Traditionally, psychological research on autism spectrum disorder (ASD) has focused on social and cognitive abilities. Vision provides an important input channel to both of these processes, and, increasingly, researchers are investigating whether observers with ASD differ from typical observers in their visual percepts. Recently, significant controversies have arisen over whether observers with ASD differ from typical observers in their visual analyses of movement. Initial studies suggested that observers with ASD experience significant deficits in their visual sensitivity to coherent motion in random dot displays but not to point-light displays of human motion. More recent evidence suggests exactly the opposite: that observers with ASD do not differ from typical observers in their visual sensitivity to coherent motion in random dot displays, but do differ from typical observers in their visual sensitivity to human motion. This review examines these apparently conflicting results, notes gaps in previous findings, suggests a potentially unifying hypothesis, and identifies areas ripe for future research.
Autism is a pervasive developmental disorder that is characterized by a triad of symptoms: qualitative impairments in social interaction, delayed or impaired communication abilities, and stereotyped patterns of behavior or restricted interests (American Psychiatric Association [APA], 2006). In the psychological sciences, autism researchers have focused largely on the identification of and treatments for social and cognitive deficits. Visual perception provides an important source of information for both social and cognitive processes. Indeed, understanding how people with autism perceive their environments may be a necessary step toward understanding the social and cognitive deficits associated with autism. Consistent with this, substantial research has examined the relationships between autism and visual perception (see reviews by Behrmann, Thomas, & Humphreys, 2006; Dakin & Frith, 2005; Happé & Frith, 2006; Pellicano, Jeffery, Burr, & Rhodes, 2007; Schultz, 2005).
Within the last decade, numerous researchers have focused on the question of how observers with autism perceive movement, specifically, the movements of dotdefined surfaces and people. Initial studies reported that observers with autism show compromised visual sensitivity to coherently moving surfaces defined by randomly located dots but not to point-light displays (PLDs) of human movement. More recently, the opposite pattern of results has emerged, with reports of typical levels of visual sensitivity to coherently moving random dot surfaces and significant deficits in visual sensitivity to coherent human motion. The goal of this review is to examine this increasingly complex literature and to suggest ways to integrate apparently divergent findings.
Because autistic disorders show substantial heterogeneity, the term autism spectrum disorder (ASD) is used to reflect this behavioral continuum. ASD is not incompatible with normal or superior intelligence (Schultz, 2005). Indeed, most studies of people with ASD are conducted with individuals whose intelligence falls within the average to above-average range. These include studies of people with Asperger's syndrome (AS) who experience deficits in social interaction and stereotyped or rigid behaviors in the absence of early language delay (APA, 2006; Wing, 2000). Many studies of ASD and AS include individuals with normal IQs, because those on the lower functioning end of the spectrum may have difficulty completing experimental tasks and often have comorbid disorders that complicate the implementation and interpretation of psychophysical performance.
Local Processing Advantage in ASD During Form Perception
Extensive research suggests that ASD is associated with a certain perceptual style that includes supranormal local processing abilities (e.g., Frith, 1989; Jolliffe & Baron-Cohen, 1997; Shah & Frith, 1983, 1993). …