Abstract
Introduction: The extent of interaction (EoI) in visual crowding is the region around the target within which the presence of flanking stimuli compromises identification of the target. Following results reported by Tripathy and Cavanagh (Vision Research, 42, 2357-69, 2002), the current study systematically investigates the influence of stimulus duration on the EoI. Methods: The target (lower visual field, 10 deg eccentricity) was a 'T' in one of four orientations. The flankers were four "square thetas" in one of two orientations. On each trial there was one flanker in each cardinal direction relative to the target, except for interleaved unflanked trials, which measured uncrowded performance. Stimulus duration (13 - 427ms) varied between blocks of trials. Target size was fixed for all stimulus durations and contrast was adjusted so that unflanked target-orientation identification was 90 - 100% correct. The flankers matched the target in width, height and contrast. Within a block, 7 equally spaced target-flanker separations were presented using the Method of Constant Stimuli to obtain full psychometric functions for 3 observers (% correct report of target orientation vs. separation) for each stimulus duration. Psychometric functions were fit by a cumulative normal function (lower asymptote = 25%; upper asymptote = unflanked performance) and EoI was estimated as the separation that yielded a drop, relative to unflanked performance, of 10% of the amplitude of the psychometric function. Results: The radius of the EoI decreased from 6.6 deg (27ms duration) to 2.8 deg (427ms duration), despite almost constant visibility of the unflanked target at all durations. Conclusions: The duration of the stimulus has a substantial influence on crowding, with spatially larger mechanisms being invoked when processing brief stimuli. Subsequent experiments investigate the characteristics of these large mechanisms and their implications for Bouma's Law.
Meeting abstract presented at VSS 2013