Login / Signup

New Methods for Unraveling Imitation Accuracy Differences Between Children with Autism and Typically Developing Peers.

Laura SparaciDomenico FormicaFrancesca Romana LasorsaLuigi RaianoPaola VenutiOlga Capirci
Published in: Perceptual and motor skills (2022)
This study applies methods used in sign language and gesture research to better understand reduced imitation accuracy (IA) of actions and gestures in children with autism spectrum disorder (ASD), and we addressed contrasting theories on IA in ASD and the role of objects and meanings in imitation. Eight male children with ASD with a mean chronological age (CA) of 86.76 months ( SD = 10.74, range 70.5-104.4) and 22 male and female peers with typical development (TD) and a mean CA of 85.44 months ( SD = 7.95, range 73.4-96.7) imitated videos of an adult performing actions with objects, representational gestures, conventional gestures and meaningless gestures. We measured accuracy as ability to effectively reproduce features (handshape, palm orientation, location, movement direction and type) and timing (speed) of observed actions/gestures, after ruling out cases of specular (i.e., mirror-like) versus anatomical imitation. Results highlighted significantly lower feature and timing accuracy in children with ASD with respect to the TD group across tasks, and these findings supported sensory-motor theories of IA in ASD. Our data also showed the different impact of objects and meanings within groups. Overall, these results suggest validity to our assessment method and suggested the importance of considering both discreet variables (i.e., variables describing action/gesture feature accuracy, e.g. handshape, movement direction) and continuous variables (i.e., kinematic variables, e.g. speed) in evaluating IA in autism.
Keyphrases
  • autism spectrum disorder
  • intellectual disability
  • attention deficit hyperactivity disorder
  • young adults
  • machine learning
  • deep learning
  • electronic health record
  • artificial intelligence