However, one important limitation of Experiment 1 is that each adaptation face
was used as either a same-identity adaptor or a different-identity
adaptor. It is therefore not clear whether the pattern of aftereffects
is due to identity or the specific adaptor–target face pairings.