Short-form video, attention and reward: a neuroscience and eye-tracking narrative review

Short-form video combines rapid audiovisual changes, personalized content and repeated choices to continue. This narrative review separates attentional capture, reward wanting and hedonic liking using a 2025 systematic review and meta-analysis of 71 studies and 98,299 participants, alongside neuroscience, psychophysiology and cognitive studies available through 14 August 2026. Greater use was associated with poorer cognitive outcomes in heterogeneous, predominantly correlational evidence. Direct experiments report changes in large-scale brain-network activity during personalized viewing and condition-dependent differences in prospective memory, gaze synchrony and retrieval. Personalized clips, random clips and continuous narratives differ in content, agency and prior context as well as duration. Cross-sectional EEG, fNIRS and MRI findings concern selected signals and samples, with some sources available only as abstracts. Sensory selection, contextual updating, personal relevance and low sampling cost are candidate contributors to continued viewing; their joint contribution remains untested. Gaze measures visual allocation, pupil and electrodermal responses inform arousal and resource allocation, and BOLD measures hemodynamic activity. None alone identifies hedonic experience or a neurochemical mechanism. Experiments that vary editing, context continuity, personalization and sampling cost independently, while measuring ratings and actual viewing choices alongside physiology, could distinguish these accounts.