Abstract
The Quick formula for response pooling cannot account for spatial summation of noncontiguous multiple patch Gabor stimuli even when the subject has full knowledge of the stimulus configuration. Thresholds for a contiguous Gabor patch with increasing radius can be accounted for by Quick summation with an exponent of 2.5 for both foveal and peripheral viewing. Thresholds for a stimulus consisting of two or more small, spatially separated Gabors viewed 2.5° in the periphery require a Quick exponent of ~6.4. This discrepancy between exponents for contiguous and noncontiguous stimuli can be explained by an attention uncertainty model in which subjects can attend to one spatial position perfectly but can attend correctly to each of the other spatial positions with a probability <1. Two interval forced choice threshold experiments were simulated by Monte Carlo methods using the uncertainty model and two models for the observer's decision process. It is shown that the uncertainty model can account for the threshold data with either decision process, but estimates of how attention probability varies with the number of patches favors one of the decision process models.
© 1989 Optical Society of America
PDF ArticleMore Like This
Allen B. Poirson and Brian A. Wandell
THK4 OSA Annual Meeting (FIO) 1989
Mark W. Cannon and Steven C. Fullenkamp
FH5 OSA Annual Meeting (FIO) 1987
C. W. CLENDENING, R. APRAHAMIAN, M. TREPKA, R. WAGNER, and H. A. SAYADIAN
FI6 Conference on Lasers and Electro-Optics (CLEO:S&I) 1989