The Recognition of Self-Conscious Emotions from Situational Contexts in Children with and without Autism Spectrum Disorders

Background:  The ability to understand and reflect upon one’s own emotions and the emotions of others is central to emotional competence. Self-conscious emotions, in particular, are thought to facilitate our social interactions and relationships by motivating us to adhere to social norms (guilt) as well as personal standards (pride).  

Objectives:  Despite the importance of self-conscious emotions, almost all studies have explored basic emotion processing in children with ASD (see Uljarevic & Hamiliton, 2013).  Moreover, recognition of emotions is often assessed through facial recognition tasks, with emotions presented at full intensity.  The purpose of this research was to assess the recognition of basic and self-conscious emotions from situational contexts that varied in intensity.  Relations between emotion recognition, ASD symptomatology, and Theory of Mind (ToM) were also explored.

Methods:   Twenty-three children with ASD and 25 neurotypical (NT) children were tested. No significant differences between groups were found for age, male:female ratio, or non-verbal reasoning (Table 1). Children were given a situational emotional test that provided 12 emotional situations for basic (happy, fear, sad) and self-conscious (pride, embarrassment, guilt) emotions. An intense and a less intense version of each emotion were presented. Children were asked to label the emotions (free, cued) and rate the intensity and duration of the emotions (1-5 scale).

Results:   Mixed-model ANOVAs and follow-up tests with Bonferroni correction were conducted. In the cued response condition, NT children were significantly better than children with ASD at recognizing intense examples of pride, embarrassment and guilt, and non-intense examples of pride and embarrassment (Table 2). NT children were better at recognizing fear from intense and non-intense situations. In the free responsecondition, NT children were better at recognizing intense examples of embarrassment and guilt, and non-intense examples of embarrassment. All children were good at labeling fear in the intense condition (Table 2). No significant group differences were seen in the recognition of happy and sad emotions. All children rated the intensity and duration of “intense” emotions as greater than that for “less intense” emotions. Children with ASD rated situations eliciting pride and embarrassment as being more intense and longer lasting than NT children (Table 2). Significant relations were found between ToM scores and the recognition of emotions in children with ASD, but not in NT children. No other significant relations were found.

Conclusions:  The present study shows that children with ASD are generally less accurate than NT children in their recognition of self-conscious emotions from situational contexts.  This is important given the significance of self-conscious emotions in terms of emotional competence.  These results conflict with a study (Tracy et al., 2011) showing that children with ASD are as accurate as NT children at recognizing pride from facial expressions—suggesting that the measurement of emotions (facial vs. situational) is important.  Consistent with facial recognition studies (Tell et al., 2014), some evidence was found that children with ASD were generally less accurate than NT children at recognizing fear.  ToM abilities appear to underlie emotion recognition in children with ASD, but not in NT children.