Individuals with ASD Imitate By Way of Mirroring More Than Their Typical Peers

Background: When performing elicited imitation, one can either mirror (e.g., use the right hand when the model is using their left, as if looking in a mirror) or transpose (i.e., imitate using the same hand as the model; see Figure). Mirroring is generally thought to be easier than transposition, potentially because mirrored movements have more visual parallels to the model’s movement, and potentially because transposition requires greater perspective taking. From a developmental perspective, typically developing children are more likely to spontaneously mirror, while adults are more likely to transpose. When prompted to mirror, typical children and adults’ performance tends to improve, suggesting that mirroring benefits imitation accuracy, a finding that has not been consistently demonstrated in ASD. Studies of school-age children have demonstrated that in goal-directed imitation, those with and without ASD do not differ in their spontaneous tendency to mirror. It is unknown whether adolescents and adults with ASD make the same shift to favoring transposition during imitation shown by typically developing adults.

Objectives: To determine whether older children, adolescents, and adults with ASD transition to a primarily mirroring strategy during elicited imitation.

Methods: Verbally fluent, right-handed children, adolescents, and adults with ASD (n=29) and age- and IQ-matched typically developing controls (TDC; n=14, see Table) completed a brief praxis battery, administered by right-handed examiners. The battery consisted of 19 items for which participants completed specified actions either by imitating the examiner (imitation condition), or following a non-imitation prompt (i.e., following verbal instruction with or without a relevant prop). Participants were explicitly instructed that they could use whichever hand they wanted to complete the actions. Each item was later coded for whether the participant used their right or left hand. Because all participants and examiners were right-handed, use of the right hand during imitation can be taken to reflect transposition, whereas use of the left hand reflects a mirroring strategy.

Results: Participants across groups primarily or exclusively used their right hand during the non-imitation condition (see Table for details). During the imitation condition, the TDC group continued to overwhelmingly select their right hand, showing the expected preference for a transposition strategy. In contrast, participants with ASD often switched to using their left hand, suggesting a stronger inclination toward mirroring compared to those without ASD (independent-samples t-test, p=.004, d=1.57).

Conclusions: The present study shows that older children, adolescents, and adults with ASD spontaneously imitate by way of mirroring more than their typical peers (who almost entirely transpose), even when doing so requires them to use their non-dominant hand. This suggests that, even beyond early childhood, people with ASD may continue to rely on a developmentally earlier strategy for imitation. Mirroring may be beneficial because actions occur in a more familiar spatial field, thereby reducing working memory load. Future research can test these potential mechanisms, and can examine whether strategy selection (i.e., mirroring vs. transposition) relates to task performance. Understanding how older individuals with ASD spontaneously imitate may shed light on self-other mappings and automatic perspective taking in this population.