No Evidence for Reduced Global and/or Enhanced Local Visual Processing in Adolescents with ASD: Evidence From Embedded Figures and Configural Superiority Tests

Background:  Atypical sensory processing is often reported in autism spectrum disorders (ASD). Two of the most prominent theories, the Weak Central Coherence (WCC) and the Enhanced Perceptual Functioning (EPF) account, postulate that individuals with ASD are characterized by a weak or absent drive for global coherence and/or a superior processing of local details. A classic example of this local processing bias in ASD is the superior performance on the embedded figures test. The well-known configural superiority task (Pomerantz, Sager & Stoever, 1977) constitutes a very simple and ‘pure’ version of this embedded figure test. In this task, subjects have to identify the oddly oriented item in a display of four items, constituted either by four single lines (part condition) or by the same four lines with identical corners added to it (whole condition).  In typically developing individuals there is overwhelming evidence for superior performance on the whole condition as compared to the part condition.

Objectives:  In the present study we aimed to determine whether performance on the configural superiority task in high functioning individuals with ASD is less influenced by the context when processing ’parts’ versus ‘wholes’.

Methods:  A sample of 20 high functioning adolescents with ASD and early language delay was recruited (mean age 16 years, range 13-21 years), as well as a sample of 20 typically developing controls, matched for age and performance IQ (mean age 16 years, range 14-20 years). All participants performed a forced-choice computerized version of the embedded figures task and the configural superiority task.

(Preliminary) Results:  Controls were significantly faster and more accurate than individuals with ASD on the embedded figures test. On the configural superiority task, both groups presented an advantage of processing wholes versus parts, both in terms of accuracy and in terms of reaction times. Yet, there was no significant group difference or group by condition interaction, implying that both groups were equally influenced by the context effects of processing ’parts’ versus ‘wholes’.

Conclusions:  Adolescents with ASD showed no superior performance on the embedded figures task. Quite the contrary, they even performed significantly more poorly. Both groups were also equally sensitive to the context when processing ‘parts’ versus ‘wholes’ in the configural superiority task. Taken together, these findings do not support the hypothesis of reduced global or enhanced local visual processing as postulated by the WCC or EPF theories.