Cross-cultural modeling of body satisfaction using social media selfies and self-descriptions
Abstract
Body satisfaction is a core component of psychological well-being and is closely intertwined with how individuals present themselves on social media. While prior research has examined body image through self-report measures or qualitative analyses of online content, it remains unclear whether body satisfaction is systematically associated with naturally occurring multimodal self-presentation behaviors, particularly across cultural contexts.In this study, we propose a cross-cultural multimodal mixed-methods framework to examine self-reported body satisfaction using social media selfies and self-descriptions. A sample of 240 participants from Western and East Asian cultural contexts provided self-reported body satisfaction measures alongside selected social media images and textual materials. Visual features were extracted using convolutional neural networks, while linguistic features were derived from multilingual transformer-based language models. These modalities were integrated through a late-fusion strategy to examine the extent to which multimodal self-presentational features were associated with self-reported body satisfaction.Results indicate that multimodal models outperform single-modality approaches in both regression and classification tasks. Moreover, feature importance and explainability analyses reveal culturally differentiated patterns of self-presentation: visual cues such as facial expression and body posture were more predictive in Western participants, whereas contextual and linguistic cues contributed more strongly among East Asian participants. Qualitative analyses of a purposively selected subset of misclassified and high-attention cases further illustrate cultural patterns in self-expression, modesty, and self-objectification strategies.Together, these findings suggest that body satisfaction is systematically associated with multimodal social media self-presentation and that cultural context plays an important role in shaping how such associations manifest online. This study illustrates the value of integrating computational modeling with qualitative interpretation to advance cross-cultural research on body image and digital mental health, while underscoring that predictive associations should not be interpreted as causal, clinical, or diagnostic evidence.
Related
- Is Insomnia a Western-centered disorder? A cross-cultural study of sleep satisfaction and perceived sleep duration
- Chronic pain and relationship satisfaction, social support, and social network size in chronic pain couples
- Examining the Relationship Between Perceived Social Satisfaction, Depression, and Anxiety for People With Multiple Sclerosis Participating in Group Treatment Modalities
- Epilepsy in children and young people as a stigma throughout history: identifying cultural, religious and educational factors to design strategies for contemporary social inclusion
- General self-efficacy, self-confidence, and satisfaction in simulation-based nursing education: a structural equation modeling and mediation analysis
BibTeX
@article{qiuyang2026crosscultural,
title = {{Cross-cultural modeling of body satisfaction using social media selfies and self-descriptions}},
author = {Huang Qiuyang and Wu Hua and Chen Zhengjun},
journal = {Scientific Reports},
year = {2026},
month = oct,
doi = {10.1038/s41598-026-74109-y},
url = {https://www.nature.com/articles/s41598-026-74109-y},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Cross-cultural modeling of body satisfaction using social media selfies and self-descriptions
AU - Huang Qiuyang
AU - Wu Hua
AU - Chen Zhengjun
PY - 2026
DA - 2026/10/04
JO - Scientific Reports
DO - 10.1038/s41598-026-74109-y
UR - https://www.nature.com/articles/s41598-026-74109-y
AB - Body satisfaction is a core component of psychological well-being and is closely intertwined with how individuals present themselves on social media. While prior research has examined body image through self-report measures or qualitative analyses of online content, it remains unclear whether body satisfaction is systematically associated with naturally occurring multimodal self-presentation behaviors, particularly across cultural contexts.In this study, we propose a cross-cultural multimodal mixed-methods framework to examine self-reported body satisfaction using social media selfies and self-descriptions. A sample of 240 participants from Western and East Asian cultural contexts provided self-reported body satisfaction measures alongside selected social media images and textual materials. Visual features were extracted using convolutional neural networks, while linguistic features were derived from multilingual transformer-based language models. These modalities were integrated through a late-fusion strategy to examine the extent to which multimodal self-presentational features were associated with self-reported body satisfaction.Results indicate that multimodal models outperform single-modality approaches in both regression and classification tasks. Moreover, feature importance and explainability analyses reveal culturally differentiated patterns of self-presentation: visual cues such as facial expression and body posture were more predictive in Western participants, whereas contextual and linguistic cues contributed more strongly among East Asian participants. Qualitative analyses of a purposively selected subset of misclassified and high-attention cases further illustrate cultural patterns in self-expression, modesty, and self-objectification strategies.Together, these findings suggest that body satisfaction is systematically associated with multimodal social media self-presentation and that cultural context plays an important role in shaping how such associations manifest online. This study illustrates the value of integrating computational modeling with qualitative interpretation to advance cross-cultural research on body image and digital mental health, while underscoring that predictive associations should not be interpreted as causal, clinical, or diagnostic evidence.
SN - 2045-2322
ER -