Effects of Contextual Affect and Face Gaze Direction in Social Cognitive Tasks: A Rigorous Mass Univariate Re-Analysis of ERP Data

Poster Session F - Tuesday, April 16, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Seth Winward¹ (swinward@uwaterloo.ca), Roxane Itier¹; ¹University of Waterloo

Recent research on how affective context and perceived gaze direction influence the neural time-course of social cognition revealed different effects of contexts on face related ERPs depending on task demands. In an empathy task, gaze modulated ERPs only in positive contexts (McCrackin & Itier, 2021a); in a theory of mind task, gaze influenced processing of both positive and negative stimuli (McCrackin & Itier, 2021b). However, these studies had average sized samples and used the FMUT toolbox to focus on known ERP components of interest. As the studies used different tasks but identical stimuli and procedures, we re-analyzed them using the robust LIMO-EEG mass univariate statistics toolbox to perform a mixed-model ANOVA (N = 86) in a data-driven way (all time points and all electrodes analyzed). We evaluated ERP responses to faces with direct or averted gaze that had been primed with positively, negatively, or neutrally valenced context sentences when participants rated either their level of empathy for, or the affect of, the neutral face targets (theory of mind). No significant effects of gaze direction or context valence was found. A significant interaction between valence and task group was found at left posterior electrodes during the LPP component, a well as a valence by gaze interaction at left posterior sites during the P1. However, none of the Bonferroni-adjusted follow-up contrasts were significant. We did not replicate the original valence and gaze effects when using data-driven rigorous statistics. Results will be discussed in the context of the replication crisis in ERP research.

Topic Area: METHODS: Electrophysiology