In the current study we presented Asian and Caucasian
participants with brief displays containing 16 faces and asked
them to judge whether there were more Asians or Caucasians
present. We varied the physical proportion of each race in the
display using the method of constant stimuli and obtained
estimates of the point of subjective equality (PSE) by fitting
cumulative normal functions to individual data. Consistent
with recent findings on “ensemble” face processing,
participants were able to make group estimates quite
accurately. However, the estimates from the two groups of
participants did not overlap, with Asian participants
appearing to weight other-race faces more heavily than
Caucasian participants. To our knowledge, this is the first
demonstration of an other-race effect in the context of groups
of faces.
