. This study investigates a novel educational approach that synchronizes physical settings with digital reality scenarios, aided by Al Stereo Cameras (e.g., Apple Vision Pro or MetaQuest), to create a more interactive learning experience. A three-group quasi-experimental design involved 103 Taiwanese students practicing hospitality tasks, such as table service in a simulated Japanese restaurant. The groups included: Group A, handling physical utensils in a real-world-like setting; Group B, controlling virtual utensils with a mouse; and a control group, learning through conventional teaching. Findings reveal that Group A reported significantly better learning outcomes and reduced cognitive load compared to Groups B and C. The reduced cognitive load was attributed to the use of real physical utensils, which minimized the need for students to learn how to navigate virtual utensils using a mouse. The findings underscore the benefits of incorporating physical objects into VR environments, bridging the gap between traditional and digital educational modalities. This approach offers a practical framework for situated learning in educational settings, incorporating embodied and multi-modal learning, and suggesting its broad applicability in improving educational practices and outcomes in the digital age.

Educational robotics research endeavors to develop companion robots that can engage students in learning activities. However, existing educational robots are not always well-suited for use outside of the classroom and may struggle to sustain compelling interactions. To address this, this study proposes a new model for educational robots that leverages interdependence theory, which explains how mutual dependence can lead to lasting relationships. The pet-owner dynamic, in which the pet relies on the care and attention of the owner and provides emotional support in return, served as inspiration for this model. The proposed model features a digital pet-like robot that students must train and empower to perform a drama presenting their final learning results in front of the class. In an experiment with 60 students in a Japanese Hospitality course in Taiwan, the study found that long-term relationships between students and pet-like robots emerged, and these relationships significantly improved learning performance compared to ordinary robots. This study highlights the potential of integrating interdependence theory with advanced technologies such as robots, mobile devices, and virtual reality to enhance student learning and foster long-term relationships between students and robots.

Educational robots have shown promise in enhancing learning performance; but, many of these robots function solely as companions or tutors and rely on novelty to attract attention, making it challenging to maintain long-term relationships. Although pet robots have been utilized to create relationships with users, their potential in education remains underexplored. This study addresses this issue by designing a robot as a pet, applying the interdependence theory to establish lasting relationships and improve learning performance. The interdependence theory suggests that relationships between individuals can be strengthened through mutual dependence and the fulfillment of each other's needs. In this study, learners were prompted to engage in continuous care for their pet robots, receiving emotional reinforcement from them seamlessly across both virtual and tangible forms enabled by the ChatGPT API. To culminate their learning process, learners presented their ultimate learning outcomes alongside their pet robots, enacting situational dramas for their classmates. To evaluate the effectiveness of this approach, a quasi-experiment was conducted with 100 undergraduate learners enrolled in a Japanese for Hospitality and Tourism course in Taiwan. Results indicate that robots designed as pets, leveraging interdependence theory, significantly yield positive effects on learners, encompassing improved learning outcomes, extended interaction rates, and heightened satisfaction with their study journey compared to conventional robots. The study concludes by discussing research limitations and offering suggestions for further enhancements to this approach.

One of the ways of thinking about how to finish the learning task is through self-dialogue. Doing so enables learners to ask themselves what should they do and evaluate whether it is appropriate from another point of view. However, learners might have difficulties imagining the scenario and interacting with an imaginary person. They may need a scaffolding tool or a person as a partner to practice the self-dialogue. Besides, the ventriloquist phenomenon in education has been noted as beneficial for learning as it could make the learners understands the idea through dialogue. This study proposed a learning approach that allows the learners to use a robot as a ventriloquist doll for practicing self-dialogue and knowledge application to enhance their learning effectiveness. The experiment was conducted by participating 104 undergraduate learners who enrolled in a Japanese Hospitality Management course. The results showed that the approach used a ventriloquist robot to facilitate learning through self-dialogue exhibited significant effects on motivation, anxiety, and better learning outcomes. This study provides a novel area of insight into the use of robots in learning, by integrating digital technology with pedagogical approaches to create an immersive intelligent learning environment and offer a self-dialogue mechanism to foster collaborative learning tasks.

Student audiences in drama-based learning classrooms must concentrate on the learning content of drama performance. Teachers spend much effort on teaching and classroom management, but less on student audience participation. To address this issue, this study proposes a digital theater environment with a robot to explore how the robot can interact with students while performing drama-based activities in the digital theater to improve students' engagement and help them to learn. A quasi-experimental study was conducted and 50 third-grade students from a local junior high school were involved in this study for six weeks. The results show that the learning performance of the experimental group is significantly better than that of the control group. Moreover, the results of questionnaires reveal that the interactive robot can enhance students' interest in learning and help them put concentration more on learning in digital learning theater.