Peer supports 135Because teachers are rarely present during recess, they tend to select children as peer
models based on classroom characteristics and behaviors. Typically, selected peers
are obedient, self-confident leaders in the classroom who are considered socially
competent, prosocial, popular, and already have some interest and connection to
children with ASD, and willingness to participate in the intervention (Jackson and
Campbell, 2009; Locke, Rotheram-Fuller and Kasari, 2012; Kamps et al., 2014).
It is thought that children comprising these qualities may be more empathetic
or understanding of children with ASD and more willing to socially engage with
them as compared to non-peer models (Locke et al., 2012). There also is anecdotal
evidence that suggests that elementary-aged girls may be more likely than boys to
take on a “care-taking role,” and be more nurturing and responsive when acting as
peer models for children with ASD (Center and Curry, 1993; Chamberlain et al.,
2007; Owen-DeSchryver et al., 2008).
Garrison-Harrell and colleagues (1997) operationalized the selection of peer
models in their study of a peer network intervention for children with ASD by
using children’s social network status, or social positioning within their classroom.
Students completed a formal peer rating system and listed their friendship nomi-
nations (who they prefer to hang out with), which teachers then reviewed for peer
model selection to determine which students might be appropriate peer mod-
els. Despite consistencies in the factors used by teachers and researchers to select
peer models across studies, these characteristics have yet to be empirically tested to
determine which qualities encompass the ideal peer model.
Social demands dramatically change between elementary and secondary school;
therefore the criteria used to select elementary school–aged peers may not be
appropriate for selecting adolescent peer models. Peer relationships undergo signif-
icant change during adolescence as the nature of friendship shifts from gameplay
to deeper relationships that focus on commitment, closeness, security, intimacy,
and acceptance (Buhrmester, 1990; Berndt and Savin-Williams, 1993; Shulman,
Laursen, Kalman and Karpovsky, 1997). Peer relationships become more complex
and interactions more nuanced, making it difficult for adolescents with ASD to
navigate this new social landscape (Adreon and Stella, 2001; Locke et al., 2010). To
compound this issue, typically developing adolescents may be more aware of stu-
dents who may have a disability and less malleable in changing their attitudes and
behaviors towards adolescents with ASD (Humphrey and Lewis, 2008). As a result,
fewer established guidelines exist for selecting suitable peer models for adolescents
with ASD compared to those for elementary school–aged children.
There also are additional logistical factors to consider when implementing
peer-mediated interventions in middle and high school settings as compared to
elementary school, which may influence how peer models are selected to partic-
ipate. For example, in middle and high school, students have different schedules
and switch classes as often as six or more times a day; therefore, there is limited
consistency in the composition of the peer group throughout the day. Additionally,
there are fewer opportunities for unstructured social interaction in middle and
high school, as classes involve more independent work (Carter and Kennedy, 2006).