Stanford Screenomics: An Open-source Platform for Unobtrusive Multimodal Digital Trace Data Collection from Android Smartphones

This is the authors' abstract. We don't add key points for this paper.

Smartphone-based digital trace data can offer powerful insights for identifying behavioral patterns and health risks. However, existing tools for comprehensive data collection lack scalability, customizability, transparency, and accessibility. To address these gaps, we developed an open-source platform that enables in-situ capture of multimodal digital traces from smartphones (e.g., moment-by-moment capture of screenshots, application usage logs, interaction histories, and phone sensor readings). The Stanford Screenomics Data Collection application allows researchers to tailor data types and quality, data transfer methods, and upload cadence. The Dashboard application supports real-time monitoring of participants' data provision, identification of data issues, and automated reactive communications to participants. The platform's back-end employs a NoSQL database for secure, and HIPAA-compliant storage. Using illustrative 24-hour digital trace data we demonstrate how the platform expands the range of possible digital phenotyping studies.