24024 WellAtlas: Mapping Digital Phenotyping Insights to Neighborhood Environment for Youth’s Health Behavioral Interventions

Project Title: 24024 WellAtlas: Mapping Digital Phenotyping Insights to Neighborhood Environment for Youth’s Health Behavioral Interventions

Project Description

Environmental influences have a direct impact on youth’s activity levels, which play crucial roles in their physical and mental health. Effective behavioral interventions are needed to motivate them to engage with their surroundings at the proper times and places. Digital phenotyping, tracking and intervening users based on their pervasive, personal data, brings promises to inform the design of such interventions, but scholars also raise concerns about inequality, data availability, privacy, and reliability of such approaches. This present work proposes a two-stage study, leveraging digital phenotyping approaches to design a data-driven behavioral intervention prototype that enhances children’s engagement with their physical environments to improve their health states. The first stage integrates data from social media and information about environmental hazards and benefits to create WellAtlas, a Google App extension that guides users on approaching comforting environments and avoiding stressors. Next, we will conduct a longitudinal user study with youth participants, tracking their usage of WellAtlas over the course of a month. We monitor variances in their physical and mental health across the study and examine whether the interventions also motivate them to reflect more on the relationship between their well-being and their local environments. The present work serves as a pilot project for a larger, multi-year research program dedicated to designing data-driven individual health interventions and public health communication approaches.

Key Words: digital mental health, behavioral intervention

Research Topics: (1) Identifying environmental stressors through spatial information; (2) Triangulating individuals' mental health states and performing digital phenotyping through personal sensing data.

Tasks and Responsibilities 

Integrating and analyzing multi-stream data

Minimum Qualifications    

Python and/or R programming

Terms of the Project: One year (students can choose to engage in 3-month, 6-month, or one-year term depending on their availability)

Mentor: Angel Hsing-Chi Hwang

Job Title: Assistant Professor

Organization: University of Southern California

Email: angel.hwang@usc.edu