Hypothesis Formation and Hypothesis Testing: Design Challenges in Self-Experimentation

James Fogarty University of Washington Seattle, WA 98195, USA [email protected] Jasmine Zia University of Washington Seattle, WA 98195, USA [email protected] Abstract Personal informatics applications often fail to effectively support diagnostic self-tracking, wherein people seek to use data in answering specific questions about themselves. Diagnostic self-tracking is therefore difficult, tedious, and error-prone. To address these shortcomings, we designed a framework for self-experimentation in personalized health that helps guide the design of applications to support diagnostic self-tracking. We instantiated multiple aspects of the framework in our work assisting people with irritable bowel syndrome (IBS) in identifying their individualized food-based triggers. In this workshop paper, we describe our experiences from two studies to highlight key findings surrounding validity and reliability of such data. We then propose discussion topics for the workshop.