Quality Matters: Differences Between Expressive and Receptive Non-Verbal Communication Skills In Children with Autism

Background: Expressive and receptive nonverbal communication skills, such as facial expressions and prosody, are interrelated in typical individuals, but very little is known about the specific relationship between these two modalities in children with autism.

Objectives: To determine patterns of qualitative and quantitative differences for receptive vs. expressive facial and vocal communication skills between children with autism and their TD peers. 

Methods: Participants were children with autism (N=7 or 11, depending on task) and typically developing children (N=5 or 6 depending on task) aged 9-18.  We consolidated data from nine measures (i-ix) taken from five studies (1-5):  1. Facial expression and prosody productions recorded during a story-retelling task and coded for naturalness of expression.  The autism group was found to be more awkward than the TD group in faces (i) and voices (ii).  2.  Length of whole-word production in a lexical stress task.  The autism group showed abnormally long productions for noun-phrases (iii) (hot dog) and compound nouns (iv) (hotdog). 3.  Auditory-visual integration of speech information in an onset-asynchrony detection task.  The autism group was as accurate as the TD group at audio delays of 10 (v) and 12 (vi) frames.  4.  Accuracy of matching emotional facial expressions to auditory-only emotional sentences. The autism group was less accurate than the TD group when prosodic expressions and face choices were more subtle (vii).  5.  Ability to place individual images taken from a facial expression video in their correct dynamic sequence.  The autism group was less accurate than the TD group when eyes were visible (viii), the groups were equal when eyes were masked (ix).  To conduct a meta-analysis of these results we calculated z-scores for each subject’s performance on every measure and produced two aggregate scores, one for the four expressive (i-iv) and one for the five receptive tasks (v-ix).

Results: A Wilcoxon Rank Sums test revealed that the autism group had significantly larger z-scores than the TD group for the expressive aggregate, but not the receptive aggregate (p<.025).  Within the autism group, a Kruskal-Wallis test showed that the z-scores for expressive tasks were significantly larger than for receptive tasks (p<.05).  No such difference was found in the TD group. 

Conclusions: Our data show that the facial and vocal affective communication skills of individuals with autism have a greater difference from the typical mean for expressive tasks, but not receptive tasks.  Based on the significant difference in z-scores for expressive vs. receptive tasks within the autism group, we also hypothesize that their expressive skills are potentially more fragile than their receptive skills.  Across tasks, receptive skills were measured in objective accuracy scores, while expressive ability was measured by qualitative methods, such as length of whole word production or perceived naturalness of facial and vocal expressions.  Taking this distinction into account, our data indicate that individuals with high functioning autism are capable of achieving typical accuracy in a variety of receptive nonverbal communication tasks, but still exhibit significant qualitative differences from their TD peers in the production of those same skills.