Independent effects of working memory loads and facial expressions on event-related potential (ERP) responses: Evidence from mass univariate analysis

Abstract

Previous studies have investigated whether event-related potential (ERP) responses to negative and positive facial expressions are modulated by perceptual and working memory loads of face-irrelevant tasks, but results have been mixed. These studies typically employed traditional analysis methods that focus on a limited number of electrodes and timepoints, which increases the risk of statistical errors. Moreover, no studies have investigated these issues for emotionally ambiguous facial expressions. Therefore, the current study aimed to investigate whether working memory loads influence ERP effects of positive, negative and importantly, emotionally ambiguous (e.g., surprised) facial expressions by using robust Mass Univariate Analysis (MUA). To address these issues, thirty-three participants were presented with angry, happy, surprised or neutral faces. Participants were required to memorize a letter (low working memory loads) or 5 letters (high loads) before face presentations and to recognize them afterwards. ERP results showed that working memory loads and facial expressions independently affected ERP responses, with no significant interaction. Working memory loads influenced ERP responses starting after the P100 peak. Regarding the effect of facial expressions, surprised facial expressions generally weakened late positive potential (LPP) responses compared to other facial expressions, and happy facial expressions weakened the responses at early time ranges, while angry facial expressions enhanced LPP responses at distinct electrode sites depending on the comparison. These findings underscore the value of employing MUA for comprehensive ERP analysis in face processing research.