Asymmetric emoji-word interactions in emotional processing: Evidence from priming and drift diffusion modeling

This study investigated the interactive effects and cognitive mechanisms involved in the integration of emotional information conveyed by emojis and emotional words. Two affective priming experiments were conducted to examine bidirectional influences. Experiment 1 assessed how emotional words modulate emoji processing, while Experiment 2 examined the influence of emojis on word processing. Behavioral performance and cognitive parameters were analyzed using linear mixed-effects models and the Drift Diffusion Model (DDM). Behavioral results showed that when emojis served as targets, their evaluation was reliably modulated by preceding word valence, with positive emojis exhibiting a robust positivity bias in both speed and accuracy. In contrast, when emojis served as primes, immediate word valence judgments were primarily driven by the words themselves, whereas emoji-related effects emerged mainly during the recognition phase. Across both experiments, emotionally congruent prime-target pairings facilitated processing, while incongruent pairings impaired performance, consistent with schema congruence accounts. Drift Diffusion Model analyses indicated that emotional congruency facilitated information accumulation, whereas incongruency reduced processing efficiency, with overall lower drift rates when emojis served as primes, suggesting reduced efficiency in sustaining emotional expectations. Together, these findings demonstrate an asymmetric interaction between emoji and verbal emotional information. Words exert a more robust and sustained influence on emoji processing, whereas emojis modulate word processing in a weaker and stage-dependent manner. Rather than functioning as mere decorative elements, emojis act as meaningful affective cues whose impact on emotional processing depends on contextual alignment and task demands.