SleepKit A Mobile Sleep Tracking Application for Promoting Healthy Sleep Behavior Supervisors: Pieter Robberechts Joris Klerkx Elena Smets http://probberechts.github.io/thesis
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
MOTIVATION Adequate sleep is a key part of a healthy lifestyle. 35 percent of Americans report their sleep quality as “poor” or “only fair” First step towards an improved sleep quality = improve sleep hygiene Engagement with data collection enhances people’s awareness and provides an opportune moment to reflect on their behaviors Self trackers are attracting a lot of users SLEEPKIT : a mobile phone sleep tracking application to help promote awareness about healthy sleep habits and improve sleep quality
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is How to Prompt good / bad? improve change
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is good / bad? Not enough? What should it look like? Is this normal?
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is good / bad? % LIGHT SLEEP DEEP SLEEP
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is good / How to bad? improve 6 HRS 5 HRS 4 HRS 3 HRS 2 HRS 1 HRS 0 HRS Stop drinking Stop drinking Finish Turn of Sleep! caffeine alcohol excercising electronics Avoid heavy Stop working, studying meals and stressing
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is good / How to Prompt bad? improve change
GOAL OF THIS THESIS How can a sleep tracking application promote healthy sleep behavior and contribute to a better sleep quality? What is good / How to Prompt bad? improve change HealthPatch sensor (IMEC): - 2-lead ECG - 3-axis accelerometer → WAKE, REM, DEEP, LIGHT
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
EXPERIMENT DESIGN Initial session In the field Final session Informed consent 10 participants Questionnaire: - SUS - Interpretability 7 nights - Awareness - Motivation - Sleep with HealthPatch + - Daily synchronize - Reflect on the tracked data Discussion Tutorial Entrance questionnaire - Demographics - Sleep quality (PSQI)
PARTICIPANTS Their sleep quality: 10 Participants Their knowledge of sleep hygiene: Know that caffeine should be avoided in the evening 1O 7 Know the effect of alcohol on sleep 0 Can name at least 6 sleep hygiene recommendations
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
USABILITY
USABILITY 5 4 3 2 1 I think that I would like to use this system frequently
USABILITY 5 4 3 2 1 I found the system unnecessarily complex
USABILITY WORST POOR OK GOOD EXCELLENT BEST IMAGINABLE IMAGINABLE DP1 DP3 Experiment Experiment A. Bangor, P. Kortum, and J. Miller, “Determining what individual SUS scores mean: Adding an adjective rating scale,” Journal of usability studies , vol. 4, no. 3, pp. 114–123, may 2009.
G1: INTERPRETABILITY 4 Q1. Easy to tell if quality of night is good / Q4. Easy to link poor sleep quality to cause bad Q2. Easy to compare quality of nights Q5. Learned all term quickly Q3. Easy to judge sleep pattern Q6. No irrelevant information
G2: AWARENESS Q1. More aware of importance of sleep Q4. Learned something new about my sleep hygiene pattern Q2. Learned something new about sleep Q5. Learned how to improve some poor hygiene elements
G3: MOTIVATION 1 3 2 4 Q1. Spent more attention on sleep hygiene Q4. Sleep hygiene tips had an effect Q2. Spent more attention on sleep pattern Q5. Motivated by seeing effect in tracked data Q3. Motivated because of tracking Q6. Bedtime notifications motivated me
Goal and Motivation Iterative Development Demo Experiment Results Conclusions
Conslusions Users are more interested in tracking their sleep than improving their sleep quality BUT: - See sleep hygiene recommendations as an added value - A reason to check the app daily - Encourages self reflection http://www.sarahvanbelle.be/The-Quantified-Self
Conslusions Risk for demotivation BECAUSE: - Effect is not always immediate - Sleep tracking is not 100% accurate → false negative data http://33minutes.net/
Conslusions SleepKit has the potential to promote healthy sleep behavior DATA: - 7 users got more aware of the importance of sleep hygiene - 8 users learned something new about sleep hygiene - 6 users spent more attention on their sleep hygiene / sleep pattern http://sleepsloth.com/
Thanks! Any questions?
Literature Implementation Sleep Stage Classification Iterative Development
SHUTEYE J. S. Bauer, S. Consolvo, B. Greenstein, J. Schooler, E. Wu, N. F. Watson, and J. a. Kientz, “ShutEye: Encouraging awareness of healthy sleep recommendations with a mobile, peripheral display,” Human Factors in Computing Systems (CHI Conference) , pp. 1401–1410, 2012.
SLEEPTIGHT E. K. Choe, B. Lee, M. Kay, W. Pratt, and J. A. Kientz, “SleepTight: Low-burden, Self-monitoring Technology for Capturing and Reflect- ing on Sleep Behaviors,” Proceedings of the ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp ’15) , pp. 121–132, 2015.
LIGHTS OUT W. V. Chen, M. Sra, and R. W. Picard. Improving Sleep-Wake Schedule Using Sleep Behavior Visualization and a Bedtime Alarm. In MobiHealth , pages 0–3. ICST, dec 2015.
COMMERCIAL APPS
Literature Implementation Sleep Stage Classification Iterative Development
IMPLEMENTATION: DATA SYNCHRONIZATION
IMPLEMENTATION: MVP
Literature Implementation Sleep Stage Classification Iterative Development
ML APPROACH
QRS PEAK DETECTION
EDR
BASELINE WANDER
BASELINE WANDER REMOVAL
Literature Implementation Sleep Stage Classification Iterative Development
PAPER PROTOTYPE
DIGITAL PROTOTYPE 1 Main ideas: - Clock based metaphor for ideal preparation of a night - A quick overview of the quality of the previous nights
DIGITAL PROTOTYPE 1 Evaluation results: - 36 possible usability problems - Glanceability problem - Confusing: two days visible on circle - What do colors mean? - What is clickable - Too dense to click on icons SUS score: 72 [64, 83]
DIGITAL PROTOTYPE 2 Main ideas: - Vertical timeline to create more space - Show by default only data most relevant at the current time + scroll for older data
DIGITAL PROTOTYPE 2 Evaluation results: - Same problem: too dense
DIGITAL PROTOTYPE 3 Main ideas: - Split sleep hygiene information and tracked data
DIGITAL PROTOTYPE 3 Evaluation results: - Horizontal scrollable areas → add shadow at borders - What is clickable → add shadow below SUS score: 81 [75, 87.5]
FINAL VERSION Main ideas: - Small improvements - Bugfixes
Recommend
More recommend