Evidence of Change Deafness with Continuously Moving Targets

Abstract

ABSTRACTChange deafness, the failure to detect above-threshold changes within complex scenes, has been considered analogous to change blindness despite the fact that auditory and visual demonstrations have relied upon different task conditions. Further, change deafness is often difficult to distinguish from errors due to stimulus uncertainty and memory encoding. These issues were addressed across experiments involving spatialized arrays of the same four simultaneously and repeatedly presented 1 s events that were separated by 40° along the azimuth. In each array, one target event oscillated across a 60º arc. In Experiment 1, listeners identified targets under three conditions: flicker across successive presentations/positions, a corresponding condition with a 3 s initial array to determine if encoding time improved performance, and a “continuous” condition characterized by gradual target movement. Responses in the continuous condition were slowest and least accurate, especially when accounting for baseline localization performance with isolated events. Experiment 2 further examined this possibility of change deafness by manipulating target velocities (80º/s, 40º/s, 24º/s, 8º/s), two of which (40º/s and 24º/s) matched rates of flicker and continuous movements, respectively, from Experiment 1. Error rates significantly exceeded those from localization of isolated events, confirming that change deafness occurred. Furthermore, the 8º/s condition produced the highest error rates, suggesting that change deafness is more likely with slow changes in perceived event locations. These findings suggest that, unlike change blindness, change deafness may be primarily limited to continuous changes once encoding concerns are eliminated, reflecting increased perceptual similarity for within-event comparisons across local regions in time.