Children with ASD Do Not Benefit From Being Oriented to the Most Informative Part of the Face When Classifying Emotions

Background:

Individuals with autistic spectrum disorders (ASD) have been observed to show atypical patterns of eye gaze during face processing. However, these findings alone do not shed light on what facial cues individuals with ASD are using to classify facial expression. Different areas of the face have been found to be critical when encoding certain facial expressions.  Happiness is predominantly signified by the mouth area, and anger through the eye and brow line (Smith et al., 2005). If a non-critical area of the face is occluded then critical areas should become more salient, resulting in enhanced recognition ability. The current study therefore aims to explore whether individuals with ASD can benefit from having their attention directed to the most informative part of the face when classifying an emotional expression. Failure to benefit would suggest a fundamental difficulty with extracting pertinent information from facial cues.

Objectives:

To determine whether individuals with low functioning ASD benefit from being directed to a part of the face that aids the identification of a specific emotion.

Methods:  

A group of eighteen children with low functioning ASD (Age: 10;8 years, VIQ: 64.5, NVIQ: 100.6) and eighteen typically developing (TD) children (Age: 8;8 years, VIQ: 92.7, NVIQ: 101.4) participated in an emotion classification task. Photographs of faces expressing angry, sad, happy and (positive) surprised expressions were presented in three conditions: full face, wearing sunglasses (eyes occluded) and wearing a mask (mouth occluded), following Roberson et al. (under review). Each condition was divided into two parts. In the first participants had to decide if each person was feeling ‘happy’ or ‘not happy’ and in the second participants had to decide if each person was feeling ‘angry’ or ‘not angry’.  

Results:  

For the baseline full face condition, the TD group performed better than the ASD group at classifying angry faces, but did not differ in their ability to classify happy faces. To assess for group differences in the relative effect of occluding parts of the face, difference scores were calculated for each group to determine the benefit of the mask or the sunglasses compared to the baseline full face condition. Compared to the ASD group, the TD group showed significantly greater benefit of the mask when classifying angry faces and significantly greater benefit of the sunglasses when classifying happy faces.

Conclusions:

When classifying emotional expressions, the TD group demonstrated greater benefit than the ASD group when oriented to the most informative part of the face. Indeed, there was no evidence that the ASD group benefited significantly from being cued to the eyes for identification of anger or to the mouth for identification of happiness. This is compatible with the hypothesis that poor emotion recognition ability in ASD is not merely a function of diminished attention to core features. Rather, it suggests a difficulty in extracting relevant information from core regions.