Working memory capacity as controlled attention: Implications for visual selective attention.

摘要

Baddeley (1993) and Engle, Kane, and Tuholsky (1999) view working memory (WM) as ‘working attention’. If this is accurate, differences in working memory span should predict performance on visual attention tasks. Bleckley, Durso, Crutchfield, Engle, and Khanna (under review) found that visual attention allocation, as tested using a modification of Egly and Homa's (1984) selective attention task differed between groups that were tested as being high or low in working memory capacity. In their task participants identified a central letter and localized the displaced letter flashed somewhere on one of three concentric rings. When the displaced letter occurred closer to fixation than the cue implied, high spans, but not low spans, showed an invalid cue cost in the letter localization task. This suggests that low working memory span participants allocated attention as a spotlight, whereas those with high spans showed flexible allocation.; This research replicated and extended the findings of Bleckley et al. that individual differences in working memory capacity correspond to differences in the allocation of visual attention. Allocating visual attention to discontiguous locations is a resource limited capacity process that is difficult for low span individuals. The high span individuals were able to allocate attention discontiguously because of the availability of greater attentional resources, and when an attention demanding secondary task was performed, high span individuals no longer allocated attention discontiguously.; The question posed by the second experiment is whether the distinction between endogenously and exogenously generated attention shifts is important to the findings of a relationship between working memory capacity and allocation of visual attention. The view of working memory capacity taken by Engle and colleagues would suggests that exogenous cues should lead to less of a difference between high and low working memory capacity individuals. This view was supported.