Abstract
Research on spatial attention has predominately studied attention effects in low precision tasks that require discrimination between relatively dissimilar stimuli. Here, we extend tests of spatially-cued attention to a high precision task with manipulations of visual noise and perceptual workload (see Dosher & Lu, 2000). An elaborated perceptual template model (ePTM; Jeon, Lu & Dosher, 2009; Lu & Dosher, 1998; Dosher & Lu, 2000) that incorporates precision of discrimination as a factor allows the consideration of feature template tuning in the response dimension as well as previously-studied attention effects of stimulus enhancement and external noise exclusion. Orientation discrimination was tested with low (±20°) or high (±4°) precision, in either no or high external noise, in relevant set sizes of 2, 4, and 8 Gabor patches. Attention was manipulated by target precuing and simultaneous cuing. Four out of five participants showed significant attention effects in contrast psychometric functions. Consistent with prior reports, large attention effects occurred in high noise, reflecting external noise exclusion. Exclusion increased with workload and was found at all precision levels.In addition, the high precision task exhibited attention effects in the asymptote of the psychometric functions, where precued and simultaneously-cued performance ceased to converge at high contrasts. This attention effect is modeled in the ePTM with a narrowing of Gaussian-shaped orientation tuning of templates. Template narrowing decreases the response to a non-matching orientation, which has differential effects on response based on the orientation difference of the stimuli: the orientation difference in the high precision task falls on the steep region of the template and so shows the largest effect. Our results extend the understanding of spatial cuing by uncovering effects of attention on asymptotic portions of the contrast psychometric functions in high precision, in addition to the known body of attention effects in low precision.
National Institute of Mental Health Grant # R01MH81018.