Pretend Play As a Predictor of Expressive and Receptive Communication Skills in Preschool Aged Children: The Relative Contributions of Functional Versus Symbolic Play and Shared Versus Solitary Play

Background:  Pretend play has been found to be correlated with language development in both typically developing (TD) children and children with Autism Spectrum Disorder (ASD), although children with ASD exhibit less play (McCune-Nicolich, 1981; McCune, 1995, Mundy, et al., 1987). However, the more complex relationship between different types of pretend play (functional and symbolic) and expressive and receptive language is not well understood in children with ASD. Symbolic pretend play in particular has been theorized to have a predictive relationship with expressive language development in children as both require a similar form of symbolic understanding that is not required in functional pretend play (Lewis et al., 2000; Toth et al., 2006).  There is also limited research testing whether symbolic pretend play is more predictive of expressive and receptive language than functional play for children with ASD.  Finally, although pretend play is often social in nature, there is a paucity of research examining pretend play that is shared compared to solitary. 

Objectives:  To examine the ability of functional pretend play and symbolic pretend play to predict expressive and receptive language and explore differences in the relationship between shared versus solitary pretend play and language in children with and without ASD. 

Methods:  Data was collected from 11 children with ASD (Mage = 60.18 months) and 11 TD children (Mage = 40.45 months) matched on cognitive ability.  Measures included the Autism Diagnostic Observation Schedule (ADOS), Psycho-Educational Profile- 3^rd Edition, Preschool Language Scale- 4^thEdition, and the Play Assessment Scale (PAS). 

Results:  Together, functional and symbolic pretend play predicted an additional 8% of the variance in expressive language, over and above cognitive ability and diagnosis (F_(2,16) = 4.94; p = .02).  Only cognitive ability (β = .43, p = .00) and symbolic pretend play (β = .41, p = .02) contributed significantly to the final model.  Functional pretend play did not contribute significantly to variance in expressive language.  Within symbolic pretend play, solitary and shared play predicted an additional 7% of the variance in expressive language, over and above cognitive ability and diagnosis (F_(4,15) = 4.57, p = .00). Only cognitive ability (β = .43, p = .00) and shared pretend play (β = .37, p= .02) contributed significantly to the final model. Neither functional versus symbolic pretend play nor shared versus solitary pretend play predicted receptive language over and above cognition and diagnosis. 

Conclusions:  Results indicate that symbolic pretend play, particularly play that is shared, is related to expressive language but not to receptive language when controlling for cognition and diagnosis in children with ASD and TD children.  These results support the theoretical link between symbolic pretend play and expressive language (Stanley, & Konstananreas, 2007).

In addition, the results suggest that play that is shared may provide opportunities for expressive language that are not present in solitary play.  Further research could explore longitudinal relationships to determine whether increases in symbolic pretend play (particularly shared play) lead to increases in expressive language for children with ASD and for TD children.