Featural versus configural face processing in a rare genetic disorder: Williams syndrome

Background Williams syndrome (WMS) is a rare genetic disorder with an estimated prevalence of 1 in 20 000 live births. Among other characteristics, WMS has a distinctive cognitive profile with spared face processing and language skills that contrasts with impairment in the cognitive domains of spatial cognition, problem solving and planning. It remains unclear whether individuals with WMS process faces using a featural strategy that focuses on features or a configural strategy that takes into consideration the contour of a face and spatial relations between features. Methods To investigate face processing in WMS, the tasks specifically probe unfamiliar face matching by using a design that includes manipulations in face presentation (thatcherised and non‐thatcherised), face orientation (upright and inverted) and face valence (happy and neutral expression) in a match‐to‐target face recognition design. The sample consisted of 20 participants with WMS, 10 participants with non‐specific developmental delay (IQ‐matched) and 10 normal control participants (chronological age‐matched). Results Similar to normal controls, WMS performed best when faces were presented upright. The results show while the WMS group did not perform as well as their typically developing counterparts, they did significantly better than the IQ‐matched developmentally delayed group. WMS did not show an accuracy advantage for inverted faces commonly understood as an index for featural face processing, nor did they perform better on thatcherised inverted face conditions whereby featural processing is forced. Furthermore, no accuracy advantage was observed for positively valenced (happy) faces in the WMS group. Conclusion These results are consistent with previous work showing a configural face processing approach in WMS, a strategy that is also utilised by normal controls.