globusjournal

Short-video platforms (e.g., Douyin, Kuaishou) have become central to media use among China’s “silver-haired” users. While algorithmic personalization increases perceived relevance and enjoyment, it may also encourage excessive use. This study proposes and tests a mediated-moderated model in which immersion (flow), mood repair/loneliness alleviation, and cue-reactivity mediate the association between perceived algorithmic personalization and excessive short-video use, while digital literacy, family support/joint media engagement, and urban-rural context moderate these pathways. Using a multi-site survey of Chinese older adults (≥60; N ≈ 800) with device-logged usage and partial least squares structural equation modeling (PLS-SEM), we estimate direct, indirect, and conditional effects; qualitative interviews (n ≈ 40) deepen interpretation. We expect personalization to predict excessive use predominantly via immersion and mood repair, with buffering by higher digital literacy and stronger family support. Findings aim to inform dignity-preserving, family-integrated interventions that rebalance use without sacrificing valued social functions.

Keywords: Older Adults, China, Algorithmic Personalization, Immersion, Loneliness, Digital Literacy, Family Support, PLS-SEM.