When Family Support Fails in Algorithmic Media: Parental Burnout,Algorithmic Personalization, and Adolescent Short-Video Dependence

<title>Abstract</title> <p>Short-video platforms have become central to adolescent digital life, yet dependence is often explained as either a self-control problem or excessive screen-time exposure. This study conceptualizes short-video dependence as a family-platform process linking depleted family support, adolescent psychological vulnerability, and perceived algorithmic personalization. Using a four-wave, multi-informant longitudinal design, we recruited 1,247 middle-school students from four schools in Shandong, Henan, and Sichuan, retaining 968 matched cases. Drawing on the Balance between Risks and Resources model, compensatory internet use theory, and the I-PACE model, we examined whether parental burnout predicted later dependence through parent-child communication, digital-rule consistency, loneliness, rumination, and fear of missing out (FoMO). Parent-reported burnout predicted weaker communication and less consistent digital rules. Communication reduced loneliness and FoMO, which in turn predicted later short-video dependence, whereas the direct effect of parental burnout became nonsignificant after accounting for these mediators. Perceived algorithmic personalization strengthened the association between FoMO and dependence. The findings support a temporally ordered, largely indirect family-platform model and identify perceived algorithmic personalization as an important platform boundary condition in adolescent digital-risk research.</p>