Speed Discrimination Abilities in Typical Development and in Children with Autism

Background:  Despite much research focusing on visual motion processing in individuals with autism, relatively little has focused on the processing of speed. Autobiographical reports suggest that the world is ‘moving too fast’ for at least some individuals with autism (e.g., Williams, 1992), and it has been shown that slowing down video presentations can aid recognition of facial expressions in individuals with autism (Gepner et al., 2001).  Children with autism might therefore show impaired speed processing abilities, particularly for fast moving stimuli.  Furthermore, previous research has shown that some individuals with autism have elevated motion coherence thresholds (e.g., Milne et al., 2002; Pellicano et al., 2005).  Since both speed processing and global motion processing rely on integrating neuronal responses in extrastriate cortex (area MT/V5), difficulties in these two processes might co-occur in autism.

Objectives:  The aims of this study were threefold: 1) to investigate age-related changes in speed discrimination abilities in typically developing children about which, at present, very little is known, 2) to address whether such speed discrimination abilities are atypical in children with autism, and 3) to examine the relationship between speed discrimination abilities and motion coherence abilities in both typical development and autism.

Methods:  To address our first objective, we measured speed discrimination thresholds using random dot stimuli in a typically developing sample of children aged 5, 7, 9 and 11 years and adults (total n = 116), with two different reference speed conditions (1.5 deg/sec and 6 deg/sec). To address our second and third objectives, we are administering both speed discrimination tasks and motion coherence tasks using the same reference speed conditions (1.5 deg/sec and 6 deg/sec) to a group of children with autism aged between 6 and 13 years, and an age- and non-verbal ability-matched typically developing comparison group.

Results:  The typically developing sample showed age-related improvements in speed discrimination sensitivity for both reference speeds, with participants being more sensitive to the faster reference speed at all ages.  Sensitivity to the slower reference speed became adult-like by 11 years, whereas sensitivity to the faster reference speed became adult-like by 9 years.  This suggests that sensitivity to slower speeds matures later than that to intermediate speeds. Data collection with individuals with autism is still ongoing.

Conclusions:  In typical development, there is a reasonably protracted development of speed discrimination abilities, with adult-like levels being reached later for the slow reference speed than the faster reference speed.  If individuals with autism show particularly pronounced impairments in processing faster speeds specifically, this would suggest a deviant rather than a delayed pattern of development.  We will examine the association between speed discrimination and motion coherence detection in both typical development and in autism.