Chameleon Effect in Autism: Decreased Mimicry of Body Movements in Social Interaction Contexts
Chameleon effect refers to nonconscious mimicry of interaction partners’ movements, postures and facial expressions (Chartrand & Bargh, 1999). Research shows that such mimicry facilitates interpersonal liking and prosocial behaviours in children and adults. It has been shown that children with autism spectrum disorders (ASD) mimic others’ emotional facial expressions less than typically-developing children (TDC). Facial mimicry deficiency might stem from broader attention and emotion recognition problems. Yet, given mounting research indicating autism-associated impairments in elicited motor imitation, spontaneous mimicry of a broader range of actions may also be decreased in ASD. Examining different action types can elucidate how mimicry contributes to social-communicative functioning in autism.
(1) To investigate whether children with ASD will mimic object-oriented and non-object-oriented body movements less frequently than TDC.
(2) To examine associations between decreased mimicry frequency and increased autism symptom severity through the Autism Diagnostic Observation Scale (ADOS-2) and parent ratings of Social Responsiveness Scale (SRS-2).
Data have been collected from 28 8-12 year-old children, with mimicry coding completed on 20 age- and IQ-matched participants to date (10 ASD:1 female, 10 TDC:2 females).
Children played a “memory game”, in which they watched a video of a narrator tell a story in five parts (2 baseline, 3 test). In the test blocks, the narrator performed four body movements per block in two categories (Figure-1): object-oriented (drinking) and non-object-oriented (yawning, arm scratching, face rubbing). Following each block, children told the story back to the narrator who was shown listening intently. Mimicry was assessed during both “listen” and “retell” phases.
A 2(TD vs ASD)*2(Baseline vs Test) ANOVA on the frequency of object-oriented actions (drinking) revealed no significant main or interaction effects. Analysis of non-object-oriented (yawning, arm scratching, face rubbing) actions revealed no main effect of diagnosis, but a main effect of phase such that overall, children mimicked more in the test phase than in baseline (F(1,18)=8.50, p=.009). Moreover, interaction effect for non-object-oriented actions was significant, showing that mimicry difference between baseline and test was higher in TDC as compared with children with ASD (Figure-2a; F(1,18)= 4.78, p=.04). Furthermore, increased autism symptom severity was associated with less mimicry of non-object-oriented actions among children with ASD (Figure-2b; ADOS-Total: r=.70, p=.03, ADOS-Social Affect subscale: r=.70, p=.02, ADOS-Restricted, Repetitive Behaviour subscale: r= .28, p>.05) and across the sample (SRS-2 Total: r=-.45, p=.05, SRS-2 Social Communication and Interaction subscale: r=-.43, p=.068, SRS-2 Restricted Interests and Repetitive Behaviour subscale: r= -.45, p=.05).
The findings reveal that children with ASD show reduced spontaneous mimicry of non-object-oriented actions and that this is associated with autism severity, particularly the social affect domain. In contrast, no significant differences between diagnostic groups were found in mimicry of object-oriented actions, potentially reflecting lesser availability of mimicry opportunities, or differences in salience of the object-oriented action. Our design provides a novel method that is both naturalistic and well-controlled to assess spontaneous mimicry in social interaction contexts. Investigating differences in mimicry of different types of actions can elucidate non-verbal factors influencing social interactions and autism-related social-communicative issues.