Abstract
While speech perception is thought of as an auditory-first process, it is clear that integrating visual speech has dramatic behavioral impacts. This audiovisual integration of speech is impaired in children with autism. We investigated the role that temporal processing has on audiovisual speech integration. We measured the temporal binding window (TBW), a probabilistic construct reflecting the interval of time within which two sensory signals may be perceptually bound, and related this to the McGurk Effect, a measure of audiovisual speech integration. Three age groups of participants, 6-9, 10-13, and 14-18yo (24 ASD, 40 typical) performed two tasks. To measure the TBW, participants completed an audiovisual simultaneity judgment task where speech tokens were presented at varying stimulus onset asynchronies, 500ms A-V to 500ms V-A. Additionally, participants completed a McGurk task in which they reported their perception to congruent /ba/ and /ga/ utterances and an illusory condition with an auditory /ba/ presented with a visual /ga/. The typical TBW (resembling a normal curve) was wide and symmetrical at the youngest age and narrowed first with the audio-first conditions, producing an asymmetry, followed by a general narrowing of the TBW. Comparatively, ASD individuals showed wider and more symmetrical TBWs, suggesting that these developmental changes failed to occur across age groups. ASD groups also showed a decrease in McGurk perceptions, and importantly, this measure was significantly correlated with the individuals’ TBW. The narrower the TBW, the greater the McGurk Effect. Thus, the development of temporal multisensory function was severely impaired in ASD. The ASD group failed to develop an asymmetrical TBW that mirrors the natural statistical relationship between auditory and visual stimuli in the environment. Individuals’ TBW widths were significantly correlated with their perceptual fusion of speech stimuli, suggesting this temporal deficit may cascade into deficits of higher-level cognitive processes such as speech perception.
Meeting abstract presented at VSS 2012