Learning Analytics for Detecting Silent Academic Disengagement

Elena Martínez

Learning Analytics for Detecting Silent Academic Disengagement

Author(s)	Dr. Elena Martínez
Country	United States
Abstract	Student disengagement is often recognized only after it becomes administratively visible through repeated absence, missed assessment, failing grades, withdrawal, or non-completion. A more difficult condition occurs when students remain formally enrolled and continue producing enough digital activity to appear active while their learning behavior gradually becomes less regular, less persistent, and less academically purposeful. This study describes this condition as silent academic disengagement, a proposed analytical construct referring to sustained deterioration in learning participation before conventional failure indicators become clearly observable. Contemporary learning-analytics research supports the importance of this temporal perspective. A 2024 systematic review of 159 higher-education studies found that learning analytics overwhelmingly operationalized engagement through observable behavioral variables and rarely captured its multidimensional character, while recent longitudinal work has demonstrated that students transition among distinct weekly engagement profiles rather than remaining consistently “engaged” or “disengaged.” The study proposes a governed temporal learning-analytics framework for identifying silent disengagement without relying on intrusive video, facial-expression, gaze, or biometric surveillance. A synthetic cohort of 900 students was modeled across a 14-week semester, producing 12,600 student-week observations. Four detection approaches were compared: attendance-and-grade thresholds, static LMS activity analytics, temporal engagement-transition analytics, and a governed multi-source early-warning model combining study regularity, assessment timing, resource progression, feedback use, participation changes, and recent academic trajectory. The simulated governed model achieved an early-detection recall of 89%, precision of 87%, a false-alert rate of 8%, and a median 4.2-week warning interval before conventional academic difficulty became visible. By comparison, the attendance-and-grade threshold approach produced 54% early-detection recall and a median warning interval of 1.4 weeks. The study emphasizes that prediction alone is insufficient. Recent empirical learning-analytics research has shown that models with AUC values above 0.8 can generalize to later cohorts and that relational feedback following identification can stimulate renewed engagement with previously unused learning activities. Consequently, the proposed framework connects detection to proportionate, supportive intervention rather than punitive labeling. It also incorporates fairness auditing, student-facing transparency, data minimization, human review, and a requirement that low digital activity should never automatically be interpreted as lack of motivation. The paper concludes that learning analytics can contribute to earlier recognition of academic disengagement when it focuses on changes in learning trajectories rather than isolated activity counts. The strongest system is not the one that monitors students most intensively, but the one that identifies meaningful deterioration early enough for instructors or support services to offer timely, respectful, and context-sensitive assistance.
Keywords	learning analytics, student disengagement, silent disengagement, early-warning systems, student engagement, higher education, educational data mining, temporal analytics, academic risk, student success
Field	Engineering
Published In	Volume 3, Issue 5, September-October 2022
Published On	2022-09-01

View / Download PDF File

About AJAMIR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+1-872-302-3851	editor@ajamir.com

American Journal of Advanced Multidisciplinary Innovation and Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Learning Analytics for Detecting Silent Academic Disengagement

Share this