Designing Self-Monitoring Technology to Promote Data Capture and Reflection

Abstract

Self-monitoring is a powerful means for self-reflection, which is important in health behavior change. More recently, researchers and companies began to offer numerous self-monitoring technologies for health. However, people—even experienced self-trackers such as Quantified-Selfers—have difficulty with continued tracking and gaining insights from their personal data. The goal of this dissertation is to provide insights on designing effective self-monitoring technology. In particular, I examine ways to support easy capturing of data and create persuasive feedback to nudge people toward positive behaviors. As part of this research, I introduce a mobile self-monitoring technology, SleepTight, a lightweight self-monitoring application widget that helps people capture and reflect on sleep behaviors. The SleepTight system was designed based on two formative studies and theories of reactivity in self-monitoring research. I leveraged the Android platform's lock screen and home screen widgets to lower the capture burden and increase awareness. I conducted a 4-week deployment study to evaluate the efficacy of the SleepTight system and found that SleepTight's widgets served as visual reminders and helped participants collect more data, more accurately. Participants were also able to reflect on sleeping patterns and relationships among the captured factors. I further examine effective ways to provide self-monitoring feedback by leveraging the Framing effects with an aim to nudge people toward positive health behaviors. The goal of self-monitoring is not simply to quantify individuals' behaviors, but to improve it. Therefore, self-monitoring feedback needs to convey information to help people make health-enhancing, self-beneficial decisions. To identify the type of framing that could best convey self-monitoring feedback, I conducted an online experiment to test the effects of three types of performance feedback framing—(1) valence, (2) presentation type, and (3) data unit—on individuals' self-efficacy. I identified that it is better to use a positive framing with data units (e.g., raw data, rate, percentage) that can increase the perception of one's performance capabilities to enhance individuals’ self-efficacy. This work provides empirical guidance for creating influential, persuasive performance feedback, thereby helping people designing self-monitoring technologies to promote healthy behaviors. In this dissertation, I discuss how we can successfully design self-monitoring technology to help people collect data easily, learn their behavioral patterns, and develop positive changes for improving health. Effective self-monitoring technology eases the capture burden, supports customization, prevents backfilling, and provides feedback in a positive light. Self-monitoring technology that adheres to these guidelines can enhance tracking adherence, data accuracy, data awareness, self-reflection, and self-efficacy. I verify these thesis statements through a mixed-method approach including formative studies, technology deployment study, and experimental study. This dissertation research expands our knowledge of how consumer health information technology should be designed to support self-monitoring and reflection. Once a motivated individual meets a well-designed self-monitoring technology, exciting possibilities will arise for gaining insights for health, wellness, and other aspects of life.