ISCAP Proceedings: Abstract Presentation
Exploring the Use of a Multi-Model AI Tutor in a Database Course
Michael Leih
Point Loma Nazarene University
Abstract
The integration of artificial intelligence (AI) tools into IS courses presents new opportunities for personalized learning. This study is utilizing a custom chatbot that uses multiple LLMs and acts as an AI tutor to present new options for student engagement. This early-stage research investigates how undergraduate students interact with an AI-powered tutoring system integrated into a database design course. Expanding on previous AI pedagogical research and teaching tips, an AI tutor was designed to guide students through conceptual understanding and problem-solving in relational database theory, SQL, normalization, and big data concepts—without directly providing solutions to graded assignments. Over the course of a semester, the system logs student-AI interactions, including prompts, the LLM selected, responses, and follow-up behavior. To evaluate student perceptions, students complete periodic surveys and reflection questions to capture their opinion of AI usefulness, trust, and influence on learning strategies. Preliminary analyses focus on AI tutor usefulness, identifying patterns in student prompting behavior, differences in interaction styles, and which LLM students prefer for various tasks. This study aims to contribute to the growing body of research on AI in education by providing insights into how students use multiple AI models as a learning partner and by informing best practices for integrating AI tutors in a database course.