Spatial Guidance Overrides Dynamic Saliency in VR: An Eye-Tracking Study on Gestalt Grouping Mechanisms and Visual Attention Patterns

(1) Background: Virtual Reality (VR) films challenge traditional visual cognition by offering novel perceptual experiences. This study investigates the applicability of Gestalt grouping principles in dynamic VR scenes, the influence of VR environments on grouping efficiency, and the relationship between viewer experience and grouping effects. (2) Methods: Eye-tracking experiments were conducted with 42 participants using the HTC Vive Pro Eye and Tobii Pro Lab. Participants watched a non-narrative VR film with fixed camera positions to eliminate narrative and auditory confounds. Eye-tracking metrics were analyzed using SPSS version 29.0.1, and data were visualized through heat maps and gaze trajectory plots. (3) Results: Viewers tended to focus on spatial nodes and continuous structures. Initial fixations were anchored near the body but shifted rapidly thereafter. Heat maps revealed a consistent concentration of fixations on the dock area. (4) Conclusions: VR reshapes visual organization, where proximity, continuity, and closure outweigh traditional saliency. Dynamic elements draw attention only when linked to user goals. Designers should prioritize spatial logic, using functional nodes as cognitive anchors and continuous paths as embodied guides. Future work should test these mechanisms in narrative VR and explore neural correlates via fNIRS or EEG.