The Effect of Perspective and Training on Imitation in Autism

Background:  

Imitation is important for learning new skills, along with social activities such as play and developing affiliations. Therefore, exploring imitation ability in autism is particularly relevant. In our previous work we showed that neurotypical participants imitated changes in hand movement speed in the absence, but not presence, of visual goals^1-2. However, an autistic group failed to modulate their movement speed under either condition, adding to evidence that autistic people are more impaired on goal-less than goal-directed imitation tasks ^3-4. In the current work, we investigated whether perspective or training would improve goal-less imitation.

Objectives:  

Usually we view people from the “other” or allocentric perspective, rather than from our “self” or egocentric perspective.  However, imitation from the former is more challenging, as visual-spatial transformation is required. Our first objective was to investigate whether imitation in the autistic group would improve in the egocentric perspective, during goal-less imitation, due to the removal of cognitively demanding spatial transforms. Our second objective was to investigate whether training would improve goal-less imitation for the autistic group by both providing practice, and by increasing the visual saliency of the movement.

Methods:  

Thirteen high functioning autistic adults and matched controls observed and then imitated video clips of hand movements, while their own hand movements were recorded. Observed movements were either directed towards visual targets (goal condition) or no targets (goal-less condition) and were of fast or slow speed. Imitation ability was characterized as the degree to which participants modulated their movement speed between fast and slow trials. Clips were presented from either the allocentric or egocentric perspective. Halfway through the experiment a training phase occurred in which the finger in the clips was painted red to increase saliency of the movement.

Results:  

Participants appeared to benefit from the egocentric perspective, as imitation speed was faster for both groups. However, modulation of imitation speed was only observed in the goal-less allocentric view for controls. Training increased imitation speed in both groups, indicating that the task became easier following training, although it did not increase the modulation of imitation speed for either group. Finally, movement variability was reduced following training in the autistic group for the goal-less condition.

Conclusions:  

The autistic group benefitted from an egocentric perspective and training, suggesting they are able to map other people’s actions onto their own motor system and that training can facilitate imitation.  The lack of effect of the egocentric perspective or training on modulation of imitation speed for both groups highlights that task familiarity affects which aspects of the movement are imitated.

1. Wild.K et al. (2010). Exp Brain Res, 204, 353-60
2. Wild.K et al. (2011). JADD under review
3. Hamilton et al. (2007) Neuropsychologia, 45, 1859-68.
4. Hobson and Hobson (2008) J Exp Child Psychol, 101, 170-185.