Abstract
The nature of selective attention was investigated for partially valid cueing and spatial filtering. The task was to judge the contrast polarity of a brief flash of light of varying contrasts. For both paradigms, a location was cued and responses to stimuli at cued locations were compared to responses to stimuli at uncued locations. For partially valid cueing, a stimulus at a cued location was more probable than a stimulus at an uncued location. However, the task was to respond to a stimulus at any location. For spatial filtering, the task was to respond to stimuli at the cued location and ignore stimuli at the uncued location. It was therefore inappropriate to respond to a stimulus at the uncued location in the spatial filtering task. We considered the predictions of two selection hypotheses: contrast gain and all-or-none mixture. For the contrast gain hypothesis, information from a stimulus at an uncued location is attenuated on all trials. Thus, increasing contrast can overcome the attenuation. For the all-or-none mixture hypothesis, information from a stimulus at an uncued location is blocked completely on some fraction of the trials. Thus, increasing contrast cannot overcome the block. Results for cueing are consistent with the contrast gain hypothesis and reject all-or-none mixture. In contrast, results for spatial filtering are consistent with the all-or-none mixture hypothesis and reject contrast gain. One interpretation is that spatial filtering is limited by the imprecise localization of the cued location. We argue that spatial filtering reveals an aspect of the selection process that was hidden with cueing paradigms.
Partially supported by NIH grant MH067793 to C. M. Moore.