Target Localisation and Identification in Rapid Visual Search

In recent models of visual search it has been proposed that, in rapid parallel search, information about the location of a target pattern among distractors and information about its identity would not be available simultaneously, but that target location is represented at earlier stages of visual processing than target identity. In the present study, the priority of location information over identity information in parallel visual search was investigated by means of one identification and two localisation tasks of different levels of difficulty. In all three tasks, the stimulus display was identical. An oblique line segment, randomly 45° or 135° in orientation, was presented randomly at one quadrant of the display. In the identification task, the observer reported the orientation of the oblique line irrespective of its location. In the easy localisation task, the observer indicated whether the oblique line was at the left or right side of the display, and in the difficult localisation task whether it was in the upper or lower part of the display. In both localisation tasks, the observer ignored the orientation of the target line. The oblique line was accompanied by one, five, seventeen, or thirty-nine vertical distractor line segments in all tasks. The results showed that the response speed in the left vs right localisation was faster than in the identification task, whereas performance in the up vs down localisation was inferior. When the response factors (stimulus–response compatibility) in the left vs right localisation were taken into account, there were no performance differences between localisation and identification. Thus, these results demonstrated that direct performance comparisons between localisation and identification may be somewhat arbitrary, and they do not solve the issue of priority of location or identity information in rapid visual search.