Can intertrial priming account for the similarity effect in visual search?

In a visual search task, a salient distractor often elongates response times (RTs) even when it is task-irrelevant. These distraction costs are larger when the irrelevant distractor is similar than when dissimilar to the target. In the present study, we tested whether this similarity effect is mostly due to more frequent oculomotor capture by target-similar versus target-dissimilar distractors (contingent capture hypothesis), or to elongated dwell times on target-similar versus dissimilar distractors (attentional disengagement hypothesis), by measuring the eye movements of the observers during visual search. The results showed that similar distractors were both selected more frequently, and produced longer dwell times than dissimilar distractors. However, attentional capture contributed more to the similarity effect than disengagement. The results of a second experiment showed that stronger capture by similar than dissimilar distractors in part reflected intertrial priming effects: distractors which had the same colour as the target on the previous trial were selected more frequently than distractors with a different colour. These priming effects were however too small to account fully for the similarity effect. More importantly, the results indicated that allegedly stimulus-driven intertrial priming effects and allegedly top-down controlled similarity effects may be mediated by the same underlying mechanism.