Abstract
Since the COVID-19 pandemic, first- and second-year mathematics success rates have declined, negatively impacting throughput rates. Aligned with Vygotsky’s scaffolding framework, students may have foundational math gaps from secondary education. These deficits vary based on academic backgrounds and pandemic-related disruptions. Meanwhile, generative AI advancements enable personalized and scalable interventions to address these gaps. This study examined a 2024 pilot intervention in a second-year calculus module using AI Assistants. A Socratic-mode AI Assistant was developed through Mindjoy’s generative AI platform, was aligned with the Calculus II curriculum and integrated with Wolfram Alpha for computational support. The intervention was voluntary for repeat students, supplementing regular instruction. A human tutor coordinated weekly tutorial cycles, selecting problems and refining AI prompts. Students engaged with the AI Assistant independently before weekly in-person sessions, where insights from Mindjoy analytics guided personalized support. Weekly reports identified knowledge gaps and informed instructional adjustments. Engagement metrics, sentiment analysis, and academic performance data were collected to assess effectiveness. Findings indicate that the AI-human hybrid tutoring model improved student engagement and helped identify critical learning gaps. This paper outlines the intervention’s design, AI Assistant prompt development, and its integration with human tutoring. The study evaluates the impact of AI-assisted learning on student performance, engagement, and self-directed problem-solving skills. Implications for scalable AI-driven interventions in engineering mathematics are discussed, informing a follow-up intervention planned for 2025.
Presenters
Gustav PotgieterStudent, PhD in Engineering Development and Management, North-West University (NWU), South Africa, North-West, South Africa
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
AI TUTORBOTS, ADAPTIVE LEARNING, ENGINEERING EDUCATION, DIGITAL LEARNING TOOLS, MATHEMATICS
