Abstract
In order to investigate the interactive effects of attention and contrast on visual processing, we compared psychophysical “threshold vs. pedestal contrast” (TvC) curves obtained under conditions of full- vs. poor-attention. The attention effect, defined as the ratio of thresholds for poor- vs. full-attention conditions, was substantial and was found to decrease significantly with increasing pedestal contrast (from approximately 10-fold at low contrasts to 3-fold at high contrasts, p < 0.005), indicating an interaction between attention and contrast. This result was observed regardless of whether the pedestal contrast was certain (pedestal contrast constant across trials) or uncertain (pedestal contrast varied across trials). This interaction is consistent with the existence of a contrast gain control mechanism for attention, as has been reported for neurons in early stages of visual processing (e.g., Reynolds, Pasternak and Desimone 2000; Martinez-Trujillo and Treue 2002). However, a pure contrast gain model predicts that attention should lower thresholds (i.e., attention effect greater than 1.0) at low pedestal contrasts, yet increase thresholds (i.e., attention effect less than 1.0) at higher pedestal contrasts. Since the attention effect in the current study was greater than 1.0 at all pedestal contrasts, this suggests the existence of a later stage in visual processing where attention imposes an additional response gain control.