Object-based attention: the effects of selective and divided attention

Abstract

Selective attention improves information processing for a subset of relevant information, usually at the expense of irrelevant information. Visual information can be selected on the basis of a particular spatial location, feature, or object. Selective attention has been shown to modulate the response of sensory neurons encoding the selected feature. Feature- and object-based attention make competing prediction when an observer is asked to attend to a single feature within an object. Feature-based attention predicts that only the task-relevant feature will be selected, and thus the neural response to the task-irrelevant features will be suppressed. Object-based attention predicts that all of the object's features will be selected, and thus attentional modulation will spread to the task-irrelevant. Using a transparent motion stimulus composed of two overlapping dot fields with unique color-motion conjunctions, we measured the independent responses to each surface-features with fMRI. Attention modulated the response to both surface features consistent with object-based selection. Our fMRI result led to the following prediction: if all features of a relevant object are selected, then it should be easier to divided attention across features within an object versus between two objects. Using a dual-task paradigm, we found no perceptual cost to dividing attention within a surface but a large perceptual cost to dividing attention between surfaces. Surprisingly, when attention was divided between surfaces, subjects became more susceptible to distractors in the other surface. Together, these results suggest that object-based attention selects all features of a relevant object, which makes it difficult to filter distractors when dividing attention across multiple objects.