Reliability and Validity of the Other Race Effect: Comparison of Two Measures

Abstract

The other race effect (ORE) is the phenomenon where people better recognise faces of their own race, as opposed to faces of other less familiar races. In this study the reliability and validity of the ORE was examined with two different measures of face perception. A total of 52 participants completed both measures, 28 of whom had previously been diagnosed as dyslexic. The two measures were the Cambridge Face Memory Test – Caucasian Asian (CFMT-CA), based on the Cambridge Face Memory Test, and the Other Race Caucasian Asian (ORCA), a task specifically designed for this study. Both tasks used a forced-choice paradigm to measure face recognition for Asian and European faces. The ORE measurement is the difference in performance on Asian vs. European faces. The results showed that both tasks were reliable measurements of face perception, but only CFMT-CA measured the ORE reliably. Validity of the tasks was also assessed with a measure of interracial contact and only the CFMT-CA correlated with interracial contact. This study illustrates the importance of having a reliable measurement of the ORE itself rather than just of face recognition. It also highlights the possible dangers of using unknown and untested tasks and stimuli to measure the ORE.