Predictive Learning Analytics for Early Identification of At-Risk Students in Online Education Programs

Abstract

The dramatic growth in online education has presented new prospects on the flexibility of learning in addition to posing a considerable challenge of student engagement, retention, and academic success. Among the most urgent concerns of the online education programmes can be outlined as the inability to recognise the at-risk-students that might perform poor in school or even drop out of it before it is too late to provide them with effective help. Predictive learning analytics has become an effective data intensive tool that facilitates the analysis of academic data of high volumes of students and finding of the predilection signs of academic risk. This paper is research about the application of predictive learning analytics to identify at-risk students during the initial stages of an online education course. The study is quantitative weighed research which relies on the data obtained through Learning Management Systems (LMS) and has indicators like frequency of logins, submission of assignment, contribution in discussion forums, and performance in the assessment. The use of predictive models and learning analytics methods is done to find trends that relate to student disengagement and academic challenge. In the findings, the behavioural and engagement indicators such as low platform activity, late submission of assignments, and declining grades in assessment are identified as robust predictors of at-risk-students of failure or dropout. The paper indicates that predictive learning analytics could be utilised successfully to complement early warning systems that allow teachers and colleges to intervene at the earliest stage by providing specific academic assistance, personalised comments, and also through mentoring students. With the provision of predictive analytics to online learning spaces, schools can achieve a considerable degree of monitoring of their students, better retention rates, as well as a robot-most-favourable learning environments. The paper emphasises the need to embrace the use of data-informed decision-making methods in enhancing the success and sustainability of online education programmes among students.