| |
Abstract:
Duncan (1984) provided some of the first evidence for
object-based selection. In his experiments, subjects were presented
briefly with two superimposed objects followed by a mask. Subjects
were cued before stimulus presentation to attend to either one or
both of the objects. Accuracy was higher when subjects reported two
attributes from one object (the within-object condition) than when
they reported one attribute from each object (the between-object
condition). Because the spatial separation of the judged properties
was equal across the within-object and the between-object
conditions, Duncan argued for an object-based model of selective
attention, in which a limited number of objects could be perceived
simultaneously. Other researchers have suggested that this effect
might be accounted for by spatial selection of the locations
occupied by the objects. Using a similar procedure, we observed a
within-object advantage that was larger when the targets were
farther apart. This effect of target separation supports a spatial
selection account, however, we were also able to observe Duncan's
within-object advantage without the use of pre-cues, when subjects
were informed of the relevant targets only after they stimuli had
been masked. Together these finding suggest that there are two
processes underlying the within-object advantage, an early spatial
component and a later object-based component.
|
