Abstract
Rensink (1999) found that visual search for the absence or conjunction of change had a capacity of about 1; this “magical number one” was taken as evidence for a nexus of attention that mediates change perception. The studies here explored this idea further, using arrays of horizontal and vertical rectangles under flicker conditions, where displays continually alternated between “on” (exposures of 160-800 ms) and “off” (blank fields of 120 ms).
For simple change involving orientation, capacity for presence was 3.5, while capacity for absence was 1.2. When the rectangles were embedded in ellipses, capacity for presence decreased by 50%, while capacity for absence remained unaffected. Likewise, when an irrelevant background grid was placed between items, capacity for presence decreased by 50%, while capacity for absence remained unaffected. This supports the idea that detection of both presence and absence of change is mediated by a nexus, with the higher capacity for presence achieved via some form of configural organization.
For compound change involving orientation and polarity, capacity for presence (i.e, conjunction of changing properties) did not exceed 1. Interesting, capacity for absence (i.e., a single change among conjunctions) was 40% higher. This indicates that the low capacities were not due to high levels of noise from the other items, but resulted from a structural limitation - viz., the nexus. It also suggests that the slight over-unity excess for both absence conditions might be due to the use of a comparator that can be applied to 2 items at a time.
Support provided by the Natural Sciences and Engineering Research Council of Canada and Nissan Motor Co., Ltd.