Abstract
Are rejected distractors marked in a visual search task and, if so, what are the limitations of this marking process? To address these questions, we introduce a technique for explicitly marking display locations visited by gaze during overt search. Each display item fixated for at least 200 ms was visibly marked following the saccade away from that item. The stimuli in this initial series of experiments were small (0.35 ) gray Os and Qs presented on a black background. Marker type was manipulated between experiments. Rejected items in Exp 1 were marked in red as observers searched for an O among Q distractors in two target (present vs. absent) and set size (31 or 46 items) conditions. Exp 2 kept the color marker but varied set size between 6 and 46 items in order to determine at what distractor memory load external marking benefits are expressed. Exp 3 used an “eraser” marker to remove rejected items, thereby enabling us to further explore the relationship between set size and marking. Exp 4 examined the effect of onsets vs. offsets by either adding or removing a box enclosing each marked item (i.e., items in the onset condition appeared initially without boxes but then box markers were added during search, whereas items in the offset condition appeared initially in boxes with these box markers then removed during search). For each experiment, search efficiency in the marked conditions was compared to a no-mark control condition. From these comparisons, we were able to estimate the effectiveness of internal marking during search. If the internal marking process was perfectly effective (i.e., perfect memory), no advantage should result from external marking. Our data clearly reject this hypothesis. Although the benefit derived from external marking varied with the marker type and the search manipulations, search was generally more efficient with marking than without. The implications of these results for models of memory during search are discussed.