Visualizing Self-Tracked Mobile Sensor and Self-Reflection Data to Help Sleep Clinicians Infer Patterns.

We present results from a design study of two novel visualization methods for self-tracked sleep data. These methods combine multivariate time series data and ordinal self-reflection data to help sleep clinicians analyze factors related to sleep quality such as noise and movement and make recommendations to improve sleep quality. We conducted an iterative design process, driven by our collaborator's domain specific goals in analyzing self-tracked sleep data. The final visualizations feature a unique spiral clock radial design and interactive controls for ordinal data. We ran a survey with three sleep clinicians to assess the effectiveness of each visualization type. In our user study a sleep clinician also performed a mock in-patient session, which demonstrated their effectiveness in a clinical setting. The new visualizations prove more effective in achieving the domain specific goals of sleep clinicians in comparison to previous efforts from similar work.