Race perception and gaze direction differently impair visual working memory for faces: An event-related potential study

Abstract

Humans are amazingly experts at processing and recognizing faces, however there are moderating factors of this ability. In the present study, we used the event-related potential technique to investigate the influence of both race and gaze direction on visual working memory (i.e., VWM) face representations. In a change detection task, we orthogonally manipulated race (own-race vs. other-race faces) and eye-gaze direction (direct gaze vs. averted gaze). Participants were required to encode identities of these faces. We quantified the amount of information encoded in VWM by monitoring the amplitude of the sustained posterior contralateral negativity (SPCN) time-locked to the faces. Notably, race and eye-gaze direction differently modulated SPCN amplitude such that other-race faces elicited reduced SPCN amplitudes compared with own-race faces only when displaying a direct gaze. On the other hand, faces displaying averted gaze, independently of their race, elicited increased SPCN amplitudes compared with faces displaying direct gaze. We interpret these findings as denoting that race and eye-gaze direction affect different face processing stages.

Keywords:

• Race
• Eye-gaze direction
• Face processing
• Visual working memory
• Event-related potentials

Notes

¹ Embracing a more socially connoted view, it is important to note that faces with direct gaze may communicate the intention to establish an approach with an observer. Interestingly, a recent study showed that other-race faces with direct gaze were rated as more threatening than own-race faces with direct gaze, whereas there were no differences when these facial stimuli were displayed with averted gaze (Richeson, Todd, Trawalter, & Baird, 2008). These results emerged, reasonably, because the intentions communicated by other-race faces with direct gaze were perceived by participants as more negative. In our view, this is perfectly congruent with both the model proposed by Hugenberg et al. (2010, 2013) and with our findings.

Additional information

Funding

Part of the present research was financially supported by the Italian Ministry of Education, University and Research (Futuro in Ricerca, 2012 grant RBFR12F0BD).