Academic Engagement of Minimally Verbal Children with ASD At School: Virtual Reality Paradigm with Secondary Students

Background:   Social attention theory has contributed to clinical advances for preschool children with autism, but has had less impact on school-aged children. In school-aged children social attention become more complex, such as in public speaking tasks where children must coordinate attention with multiple social partners while also attending too and referring to their internal thoughts and representations. Recent data suggests that more secondary school-aged children with ASD in inclusive setting rarely engage in this type of public speaking in frontal of a class (51% to 71%) compared to peers (32%, IES, 2007).  We have developed a virtual reality classroom public speaking paradigm to better understand the impediments to development of this type of complex social attention in secondary students with ASD.

Objectives: The proposed study compares the development of complex social attention to avatar peers versus complex attention to non-social stimuli in a virtual classroom paradigm in order to understand differences in age effects and associations with measures of behavior, learning and cognition specific to social attention in students with ASD.

Methods: Forty students with ASD, IQ = 40, and age-matched typical 8- to 16-year-old control children participated in this study. Participants were presented with a battery that included measures of behavior and learning problems (Connors Parent Report and Multidimensional Scale for Anxiety in Children), IQ (WISC-IV), working memory (WRAML), Academic Achievement (WIAT-II) and two virtual reality paradigms assessing complex social and non-social attention. These required children to answer self-referenced questions while directing their attention to each of nine avatar peers (Social Condition), or nine inanimate obelisques (Non-Social Condition) positioned around the virtual classroom.

Results: Preliminary data are currently available on 15 children in each diagnostic group. Analyses of these data revealed evidence that children with typical development display greater age related advances in social attention than children in the ASD sample. In addition social attention measures were related to cognition and behavior to a greater degree than were non-social measures; however the pattern of these differences varied across diagnostic groups. In the ASD sample, frequency of social orienting was negatively related to parents ratings of executive function problems (-.51, p < .075) and an overall problems score on the Conners (-.71 p < .05). Conversely, in the typical sample social attention was positively related to visual working memory on the WRAML (.72, p < .05) but negatively related to self reports of social anxiety (-.71, p < .05). Additional data on the full sample, including data on the relations between social-attention for both groups will be presented.

Conclusions: This study provides additional support for the validity of virtual classroom measures for examining developmental differences in attention that may impact classroom performance in school aged children with autism. The intervention potential of VR classroom paradigms will also be discussed.