SIAH Chiew Jiat, Rosalind*, LAU Siew Tiang, Lydia, KOH Siew Lin, Serena, Laura THAM-SCHMIDT, and Mary Jeanette IGNACIO
Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine (YLLSOM),
National University of Singapore (NUS)
Sub-Theme
Building Technological and Community Relationships
Keywords
LLMs, critical thinking, nursing undergraduates, GenAI, clinical reasoning
Category
Lightning Talks
Background
Cognitivism is a fundamental theoretical framework that highlights the importance of engaging the mind’s internal processes for meaningful learning and the development of critical thinking skills. However, in clinical settings, students often rely on search engines and AI for quick answers, risking fragmented understanding instead of comprehensive learning. This can weaken the learner’s relationship with authentic clinical environments, professional communities, and reflective practice, thereby preventing deeper cognitive analysis and potentially hindering students’ abilities to think critically and solve problems that ultimately affect quality of care. To mitigate this, we propose a technology-enhanced learning strategy that embeds critical thinking and reasoning frameworks into AI to encourage questioning, problem-solving, and real-world simulations, thereby strengthening relationships between learners, technology, and their practice environments in clinical settings.
Aim
Accordingly, the project aims to weave the Year 1–Year 3 curriculum into an AI platform that structures learner interactions with their environments, connecting educators, clinical placements, and professional communities by using Socratic methods to cultivate critical thinking and reasoning beyond the classroom and to align decision-making with shared clinical frameworks and community standards.
Methods
The project began with a systematic review and meta-analysis of 14 studies from six databases, revealing that AI tools often lack integration with learning theories, frameworks, and self-directed learning activities, and do not significantly improve understanding or critical thinking skills. Building on these insights, a mixed-method study with 204 second- and third-year nursing undergraduates assessed the impact of large language models on attitudes toward self-directed learning and metacognitive skills. The study found that AI tools enhanced learning experiences by providing personalised and interactive support, fostering metacognitive development. However, it emphasised the need for structured integration of large language models into the curriculum and into the broader learning ecosystem. This underscores the importance of developing a customised platform that is pedagogically grounded, co-designed with educators and students, and integrated within institutional infrastructure to strengthen learner environment relationships.
Proof of Concept
Guided by these findings, a prototype (Figure 1) of LLM platform was created and integrated into the curriculum to address the concern. Using the Socratic method, the platform emphasises active questioning and engagement, moving beyond memorisation to analytical thinking. It utilises frameworks like ADPIE (Assessment, Diagnosis, Planning, Intervention, Evaluation) and the Clinical Reasoning Cycle to help students apply knowledge to real-world scenarios (Figure 1).
.
The NUS AI Learning Buddy was trained through direct ingestion and fine-tuning of local nursing curriculum resources and evidence-based frameworks. Its prompts were deliberately structured to simulate Socratic questioning, fostering deep engagement, higher-order thinking, and application of critical frameworks like ADPIE and the Clinical Reasoning Cycle. This approach moves beyond rote memorisation, encouraging students to reflect, analyse, and justify their clinical decisions—preparing them for real-world nursing practice.
A preliminary study involving 142 participants (mean age=24.3, SD=5.085) indicates that users generally view the tool as intuitive and easy to adopt. Participants reported a mean score of 3.70 (SD=0.79) out of 4 in agreement that the AI Learning Buddy enhances their clinical learning experience, and a mean score of 3.73 (SD=0.74) that it complements their learning. The findings further underscore the tool’s perceived effectiveness in guiding users through ADPIE, a shared framework widely used in nursing and healthcare education to foster critical thinking and to align learning with community practice.
The findings also show that students primarily use the AI Learning Buddy to review clinical procedures/skills (highest), followed by reviewing diseases/conditions and supporting case study work, preparing for patient education, and answering queries during clinical placements (Figure 2). This usage pattern indicates the tool is engaged at authentic points of learning and care, functioning as a bridge between technology and the learner’s environments. It also connects classroom knowledge to ward-based practice, aligns student actions with shared clinical frameworks and preceptor expectations (procedures/skills), and strengthens ties to the community by helping students translate knowledge for patients and families (patient education). In doing so, the AI Learning Buddy builds technological and community relationships between learners and their environments by mediating meaningful interactions rather than merely supplying decontextualised answers.
.
Significance of the Project
The results highlight the AI Learning Buddy’s positive impact on student learning, with above-average scores in usability, feasibility, cognitive support, and clinical reasoning competency. Importantly, the system promotes critical thinking and reasoning by guiding students through structured decision-making processes that connect learners to authentic clinical contexts and community standards, rather than simply delivering information for passive consumption. To further strengthen the project’s alignment with “Building Technological and Community Relationships between learner and environment,” we recommend collaboration with Elsevier and major acute hospitals (NUH, SGH, TTSH) to co-develop content and cases reflective of local practice, and migration to NUS IT infrastructure for secure, equitable, and integrated deployment. These steps will deepen ties between students, technology, educators, and clinical partners; enhance real-world relevance; and support a sustainable learning ecosystem that bridges the classroom, the clinic, and societal expectations of safe, high-quality care.
References
Dergaa, I., Ben Saad, H., Glenn, J. M., Amamou, B., Ben Aissa, M., Guelmami, N., Fekih-Romdhane, F., & Chamari, K. (2024). From tools to threats: a reflection on the impact of artificial-intelligence chatbots on cognitive health. Frontiers in Psychology, 15, 1259845. https://doi.org/10.3389/fpsyg.2024.1259845
Diamond-Fox, S., & Bone, H. (2021). Advanced practice: critical thinking and clinical reasoning. British Journal of Nursing, 30(9), 526–532. https://doi.org/10.12968/bjon.2021.30.9.526
McSparron, J. I., Vanka, A., & Smith, C. C. (2019). Cognitive learning theory for clinical teaching. The Clinical Teacher, 16(2), 96–100. https://doi.org/10.1111/tct.12781
Negm, L. M. M. A., Mersal, F. A., Fawzy, M. S., Rajennal, A. T., Alanazi, R. S., & Alanazi, L. O. (2024). Challenges of nursing students during clinical training: A nursing perspective. AIMS Public Health, 11(2), 379–398. https://doi.org/10.3934/publichealth.2024019
Oh, J. A., Kim, E. A., & Kim, H. R. (2025). Influence of Critical thinking disposition, clinical reasoning competence, and nursing practice environment on medication safety competence of hospital nurses. Healthcare, 13(5), 542. https://doi.org/10.3390/healthcare13050542
Weidener, L., & Fischer, M. (2023). Artificial intelligence teaching as part of medical education: Qualitative analysis of expert interviews. JMIR Medical Education, 9, e46428. https://doi.org/10.2196/46428