Analyzing factors impacting upper-level information technology course grades using machine learning

Identifying the factors that contribute to academic success in information technology (IT)-related courses in higher education remains a major challenge for educators and researchers. A systematic analysis of class attendance, homework performance, and exam results reveals a significant correlation between attendance and academic achievement. Notably, our analysis shows that students who attended the first class session achieved final scores 23.54% higher on average. To further explore the key factors influencing students’ overall grades, this study applies machine learning techniques to data from multiple advanced undergraduate IT courses. The predictive capabilities of the models are evaluated, with the Artificial Neural Network demonstrating the highest overall accuracy (79.01%), followed by the Decision Tree (69.17%) and Naïve Bayes (65.43%), all significantly outperforming the ZeroR baseline (34.57%). Accordingly, our findings strongly suggest the importance of class attendance for academic success and advocate for implementing attendance policies to identify struggling students early and improve learning outcomes.