Recognition of faces and the individual with autism: The saliency of facial features

The current study was designed to assess recognition memory for unfamiliar faces in masked and unmasked conditions in individuals with autism and typical controls. Each diagnostic group was divided into three different age groups (6 to 10 years, 11 to 15 year, and 16 years and older). It was hypothesized that individuals with autism would have impaired memory for faces but not for buildings relative to control groups. It was also hypothesized that individuals with autism would perform better on masked conditions (upper region masked/lower region masked) than controls. Finally, it was hypothesized that there would be an age by diagnostic group interaction, with younger children with autism recognizing better by lower features (when the eyes were masked) compared to the upper features (when the mouth was masked). Results of this study revealed a main effect of diagnosis, with individuals with autism performing more poorly on recognition conditions. There was also a main effect of age, with the younger group (6 to 10 years) and the middle group (11 to 15 years) performing significantly more poorly than the oldest group. An analysis of the pattern of errors revealed that individuals with autism had more frequent false alarms, lower sensitivity and set lower criteria than typical groups. The younger group had a lower rate of hits, lower sensitivity, and set lower criteria than the older group and the middle group. A final study investigated the hypothesis that there would be a bias to match based on eyes in the middle and older groups, but that the younger autistic group would match based on the mouth. Results revealed that the younger and middle groups matched equally by mouth and eyes. The older group showed a preference for the eye region.