The Development of Perceptual Expertise for Faces and Objects in Autism Spectrum Conditions

Background: The high perceptual demands of human social interaction necessitate expertise for the rapid processing of faces. However, previous research indicates that this expertise for faces may be impaired in individuals with autism spectrum conditions (ASC). It is unclear if this deficit is related to a lack of social motivation or to a more general cognitive impairment associated with the process of gaining expertise. Furthermore, there is very little empirical evidence describing how individuals with ASC acquire perceptual expertise or how this process compares to expertise development in typical individuals.

Objectives: (1)To determine if face processing deficits in ASC are attributable to a lack of experience-dependent expertise with faces or to a more general deficit in gaining expertise. (2) To examine the extent to which individuals with ASC can develop expertise with a set of novel objects. (3) To compare the perceptual strategies developed for expertise processing to those involved in face processing, to determine if these mechanisms are functionally independent in individuals with and without ASC. 

Methods: This experiment was comprised of three parts: an initial session in the laboratory, ten sessions of expertise training, and a second laboratory session following training. In the laboratory sessions, participants performed two tasks: an inversion effect task and a composite effect task. As a measure of performance, response time and accuracy data were collected for both tasks. Between laboratory sessions, participants completed ten sessions of a training program with a set of complex, novel objects, known as Greebles. Like faces, Greebles can be categorized on basic and subordinate levels according to configural relationships among parts. Expertise was empirically defined as automatic subordinate level processing of Greebles.

Results: Individuals with ASC successfully developed perceptual expertise with Greebles, requiring the same quantity of experience to reach expertise as typical individuals. As experts, individuals with ASC showed a Greeble composite effect, but failed to show a significant Greeble inversion effect. Typical individuals, on the other hand, showed both composite and inversion effects. Furthermore, while typical individuals showed improvement specific to upright faces, individuals with ASC showed significant improvement on both upright and inverted faces.

Conclusions: The face processing deficit in ASC is likely to be attributable to a face-specific deficit in holistic processing (indexed by the composite effect), along with a more pervasive deficit in configural processing (indexed by the inversion effect). Therefore, both social and more general cognitive or perceptual models can be used to explain this impairment. Because Greeble training had an effect on face processing, these results suggest that at least some of the same neural substrates underlie face and expertise processing, providing further support for the expertise hypothesis. These findings indicate that with sufficient motivation and training, individuals with ASC show significant gains in expertise processing. However, these gains may be limited by general perceptual impairments that cannot be ameliorated even with extensive experience. Future face processing interventions should focus on encouraging greater social interest and on enhancing compensatory strategies to take into account the perceptual deficit in ASC.