AI Insight
This conceptual review proposes the Reinforcement-Narrowing-Shift (RNS) framework to explain how problematic short-form video use develops through interactions between platform features (auto-play, infinite scroll, personalized recommendations) and user vulnerabilities. The framework suggests that repeated engagement can lead to informational narrowing in content diversity and cognitive demand, which through processing changes and metacognitive erosion may create cognitive-affective shifts that drive continued use. The authors present nine testable propositions and emphasize that personalization does not necessarily cause narrowing, as user exploration, changing preferences, and external events can interrupt the cycle.
Why it matters
This framework shifts focus from individual pathology to the role of platform design and recommendation systems in problematic media use, providing specific testable hypotheses for researchers and potential intervention points for platform designers and policymakers concerned with digital wellbeing.
Understand the Science
Short-form video (SFV) platforms have become a central component of contemporary digital media use. Although research has examined problematic short-form video use (PSVU), existing accounts often emphasize addictive-like symptoms or individual self-control while giving less attention to the exposure environment that recommendation systems continually update. This structured integrative conceptual review develops a recursive Reinforcement–Narrowing–Shift (RNS) framework to explain how PSVU may emerge and become self-maintaining. The framework proposes that interactions between platform affordances—including auto play, infinite scroll, full-screen presentation, rapid switching, and fine-grained behavioral feedback—and user vulnerabilities can reinforce repeated engagement. Under identifiable boundary conditions, including engagement-dominant optimization, stable behavioral signals, weak diversity constraints, and predominantly passive consumption, repeated engagement may contribute to multidimensional informational narrowing. We distinguish exposure-level narrowing in content and affective diversity from experience-level narrowing in cognitive demand and temporal horizons. These changes may be translated into cognitive-affective shifts through processing fluency and metacognitive erosion, affective habituation and escalation, and attentional capture and automaticity, thereby increasing renewed engagement. The framework does not assume that personalization necessarily narrows exposure: exploration, serendipity objectives, active user search, changing preferences, and external life events can stabilize, diversify, or interrupt the cycle. We develop nine falsifiable propositions, organize them into feasible research modules, and identify executive control, developmental stage, and cultural-institutional context as moderators. RNS is presented as a new platform-sensitive research framework rather than an extension of an existing clinical model, offering testable directions for longitudinal, experimental, computational, and intervention research.