Social Preference In Children with ASD: Exploring the Gray Area

Background:  Children with ASD experience complex social difficulties in school. Research suggests that inclusion may not be sufficient to fully integrate them into the classroom social structure. This often results in higher rates of isolation, and less reciprocal friendships when compared to typically developing peers.

Objectives:  This paper analyzed the social network status of 60 fully included elementary school children with ASD and 60 of their matched peers. The study aim was to examine the reliability of peer nomination methodology and compare social impact, social preference and social network scores.

Methods:  Children were asked to list all children in their class that they liked to play with as well as children they did not like to play with. Indegree scores were obtained for each child with ASD and a matched peer from their class, by tallying how many times the child was nominated as a friend. Rejections were tabulated in the same fashion, using the responses from the “children I do not like to play with” question. Social impact score (SI) was determined by standardizing the indegrees and rejections, and adding the two. Social preference score (SP) was determined by subtracting the standardized rejection score from the standardized indegree score. Asking children to list groups of children that play together in their class, and then analyzing the co-occurrences for statistical significance determined social Network Centrality score (SNC). Children were grouped into four categories: isolate, peripheral, secondary, and nuclear. 

Results:  Children with ASD differed in their SNC, SI, and SP scores than matched peers (p<.05). The greatest difference was found in SNC and SP scores (p<.001). Further analyses indicated that across the full sample, SI and SP accounted for 29% of the variance in SNC score, with varying contributions (SI r²=.252; SP r²=.334). However, when children with ASD were compared to their matched peers, it was found that SI and SP were only predictors of SNC in children with ASD (p<.001). 

Conclusions:  Examining the relationship between SI, SP and SNC scores gives a complete picture of classroom social structures. Our data indicate that for children with ASD, being well liked by classmates was an important predictor of group centrality, or connection with peers. It appears that typically developing children may have more ways to connect with peers, beyond being well liked, than children with ASD. Therefore it is important for school-based interventions for children with ASD to facilitate connection and engagement with peers, which could potentially compensate for lower social preference.  Additionally, our findings highlight the social roles of colleagues and acquaintances, which may be underdeveloped in children with ASD. These data indicate that social structures in elementary school can be multifaceted, and perhaps children with ASD falter by missing the gray areas of friendships.