Objectives: The aim was to explore the language trajectories of children on the autism spectrum (including autism, Asperger’s, and PDD-NOS), a group of children and a typical group of children using social affect as a covariate within the models to identify whether social affect contributes to the patterns of language development observed.
Methods: The present study utilizes previously collected data gathered as part of three longitudinal investigations. Ninety children were assessed at multiple time points between 12 and 48 months of age. Language development was measured with the total number of nouns and verbs produced as measured by one of three versions of the MacArthur-Bates Communicative Development Inventory (CDI: Fenson, Dale, Reznick et al., 1993). Social affect was measured with the Autism Diagnostic Observation Schedule (ADOS; Lord, C., Risi et al., 2000).
Results: To assess the language trajectories of verb and noun production among children in each group (autism, PDD-NOS, non-spectrum, typical) HLM 6 multilevel modeling was utilized. Chronological age was used as the primary predictor in the language models. Chronological age was found to significantly predict the number of verbs produced. When added into the basic model, non-verbal IQ (NVIQ), and diagnosis accounted for a portion of the variance explained in the basic model. Social affect did not contribute to the model when included.
Conclusions: The number of nouns and verbs produced by children in the present study was significantly predicted by NVIQ and diagnosis. The language models were similar in their patterns and social affect did not differentiate them. Contrary to previous research (Rapin & Dunn, 2003), the present study indicates that children with autism and with developmental delays are significantly different based on the main effects of diagnosis. Thus, clinical implications include the importance of verb development among children that may be indicative of differential diagnosis in cases that appear similar on the surface.