Unveiling the Needs: Developing AI-Based Chatbots in Blended Learning to Enhance Computational Thinking and Self-Regulated Learning in Mathematics Education

Abstract

The advancement of Artificial Intelligence (AI) in education has introduced innovative learning tools, including AI-based chatbots, which have the potential to enhance student engagement and personalized learning. This study focuses on the Analysis phase of the ADDIE model to investigate the needs and feasibility of developing an AI chatbot for blended learning, specifically to improve Computational Thinking (CT) and Self-Regulated Learning (SRL) among undergraduate mathematics education students. A mixed-method approach was employed, utilizing SRL questionnaires, computer literacy questionnaires, CT tests, and preliminary interviews with students and lecturers. The SRL questionnaire assessed students' challenges in self-regulation and their preferences for technology-assisted learning. The computer literacy questionnaire explored students' familiarity with digital tools. The CT test provided insights into students' computational thinking skills, while interviews offered qualitative perspectives on learning difficulties and expectations regarding AI chatbots. Findings from this analysis phase reveal that students require adaptive and interactive learning support to enhance their computational thinking skills and autonomous learning. Moreover, while students express strong interest in AI-based chatbots, concerns regarding usability, integration, and effectiveness in mathematical problem-solving remain. The results of this study lay the groundwork for designing an AI chatbot framework that aligns with students' academic needs and technological readiness, providing a foundation for future development and implementation.