Personal Informatics and HCI, Ian Li, June 2010

Personal Informatics & HCI

Ian Li
HCI Institute
Carnegie Mellon University

Ian Li Personal Informatics & HCI June 2010

Gnothi seauton.

2

Know thyself.

3

4

Personal Informatics
A class of systems that help people  
collect and reﬂect on their behavior  
to gain self-knowledge

5

http://personalinformatics.org/tools

Physical Activity

Finance

Electricity

Diabetes

Health

Mood

6

HCI can help with the design of
personal informatics systems.

7

Human-Computer Interaction
The study of interaction between people and
computers

Intersection of computer science, behavioral
sciences, and design

8

Why HCI?
Personal informatics systems must provide
users with different kinds of support.
•  Make daily usage easier and facilitate use
over a long period of time.
•  Organize the data.
•  Help users understand the data.
•  Help users act on new knowledge.

9

Alice
•  20 years old
•  Family history of heart
disease
•  Wants to be more active,
but doesnʼt know how
because sheʼs busy

10

1. Alice prepares.

11

2. Alice collects data.
Mon
1573
Tue
4392
Wed
4537
Thu
5842
Fri
10258
Sat
7528
Sun
1368
Mon
1497
Tue
1837

12

3. Alice transcribes data.
Transcribe to Excel

M
T
W Th
F
Sa
Su
M
T

13

4. Alice reﬂects on the data.

Active

Inactive
Inactive

M
T
W Th
F
Sa
Su
M
T

14

5. Alice takes action.
Walk in the park
instead of
watching TV

M
T
W Th
F
Sa
Su
M
T

15

PREPARATION COLLECTION INTEGRATION REFLECTION ACTION

16


Collection
The stage when people collect information
about themselves (e.g., inner thoughts, behavior,
social interactions, and their immediate environment).

17


Collection Barriers
•  Using the tool
•  Remembering
•  Lack of time
•  Motivation
•  Finding data
•  Accuracy
18


Collection Barriers
•  Using the tool
One problem is: 
•  Remembering
“Keeping up the
•  Lack of time
motivation to do so;
•  Motivation
like ﬁnding payback
for the investment of
•  Finding data
time and effort.”
•  Accuracy
19


Reflection
The stage when people reflect on their
personal information.
•  Users may reflect immediately (short-term)
•  Or after several days or weeks (long-term)

20


Reﬂection Barriers
•  Lack of time
•  Self-criticism
•  Visualization
•  Interpretation
•  Sparse data
•  No context
21


Reﬂection Barriers
“Itʼs hard to get a
•  Lack of time
holistic view of the
•  Self-criticism
data since the time
ﬁlters are at most one
•  Visualization
month and Iʼd like to
•  Interpretation
look at several
•  Sparse data
months at once.”
•  No context
22

Properties of the Stages
1.  Barriers cascade
2.  Stages are iterative
3.  User- vs. System-driven
4.  Facets

23

3. User- vs. System-driven

User-driven
System-driven

24

Mon
1573
Tue
4392
Wed
4537
Thu
5842
Fri
10258
Sat
7528
Sun
1368
Mon
1497
Tue
1837

User-driven
System-driven

Collection

Integration

Reﬂection

25


User-driven
System-driven

Collection

Integration

Reﬂection

26

Design Guideline
Consider the tradeoffs between user-driven
and system-driven stages.

27

Introduction
Model of Personal Informatics
Collection vs. Reﬂection
Maintaining Compliance
Increase Use by Sharing
Easing Data Integration
Other Projects

28


29

IMPACT
Different from most personal informatics
systems for physical activity:

•  Collects physical activity information  
and context (e.g., type of activity, location, people)

•  Visualizations to help users become aware 
of factors in their lives that affect their
physical activity.
30

Two prototypes – Two studies
IMPACT 1.0
IMPACT 2.0
Manual collection
Semi-automated
collection

31

IMPACT 1.0

32

Awareness of opportunities
“It turns out I get the most walking done to
and from work…and walking around my
neighborhood for an hour or two made a
difference.” P35

33

IMPACT 2.0

34

35

36

No difference
Compared to other systems:
•  No visualizations
•  Steps-only visualizations

37

Long-term value
Six months later, users who collected
contextual information reﬂected on their data
better.

