Abstract
People search targets more efficiently in repeated configurations than in novel configurations (contextual cueing; Chun & Jiang, 1998). Most studies suggest that target-distractor associations could guide attention to the target location. However, studies have found conflicting results on whether repeated configurations without a target could produce contextual cueing effects (Beesley et al., 2015) or not (Kunar & Wolfe, 2011). We outlined three major differences between the two studies that might influence the formation of distractor-distractor associations, and investigated which factor produced contextual cueing effects in the target-absent condition. First, the number of distractors in each configuration may influence the formation of distractor-distractor associations by changing the average inter-distractor distance and the number of possible distractor-distractor associations. Second, the number of repeated configurations that people could learn simultaneously may be limited. Finally, distractor specificity in terms of visual features may increase the strength of distractor-distractor associations. In Experiment 1, we manipulated the number of distractors within participants and the number of repeated configurations between participants. Participants searched for a T shape (target) among L shapes (distractors). We found contextual cueing effects only in the target-present condition regardless of the number of distractors and repeated configurations. In Experiment 2, we increased distractor specificity by adding a second, task-irrelevant feature (i.e. color) to the target and distractors, thereby rendering distractor-distractor associations more specific. By increasing the distractor specificity, we revealed contextual cueing effects in both target-present and target-absent conditions. These results suggest that distractors with enough feature specificity can be associated with other distractors and contribute to the contextual cueing effect. Meanwhile, the number of distractors and the number of repeated configurations do not play a key role in building distractor-distractor associations.
Meeting abstract presented at VSS 2018