Examination of Script-Based and Non-Script Based Narrative Retellings in Children with Autism Spectrum Disorder

Background:   Narrative production is often challenging for children with Autism Spectrum Disorders (ASD), particularly in terms of creating coherent and cohesive stories (see Stirling et al., 2014).  However, differences between the narratives of children with and without ASD may be minimized when the cognitive and linguistic demands are reduced (e.g.,Losh & Capps, 2003; Novogrodsky et al., 2013).  Research on neurotypical (NT) children indicates that incorporating a script-framework into a story can improve narrative coherence and cohesion because having prior event knowledge can help children predict what will happen in a story (Davidson, 2006; Shapiro & Hudson, 1991). It is not known, however, how event knowledge might improve narratives in children with ASD. 

Objectives:   The current study was designed to 1) gain a comprehensive understanding of the pervasiveness of narrative difficulties in children with ASD by examining whether providing more structure, in the form of a script-framework, would allow children with ASD to produce more well-formed narratives, and 2) examine how Theory of Mind (ToM) ability may predict narrative production in children with ASD, regardless of script or non-script framework.  

Methods:   This study compared the narrative abilities of 19 children with ASD and 26 NT children (M= 10;0, SD= 1.6), on two story-retelling tasks: a script-based and a non-script based story.  Children were matched on age, nonverbal-reasoning, and receptive language (Table 1).  Narratives were coded for a number of microstructure and macrostructure variables.  Script-based stories were examined for the types of events included (script, non-script) to gain an understanding of children’s script knowledge. ToM was assessed using a battery of measures (e.g., Birthday Puppy Task, Strange Stories Test). 

Results:   Lexical diversity and productivity were relative strengths for children with ASD.  Unexpectedly, the narration of both stories was equally difficult for children with ASD for the majority of narrative variables assessed, including grammatical complexity, appropriate use of references, adverbials, and connectivity, structure, content, which resulted in narratives that were less cohesive and coherent than the NT group (Table 2).   All children told non-script based stories that were more semantically complex. Closer examination of the script-based story revealed that children with ASD were including the same number of script details as the NT children, but were less likely to include the non-script details that were essential to the plot of the story.  ToM ability predicted narrative coherence and cohesion for both story types, but only for the children with ASD. 

Conclusions:  These findings provide evidence that impairments in narrative microstructure and macrostructure may be pervasive across narrative type for children with ASD, even when using a retelling task.  Although children with ASD did not use the script-framework to produce more well-formed script-based narratives, the apparent difficulties with the script-based story appear to reflect more general narrative impairments, instead of abnormalities in their representation of script knowledge. These findings confirm the importance of considering socio-cognitive (ToM) impairments when studying the narrative abilities of children with ASD.