Cognitive relevance is not enough to facilitate search when a distractor becomes a target.

Repeated visual search tasks are useful to investigate under which conditions incidental memory for targets and distractors improves search. In a previous study, using letters as targets, we analyzed whether previous fixations facilitated search. Two groups of participants searched for the same 12 letters six times on the same display (72 trials). Targets were presented in a different order to each group, so that a ‘critical’ letter appeared either in each block, or for the first time after 55 trials. Surprisingly, first search RTs for that letter were similar in both groups, even though they differed considerably in the number of previous incidental fixations. Since incidental retention of visual information is higher for natural objects (Williams, Henderson & Zacks, 2005) we decided to replicate the study using pictures instead of letters. A single search display made of 24 different colored objects was used. Twelve objects (fruits and toys) were presented as targets and each target was searched for 6 times. We compared the results of three groups of participants. The first group found all 12 objects in each block of trials. The second group had to find the ‘critical’ object (an apple) for the first time after 55 trials. The third group also first searched for the apple after 55 trials, but saw a slightly different display, with a raspberry in the location of the apple, up to that point. Therefore, the three groups differed in the number of previous incidental fixations on this target object. Interestingly, we found that RTs for the first search of the apple were similar in the three groups, even though participants often made direct saccades to the target objects and could recall many details of the search display. Even when target objects were cognitively more relevant, incidental memory did not seem to improve search.