Psychosocial boundaries of digital well-being: Exploring smartphone addiction, information overload, and sense of belonging in mobile ecosystem.

In the digital era, mobile applications have become ubiquitous in China's organizational and daily life, creating a condition in which constant connectivity coexists with social disconnection, posing critical challenges to user well-being management. This study investigates how smartphone addiction undermines sense of belonging, treated here as one key dimension of the broader digital well-being construct rather than as digital well-being in full, through the mediating role of information overload, with self-control and social media fatigue as boundary conditions. Based on a sample of 326 Chinese mobile app users, we adopt a mixed-methods approach integrating structural equation modeling (SEM) and a feed-forward neural network to test a moderated mediation model. The results reveal that smartphone addiction significantly predicts information overload, which in turn erodes sense of belonging; information overload mediates the majority of the total effect. Self-control buffers the addiction-to-overload link, while social media fatigue amplifies the overload-to-belonging erosion. The neural network model achieves high accuracy in predicting belonging deficits and uncovers non-linear threshold effects not modelled by the linear SEM specification. Because the predictive comparison rests on a small hold-out test set (n = 49), between-model performance differences are interpreted with caution. Theoretically, this study integrates Cognitive Load Theory, Self-Regulation Theory, and Conservation of Resources Theory to advance digital well-being research in management. Practically, it provides actionable insights for platform management through ethical design to reduce overload, organizational policy through digital boundary setting, and user behavior management through self-control training and fatigue mitigation.