Goal-Centered Personal Informatics Tools

Abstract

Personal data holds considerable potential for improving health and well-being: it can enable health understanding, management, and behavior change. However and despite the increasing abundance of personal data, a myriad of challenges come in the way of realizing this potential. Available systems often force recording data that is poorly connected to care needs and offer no or limited means of adjusting recording and aligning the data to the needs. End-users are on their own in working with the data, even though they frequently lack time or skills to implement their own data collection and analysis workflows. Difficulties linking data to effective actions further limit care opportunities. My dissertation demonstrates that health informatics tools with explicit representation of an individual's goals can help overcome these challenges in collecting, analyzing, interpreting, and acting on personal data. It does so through three systems: In MigraineTracker, I show tools which elicit an individual's goals can support highly personalized tracking (e.g., what, when, and how to track) and sense-making (e.g., what data to present and how). In Analyticons, I introduce constraint-embedded goal-aligned visual objects that scaffold end-user analysis of heterogeneous personal data. In WoNoB, I leverage goal pursuit techniques to enable reflection on personal data toward effective actions, despite changing contexts and competing needs. Together, my work establishes goal-directed design as a principled approach to giving individuals the agency to control systems and align them to their needs, while complementing their efforts with relevant computational and theoretical scaffolding.