FIGURE 3.

The adults’ mean performances with the four different manipulations for each race/ethnicity. The abscissa represents the four racial/ethnic groups (Taiwanese, Filipino, Caucasian, and African), while the ordinate depicts the mean accuracies for correctly discriminating the new image from the familiar ones. The color bars represent recognition accuracies with the four stimulus manipulations (blue: change identity; red: change eyes; green: widen eye spacing; purple: move mouth up). Error bars represent the standard errors (±SE) of the group means. The pattern of results clearly indicates that the adults’ response accuracies decrease as the manipulations change from identity, eyes only, eye spacing, to mouth spacing, and such tendency is present for each of the four race/ethnicity conditions.