Children's face identity representations are no more view specific than those of adults.

Abstract

Face recognition performance improves during childhood, not reaching adult levels until late adolescence, yet the source of this improvement is unclear. Recognition of faces across changes in viewpoint appears particularly slow to develop. Poor cross-view recognition suggests that children's face representations may be more view specific than those of adults and is consistent with arguments that extensive experience with faces may be required to form representations that are robust to view changes. We conducted the first direct test of the view specificity of children's face representations by using face aftereffects to investigate whether children's face aftereffects transfer across changes in viewpoint. Using both the figural aftereffect (E1) and the identity aftereffect (E2) we showed that 8-year-old children's aftereffects transferred across substantial changes in viewpoint and that children did not differ from adults in the amount of transfer across viewpoint. These results suggest that children's coding of identity is no more view specific than that of adults and are consistent with a growing body of evidence indicating that the key perceptual mechanisms of face recognition emerge early in life.