Drawing Out Inner Feelings: Visual Expression and Recognition of Emotions in Drawings by Children with ASD

Background:  A lack of emotional reciprocity is a cardinal feature of Autism Spectrum Disorders (ASDs). Affected individuals have difficulty identifying emotions expressed by others (e.g., Hobson, 1986), and they, themselves, tend to be emotionally non-expressive (e.g. Czapinski & Bryson, 2003; Snow et al., 1987; Yirmiya et al., 1989). One outstanding question is whether difficulties with the expression and recognition of emotion in ASD extend to visual arts. We addressed this question by asking children with ASD and typical controls to depict various emotions in their drawings, and to identify the emotion depicted in the drawings of other children.

Objectives:  The primary objective of this study was to determine whether young school-aged children with ASD could depict identifiable emotions in their artwork, and recognize the emotions depicted in the artwork of other children.

Methods:  Forty children aged 4 to 9 years participated in this study. Children with ASD were matched to typical controls based on expressive and receptive language skills, using the Oral and Written Language Scale (Carrow-Woolfolk, 1996). This study was conducted in two phases. The first phase, which involved expressing emotions, included a total of 8 children, half of whom had ASD. Modelled after Driessnack (2006), children were asked to make four drawings depicting themselves in a situation in which they felt one of four different emotions: happiness, sadness, anger, and fear. This yielded a total of 32 drawings, which were used as stimuli in the second phase. The second phase of the study, which involved recognizing emotions, included a total of 32 new participants, half of whom had ASD. Following Misalidi and Bonoti (2008), each child was asked to look at the drawings and choose which of the four emotions they recognized in each one, with the aid of small cue cards with the labels “happy”, “sad”, “angry”, and “scared”. Responses judged correct (i.e., those with a match between the emotion the child artist intended to express and the emotion selected by the participant) were given 1 point and all other answers were given a score of 0.

Results: The number of correct responses in the recognition phase of the study was analyzed using a 2 (Artist: ASD vs. Typical children) by 2 (Viewer: ASD vs. Typical children) by 4 (Emotion: Happy, Sad, Angry, and Scared) mixed analysis of variance (ANOVA). Interestingly, drawings made by children with ASD received significantly higher ratings of agreement (50.59%) than drawings made by control children (36.33%), F(2,31) = 45.547, p < .001. Additionally, there was a non-significant trend toward better recognition scores in the typical control children, F(2,31) = 3.677, p >.065. 

Conclusions:  While children with ASD may have some difficulty recognizing emotion in the drawings of other children, we provide preliminary evidence that the representation of emotion in drawings may be a relative strength in some children with ASD (examples of which will be provided). Given the potential this latter finding has for better understanding the inner feeling states of children with ASD, this line of inquiry warrants further study.