Impact of the proportion of emotional faces on ensemble perception of group emotion.
Abstract
In everyday life, we frequently interact with crowds with faces showing various emotional expressions. Ensemble coding allows the visual system to efficiently extract average group emotion in groups of faces. However, this process is biased as emotional intensity of the group is often overestimated. This amplification effect is reported to increase as a function of group size. Yet, it remains unclear whether this effect is driven only by the number of faces or could be influenced by another factor, the proportion of emotional faces in the group. The present study tested whether the amplification effect is driven by group size or the proportion of emotional faces within the group. Emotional valence (anger vs. happiness) was also manipulated as a potential influence of the amplification effect. Twenty-one participants estimated the average emotion intensity in groups of 4, 8, or 16 male faces, with 25%, 50%, or 75% of faces showing either happiness or anger. The amplification effect was operationalized as the proportional estimation error between the actual and estimated proportion of emotional faces. The results showed that larger groups were associated with greater amplification than small groups, that amplification increased when the proportion of emotional faces was higher, and that it was consistently stronger for anger than for happiness. These findings suggest that amplification may not depend only on group size, but also on the proportion of emotional faces, and that valence could modulate this effect. Together, the results highlight the role of structural and emotional factors in shaping how people perceive group emotions.
Related
- Glabellar Rhytids in Thyroid Eye Disease: Severity and Their Impact on Emotional Perception
- Impact of emotion regulation on psychosocial and emotional problems in Dutch adolescents with inflammatory bowel disease
- Emotional Reactivity and Emotion Regulation in Relation to Health-Related Quality of Life After Burn Injury: A Cross-Sectional Observational Study.
- Comparing the effectiveness of emotion-focused therapy and cognitive rehabilitation based on RehaCom software on emotional and cognitive components in female migraine patients: A randomized control trial.
- Integrated Emotional Schema Therapy and the Unified Protocol for Executive Functioning Difficulties and Cognitive Emotion Regulation in Generalized Anxiety Disorder
BibTeX
@misc{mourgues2026impact,
title = {{Impact of the proportion of emotional faces on ensemble perception of group emotion.}},
author = {MOURGUES and Guillaume T. Vallet and Simon Rigoulot},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/vqu8y_v1},
url = {https://osf.io/preprints/psyarxiv/vqu8y_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Impact of the proportion of emotional faces on ensemble perception of group emotion.
AU - MOURGUES
AU - Guillaume T. Vallet
AU - Simon Rigoulot
PY - 2026
DA - 2026/09/19
PB - PsyArXiv
DO - 10.31234/osf.io/vqu8y_v1
UR - https://osf.io/preprints/psyarxiv/vqu8y_v1/
AB - In everyday life, we frequently interact with crowds with faces showing various emotional expressions. Ensemble coding allows the visual system to efficiently extract average group emotion in groups of faces. However, this process is biased as emotional intensity of the group is often overestimated. This amplification effect is reported to increase as a function of group size. Yet, it remains unclear whether this effect is driven only by the number of faces or could be influenced by another factor, the proportion of emotional faces in the group. The present study tested whether the amplification effect is driven by group size or the proportion of emotional faces within the group. Emotional valence (anger vs. happiness) was also manipulated as a potential influence of the amplification effect. Twenty-one participants estimated the average emotion intensity in groups of 4, 8, or 16 male faces, with 25%, 50%, or 75% of faces showing either happiness or anger. The amplification effect was operationalized as the proportional estimation error between the actual and estimated proportion of emotional faces. The results showed that larger groups were associated with greater amplification than small groups, that amplification increased when the proportion of emotional faces was higher, and that it was consistently stronger for anger than for happiness. These findings suggest that amplification may not depend only on group size, but also on the proportion of emotional faces, and that valence could modulate this effect. Together, the results highlight the role of structural and emotional factors in shaping how people perceive group emotions.
ER -