Abstract
Our visual system deals with vast amounts of information at any given moment. In order to be able to face this monumental challenge, our brains use the world's statistical regularities to process information. Local features are collapsed down and used to represent the gist, a phenomenon known as summary representation or ensemble perception. Recently, researchers studying enhanced local or reduced global visual processing in autism spectrum disorder (ASD) have started to explore ensemble perception, as it brings about a new way of tapping into these atypical processing styles. The current study is the first to investigate ensemble perception in children with ASD by (1) administering a pair of tasks pioneered by Ariely (2001), evaluating both member identification ("Was this item a member of the set?) and mean discrimination ("Was this item larger or smaller than the mean size of the set?"), and (2) manipulating the amount of external noise to allow a more detailed evaluation of task performance. Age, intellectual abilities, sensory profiles and ASD phenomenology of both the ASD group and the matched, typically developing (TD) group were taken into account (N= 46, 8-15 years). Results show that both groups were equally accurate on the member identification task, a test of "local perception", while the ASD group performed better than the TD group on the mean discrimination task, a test of "gist perception". In addition, in both tasks, performance of the TD group proved more sensitive to the degree of external noise present in the display, than performance of the ASD group (p < .001). Somewhat surprisingly, these results suggest that individuals with ASD are able to extract information, both at the local and global level, and information processing in ASD is more robust against external noise in the displays than in TD individuals.
Meeting abstract presented at VSS 2016