Abstract
An attentional blink occurs when observers are asked to identify two targets embedded in an RSVP sequence of distractors: Performance is severely impaired for second targets that occur within less than 500 ms of the first target. Last year, we reported that this impairment does not occur when observers are asked to report all items in a sequence (Nieuwenstein & Potter, 2006). This whole-report advantage indicates that performance in dual-target partial report tasks is primarily constrained by the requirement to select the targets and little affected by the processing that occurs following their selection.
To further examine which task differences are critical to the selection cost, we contrasted the effects of first-target selection and distractor filtering on recall of subsequent items. RSVP sequences consisted of six characters and observers were instructed to report either as many items as possible (whole report), only two items designated by color or category (partial report), or all items starting with a designated item (mixed report). The comparison of second-target recall in partial report with recall of nominally the same items in mixed report revealed that the attentional blink is specific to partial report, and thus to the requirement to filter out inter-target items. Mixed report performance for items that matched the first target's color showed no indication of an attentional blink and closely resembled performance in whole report. The same comparison did show a mixed-report cost for items not matching the first target's color, replicating earlier findings (Weichselgartner & Sperling, 1987). Together, these results show that the requirement - or tendency - to suppress items directly following a first target is a critical determinant of the occurrence of an attentional blink effect in recall of ensuing items.
NieuwensteinPotter
(2006). Psychological Science, 17, 471–475. WeichselgartnerSperling
(1987). Science, 238, 778–780.
This work was supported by NIMH Grant MH47432.