Abstract
Introduction: Previously, the temporal dynamics were assessed in a cueing task across two locations using classification images to measure information use (Shimozaki, Vision Sciences Society, 2007). Results suggested parallel processing of the cued and uncued locations with no delay at the uncued location, inconsistent with serial descriptions of visual attention (e.g., Posner, 1980). It is assumed that attentional costs increase with the number of locations (set size effects); thus, the goal of this study is assessing the temporal dynamics with more than one uncued location. Method: Two observers participated in a yes/no contrast discrimination task of vertical Gabors (1 cpd, 1 octave bandwidth, full-width, half-height) appearing 7.5° from central fixation across 2, 3, or 4 locations (set sizes). Half the trials contained a high-contrast Gabor signal (15.6% mean peak contrast on a 23.4% mean peak contrast pedestal) appearing at one location and the pedestal at the other locations, half the trials contained the pedestal at all locations, and participants judged for signal presence. The stimuli were presented for 272 ms divided into 12 intervals (22.7 ms), with the contrast of each Gabor varying randomly from interval to interval (Gaussian, sd = 11.7% contrast). A dark 4° square cue appeared simultaneously with the stimuli and indicated the location of the signal with 70% validity; therefore, the signal appeared at each uncued location on 30%, 15%, or 10% of the signal present trials for set sizes of 2, 3, and 4, respectively. Results: Set size effects for percent correct were modest, while cueing effects (valid hit rate - invalid hit rate) increased with set size, reflecting the decreasing validity of each uncued location with set size. Across set size, results from the classification images found no evidence of a delay of information use at the uncued locations relative to the cued location.