Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Wearable Meets LLM for Stress Management: A Duoethnographic Study Integrating Wearable-Triggered Stressors and LLM Chatbots for Personalized Interventions

Sameer Neupane, Poorvesh Dongre, Denis Gracanin, Santosh Kumar

TL;DR

The paper investigates integrating wearable-triggered stressor journaling with large language model chatbots to deliver personalized, real-time stress interventions. Using a duoethnographic design with two researchers over 22 days, it compares just-in-time one-shot interventions (DeStressify) against end-of-day, multi-turn conversations (StressGPT). The study finds that only about 1 in 5 detected stress events warranted interventions, and that interventions informed by brief event descriptions were more effective than generic prompts. It discusses design implications for continuity, precision, and privacy, highlighting the potential of physiology-informed, memory-enabled chatbots for real-time stress relief and behavior change, while noting methodological limitations. It points to future work with larger samples, memory integration, and strengthened privacy controls.

Abstract

We use a duoethnographic approach to study how wearable-integrated LLM chatbots can assist with personalized stress management, addressing the growing need for immediacy and tailored interventions. Two researchers interacted with custom chatbots over 22 days, responding to wearable-detected physiological prompts, recording stressor phrases, and using them to seek tailored interventions from their LLM-powered chatbots. They recorded their experiences in autoethnographic diaries and analyzed them during weekly discussions, focusing on the relevance, clarity, and impact of chatbot-generated interventions. Results showed that even though most events triggered by the wearable were meaningful, only one in five warranted an intervention. It also showed that interventions tailored with brief event descriptions were more effective than generic ones. By examining the intersection of wearables and LLM, this research contributes to developing more effective, user-centric mental health tools for real-time stress relief and behavior change.

Wearable Meets LLM for Stress Management: A Duoethnographic Study Integrating Wearable-Triggered Stressors and LLM Chatbots for Personalized Interventions

TL;DR

The paper investigates integrating wearable-triggered stressor journaling with large language model chatbots to deliver personalized, real-time stress interventions. Using a duoethnographic design with two researchers over 22 days, it compares just-in-time one-shot interventions (DeStressify) against end-of-day, multi-turn conversations (StressGPT). The study finds that only about 1 in 5 detected stress events warranted interventions, and that interventions informed by brief event descriptions were more effective than generic prompts. It discusses design implications for continuity, precision, and privacy, highlighting the potential of physiology-informed, memory-enabled chatbots for real-time stress relief and behavior change, while noting methodological limitations. It points to future work with larger samples, memory integration, and strengthened privacy controls.

Abstract

We use a duoethnographic approach to study how wearable-integrated LLM chatbots can assist with personalized stress management, addressing the growing need for immediacy and tailored interventions. Two researchers interacted with custom chatbots over 22 days, responding to wearable-detected physiological prompts, recording stressor phrases, and using them to seek tailored interventions from their LLM-powered chatbots. They recorded their experiences in autoethnographic diaries and analyzed them during weekly discussions, focusing on the relevance, clarity, and impact of chatbot-generated interventions. Results showed that even though most events triggered by the wearable were meaningful, only one in five warranted an intervention. It also showed that interventions tailored with brief event descriptions were more effective than generic ones. By examining the intersection of wearables and LLM, this research contributes to developing more effective, user-centric mental health tools for real-time stress relief and behavior change.

Paper Structure

This paper contains 31 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Background and Related Works
  3. Wearables for Stress Tracking, Stressor Journaling, and Interventions
  4. LLM for Mental Health Support
  5. Methods
  6. Duoethnography
  7. Wearable for Stressor Journaling
  8. Custom GPT for Stress Intervention
  9. Data Collection and Analysis
  10. Findings
  11. Diversity in Stressors Detected and the Need for Interventions
  12. Diversity in Stressors Detected
  13. Proportion of stress events that require interventions
  14. Common Themes in Our Experience with DeStressify and StressGPT
  15. Need for Continuity and Stressor Awareness in Chatbot Interactions
  16. ...and 16 more sections

Figures (2)

  • Figure 1: CuesHub app screenshots for valence and event descriptions
  • Figure 2: Number of Stressors and Interventions per day and Valence Proportion of Events Detected by CuesHub for each researcher