38

Short-term Long-term
Reﬂection
Reﬂection

IMPACT 1.0
Manual GOOD
NOT GOOD
Collection

IMPACT 2.0
Automated NOT GOOD
GOOD
Collection

39

Introduction
Other Projects

40

Barriers Example Quote

Tool (13/67) “not having ready access to a
computer at the time symptoms
happen” P6

Memory “Forgetting to record it. Because I
(12/67) am often not at my personal
computer.” P57

Lack of time “not difficult, time consuming at
(11/67) times.” P16
Finding “Sometimes life isnʼt interesting
data (7/67) enough to make me want to write
it down, other times I canʼt find
any worthy writing material.” P54

Accuracy “Guestimating mass of food
(6/67) matching homemade or
restaurant foods against
database entries” P5
Motivation “keeping up the motivation to do
(5/67) so, finding payback for the
investment of time and effort.” P4

41

Experience Sampling
Collect self-reports from users
•  3 questions per self-report
•  10 self-reports per day
•  28 days
•  840 pieces of information from user

42

43

44

45

46

Experience Sampling


Experience Sampling+Feedback


49

Increased compliance by 23%


Maintained compliance
Control
A+I
A+M

1
Compliance Rate

0.8

0.6

0.4

0.2

0
P1
P2
P3
P4
P5

51

Personal Insights
“When I was doing something productive my
mood was not high. My mood was highest
when I was playing games.”

“Apparently, I am least productive with work
(Word, Excel) at night when I use AIM. I
never use AIM in the morning, thus much
more productive with work.”

52

Introduction
Other Projects

53


54

System Issues
Value of self-reﬂection is not immediate

Low participation by users

55

System Issues
Value of self-reﬂection is not immediate
Use sharing as initial motivation
Low participation by users
Sharing increases participation

56

http://moodjam.org

57

58

59

Introduction
Other Projects

60

Barriers Example Quote

Tool (13/67) “not having ready access to a
computer at the time symptoms
happen” P6

Memory “Forgetting to record it. Because I
(12/67) am often not at my personal
computer.” P57

PREPARATION COLLECTION
Lack of timeREFLECTION
INTEGRATION
“not difficult, ACTION
time consuming at
(11/67) times.” P16
Finding “Sometimes life isnʼt interesting
data (7/67) enough to make me want to write
it down, other times I canʼt find
any worthy writing material.” P54

Accuracy “Guestimating mass of food
(6/67) matching homemade or
restaurant foods against
database entries” P5
Motivation “keeping up the motivation to do
(5/67) so, finding payback for the
investment of time and effort.” P4

61

System Issues
Access to collection tool

Lack of integration of data sources

Most tools collect only one type of data

62

Graﬁtter
Access to collection tool
Support multiple social media
Lack of integration of data sources
Cull data automatically
Most tools collect only one type of data
Flexible and easy-to-enter format

63

Supports multiple social media
Twitter
Delicious
Instant Messenger
Blogger

64

Flexible and easy-to-enter
Extension of Twitter hashtags
#mood(happy) Just had ice cream.
A long day #work(11) and itʼs not over.
Lunch with friends #lunch(amy, bob)

65

http://graﬁtter.com

66

Graﬁtter
Up since February 2009
Continuing development
•  Support privacy
•  Additional visualizations
•  Multi-faceted reﬂection

67

Introduction
Other Projects

68

http://ianli.com/dd/

69

http://tinyurl.com/actionshot/

70

http://belikeben.com
Be Like Ben

71

http://ianli.com/tlc/

72

Thank you!
ianli@cmu.edu 

http://ianli.com 
http://personalinformatics.org
http://personalinformatics.org/lab/model/

Funded by

73

Personal Informatics and HCI, Ian Li, June 2010

More Related Content

Similar to Personal Informatics and HCI, Ian Li, June 2010 (20)

More from Ian Li (7)

Recently uploaded (20)

Personal Informatics and HCI, Ian Li, June 2010