Identification of Distinctive Faces in Individuals with Autism

Background: Research indicates that typically developing children as young as 4 years of age discriminate distinctive faces from average looking faces. (McKone & Boyer, 2006).  Additional research with both children and adults has shown that distinctive faces are remembered better  than less distinctive faces (i.e., the distinctiveness effect).  The distinctiveness effect is frequently explained using Valentine’s theoretical multidimensional face space (1991). In this model, typical or average faces, which are more prototypical, are stored at the center of the face space.  In contrast, distinctive faces which deviate more from the prototypical face are stored toward the perimeter of the face space.  Since the majority of faces are stored near the center of the face space, distinctive faces are easier to recall.  Preliminary evidence suggests that face recognition in children and adults with autism is not aided by distinctiveness.   However, it is unclear whether individuals with autism are unable to identify distinctive faces or whether distinctive faces simply do not enhance their memory recall.

Objectives: The current study examined adults with autism’s ability to identify distinctive faces in order to expand on research which indicates that individuals with autism do not demonstrate a distinctiveness effect.  Thus, this study explored whether individuals with autism are able to choose the more distinctive face.  In addition, this study manipulated the difficulty of the task to determine the degree to which individuals with autism have an understanding of face distinctiveness.  

Methods: Participants included high-functioning adults with autism who ranged in age from 18 years to 42 years of age.  Control participants included adults without autism matched on age and gender. The study consisted of 30 trials, in which participants were asked to choose which of two faces was more distinctive.  Facial stimuli consisted of 60 female faces that had been rated by a separate sample of adults without autism on distinctiveness using a 7-point scale.  Trials were divided into three difficulty groups based on the magnitude of the change in distinctiveness between the two faces: hard, moderate, and easy.

Results: Preliminary analysis indicated that overall, individuals with autism were less accurate at identifying the more distinctive face compared to control individuals. Furthermore, while adults in the control group performed better than chance on both the easy and the moderate difficulty trials, adults in the autism group performed better than chance only on the easy trials. As expected, accuracy for both groups decreased as trial difficulty increased. 

Conclusions: These results indicate that adults with autism are impaired in their ability to identify distinctive faces.  While individuals with autism were able to identify distinctive faces on the easy trials, they were not able to on more difficult trials. Thus, the lack of a distinctiveness effect in individuals with autism may arise from a failure to discriminate between distinctive and typical faces.  More generally, these results suggest that individuals with autism are not learning and storing faces in an organized “face space” as do typically developing individuals.