Directing covert visual attention to a particular region of visual space improves performance at that location. Three main mechanisms have been postulated to account for this effect - signal enhancement, distractor exclusion and noise reduction. The experiments described in this proposal will examine performance on an orientation discrimination task while manipulating attention conditions and target predictability. One powerful way in which visual attention has been manipulated is with the use of so-called peripheral, central and neutral cues. While peripheral and central cues indicate to the observer both the time and spatial location of the upcoming target, neutral cues indicate the time but not the spatial location of the upcoming target. Peripheral and central cues are thought to be subserved by different attentional mechanisms that operate over different time courses. In the first set of experiments we will compare performance among these three cuing conditions in a discrimination task. We will measure performance when targets are either presented alone, in order to assess signal enhancement, or in the presence of """"""""distractors"""""""", in order to measure distractor exclusion. Our experiments are designed to measure the contribution of signal enhancement and distractor exclusion to performance improvements brought about by attentional manipulations. We will also examine the effect of target eccentricity. A critical manipulation in our experiments will be to vary the spatial frequency content of the targets and the distractors. The spatial frequency of the target will either be fixed or variable within a block of trials. The spatial frequency of the distractors will either be homogeneous (same spatial frequency) or heterogeneous (different spatial frequencies). Whereas visual discrimination tasks are thought to tap relatively high level mechanisms, detection tasks with simple stimuli appear to tap mechanisms early in the visual system. In the second set of experiments we will examine these same parameters in a detection task, in order to compare the effect of attention at different levels of visual processing. Finally, we plan to model the contribution of signal enhancement and distractor exclusion in any cueing effects we observe.
