Abstract
Abrupt onsets disrupt visual search with increased response times to the target as well as attentional capture of eye movements towards the irrelevant location. However, onsets may fail to capture attention when irrelevant to search or if a saccade has already been planned to another location. Contextual cueing in scenes has shown robust effects of explicit allocation to a target with faster search times and fewer eye movements. This study used a contextual cueing paradigm with repeated and novel search scenes. The final two search blocks introduced a 1o red square onset that appeared 100ms after central fixation or during a saccade landing more than 2o from center. The onset would appear 4o left or right of center on the same or opposite side of the target. Eye movements were monitored to determine whether an irrelevant onset would capture attention automatically despite knowledge of a target location from repeated scene contexts. Results showed both fixation and saccade onsets were highly significant compared to baseline measures of the area in previous blocks without the onset. A fixation onset on either side of the target within novel scenes was similar to onsets on the same side of a repeated scene target. However, a fixation onset on the opposite side of repeated scene targets was significantly different from the other fixation onsets suggesting less attentional capture when onsets are further from known target locations. Examination of ordinal fixations after fixation onsets showed significant proportions of looking by the first fixation for novel and repeated searches compared to baseline. Saccade onsets comparisons were not significantly different from baseline. Similar results across novel and repeated searches suggest memory for a known target location does not eliminate attentional capture of an irrelevant onset.
Meeting abstract presented at VSS 2016