Abstract: Journal of Digital Marketing, E-Commerce, and Consumer Analytics

Abstract:
This study examines the effects of implementing artificial intelligence (AI) in IB Mathematics: Analysis and Approaches Higher Level at Kolej MARA Banting on student learning when guided by educators. The inter vention aimed to incorporate AI technologies and assist students, particularly those requiring further support, in learning to assess AI-generated responses critically and utilise them responsibly to enhance their mathe matical skills beyond the classroom. In the third semester of Year 2, students either went to regular classes with no scheduled AI support or structured activities led by a teacher with AI support on five specific sub topics. Independent-samples t-tests showed no statistically significant differences between groups on four of the five quizzes. However, there was a significant improvement in the Maclaurin Series topic (p < .05), which suggests that AI integration may work better for some topics than others. Subgroup analysis indicated that lower-achieving students exhibited significant improvement after AI-supported intervention, as evidenced by significant improvement on subsequent similar evaluations following guided elucidations of solution meth odologies. Additionally, the AI-integrated group exhibited markedly superior performance in the Semester 3 examination, t (90) = −3.15, p = .002, despite similar baseline performance in Semester 2. In Semester 3, structured AI guidance was used to help students understand the pros and cons of AI. In Semester 4, however, the focus was on independent thinking and on talking to teachers rather than using AI tools. Data from students showed that AI-supported activities prompted them to determine whether their answers were correct and to reflect on how to use algebraic methods or graphical display calculators (GDCs) correctly. This aligns with the IB learner profile, which includes critical thinking and reflective learning. A comparison of the IB May 2024 and May 2025 exam results shows that AI-assisted instruction may help students learn more in the short term, but its overall effect on their performance on external exams is only moderate and varies by subject. The findings indicate that teacher-facilitated AI integration can serve as an effective educational support tool, especially for structured procedural subjects, although its efficacy varies across mathematical domains and student proficiency levels.