A Comprehensive Analysis on the Student Behaviour in Open.uom.lk: A Large-scale Asynchronous Open Online Platform
Keywords:e-learning, e-learning design, online asynchronous learning, student behaviour, Course redesign, data analysis
The open learning platform (open.uom.lk) of the University of Moratuwa, Sri Lanka has attracted over 180,000 registered students in just over one year of its launch. This platform offers the Trainee Full Stack Developer (TFSD) programme, enabling the participants to enter the Information Technology (IT) industry to address the much needed human resources for the growing IT industry of the country. The programme consists of six courses related to IT covering Python Programming, Web Development and Professional Practice in Software Development. The platform operates with minimal restrictions for registrants and has students from all parts of the country with some foreign students with an equitable gender distribution. The steady growth of the registration numbers shows a high level of enthusiasm from the community to explore the potential opportunities in the IT industry. While the platform is being used actively by thousands of participants and new users are registered on a daily basis, it is also observed that some of the participants have shown slow progress at different stages. This study presents analyses performed at different stages of the programme to study the student behaviour and identify the possible causes for the participants not being able to achieve steady progress. The findings of the study indicate that the participants generally find it difficult to get through the programming exercises and assignments. Correlating learning patterns of the students help understanding the overall learning strategies which can be adopted by developers of similar asynchronous learning programmes. Furthermore, the study goes on to suggest and discuss possible solutions to clear the bottlenecks identified at different stages of the programme. The subsequent analyses allow the prediction of completions by participants leading to a machine learning based model for predictive analytics.