Abstract
When a precue appears in an object, subjects normally identify a subsequent target in the same object faster than one in different object (targets are equidistant from the cue). This response pattern has been intererpreted as evidence for automatic object-based attention (OBA). We failed to find OBA effects for two central objects with two outer “flanking” objects (never containing cues or targets) present in the display (Mojica et al., VSS 2009). We hypothesized that with four objects attention may be allocated more widely than with two, and subjects may not search for the target in the cued object first as they do with only two objects. The four objects in the previous experiment were identical rectangles presented on the same depth plane; they may therefore have been grouped into a single surface. Is grouping into a single surface is necessary or is simply having four objects in the display sufficient to eliminate OBA effects? We ungrouped the four objects by using shadows to indicate that the flankers were nearer to the viewer than the central rectangles, and by coloring flankers and central rectangles differently. Again, OBA effects were not observed with flankers present (p > 0.74), although they were found with only two central rectangles (p < 0.01). Thus, simply having more objects in the display may be sufficient to eliminate OBA effects; grouping them into a single surface is not necessary. Because all rectangles were the same shape, however, grouping may still cause the loss of OBA effects. A second experiment tests whether the presence of different shape flankers is sufficient to eliminate OBA. Our evidence that adding two irrelevant objects to the display eliminates OBA supports the view that OBA effects are due to strategic rather than automatic allocation of attention (Shomstein & Behrmann, 2008).
MAP NSF BSC 0425650 & 0418179.