Abstract
BACKGROUND: Configural asymmetries refer to differences in visual search performance caused by the orientation of otherwise identical search elements (Junge, VSS 2008). For example, bicolor elements split horizontally (up/down configuration) are searched more efficiently than identical elements split vertically (left/right configuration). This study tested the robustness of these asymmetries to manipulations of the spatial arrangement of the search elements. The spatial arrangement of the elements (pseudo-random vs. regularly positioned) and spatial separation between search elements were manipulated to influence global vs. local processing of the search displays. Specifically, we tested the hypothesis that the magnitude of the asymmetry might be attenuated with displays favoring global processing of the search elements (e.g., regularly positioned elements in close proximity).
METHODS: Three observers participated in two latency visual search experiments. Experiment 1 investigated the effects of set size and the spatial arrangement of the search elements (regularly positioned, slightly irregular, moderately irregular, and pseudo-randomly positioned) on search performance for horizontally and vertically split elements. Experiment 2 tested the effect of element separation on search performance for both horizontally and vertically split elements.
RESULTS: More efficient processing of horizontally split elements compared to vertically split elements was replicated. The overall spatial arrangement of the search elements affected search performance; regularly positioned elements were searched more efficiently than pseudo-randomly arranged elements. Surprisingly, the elements' spatial arrangement had little effect on the magnitude of the configural asymmetry. Similarly, element separation had little effect on the magnitude of the asymmetry, although a trend was observed in which displays composed of regularly positioned elements in close proximity exhibited a reduced configural asymmetry.
CONCLUSION: The configural asymmetries were remarkably robust to manipulations of the spatial arrangement of the search elements. The measurements point to a fundamental difference in how objects requiring left/right vs. up/down comparisons are processed visually.