Impact of Large Language Models in Strengthening Mathematical Foundations for Engineering and Energy Education in Nigeria

Abstract

Proficiency in mathematics is a key factor to students’ success in engineering and energy-related disciplines. Many students in Nigeria struggle with foundational concepts in algebra and calculus. With the rapid advancement of Artificial Intelligence (AI), particularly Large Language Models (LLMs) like ChatGPT, new opportunities are emerging to bridge these educational gaps through AI-assisted instruction. This study investigates the impact of LLMs in enhancing students’ conceptual understanding of mathematics within the context of engineering and energy education in Nigeria. Using a mixed-methods design, the research involves administering pre-test and post-tests, conducting student and lecturers’ interviews, and facilitating classroom interventions using ChatGPT.  The sample size for this study includes 200 students from three different universities, spread across three geopolitical zones in Nigeria; north, east and south.  An independent samples t-test carried out revealed that the LLM-supported group performed significantly better than the traditional group ((198) = 10.35, ???? < 0.0001), with a large effect size (Cohen’s d ≈ 1.46). These results suggest that LLM integration has a substantial impact on student learning outcomes, improving both test performance and classroom engagement. Based on these findings, we recommended that STEM curricula be revised to include AI-assisted learning activities and that a formal policy framework on the use of LLMs by students and faculty be developed.