Abstract:

Student retention remains a persistent challenge in large-scale online education, particularly in programs characterized by diverse learner profiles and limited direct interaction. While the flexibility of online learning expands access, it may also lead to reduced engagement, especially in early stages of study. In this context, higher education institutions increasingly use learning analytics to identify patterns of student behavior and support retention. This paper addresses the following research question: To what extent can learning analytics indicators derived from student activity data predict student persistence and identify early signs of disengagement in fully online higher education programs? The study focuses on the predictive and descriptive potential of analytics rather than on establishing causal effects of interventions. The research adopts a quantitative longitudinal design based on institutional data from 4,200 students enrolled in fully online undergraduate and postgraduate programs at a national open university over six consecutive semesters (Spring 2022 to Autumn 2024). The dataset includes variables such as frequency of learning management system access, timing and regularity of assessment submissions, time spent on course materials, and participation in online communication tools. Predictive modeling techniques (logistic regression and decision tree analysis) were used to identify patterns associated with student persistence and dropout risk. Model performance was evaluated using classification accuracy and recall indicators for at-risk students. Results indicate that consistent engagement patterns are strongly associated with higher probabilities of course completion, and the models demonstrated moderate predictive accuracy (approximately 72%). The paper argues that learning analytics should be integrated into broader, human-centered support systems combining data-informed identification of risk patterns with academic advising and mentoring.

Keywords:

Learning analytics, student retention, online education, institutional strategy, higher education management.