Abstract
Previous research has shown that comparisons among object-properties are more efficient within a perceptual “object” than between distinct objects, a result referred to as the “object-superiority effect.” We tested this effect in a stimulus configuration in which the global organization into objects was induced by an isolated luminance change in a region remote from the locations of the critical properties. Observers were presented with a checkerboard pattern, including four grid squares (a), (b), (c) and (d) (see Figure at http://ruccs.rutgers.edu/~elanbz/checkerboard.html), on which two probes could appear. It also contained two variable squares (x) and (y) whose luminance was varied across conditions. Depending on the luminance of (x) and (y), squares (a) and (b) were perceived to be either separate squares in the checkerboard pattern or part of the same continuous object (a bar) in one of three different ways: in front of an occluder, behind an opaque occluder, or behind a transparent occluder. In the “separate squares” organization, the comparison between (a) and (b) and (a) and (c) is always between objects. However in the three “bar” organizations, the comparison between (a) and (b) is within-object, but (a) to (c) is between objects. Results showed a significant object-superiority effect just in those cases where the luminance manipulation induced a continuous object (bar) organization; subjects' performance was better on within than between trials, but there was no difference on the “separate squares” case. These results demonstrate the potential influence of small local manipulations on the global perceptual organization of the image, and also highlight the importance of transparency in the determination of coherent surfaces.