Keyword: gender

ABSTRACT: This study examined the relationship between AI-powered learning tools, student engagement, and academic performance in higher education, with a focus on differences across academic disciplines, age groups, and gender. The study employed a quantitative, correlational, and causal-comparative research design, involving undergraduate students from both STEM and non-STEM disciplines through a multi-stage sampling approach. Data were obtained from AI-generated learning metrics, specifically Time-on-Task, Interaction Frequency, and Knowledge Mastery, alongside a structured questionnaire measuring behavioral, cognitive, and emotional aspects of student engagement, as well as students’ self-reported academic performance. The findings revealed that student engagement varied according to the type of AI learning tool utilized. Tools designed to support knowledge mastery were associated with higher levels of engagement compared to those focused primarily on interaction frequency or time spent on tasks. Students in STEM-related disciplines generally demonstrated stronger engagement than those in non-STEM fields, although the pattern of association between AI tool use and engagement was consistent across disciplines. Knowledge Mastery also emerged as the most influential factor in predicting academic performance across different age groups, with older students tending to achieve better academic outcomes. Additionally, gender differences were observed in how students benefited from specific AI tools, suggesting varying learning preferences and responses to AI-supported instruction. Overall, the study highlights the significant role of AI-powered learning tools in shaping student engagement and academic performance. It emphasizes the need for mastery-oriented, learner-sensitive, and discipline-responsive AI interventions to optimize learning outcomes in higher education.