Published on March 10, 2014
Using Smartphones to Research Daily Life Dr. Neal Lathia Computer Laboratory, University of Cambridge firstname.lastname@example.org
“First were mainframes, each shared by lots of people. Now we are in the personal computing era, person and machine staring uneasily at each other across the desktop. Next comes ubiquitous computing [...] when technology recedes into the background of our lives.” - Weiser
Android: 1.5 million daily activations
Phone Calls Text Messages Social Media E-mail Camera Calendars Maps Transport Games News Videos
Phone Calls Text Messages Social Media E-mail Camera Calendars Maps Transport Games News Music Videos Communication Patterns; Social Networks and Sharing; Information Diffusion; Points of Interest; Mobility; App Usage ...
Accelerometer Microphone Camera GPS Compass Gyroscope Wi-Fi Bluetooth Proximity NFC Light
Accelerometer Microphone Camera GPS Compass Gyroscope Wi-Fi Bluetooth Proximity NFC Light Activities; Transportation Modes; Co-location; Mobility; Speaker Detection; Speech Inferences; Context Identification
How can we sense, identify, and capture high-level behaviours (e.g., walking, running) from low-level sensors (e.g., accelerometer)? “Classic” Research Question (1): For ever higher level behaviours Given energy, computation constraints
How can we verify or understand qualitative aspects of behaviour that can be sensed (i.e., the why)? “Classic” Research Question (2): Choices, feelings, reasons, semantics (e.g., home vs. lat/lon).
Experience Sampling Mood; Location Semantics Focus/Engagement; Transportation Choices; Physical Activity; Health; Reactions to Music
Researching Daily Life: Reflecting on the 'scientific' method
Hypothesis Measure Test Tradition...
Hypothesis Measure Test ESM Design Sensor-ESM studies... Sensor-Sampling Design
Hypothesis Measure Test ESM Design Ready for the public... Sensor-Sampling Design User Experience Design Build App Test App Deploy App Market App Debug App Update App Re-design UX User Support De-noise data
Emotion Sense Basics: (a) Complete a “long” survey about how you feel & your current context (b) Receive notifications that ask you to complete a “short” survey (c) View how your survey responses compare to the sensor data your phone has collected.
Emotion Sense Surveys (momentary) Positive/Negative Affect Day “Goodness” Current Location Texting/Calling Habits Current Activity Device Interaction Recent Speech Personality Sociability Connectedness Engagement (non-momentary) Demographics Satisfaction with Life Personality Gratitude Health Sociability Job Satisfaction Life Aspirations Connectedness
Emotion Sense Sensors Location Texting Logs Call Logs Accelerometer Microphone Screen On/Off Samples Wi-Fi Battery Levels
● How can we keep users engaged in a seemingly repetitive task? ● Surveys are “too long” … “too obtrusive” .. “too repetitive” ● ESM experiments “don't beneft me” … “too boring” Design Challenges
A “Non”-Game designed like a game
How many questions to ask...? When the user is “volunteering” a response (10) vs. When the user is responding to a notification (3)
Limited feedback – avoiding judging
How are people interacting? – 12,000 interacting users; – 650,000+ survey responses – ~6% answer only 1 survey – ~800 daily user responses – Average of 60+ days responding to surveys per user – ~15% users > 3 months – App demographics ~ published Android user demographics
angry anxious lonely relaxedenthusiasticcalm
Emotion Sense: Ongoing Work – Making sense of noisy sensor data: various devices, time zones, participation lengths – Analysing quality, content of survey responses – Investigating the future of these tools: generalising (for researchers) vs. specialising (for a particular context).
Easy M: Can I run a study like Emotion Sense? (Without going back to the drawing board) Generalised sensor- enhanced experience sampling tool.
Easy M Basics: (a) Enter a “participant code” to join an experiment; app reconfigures itself (with timeout) (b) Receive notifications that ask you to complete your researchers' survey (c) Volunteer additional responses
Easy M Basics: Allows for both “momentary” and “volunteered” survey types. Questions include all the same types as Emotion Sense. User feedback is limited to response time and compliance.
Where could Easy M be used?
Using Smartphones to Research Daily Life Dr. Neal Lathia Computer Laboratory, University of Cambridge email@example.com http://emotionsense.org/ http://www.cl.cam.ac.uk/~nkl25/easym/
How Smartphones are Changing Consumers’ Daily Routines Around the ... Americans spent 34 hours using smartphone apps and on the mobile Web in December ...
Research: Using a Smartphone After 9 pm Leaves Workers ... As smartphones become more embedded in our daily ... This article is about WORK-LIFE ...
Evaluating Daily Life Activity Using Smartphones as Novel Outcome Measure for Surgical Pain Therapy Julia Seiter, Sebastian Feese, Bert Arnrich, Gerhard ...
The effect of Smartphones on Work-Life Balance ... healthy work life balance alongside some research into ... who have been using the ...
Where and what: Using smartphones to predict next locations and applications in daily life on ResearchGate, the professional network for scientists.
Texting at work can make employees more productive. ... between using smartphones to take short breaks for things like ... Business News Daily.
6 facts about Americans and their smartphones. ... only using their smartphone to find ... to your life? Fully 46% of smartphone owners say their ...
It's one of the first attempts to quantify the moral landscape of daily life, ... Using Smartphones to Track Our Everyday Moral ... research out ...