Learning analytics in computer programming education: A bibliometric scoping review

There are often high failure rates and student attrition in programming education due to challenges with syntax, debugging, and abstract concepts. Traditional teaching approaches have struggled to meet the diverse learning needs of students. This paper presents a scoping review incorporating bibliometric analysis that examines Learning Analytics (LA) research in programming education within Computer Science, Engineering, and Mathematics. The study identifies thematic trends, research gaps, and instructional implications. A bibliometric scoping review was conducted on documents published from 2014 to 2023, retrieved from Scopus and Web of Science. After screening, 1,208 documents were analysed. The review reveals a growing focus on data mining, predictive modelling, and student-centred learning. Most research outputs emerge from Europe and North America, while Africa shows a growing contribution. However, programming-specific applications such as debugging and formative feedback remain underexplored. The study highlights the limited integration of learning theories in LA applications. It also suggests that aligning LA with frameworks like cognitive load theory can foster personalised learning, enhance engagement, and support skill acquisition. These findings provide evidence-based insights to guide instructional innovation, research collaboration, and the development of adaptive programming education systems.