“They all still look the same to me”: Navon processing and the cross-race effect

ABSTRACT

Research suggests the Cross-Race Effect (CRE), the difficulty to remember faces from a different race relative to their own, is due to other race faces being processed featurally. It has been theorised that if individuals process other race faces the same as they do same race faces (globally) the CRE could be attenuated. Research inducing global processing, through the use of a Navon processing task, has improved memory for same race faces. Across two experiments we explored if inducing global processing through the Navon processing task prior to encoding (Experiment 1) and at both encoding and retrieval (Experiment 2) would reduce the CRE. Results revealed that biasing White participants’ processing orientations failed to improve memory for Black faces. Furthermore, in Experiment 2, global processing impaired participants’ memory for identifying faces relative to a control group. Practical implications and the need to address inconsistencies in the Navon processing literature are discussed.

KEYWORDS:

• Cross-race effect
• face memory
• global processing
• local processing
• Navon figures

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The Navon stimuli used in this experiment typically give rise to global precedence effects; more efficient processing of global than local level information

2 We used an all-male participants and male stimuli primarily to control for gender. With all male participants and all male photos, gender biases cannot come into play. Additionally, men, typically do not show an own gender bias relative to women (Lovén, Herlitz, & Rehnman, 2011), we decided to recruit men only for the study. Second, because we were focused on racial influences of memory, and there have been no studies to date that suggest White individuals remember Black men’s faces better than women’s or vice versa, we used only men’s faces as stimuli.

3 The difference in effect sizes between Experiment 1 and 2 for the CRE are consistent with previous CRE research (for a review see Meissner & Brigham, 2001). Thus, it is unlikely that difference across the two samples is due to Navon processing but rather reflects generally variability in memory for cross faces across samples.

4 Although holistic processing and global processing are considered to be the same by some researchers, others have debated over whether these terms refer to different processes and if they can be used interchangeably (see Hills & Lewis, 2008)