Assessing Gesture In Young Children with Autism Spectrum Disorders

Background: Gesture has been correlated with language outcomes in young typically developing children and children with autism spectrum disorders (ASD).   Given the importance of gestures as a predictor of later language outcomes, researchers have posited that gesture profiling should be incorporated into communication skills assessment and intervention procedures (Crais, Watson, & Baraneck, 2009).

Objectives: The objective of this study was to determine whether performance of three gestures (point, give, show) was correlated across three different assessment instruments routinely used to evaluate children with ASD.  The methodology of two of the instruments, the Autism Diagnostic Observation Schedule (ADOS) and the Early Social Communication Scales (ESCS), involved children’s participation in tasks designed to elicit behaviors, whereas the MacArthur Bates Developmental Communication Inventory – Words and Gestures (MCDI-WG) is a parent checklist. 

Methods: Participants were children with ASD (mean chronological age = 30 months, SD = 3.77 months) without phrasal speech (n=78) who were participating in a larger ongoing longitudinal study.  The gestures point, give and show were assessed using three different instruments.  To equate scoring across the ADOS and ESCS children received a score of 0 if they demonstrated the behavior more than one time, a score of 1.5 if the behavior was observed once, and a score of 3 if the behavior was never observed.  Ratings on the MCDI-WG were transformed according to the same guidelines.  Children received a score of 0 if the behavior was reported to occur often, a score of 1.5 if the behavior was reported to occur sometimes, and a score of 3 if they parent reported that the child did not yet demonstrate the behavior.

Results: Correlations were run for each gesture.  The correlations between pointing on the ESCS and ADOS (r=.47), the ADOS and MCDI-WG (r=.38), and MCDI-WG and ESCS (r=.30) had medium effect sizes (Cicchetti et al., 2010).  The correlation between showing on the ESCS and ADOS (r=.449) also had a medium effect size.  There was not a significant correlation between showing on the MCDI-WG and ADOS (r=.107) or the MCDI-WG and ESCS (r=-.032).  The correlation between giving on the ESCS and the MCDI-WG (r=.261) had a small effect size.  There was not a significant correlation between giving on the ADOS and the ESCS (r=.065) or the ADOS and the MCDI-WG (r=.072).

Conclusions:  The results from this study suggest that gesture performance in young children with ASD is not entirely consistent across different assessment instruments.  This may be attributed, in part, to the design of the instruments.  For example, pointing is defined much more stringently on the ADOS and ESCS as compared to the MCDI-WG.  In addition, many of the tasks in the ESCS are designed to elicit pointing gestures rather than giving gestures.  Thus, when profiling the gestures of children with ASD it is important to gather data from multiple sources and consider performance in light of the instrument being used.