Researchers have learned a great deal about the processes involved in selective attention by studying how humans covertly orient their attention to different locations in visual space (for reviews, see Klein, Kingstone, & Pontefact, 1992; Posner, 1980; Wright & Ward, in press). Laboratory studies of visual spatial attention have commonly used the spatial-cueing paradigm to examine the effects of orienting attention on stimulus processing. In this paradigm, a cue stimulus is used to direct attention to a specific location prior to the appearance of a target stimulus. In studies of goal-driven attention, the cue is typically presented at fixation and is informative about the likely location of the forthcoming target. Studies of this kind demonstrate that providing information about the location of a forthcoming visual stimulus can influence the speed with which the stimulus is detected, identified or localized. The general finding is that subjects respond more quickly and more accurately to the target when it is presented at the validly-cued (expected) location compared to when it is presented at an invalidly-cued (unexpected) location. Performance is relatively enhanced on valid-cue trials even in the absence of eye movements indicating that covert attention-orienting mechanisms are involved.

In studies of stimulus-driven attention, the cue is presented at or near a potential target location. If a target stimulus appears at the cued location within about 200 ms of its onset, responses are faster and more accurate relative to responses to targets presented at different locations. The effects of such direct spatial cues are said to be stimulus-driven, rather than goal-driven, because they occur even when the cue is non-predictive as to the location of the forthcoming target. Importantly, non-predictive direct cue effects differ in several important ways from symbolic cues indicating that separate mechanisms are involved in stimulus-driven and goal-driven attention orienting (Wright & Ward, in press; Klein et al., 1992). One difference, first demonstrated by Jonides (1981), is that spatial attention effects develop more rapidly with direct cueing than with symbolic cueing. Spatial attention effects also vanish more rapidly with direct cueing than with symbolic cueing, particularly when the direct cue is non-predictive. Rapidly appearing but sustained benefits can arise from predictive direct cueing, indicating that both stimulus-driven and goal-driven mechanisms of spatial attention are invoked under such conditions.

An important question is whether these mechanisms of attention influence the transmission of visual information at early stages where stimulus features are registered and perceptual representations are constructed, or at later stages where categorization, decision-making, and response selection occurs. Theories of early selection propose that selective attention controls the transmission of visual information at sensory/perceptual stages of processing. By comparison, theories of late selection propose that transmission of visual information at these initial stages proceeds in parallel across the visual field while selective attention controls access to more central processing stages. There is now convincing evidence from psychophysical and electrophysiological studies that spatial attention modulates the transmission of visual information at relatively early stages of processing (e.g., Luck, Hillyard, Mouloua, Woldorff, Clark, & Hawkins, 1994).