From Data to Decisions: AI-Supported Learning Analytics for Self-Regulated Learning and Academic Decision-Making

Abstract

This study investigated the effect of a learning environment based on AI-supported learning analytics on self-regulated learning and academic decision-making skills among first-level students at the Faculty of Education, Damietta University. A quasi-experimental pre-test–post-test control-group design was employed. The sample comprised 100 randomly selected male and female students aged 18–20 years, divided equally into an experimental group and a control group. Both groups studied the same Technology of Education course and completed equivalent learning and assessment activities over ten weeks. The experimental group additionally used a custom Moodle plugin that provided a student-facing analytics dashboard, diagnostic information, risk predictions, intelligent alerts, personalized recommendations, and alternative academic actions generated through a fuzzy rule-based expert system. Data were collected using an achievement test, a self-regulated learning scale, a situational academic decision-making test, a digital learning-record analysis rubric, and a usability and satisfaction questionnaire. Because the assumptions of homogeneity of regression slopes and equality of gain-score variances were not satisfied, Welch’s tests were employed for the principal comparisons. The experimental group achieved significantly greater gains than the control group in achievement, t(63.48) = 15.19, p < .001, g = 3.01; academic decision-making, t(52.28) = 11.59, p < .001, g = 2.30; and self-regulated learning, t(49.37) = 12.47, p < .001, g = 2.48. It also obtained significantly higher digital learning-record rubric scores, t(89.65) = 10.64, p < .001, g = 2.11. Experimental-group usability and satisfaction significantly exceeded the neutral criterion, t(49) = 16.53, p < .001. The findings indicate that pedagogically grounded and explainable learning analytics can strengthen students’ regulation and academic decisions when analytical information is linked to actionable alternatives and students retain control over the final decision.