Influences of closure, occlusion, and size on the perception of fragmented pictures

Two experiments examined the ability to recognize a repeated figure in stimuli where only fragments of the figures were presented. Fragments were either closed or open regions presented either with or without an occluder that filled the spaces between the fragments (Experiment 1). Recognition performance was significantly better for open fragments, with or without an occluder, contrary to previous assertions that the occluder is the important factor for recognition. When stimulus size was varied, an occluder hindered recognition performance in the smallest size condition relative to unoccluded stimuli, which were equally easy to recognize across all size conditions (Experiment 2). The size results support the idea of size invariance in object naming. The difficulty in the smallest occluded condition is discussed in terms of differences in global and local processing depending on visual angle and sparsity.