An Examination of Effectiveness and Visual Attention to Visual Instructional Supports for Children with Autism Spectrum Disorder

Abstract

Individuals with autism spectrum disorder (ASD) display widespread motor deficits in addition to the core social-communication and behavioral deficits. Children and adolescents have consistently demonstrated poor performance on commonly used clinical and standardized batteries of motor performance. It is unclear to what extent poor performance is related to administration instruction and/or known impairments in sensory information processing in this population. Individuals with ASD have difficulty processing sensory information, but demonstrate a relative strength in communicating through visual versus auditory means such as using pictures to help express and receive information. For this reason, visual supports such as picture cards, schedules, and videos are frequently implemented to aid communication and instruction in this population. Studies show that video modeling is an effective intervention to improve communication, social, and cognitive skills in individuals with ASD. Video modeling may have promise as a support for motor skill instruction in ASD. In this dissertation, we conducted two studies to address this issue. First, we conducted a meta-analysis examining the efficacy of video modeling to teach movement-related tasks in individuals with ASD. The results indicate that video modeling is associated with significant improvement in movement performance. Results also indicate study bias, such that studies that utilized a smaller number of samples/observations yielded larger effects of video modeling. The findings from this meta-analysis support the use of video modeling. The second study explored visual attention using of eye tracking to monitor gaze behavior participants with and without ASD in two visual support conditions (video and task card). Participants watched one series of skills presented as a video and another series presented as series of static images at different phases of the movement (i.e., task cards) while wearing eye-tracking glasses. Visual attention, as measured by horizontal gaze predictability and normalized fixations on the model in visual support conditions, did not differ by group, nor were there any group by condition interactions. A main effect of condition indicated that horizontal gaze was more predictable in the video condition. These results suggest that visual attention to the supports used in this study did not differ by group and that the dynamic nature of video condition may guide attention more than static presentation.