Related Experiment Video
Updated: Feb 28, 2026
Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
Published on: July 27, 2018
Barriers to Designing Inclusive Ecological Momentary Assessment and Wearable Data Collection Protocols for AI-Driven
Yinan Sun1, Aditi Jaiswal2, Ali Kargarandehkordi3
1The Information Systems and Technology (IST) Department, College of Engineering and Computing, George Mason University, Fairfax, VA 22030, United States2Department of Information and Computer Science, University of Hawaii, Honolulu, HI 96822, United States, sunyinan@hawaii.edu.
Wearable sensors and ecological momentary assessment (EMA) can track substance use, but diverse groups face participation barriers. Addressing these challenges is key for inclusive digital health research.
Area of Science:
- Digital Health
- Behavioral Science
- Substance Use Research
Background:
- Ecological momentary assessment (EMA) and wearable sensors provide real-time data for substance use dynamics.
- AI/ML models for digital phenotyping raise concerns about fairness and inclusivity, especially for marginalized groups.
Purpose of the Study:
- To investigate barriers and facilitators to participation in wearable-based EMA studies among diverse populations.
- To develop design guidelines for more inclusive and equitable digital health research.
Main Methods:
- A four-week observational study in Hawai'i combining Fitbit monitoring and daily EMA surveys.
- Semi-structured interviews and grounded theory analysis to identify participation challenges.
Main Results:
- Six primary barriers identified: routine disruption, device discomfort, aesthetics, phone issues, substance use challenges, and sensitive contexts.
- Facilitators included participant-driven scheduling, motivational feedback, and adaptive protocols.
Conclusions:
- Wearable-based EMA research faces significant equity and inclusivity challenges.
- Proposed design guidelines aim to enhance engagement and fairness in digital phenotyping studies.
