Abstract
Individuals adjust preparatory attentional control settings according to environmental statistical regularities such that the behavioral cost in response time associated with shifting attention is reduced when shifting is likely (Sali, Anderson, & Yantis, under review). This learned flexibility may reflect a readiness to update the spatial locus of attention. Alternatively, the spread of attention may be variable such that when a shift of attention is anticipated, the region of space encompassed by attentional selection broadens. We, therefore, investigated whether the breadth of selection varied based on block-by-block changes in the likelihood of shifting attention. Participants held and shifted attention between two rapid serial visual presentation (RSVP) streams of alphanumeric characters in response to visual cues and made speeded parity judgments for target stimuli appearing in the cued stream. Three distractor streams flanked each of the two task-relevant streams. We manipulated the frequency of attention shift and hold cues across blocks of 40 trials each. To test whether these changes in cue frequency modulated the breadth of attentional selection, we varied the response congruency of stimuli appearing in the three distractor RSVP streams surrounding the to-be-attended location. Consistent with previous studies, the behavioral cost in RT associated with shifting attention was smallest for blocks in which shifting was most frequent. When examining shift trials alone, there was a significant interaction of block type and flanker congruency. Participants were slower for incongruent shift trials than for congruent shift trials only in blocks with frequent shift cues. A similar trend existed such that the congruency effect for hold trials was larger when holding was frequent. Our results suggest that increased likelihood of shifting attention is associated with increases in the breadth of attentional selection to include the to-be-attended spatial location, not simply a faster shift at the time of the cue.
Meeting abstract presented at VSS 2015