Modified COSMIC 5
Although providing valuable information on social communication skills, established
instruments such as those listed above do not address children’s functioning within real-world
settings. As such, they have limited utility in informing the extent to which progress achieved in
treatment might generalise outside of the specific intervention context (National Research
Council, 2001). Moreover, at all levels of functioning, individuals with ASD have been found to
show greater impairment during real-world social interactions than during assessment situations
which provide structure and scaffolding (e.g., Wimpory, Hobson & Nash, 2007). Existing
instruments therefore evaluate the skills of children with ASD in contexts that are more likely to
be facilitative, rather than in contexts which might capture the true extent of functional
impairment (e.g., during everyday interactions with peers). The need to develop such direct and
ecologically valid measures is therefore clear.
A small number of such measures do exist, but are too limited to successfully capture the
full range of skills pertinent to the evaluation of social communication in children with ASD. For
example, Carr and Felce (2007) developed a naturalistic measure to assess effectiveness of a
Picture Exchange Communication System (PECS) intervention. However, only interactions
between child-adult dyads were evaluated, and the researchers did not delineate the functions of
communication bids observed. Roos, McDuffie, Weismer and Gernbacher (2008) conducted a
comparative study of child skills used across a structured clinic assessment and a naturalistic
home play session. Again, only examiner-child interactions were observed, and this evaluation
pertained only to non-verbal joint attention bids rather than a broader range of possible social
communication behaviours. Using a more comprehensive measure designed to evaluate various
forms, targets, and functions of communication bid (Watson, Lord, Schaffer, & Schopler, 1987),
Stone and Caro-Martinez (1990) found children with autism to communicate only 3 to 4 times