Psychological Profiling of Digital Dependency Using Machine Learning and Behavioral Clustering Techniques

Mental, physical, and emotional well-being are critical for sustaining productive engineering careers, yet the integration of mobile devices into every aspect of life has blurred work–life boundaries. Using a 268 -participant survey where these people are engineers pursuing the electronics field, this study investigates the psychological footprint of digital dependency— an emerging behavioural health challenge in high-pressure technical fields like electronics engineering. This paper extracts latent factors that underlie compulsive device interaction through exploratory factor analysis, constructs an interpretable supervised learning pipeline that predicts an individual’s likelihood of moderate or severe dependency, and leverages K-Means clustering with silhouette-guided K selection to derive five distinct usage personas. Our findings reveal that sleep disruption, diminished academic satisfaction, and anxiety-mediated impulsivity dominate the dependency profile. Practically, the proposed pipeline supports early-warning diagnostics for university wellness services, while the personas offer actionable design cues for persuasive digital-wellbeing interventions.