Modeling the Impact of Screen Time on Mental Wellness Through Computational Analysis

Background: The rapid integration of digital technologies into everyday life has raised widespread concerns regarding the psychological consequences of prolonged screen exposure. While prior studies have shown associations between screen time and mental health issues, findings have often been inconsistent, largely due to simplistic linear models and limited behavioral contextualization. This study aims to address these gaps by evaluating how various forms of screen use, lifestyle behaviors, and psychological indicators jointly influence mental wellness in a population of digitally active individuals. Methods: We analyzed a self-reported dataset of 400 participants containing detailed metrics on screen use (mobile, TV, laptop), sleep quality, stress, productivity, mood, and other lifestyle behaviors. Statistical analyses included Pearson correlation, multivariate linear regression, and K-means clustering with Principal Component Analysis (PCA) to uncover behavioral subtypes. Predictors of mental wellness were identified through standardized regression coefficients, and clusters were interpreted based on their psychological and digital usage profiles. Results: Stress emerged as the strongest negative predictor of mental wellness (β = −10.69), followed by sleep quality (β = +5.92) and productivity (β = +4.72). Contrary to prevailing assumptions, total screen time and leisure screen use had minimal direct impact on wellness once mediating variables were included. Clustering revealed three distinct digital behavior phenotypes: (1) Balanced and Active Users, (2) Leisure-Heavy High-Stress Users, and (3) Burnout-Prone Professionals. These profiles showed differing wellness outcomes sharply and validated the multidimensional nature of digital health risk. Conclusion: Mental wellness in digital contexts is best understood through a multivariable lens that accounts for stress, sleep, and self-regulatory behaviors rather than raw screen time alone. These findings challenge traditional screen time metrics and highlight the need for personalized, context-aware interventions. This study offers a replicable computational framework for identifying behavioral risk profiles and supports a paradigm shift from screen avoidance to digital self-optimization.