Abstract:
As the integration of Generative AI (GenAI) into education gains prominence, it also raises critical concerns. Among the most significant challenges is the risk of learners becoming “metacognitively lazy,” which can lead to low AI literacy and an over-reliance on AI tools. To address this issue, fostering students’ metacognitive abilities is essential, as it supports AI literacy and promotes the responsible use of GenAI. Students with strong metacognitive skills can purposefully employ AI, critically assess its outputs, and align its use with their learning objectives while avoiding over-reliance or unethical practices such as plagiarism.
This project introduces an innovative pedagogical approach called student-centered GenAI-supported metacognitive learning (SCAIML), designed to enhance students’ metacognitive abilities in the context of GenAI. The SCAIML approach empowers students to collaboratively create, evaluate, and refine GenAI agents as metacognitive learning supports, while learning from these agents to enhance their metacognitive awareness, knowledge, and skills. By emphasizing student autonomy, this approach enables learners to shift from passive consumers to active participants in GenAI-based learning environments.
The SCAIML approach will be implemented in Bachelor and Master level courses at the Education University of Hong Kong, targeting approximately 600 students. A quasi-experimental design will be employed, with 300 students will be assigned in the experimental group using the SCAIML approach over six weeks, while the control group uses traditional metacognitive supports (e.g., metacognitive self-check tables). Mixed methods will be used to evaluate the impact of the approach. Pre- and post- surveys will measure students’ metacognition and AI literacy. Since SCAIML adopts the collaborative problem-solving method, two Generic Intended Learning Outcomes (GILOs) (i.e., problem-solving and social interaction skills), will also be included as outcome variables in the surveys to examine the impact of SCAIML. Additionally, individual interviews with students and teachers will provide deeper insights into their perceptions, experiences, and feedback on the SCAIML approach.
Tangible outcomes of the project include evidence of the SCAIML approach’s impacts on metacognitive abilities, AI literacy, and the two GILOs. Additional deliverables include a video-based tutorial, a guidebook for implementing SCAIML in authentic classrooms, rubrics for evaluating the functionality of GenAI agents, and showcases of student-designed GenAI metacognitive agents. These outcomes will be synthesized into a comprehensive framework for the long-term implementation of SCAIML. By empowering students as AI innovators, this initiative cultivates metacognitively aware learners, equipping them to thrive in an AI-driven future. The project’s long-term impact extends to other universities, fostering sustainable GenAI education.
Code:
T0303
Principal Project Supervisors:
Keywords provided by authors:
- Metacognitive abilities
- GenAI integration
- AI literacy
- Student-centered design
- Collaborative problem-solving
Start Date:
01 Feb 2026
End Date:
13 Jul 2027
Status:
Ongoing
Result:
The project will generate diverse outcomes.
- A 3-hour video-based tutorial will be developed for students’ self-directed learning, consisting of three modules: an introduction to GenAI agents, the creation of personalized GenAI agents. The tutorial will offer students an engaging and comprehensive learning experience.
- Second, open education resources will be developed, that includes a guidebook offering step-by-step instructions on how to implement the SCAIML approach in authentic classrooms, rubrics for evaluating the functions of the GenAI agents and showcases of student designed GenAI metacognitive agents. These resources will be uploaded to Moodle and available for teachers and students who show interest in the SCAIML approach and want to implement the approach in their teaching and learning.
- Moreover, the project team will also disseminate the project outcomes through conferences, workshops, or journal publications, which will extend our scholarship of teaching in and beyond EdUHK.
Impact:
The project will generate diverse outcomes.
Impact. The direct impact of this project is the enhancement of students’ metacognitive abilities, which is the primary goal of the project. Metacognitive abilities play a pivotal role in the responsible use of GenAI tools for learning. However, research points out that learners often do not spontaneously apply such metacognitive skills during learning, leading to lower learning performance (Bannert et al., 2015; Fan et al., 2025). Thus, support of students’ metacognitive abilities by adequate metacognitive support is always in demand, especially in the context of GenAI. The project will foster students’ metacognitive abilities through the SCAIML approach in higher education against the background of the widespread adoption of GenAI. The GenAI metacognitive agents can prompt students to engage in metacognitive activities while using GenAI in learning, which can produce short- and long-term impact with respect to metacognitive abilities in learning. Additionally, through this project, students will acquire skills to collaboratively design, evaluate, and refine GenAI agents. This process not only enhances their technical understanding and hands-on experience with GenAI but also significantly contributes to their AI literacy by helping them critically assess and improve the functionality and responsive use of AI tools. Furthermore, the collaborative problem-solving nature of this work fosters problem-solving skills as students collectively address challenges in agent development, troubleshooting, and optimization. At the same time, it strengthens their social interaction skills by encouraging teamwork, communication, and the exchange of diverse perspectives. By engaging in these collaborative activities, students gain not only the technical expertise but also the interpersonal and cognitive skills necessary to thrive in AI-integrated learning environments and professional settings. The project’s impact will extend beyond students. For teachers, the SCAIML approach provides an innovative pedagogical framework that equips them to integrate GenAI into their teaching more effectively. By using GenAI agents as tools for metacognitive support, teachers can foster deeper student engagement, promote active learning, and tailor instructional strategies to better meet the diverse needs of students. This not only enhances teaching effectiveness but also encourages teachers to embrace emerging technologies as allies in education rather than mere tools for automation. At the level of higher education, this project can offer a scalable and evidence-based framework for integrating GenAI into curricula while addressing critical challenges such as AI literacy and metacognitive ability development. By embedding the SCAIML approach into existing programs, institutions can position themselves as leaders in the adoption of GenAI for pedagogical innovation. The inclusion of open educational resources will enable widespread adoption and adaptation of the SCAIML approach across institutions, fostering a culture of continuous improvement and collaboration in educational technology. Overall, the project contributes to the sustainable integration of AI into education by promoting responsible use of GenAI tools. By equipping students with metacognitive abilities, it prepares them to navigate the complexities of an AI-driven world and encourages them to become innovators and critical thinkers. Moreover, the project highlights the importance of aligning AI integration with educational outcomes, paving the way for policies and practices that prioritize both technological advancement and the holistic development of learners. This approach has the potential to influence educational systems globally, creating a ripple effect that transforms how AI is utilized in classrooms, universities, and lifelong learning contexts. |
Financial Year:
2025-26
Type:
TDG