The Impact of Perceptual and Social Mechanisms on Human vs. Cartoon Emotion Processing in Children with High-Functioning Autism

Background:

Atypical facial emotion processing is one of the main features of autism spectrum disorders. Interestingly, recent evidences suggest that this might be specific to the processing of facial expressions displayed in human faces and might be spared in cartoon face processing. However, to date, the reasons why children with autism process emotions in human and cartoon faces differently remain unclear.

Objectives:

The aim of the present study was to examine the influence of perceptual and social mechanisms on the processing of human vs. cartoon faces in children with high-functioning autism (HFA).

Methods:

This study included 20 children with HFA matched to 20 typically developing controls. Stimuli comprised digitized faces morphed (20% signal increment) to produce a linear continuum of images between the two endpoints (cartoon and human faces). These faces were presented either upright or inverted in order to assess the perceptual strategy used in face processing. Children were first asked to judge the emotion displayed on the faces (session 1) and then asked to categorize the faces into categories – cartoon or human faces (session 2).

Results:

In the first session, a significant inversion effect was found for both HFA and control groups, independently of the morphed face. In the second session, both groups showed a similar categorical boundary, although performance at the endpoints was lower for the HFA than the control group. Importantly, this group difference was found for human faces but not for cartoon faces.

Conclusions:

Findings of this study revealed that both children with HFA and controls used similar perceptual strategies to recognize emotions on human and cartoon faces. In addition, and in line with results for controls, children with HFA were found to rely on the same perceptual criteria to categorize human and cartoon faces. These findings suggest that perceptual mechanisms do not underlie atypical human vs. cartoon face processing in children with HFA. By contrast, poor performance of children with HFA relative to controls on the categorization of human faces suggest a top-down influence of social mechanisms on facial emotion processing in children with HFA. These findings are in agreement with clinical and parental reports of great interest for cartoon characters relative to reduced interest for human faces in children with autism.