Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Understanding My Data, Myself: Supporting Self-Reflection with Ubicomp Technologies
1
Ian Li, Anind Dey, Jodi Forlizzi
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Supporting Self-Reflection
2
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Personal Informatics Tools
3
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China 4
Pedometer
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China 5
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China 6
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China 7
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Number of steps
Sleep quality
Blood glucose level
Location
Electricity usage
Web history
Food consumption
Expenditures
Weight
Driving habits
8
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
What kinds of questions do people ask about their data?
When do people ask these questions?
9
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
6 Kinds of Questions
2 Phases
People transition between phases
10
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Study Recruitment Advertised on: • Craigslist • Study recruitment site • Quantified Self blog http://quantifiedself.com
11
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Participants 91 people expressed interest.
Selected 15 participants (P1-15) to interview.
• Currently using a personal informatics tool
• Used tool for more than a month
12
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Types of Data Collected Financial expenditures (3 participants)
Sleep (3)
Weight (2)
Blood glucose level (2)
Exercise (1)
Productivity (1), web history (1), books read (1), and general life events (1)
13
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Procedure 1-hour interviews in the lab or on Skype
Observed participants using their personal informatics tool.
14
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Procedure Asked how they reflected on their data. • What were they looking for? • What questions did they ask? • How did they answer their questions? • What problems did they encounter?
15
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Analysis Identified the questions people asked about their data.
Affinity diagrams to identify themes.
16
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
6 Kinds of Questions
2 Phases
17
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
6 Kinds of Questions Status History Goals
Discrepancy Context Factors
18
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Status What is my current status?
P7 looked at her Fitbit pedometer several times a day to check her physical activity level.
19
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
History What happened in the past?
P5 looked at her spending history to check if she was reducing her expenditures.
20
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Goal What goals should I pursue?
P9 tracked what she ate and her physical activity level to find the cause of her leg cramps, so she can avoid certain foods or not do certain exercises.
21
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discrepancy How does my behavior compare to my goals?
P11 compared her current food consumption with the goal she had set for herself.
22
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Context What other things happened during a particular point in time?
P3 who has diabetes: • Measured her blood sugar level. • Kept in mind the food she ate
and her physical activity.
23
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Factors What influences my behavior over a long period of time?
P2 tracked many types of data over time to control her weight: • Exercises • What she ate • Total calories
24
Status History Goal Discrepancy Context Factors
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Importance of the Questions Questions are NOT important all the time.
Some questions are more important than others as people’s information needs change.
25
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase
Discovery Phase
26
&
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase Participants already know how different factors affect their behavior, so they just want to know what their current status is.
Participants already identified their goals. They are only concerned with whether they are meeting their goal.
27
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase Current Status
P13 just tracks the minutes that he spends on Facebook, Twitter, and other social media sites, because he already knows how these sites affect his productivity.
28
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase Discrepancy
P1 uses Mint to keep track of her expenditures to see whether she is meeting the budget that she had set for herself.
29
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase Most important kinds of questions: Status History Goals Discrepancy Context Factors
30
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase Most important kinds of questions: Status History Goals Discrepancy Context Factors
31
} Collecting data is easier because fewer types of data are collected.
Tools could provide alerts and suggestions.
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Participants are trying to figure out what their goals are.
Participants collect several types of data to find out what factors affect their behavior.
32
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Figuring out goals
P8 tracks the quality of her sleep so that she is better rested. She explores her sleep data to “spot trends for which I can take corrective action.”
33
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Finding factors that affect their behavior.
P3 has diabetes and she tracked her blood glucose levels and her food consumption to find out their interaction.
34
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Most important kinds of questions: Status History Goals Discrepancy Context Factors
35
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Most important kinds of questions: Status History Goals Discrepancy Context Factors
36
} Help users collect data over a long period of time.
Help users select the goals that they should pursue.
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Discovery Phase Most important kinds of questions: Status History Goals Discrepancy Context Factors
37
} Help users track multiple types of data.
Help users manage all the data that they collect.
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Maintenance Phase
Discovery Phase
38
&
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Transitions Discovery Maintenance
Discovery P6 initially tracked his blood sugar level, food consumption, and mood.
Maintenance Eventually, he transitioned into just collecting his blood sugar level because he felt like he had control of his diabetes.
39
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Transitions Maintenance Discovery
Maintenance P7 who has recurrent insomnia tracks her sleep using Fitbit.
Discovery When her insomnia returns, she records the factors that may be causing the problem.
40
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Transition Problems Some people might have problems transitioning.
P9 tracked multiple types of data to figure out what causes her leg cramps when she sleeps (Discovery).
She still doesn’t know the causes, so she continues to collect data about her sleep and possible causes.
41
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Summary Identified 6 kinds of questions that people ask about their data.
Described 2 phases when people ask different kinds of questions about their data.
Discussed the transitions between the phases.
42
Ian Li, Anind K. Dey, Jodi Forlizzi | Ubicomp 2011, Beijing, China
Thank you!
[email protected] http://ianli.com
Funded by
43