Impaired top-down control of visual search in schizophrenia

James M Gold, Rebecca L Fuller, Benjamin M Robinson, Elsie L Braun, Steven J Luck, James M Gold, Rebecca L Fuller, Benjamin M Robinson, Elsie L Braun, Steven J Luck

Abstract

This study examined top-down and bottom-up control of attention in a group of 24 patients with schizophrenia and 16 healthy volunteers. Participants completed a visual search task in which they reported whether a target oval contained a gap. The target was accompanied by 5, 11, or 17 distractors. On some trials, the target was identified by a highly salient feature that was shared by only 2 distractors, causing this feature to "pop out" from the display. This feature provided strong bottom-up information that could be used to direct attention to the target. On other trials, half of the distractors contained this feature making these distractors no more salient than the other distractors requiring greater use of top-down control to restrict processing to items containing this feature. Patient visual search efficiency closely approximated control performance in the first trial type. In contrast, patients demonstrated significant slowing of search in the second trial type, which required top-down control. These results suggest that schizophrenia does not impair the ability to implement the selection of a target when attention can be guided by bottom-up information, but it does impair the ability to use top-down control mechanisms to guide attention. These results extend prior studies that have focused on aspects of executive control in complex tasks and suggest that a similar underlying deficit may also impact the performance of perceptual systems.

Figures

Figure 1
Figure 1
In each panel the target is the red (gray) vertical oval. Subjects indicated the presence or absence of a gap in the target. In panels A and C, participants were instructed to search the red (gray) items. In panels B and D, participants were instructed to search the vertical items. Panels A and B are examples of a 3-attended trials. Panels C and D are examples of half-attended trials.
Figure 2
Figure 2
Mean reaction times observed for the 3-attended (A) and half-attended trials (B), along with mean slope values (C). Error bars show the standard error of the mean.

Source: PubMed

3
Abonnieren