On the nature of input channels in visual processing.

The research reported herein was designed to assess whether the presence of noise elements in a visual display affects the detection of target letters at the perceptual or feature extraction level of processing, as well as at the decision level, and more specifically, whether (a) input or processing channels operate in an independent or interactive fashion and (b) how the spatial relation between signal and noise items affects detection performance. In order to distinguish among current theories proposed to account for the influence of noise items on visual processing, a forced-choice detection task was modified to incorporate a cueing procedure, which permitted the independent variation of signal-noise similarity, confusability, and proximity. The results provide evidence for feature-specific inhibition at the perceptual level, and a theory is proposed that assumes hierarchically organized, limited-capacity feature detectors and feature-specific inhibitory channels.