Abstract
Introduction: The current study asked whether face discrimination abilities differ for unfamiliar vs. familiar faces and whether familiarity influences the face-inversion and the other-race effects. Methods: The stimuli were pictures of faces of females (East Asian and Caucasian) from two different labs. Subjects were twenty-five females of East Asian or Caucasian descent from the two labs. This resulted in two groups of participants who were tested with the same stimulus set, yet different assignments of familiar vs. unfamiliar, as well as different assignments of same-race vs. other-race. The face discrimination task consisted of participants indicating whether a test face, presented for 500 ms, was more similar to one of two sample faces presented simultaneously after the test face. Sample faces were pairs of faces of the same race and lab. Test faces were morphed stimuli that varied in 100 even step sizes from 100% Sample Face A to 100% Sample Face B. Threshold was defined as the average of two thresholds: % Sample Face A (and B) needed to correctly match the morphed test face to Sample Face A (and B) on 80% of trials. Threshold was obtained for each of eight conditions: Races (same/other) x Familiarity (familiar/unfamiliar) x Orientation (upright/inverted). Results: An ANOVA revealed superior performance on familiar vs. unfamiliar faces (p<0.05) and an interaction between familiarity and orientation (p=0.05), which was driven by a greater inversion effect for familiar faces (p<0.05). No statistically significant other-race effects were found; however, there was a trend for a smaller effect of race for familiar upright than unfamiliar upright faces. Conclusions: Inversion effects are greater for familiar faces, which may be due to familiar faces being processed more holisticly than unfamiliar faces. Further work will be needed to ascertain whether familiarity mediates the other-race effect.
Meeting abstract presented at VSS 2012