Abstract
PURPOSE: Children's ability to perform contour integration of individual elements in the presence of stimulus noise develops slowly (Káldy & Kovács, 2003; Kov ács, 2000). A recent study with 3-month-old infants has demonstrated a stark immaturity in their contour integration mechanisms (Gerhardstein, Kovacs, Ditre, & Feher, 2004). The goal of this study was to further investigate the developmental trend in infants' ability to integrate individual elements into whole contours. METHOD: Six-month-olds' discrimination of differently shaped contours was tested via a cueing paradigm in which a circle or square contour defined by the alignment of oriented Gabor patches, presented centrally and embedded in a background noise of randomly oriented Gabor patches, cued the subsequent presentation of a target on either the right or left. The relation of background noise spacing over contours spacing was set to 1.0 and 0.9. Eye movements were analyzed for correct anticipatory eye movements to the targets in response to which contour cue had been presented. RESULTS: All results were compared to random performance (no discrimination of contour cues) of 50% correct. Infants correctly anticipated for circles 70.8% and correctly for squares 69.5% in the 1.0 condition. For the 0.9 condition, infants correctly anticipated only 39.8% for circles and 60.1% for squares. CONCLUSIONS: These results indicate that infants are able to discriminate between circle and square contours better than chance in the 1.0 condition, but not for the 0.9 condition. These data suggest that 6-month-old infants' contour integration ability is not much different from that of 3-month-olds'.
Supported by Natural Science and Engineering Research Council grant 503860 to SAA