Everyday Personal Informatics

Thesis (Ph.D.)--University of Washington, 2018The increase in ubiquity of personal tracking tools has resulted in more people tracking their actions and behaviors, bringing more varied expertise and more diverse goals to the process. Although this increase in access provides great opportunity for more people to benefit from self monitoring, tracking tools often fail to acknowledge and account for the realities of what it means for this more diverse group to use track in their everyday life. My dissertation provides evidence that accounting for the challenges of everyday life can help people find more value in their data and find more support through their data. Toward helping people find value in their data, I first describe the Lived Informatics Model, a new conceptual model which describes people’s use of tracking technology in everyday life. In addition to behavior change, people are also motivated to track out of goals of curiosity and a desire to have a record. I suggest that collecting data and acting on it is part of a larger process of deciding to track, selecting a tool, and lapsing in tracking. I also surface that people desire additional insight from their data while they are tracking, which motivates the design of the Visual Cuts system. Through Visual Cuts, I demonstrate that surfacing correlations between aspects of multi-dimensional tracked data to better help people understand their habits and identify ways they can improve them. Examining how designs can better help people find support through their data, I describe the Design Framework for Sharing, distilling the body of research on sharing tracked data to six dimensions key to creating positive sharing experiences. I analyze and vary Tweets generated by the RunKeeper app to understand how one of these dimensions, post content, can be improved. I show that posts receive more response and interest when they explain a moment’s importance to the audience, which motivates the design of the Yarn app. Through the design and evaluation of Yarn, I demonstrate that a structured experience for authoring content can help people create posts from their tracked data which explain a moment’s importance. Across these projects, I argue that the common approaches in today’s tracking tools for summarizing and presenting data do not provide the benefits promised and the sharing mechanisms do not help people get advice and encouragement they desire. I suggest that accounting for the varied reasons why people track and designing for varied levels of expertise results in designs which better help people understand their habits and get support from others