A cross-cultural investigation into the influence of eye gaze on working memory for happy and angry faces

ABSTRACT

Previous long-term memory (LTM) research found that angry faces were more poorly recognised when encoded with averted vs. direct gaze, while memory for happy faces was unaffected by gaze. Contrastingly, working memory (WM) accuracy for angry faces was unaffected by gaze, but WM was enhanced for happy faces with averted vs. direct gaze. Because the LTM study was conducted in an Eastern culture (Japan) with Japanese faces, while the WM study was conducted in a Western culture (UK) with Caucasian faces, here we investigated WM further to examine whether gaze effects diverge due to cultural variation between the faces and participants. When Western participants viewed Japanese faces (Experiment 1), the happy-averted gaze advantage in WM was replicated. In contrast, Japanese participants viewing Caucasian faces (Experiment 2a) showed poorer WM for angry faces with averted vs. direct gaze, and no influence of gaze on WM for happy faces. When Japanese participants viewed Japanese faces (Experiment 2b), gaze did not modulate WM. Therefore, the way in which expression and gaze interact to influence face WM does not appear to rely on the specific memory system engaged, but instead may be attributed to cultural differences in display rules between Eastern and Western cultures.

KEYWORDS:

• Emotion
• faces
• gaze
• working memory
• cross-cultural

Acknowledgements

Thanks go to Yoshiyuki Ueda, Minako Osami, and Kazusa Minemoto for translating the study materials and helping with data collection in Japan.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Experiment 1, average response time = 986 ms, 0.22% of trials timed out. Experiment 2a, average response time = 1149 ms, 2.6% of trials times out. Experiment 2b, average response time = 892 ms, 1.02% of trials timed out. Timed out trials were not analysed.

2 For reference, despite uneven sample sizes a mixed ANOVA was conducted with Experiment 2a vs. 2b as a between factor, and confirms a non-significant 3-way interaction between emotion, gaze, and experiment, F(1, 61) = 1.705, p = .197.

Additional information

Funding

The research conducted in Japan was supported by summer programme funding from the Japan Society for the Promotion of Science (JSPS) and the British Council.