Predicting student success: leveraging people analytics in canvas to forecast student performance.

This study explores the predictive potential of learning analytics within the Canvas Learning Management System (LMS) to forecast student academic performance and inform an actionable model to support student success. Using multipleregression analysis, we examine overall course grades as the dependent variable and assess the influence of key behavioural indicators available in Canvas, including on-time assignment submission, last page views, participation time, and total page views. Prior research highlights both the promise and ethical complexities of predictive analytics in education. Marachi and Quill (2020) call for increased critical awareness within higher education communities regarding the use of predictive behavioural and learning analytics, while Oudat and Othman (2024) emphasise Canvas’s ability to reveal meaningful engagement patterns. Our findings contribute to a conceptual framework for predicting student success and guiding timely academic interventions. However, we acknowledge the limitations of LMS data, particularly its inability to capture critical socioemotional, contextual, and cultural dimensions of learning. Therefore, we position Canvas analytics not as a definitive predictor of success, but as one valuable component to student learning and success. The study concludes with practical implications and recommendations for future research to enhance ethical and effective use of learning analytics in higher education.