136 Jill Locke and Colleen Harker
Consequently, adolescents must take advantage of other opportunities to socialize at
school (e.g., in the hallways between classes, at lunch, during extra curricular activ-
ities), which may be difficult for students with ASD to navigate. Consideration of
these logistical factors and others is important as they may affect the feasibility and
effectiveness of peer-mediated interventions (Carter and Kennedy, 2006).
To date, studies of peer-mediated interventions for adolescents with ASD have
relied on both logistical considerations (e.g., shared classes) and social characteristics
(e.g., social competence) when selecting peers. Schmidt and Stichter (2012) based
their peer selection criteria on those used by McEvoy and colleagues (1990) and
Kerr and Nelson (1983), which ultimately included: willingness to participate; good
attendance record; history of compliance with instructions; ability to make up any
missed school work; frequent and appropriate interactions with peers; and task
follow-through. Haring and Breen (1992) required that adolescents with ASD and
peer models had at least one mainstreamed class together and the same on-campus
job. They also required that the adolescents had interacted before in the past and
that they shared common hobbies and interests. Lastly, the authors ensured that the
adolescents with ASD agreed with the selection. After peers were recruited to par-
ticipate, they were then asked to identify two to four friends to participate. Similar
to elementary age children with ASD, the selection criteria for adolescent peer
models has yet to be empirically tested and the ideal peer model may likely vary by
context (e.g., classroom period, school, etc.).
Impact of being a peer model
Even though the extant literature points to many ways in which peer models have
helped students with ASD, parents, teachers, and school administrators often raise
concerns about possible social, behavioral, and academic consequences of being a
peer model (Ferraioli and Harris, 2011). Concerns have arisen that peer models
will miss valuable instruction time, emulate the inappropriate behaviors of children
with ASD, and/or experience negative social consequences from their peers for
associating with children with ASD (Ferraioli and Harris, 2011; Locke et al., 2012).
However, these concerns have not been substantiated in the literature. In fact, peer
models are more socially well connected with peers in their classroom’s social struc-
ture, have higher quality friendships, receive more friendship nominations, and are
more connected to children with ASD as compared to non-peer models at both
pre- and post-intervention (Locke et al., 2012). These findings suggest that the
social status of peer models is not negatively impacted by their participation in
peer-mediated interventions. Research also indicates that peer-tutoring interven-
tions may actually improve the academic performance of peer models that were
initially performing below grade level pre-intervention (Cushing and Kennedy,
1997; Shukla, Kennedy and Cushing, 1998; Shukla, Kennedy and Cushing, 1999).
Few studies have systematically examined outcomes for peer models partici-
pating in peer-mediated interventions for children with ASD. In one of the few
published studies on this topic, Locke and colleagues (2012) used empirical data