Development of the own-race advantage in school-age children: a morphing face paradigm

We investigated perceptual discriminability for Asian-like versus Caucasian-like morph faces in school-age Taiwanese children and adults. One hundred-and-two 5- to 12-year-old children and twenty-three adults performed a sequential same/different face matching task, where they viewed an Asian- or a Caucasian-parent face followed by either the same parent face or a different morphed face (containing 15%, 30%, 45%, or 60% contribution from the other parent face) and judged whether the two faces looked the same. We applied signal detection theory and computed the group mean d’s based on the “physically same” trials. We also analyzed the group mean rejection rates as a function of the morph level and fitted with a cumulative normal distribution function. The adults exhibited a smaller threshold and a sharper slope in the Asian-parent condition than those in the Caucasian-parent condition, indicating the presence of an own-race advantage. Children aged between 5 and 10 showed equal discriminability for both conditions, indicating an absence of the own-race advantage. The 11- to 12-year-olds exhibited a smaller threshold and a sharper slope for the Asian-parent condition, indicating an own-race advantage. Although the oldest children’s performance began to look adult-like but still not reach the adult’s level of proficiency. In summary, children from age 5 to 12 showed a gradual and continuous refinement in the perceptual discriminability for the Asian-parent condition, but not in the Caucasian-parent condition. The present study demonstrated that expertise in face processing takes the entire childhood to develop, and supports the perceptual expertise hypothesis or perceptual learning view of ORE-- the own-race advantage seen in adulthood likely reflects a result of prolonged learning specific to faces most commonly seen in one’s visual environment such as own-race faces.