Abstract
We examined the neural correlates of both the configural superiority effect (CSE) and specific individual emergent features (EFs) via a 128-channel EEG system. We tested 20 healthy subjects in an odd quadrant task where subjects searched for the one unique stimulus among three homogeneous distractors. Experiment 1 used three display types: Base displays composed of negative and positive diagonal lines; CSE displays composed of those same four diagonal lines combined with homogeneous L-contexts to form arrows and triangles containing multiple EFs such as closure and intersection differences known to create large CSEs; and Configural Inferiority Effect (CIE) displays again composed of the Base display lines combined with homogeneous contexts to form stimuli lacking EF (configural) differences. Significant differences between displays were found for all ERP components (as early as the P100), as revealed by larger amplitudes at lateral occipital sites and faster responses at frontal sites for CSE displays, suggesting a unique time course of neural processing for CSE displays. In a second experiment, we used odd quadrant displays with dot stimuli producing the EFs of Proximity, Linearity, Orientation, and Surroundedness (Pomerantz & Portillo, 2011) to further investigate whether these EFs exhibit distinct neural correlates. Significant differences among EFs were found for the EEG but not for the behavioral data. EFs did not lead to significant differences before the N170 ERP component. Orientation, Linearity, and Proximity significantly differed from each other at the N170 and at later time windows (300-500ms after stimulus onset). In general, Orientation responses evoked larger ERPs and some hemispherical asymmetries. Our results are in accordance with the previous literature regarding the nature of the CSE as well as the hierarchical structure of EFs (Pomerantz et al., 1977, Pomerantz and Portillo, 2011) and are the first demonstrations of the possible neural correlates of these effects using EEG.
Meeting abstract presented at VSS 2016