A replication of the face-superiority effect

Schematic faces, scrambled faces (stimuli differing from schematic faces in that the positions of the eyes, noses, and mouths were anomalous, but constant from trial to trial), and individual face components were presented in a tachistoscope followed by a patterned mask. Following the mask, either four alternative mouths or four alternative pairs of eyes were presented, and subject3 were required to choose among the alternatives for the feature that had been presented in the tachistoscopic display. Performance was best when the original display contained only individual component features, and it was worst when scrambled faces were stimuli, thus replicating a similar experiment by Homa, Haver, and Schwartz (1976). Contrary to their findings, however, the effect of feature scrambling was greater on recognition of eyes than on mouths.