Abstract
Face identification is thought to involve implicit evaluation of how an individual face differs from a face prototype (norm-based coding) (Rhodes et al., 2005). The characteristics of the prototype are thought to be influenced by experience with faces. People with autism spectrum disorders (ASD) appear to have both deficits and areas of preserved processing in face perception (e.g., Weigelt, Koldewyn, & Kanwisher, 2012). We examined the extent to which adults with ASD show evidence of norm-based coding of facial identity. Participants were adapted to faces that differed from the average in the physically opposite way that the target face did (an anti-identity), and then asked to categorize the average face. The norm-based model would predict that the average face would be perceived as more like the target identity; a phenomenon known as a face identity aftereffect. The adapting faces were either very different from the average face (80% anti-identity strength) or close to the average face (40% anti-identity strength). The norm-based coding model predicts that more extreme adapting faces should produce a larger aftereffect than less extreme adapting faces. Both the ASD and control groups displayed larger identity aftereffects for more extreme adapting faces compared to weaker adapting faces, evidence that both groups use norm-based coding. There was no group difference in aftereffect size. This pattern is in contrast with previous findings of abnormally small identity aftereffects in children with ASD (Pellicano et al., 2007). The current results provide the first evidence of intact norm-based coding of facial identity in high-functioning adults with ASD, and suggest that deficits in face processing observed in this population may arise from another source.
Meeting abstract presented at VSS 2014