Abstract
Most studies of attention capture consider the effects of an anti-predictive or irrelevant stimulus on visual search performance. If the attentional prioritization revealed in such tasks is entirely stimulus-driven, it should be immune to variations in the attentional resources available for search. However, if prioritization depends on the availability of attentional resources and is not entirely stimulus-driven, then capture should be influenced by the division of attention.
We used two capture paradigms to measure the effects of a secondary task on attentional prioritization. In one task, the critical item was a color singleton and was never the target of search. Capture is revealed by slowed search in its presence. In the other task, the critical item was a statistically irrelevant abrupt onset—it was just as likely to be the target of the search as any of the other items. Capture is revealed by a reduced influence of distractors on search speed when the critical item happens to be the target of search. Subjects performed these tasks while listening to an auditory stream of digits and counting repetitions.
When subjects ignored the auditory stream and only performed the search task, both anti-predictive color singletons and irrelevant onsets captured attention. However, when performing the secondary task, attention capture by an anti-predictive color singleton increased, but capture by the irrelevant onset was eliminated.
When an anti-predictive critical feature is present during search, attentional resources are needed to ignore it. Consequently, a dual task leads to increased attention capture. When a critical feature provides no information about target location, the dual task reduces the salience of that feature and thereby reduces capture. In both cases, capture is modulated by the attentional resources available during search, suggesting that attentional prioritization is not automatic and that the two tasks measure different aspects of prioritization.