Moral Framing of Mental Health Discourse and Its Relationship to Stigma: A Comparison of Social Media and News

Title of the Paper

Moral Framing of Mental Health Discourse and Its Relationship to Stigma: A Comparison of Social Media and News

Paper Information

• Subject Area: Media Ethics and Mental Health
• Keywords: Moral Foundation Theory, Mental Health Narratives, Twitter, News Media, BERT, Stigma

Research Background and Problem

• What issues or challenges did the authors identify?: The portrayal of mental health in mass media and social media may exacerbate the stigma surrounding mental illness. Negative or hostile narratives in media, particularly those that "other" individuals, can lead to societal misunderstanding and antagonistic attitudes toward mental health.
• Why is this issue important?: Negative framing of mental health issues directly impacts individuals' willingness to disclose their condition and seek treatment, further contributing to systemic inequities in employment, education, and healthcare.
• Motivation and related work: Previous studies indicate that news media often associate mental illness with violent criminal behavior, while social media also contains dismissive or sarcastic narratives. These behaviors may stem from the implicit moral values embedded in the narratives, but this has not been systematically studied before.

Proposed Solution

• What methods or solutions did the authors propose?:
  • Applying Moral Foundation Theory (MFT) to analyze narratives related to mental health.
  • Comparing mental health content on two major public platforms: Twitter and news media.
  • Using a BERT-based language representation framework to quantify moral foundations and stigma in mental health narratives.
• What is innovative about this solution?:
  • The first systematic application of Moral Foundation Theory to mental health narratives, uncovering the relationship between moral framing and stigma.
  • Development and validation of a stigma-related language dictionary to measure the degree of stigma in mental health narratives.
  • Cross-platform comparative analysis.
• What are the implementation steps and key technologies used?:
  1. Data collection: Gathering data from Twitter (over 13 million tweets) and news media (over 21,000 articles) using predefined mental health keywords (e.g., anxiety, depression).
  2. Representation generation: Using the BERT model to generate contextual embeddings and constructing representative moral dimension vectors through the Moral Foundation Theory dictionary.
  3. Analysis: Calculating similarity scores between content and moral dimensions to quantitatively evaluate mental health narratives.
  4. Statistical testing: Using Kruskal-Wallis tests and linear models to assess the relationship between moral framing and stigma.

Research Findings

• What specific findings were obtained?:
  • Mental health content on Twitter reflects more positive moral dimensions (e.g., care, fairness), while news media employs more negative moral framing.
  • Negative moral framing on both Twitter and news media is often accompanied by higher usage of stigmatizing language, with more pronounced stigma.
  • In-depth analysis revealed stronger associations between care/harm and fairness/cheating dimensions and anti-stigma language, while loyalty/betrayal dimensions showed weaker associations.
• What advantages does it have compared to existing solutions?:
  • Provides a quantitative analytical framework that not only reveals the moral underpinnings of narratives but also systematically measures stigmatizing language.
  • Utilizes large-scale datasets and automated analysis methods, surpassing the limitations of previous qualitative analyses.
• What were the experimental or evaluation results?:
  • Positive moral framing in Twitter and news media content is less associated with stigmatizing language, whereas negative framing significantly increases the likelihood of stigmatizing language use.
  • Narratives centered on care/harm and fairness/cheating moral dimensions are more likely to avoid stigma.
• Limitations and future directions:
  • Limitations: The study did not include other social media platforms (e.g., Facebook) or examine mental health narratives across different cultural contexts.
  • Future directions: Expanding to more platforms and diverse cultural settings. Further analysis of narrative dissemination and its impact on audiences.

Conclusion

This study systematically reveals the moral framing in mental health narratives and its relationship with stigmatizing language, providing both theoretical foundations and empirical support for improving mental health discourse. The findings not only help news media develop safer reporting guidelines but also offer design insights for social media platforms to optimize content moderation and foster empathetic conversational environments.