22082
Empirical Support for a Transactional Model of Spoken Language Acquisition in Preschoolers with Autism Spectrum Disorder
There is a pressing need to explain individual differences in “useful speech” of young children with autism spectrum disorder (ASD) because learning to use spoken language to communicate early in life has been repeatedly linked with long-term outcomes in this population. Transactional theory suggests that models considering both parent and child factors, and the dynamic way in which such factors interact, will best account for individual differences in spoken language acquisition. One instantiation of transactional theory specifically suggests that early child vocalizations that are more “speech-like” or “complex” elicit parent linguistic responses, which in turn scaffold spoken language learning. We have previously confirmed this model in a group of young children with developmental disabilities not due to autism (Woynaroski, Yoder, Fey, & Warren, 2014).
Objectives:
The present study tests whether the transactional model of spoken language acquisition additionally applies to preschoolers with ASD.
Methods:
We drew on extant data from a recent longitudinal investigation of useful speech development involving 87 preschoolers with ASD (Yoder, Watson, & Lambert, 2015). An index of the complexity of children’s vocalizations (i.e., consonant inventory in communicative vocalizations) was measured at Time 1 from the Communication and Symbolic Behavior Scales-Developmental Profile (CSBS-DP). Children’s spoken language outcomes were measured 8 months later at Time 3 using an aggregate of scores from the CSBS-DP word scale, the Macarthur-Bates Communicative Development Inventories: Words and Gestures expressive vocabulary checklist, and the number of different words produced in a semi-structured communication sample with an examiner. Midpoint parental linguistic mapping was measured at Time 2 (4 months after Time 1) in a 10 minute parent-child interaction. Mediation analysis with stochastic regression handling of any missing data points was used to test the statistical significance of the indirect effect of early child vocal complexity on later spoken language outcomes through mid-point parental linguistic mapping (Hayes, 2009).
Results:
As expected, children’s early vocal complexity predicted their later spoken language outcomes (r = .61; p < .001). Additionally, early child vocal complexity was positively associated with mid-point parent linguistic mapping (r = .33; p < .001), and mid-point parent linguistic mapping predicted later spoken language outcomes, controlling for early child vocal complexity (r = .24; p < .01). The indirect effect of early child vocal complexity on later spoken language outcomes through mid-point parent linguistic mapping was statistically significant. Thus, mid-point parent linguistic mapping mediated the relation between early vocal complexity and later spoken language in our sample.
Conclusions:
These findings lend empirical support to the transactional theory of spoken language development in preschoolers with ASD. We confirmed that one child factor (early vocal complexity) and one parent factor (linguistic mapping) contribute in a dynamic manner to impact spoken language outcomes in children with ASD. This result suggests that we might best support spoken language acquisition in this population by targeting both the complexity of children’s vocalizations and parents’ responses to such prelinguistic child productions. Implications for theory, research, and practice will be discussed.
