"Personalized Behavior-Powered Systems for Guilding Self-Experiments"
Thursday, May 24, 2018 at 1:00 P.M.
Room 368 (CIT 3rd Floor)
The current paradigm in health tracking research, as performed in fields such as public health, social sciences, and research initiatives like mHealth, is to find generalizable effects that can be disseminated to the public. However, by definition, there only has to be a small effect on a subset of that population for those studies to claim a positive result. In order to avoid following general advice and to find out what works specifically for them as individuals, some people perform experiments on themselves (self-experiments). However, not everyone is trained to conduct such experiments.
This thesis proposal outlines research which aims to study how individuals perform self-experiments and to build behavior-powered systems that help them run such experiments. First, I present SleepCoacher: a system for self-experiments in sleep that includes a sleep-tracking smartphone application that collects data from sensors and user input to provide and evaluate the effect of actionable personalized recommendations for improving sleep. Next, building on completed studies with the system, I propose to modify the SleepCoacher application to give users more control over the way they run their self-experiments. Finally, extending this line of research, I also propose to build Self-E, a system for broader self-experiments, which would let the user choose which behaviors to change and automatically break down the behavior change experiment into a series of steps, and communicate them to the user through actionable messages. The goal is to make self-experimentation more accessible by providing easy to understand interventions and results. Together, these systems will lead us towards a vision of perpetual self-experiments, where users can continuously receive recommendations and change little snippets of their behavior to constantly improve their well-being.
Host: Professor Jeff Huang