Towards Safer Online Spaces: Simulating and Assessing Intervention Strategies for Eating Disorder Discussions
Abstract: Eating disorders are complex mental health conditions that affect millions of people around the world. Effective interventions on social media platforms are crucial, yet testing strategies in situ can be risky. We present a novel LLM-driven experimental testbed for simulating and assessing intervention strategies in ED-related discussions. Our framework generates synthetic conversations across multiple platforms, models, and ED-related topics, allowing for controlled experimentation with diverse intervention approaches. We analyze the impact of various intervention strategies on conversation dynamics across four dimensions: intervention type, generative model, social media platform, and ED-related community/topic. We employ cognitive domain analysis metrics, including sentiment, emotions, etc., to evaluate the effectiveness of interventions. Our findings reveal that civility-focused interventions consistently improve positive sentiment and emotional tone across all dimensions, while insight-resetting approaches tend to increase negative emotions. We also uncover significant biases in LLM-generated conversations, with cognitive metrics varying notably between models (Claude-3 Haiku $>$ Mistral $>$ GPT-3.5-turbo $>$ LLaMA3) and even between versions of the same model. These variations highlight the importance of model selection in simulating realistic discussions related to ED. Our work provides valuable information on the complex dynamics of ED-related discussions and the effectiveness of various intervention strategies.
- American Psychiatry Assoication. 2023. What are eating disorders? Eating Disorders.
- Amanda J Antczak and Teresa L Brininger. 2008. Diagnosed eating disorders in the us military: A nine year review. Eating Disorders, 16(5):363–377.
- Incidence and prevalence of eating disorders among us military service members, 2016–2021. International Journal of Eating Disorders.
- Consequences of making weight: A review of eating disorder symptoms and diagnoses in the u nited s tates military. Clinical Psychology: Science and Practice, 21(4):398–409.
- Centers for Disease Control and Prevention. 2022. Pediatric emergency department visits associated with mental health conditions before and during the covid 19 pandemic united states, january 2019 to january 2022. Morbidity and Mortality Weekly Report.
- BERT: pre-training of deep bidirectional transformers for language understanding. CoRR, abs/1810.04805.
- Morality between the lines: Detecting moral sentiment in text. In Proceedings of IJCAI 2016 workshop on Computational Modeling of Attitudes.
- Laura Hanu and Unitary team. 2020. Detoxify. Github. https://github.com/unitaryai/detoxify.
- Huggingface. Distilbert base uncased finetuned sst-2 model. https://huggingface.co/distilbert/distilbert-base-uncased-finetuned-sst-2-english.
- Machine learning insights into eating disorder twitter communities. In Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems, CHI EA ’24, New York, NY, USA. Association for Computing Machinery.
- How a moderated online discussion forum facilitates support for young people with eating disorders. Health Expectations, 20(1):98–111.
- Does gpt-3 generate empathetic dialogues? a novel in-context example selection method and automatic evaluation metric for empathetic dialogue generation. In Proceedings of the 29th International Conference on Computational Linguistics, pages 669–683.
- Radicalized by thinness: Using a model of radicalization to understand pro-anorexia communities on twitter. Preprint, arXiv:2305.11316.
- Investigating bias in llm-based bias detection: Disparities between llms and human perception. arXiv preprint arXiv:2403.14896.
- Mayo Clinic. 2023. Eating disorders. Diseases and Conditions.
- The impact of the covid-19 pandemic on eating disorders in us military veterans. International Journal of Eating Disorders, 56(1):108–117.
- National Institute of Mental Health. 2023. Eating disorders. Health Topics.
- What topics do members of the eating disorder online community discuss and empathize with? an application of big data analytics. In Healthcare, volume 10, page 928. MDPI.
- Interpersonal interactions on online forums addressing eating concerns. International Journal of Eating Disorders, 43(2):161–170.
- Truth of varying shades: Analyzing language in fake news and political fact-checking. In Proceedings of the 2017 conference on empirical methods in natural language processing, pages 2931–2937.
- Connotation frames: A data-driven investigation. arXiv preprint arXiv:1506.02739.
- Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. arXiv preprint arXiv:1908.10084.
- Evaluation and validation approaches for simulation of social behavior: challenges and opportunities. Social-Behavioral Modeling for Complex Systems, pages 495–519.
- Prevalence of eating disorder risk and body image dissatisfaction among rotc cadets. International Journal of Environmental Research and Public Health, 17(21):8137.
- Feelings about bodies: Emotions on diet and fitness forums reveal gendered stereotypes and body image concerns. Preprint, arXiv:2407.03551.
- Systematic biases in llm simulations of debates. arXiv preprint arXiv:2402.04049.
- The White House. 2023. Fact sheet: Biden harris administration announces actions to protect youth mental health, safety, and privacy online. Briefing Room.
- US Department of Health and Human Services. 2023. Surgeon general issues new advisory about effects social media use has on youth mental health. News.
- Why are so many american youth in a mental health crisis? exploring causes and solutions. Full Committee Hearing.
- Daphne van Hoeken and Hans Hoek. 2020. Review of the burden of eating disorders: mortality, disability, costs, quality of life, and family burden. Current opinion in psychiatry, 33, 6:521–527.
- Explaining and predicting human behavior and social dynamics in simulated virtual worlds: reproducibility, generalizability, and robustness of causal discovery methods. Computational and Mathematical Organization Theory, 29(1):220–241.
- " kelly is a warm person, joseph is a role model": Gender biases in llm-generated reference letters. arXiv preprint arXiv:2310.09219.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.