Beyond Sentences: Using the Expression, Reception and Recall Instrument (ERRNI) in School-Aged Children with Autism Spectrum Disorder (ASD)

Background: For school-aged children, most standardized language tests do not extend beyond evaluating the individual word or sentence level. Such measures underestimate the extent of language difficulties that arise when a person is required to generate utterances that hang together coherently to describe an event. Thus, narratives have emerged as a tool to examine language beyond the sentence level. Research on narratives and ASD has focused on detailed analysis of narrative productions in small samples, across wide age ranges, of high-functioning children, adolescents, and adults. Sampling across widely discrepant ages likely masks important language characteristics because different developmental levels of skill are averaged.  Plus, the labor-intensive nature of detailed narrative analysis precludes widespread use both in research with large samples and in many clinical settings. In order to characterize the language skills of school-aged speakers with ASD, a standardized test of narratives may be useful. The Expression, Reception and Recall of Narratives Instrument (ERRNI) is one possibility. It assesses narrative production and comprehension, providing standard scores (SS) on the amount of relevant story content or “gist” (Ideas) and, understanding of a story (Comprehension). 

Objectives: Objectives were to examine the language of a large cohort of 8- and 9-year-old children with ASD using the ERRNI. We aimed to determine (1) if performance on the ERRNI was different from performance on other standardized language measures, and (2) the ERRNI’s relationship to other indices of cognitive and communicative functioning. 

Methods:  Participants (n=74, 63 males) were selected from Pathways in ASD, a multisite Canadian longitudinal study which has followed an inception cohort of rigorously diagnosed children with ASD from time of diagnosis.  At a mean CA of 8.6 years, participants were administered a battery of standardized tests, including the Wechsler Intelligence Scale for Children, 4^th ed. (WISC-4), Clinical Evaluation of Language Fundamentals-4 (CELF-4), the Children’s Communication Checklist – Version 2 (CCC-2), Vineland Adaptive Behavior Scales, 2^nd ed (VABS II) and the ERRNI.  The Semantic-Pragmatic Profile (SPP; Yitzhak, et al., 2011) was generated from ADOS data collected at the age 6 assessment.

Results: Average performance on ERRNI Ideas (SS=78.7) was significantly lower than average performance on the CELF-4 Core Language Composite (CLC) (SS=89.5) (t(73)=-5.64, p <.01).  ERRNI Comprehension SS was comparable to CELF-4 CLC SS.  ERRNI Ideas  and Comprehension standard scores were significantly correlated with the WISC Perceptual Reasoning Index (Ideas r=.237, p =.042; Comprehension r=.407, p <.01), CELF-4 CLC SS (Ideas r=.422, p<.01; Comprehension r=.567, p<.01), VABS Communication Domain SS (Ideas r=.237, p<.05, Comprehension r=.334, p<.041) and negatively correlated with SPP raw scores (Ideas r=-.317, p=.003, Comprehension r=-.207, p=.039).  Ideas SS was also negatively correlated with CCC-2 Social Interaction Deviance Composite (Ideas r=-.216, p=.03), and Comprehension SS was correlated with CCC-2 Global Communication Composite (r=.261, p=.012).

Conclusions: These findings suggest that the ERRNI Ideas index measures skills other than those captured on traditional standardized tests of language structure and content.  In addition, ERRNI scores were positively related to adaptive communication and negatively related to scores on the SPP, where higher scores indicate more impairment.