Storytelling With Data

Storytelling with Data
Credit: prof. Nam Wook Kim 1

Goal
Beyond exploratory analysis: Visualization for
communication
TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

School of Information and Communication Technology
2
Perceptual
Data Variables Visualization
X Y
10.0 8.04
15
8.0 6.95
11
13.0 7.58
8
Y
9.0 8.81
4
11.0 8.33 0
14.0 9.96 0 4 8 11 15
6.0 7.24 X
4.0 4.26
Position (x, y)
12.0 10.84
7.0 4.82
5.0 5.68
3
• Expressiveness
• A set of facts is expressible in a visualization if it
expresses all the facts and only the facts in the data.
• Eﬀectiveness
• A visualization is more eﬀective than another one if
the information conveyed is more readily
perceived.

4
Data Insights ?
Exploratory
• Data Centered
• Domain Experts
• Analysis
• Desktop
• In-Lab
5
Need to communicate
insights found from data
6
Infographics
7
Storytelling Hans Rosling
8
Data Insights Messages
Exploratory Beyond Exploratory
• Data Centered • Human Centered
• Domain Experts • General Audience
• Analysis • Communication
• Desktop • Off-Desktop
• In-Lab • In-the-Wild
9
Practical guidance to better communicate visually with data
1. Understand the context

2. Choose an appropriate visual display
3. Use redundant encoding
4. Eliminate clutter
5. Focus your audience’s attention
6. Add explanations
7. Tell a story

10
Understand the context

11
Explanatory analysis
• The point of communicating our analysis to our

audience
• Be in the explanatory space
• You have a specific thing you want to explain, a
specific story you want to tell

12
Explanatory analysis
• A few things to think about and be extremely

clear on before visualizing any data or creating
content.
• Who
• First, To whom are you communicating?
• What
• Second, What do you want your audience to know or do?
• Mechanism
• How will you communicate to your audience?
• Tone
• What tone do you want your communication to set
• How
• What data is available that will help make my point?

13
14
Choosing an effective visual

15
Simple text
• When you have just a number or two to share,

simple text can be a great way to communicate
• Think about solely using the number— making it as
prominent as possible—and a few supporting words
to clearly make your point.

16
Example: April 2014 Pew Research Center report on
stay-at-home moms

17
Heatmap
• A way to visualize data in tabular format

• Leverage colored cells that convey the
relativemagnitude of the numbers
• Use color saturation to provide visual cues, helping
our eyes and brains more quickly target the
potential points of interest

19
Use color to highlight
• Focus primarily on those cases where cost per

mile is above average

20
Slopegraph
• Show relative increases and

decreases or differences
across various categories
between the two data point.
• Slopegraph shows
• Absolute values
• The visual increase or
decrease in rate of change.

21
Draw attention to the single category that decreased
over time

22
Bar charts

23
Redundant encoding

25
26
going up!
Message
Redundancy
Annotation
Human recognizable
objects

27
going up! 100%
79%
Percent
Message Data
Redundancy Redundancy
Bins
Annotation Data Labels
100%
79%
Percent
Bins
Human recognizable Lightness
objects

28
Example 1
G ENDER EQUALITY IN LABOR FORCE PARTICIPATION
100%
(%)
TO MALE
LABOR FORCE PARTICIPATION
RATIO OF FEMALE 75%
50%
25%
0%
CHINA KOREA JAPAN INDIA
Source: Gender Statistics 2013, World Bank
ORIGINAL
29
Example 1
100%
(%)
TO MALE
75% 82%
70% 69%
RATIO OF FEMALE
50%
25%
34%
0%
DATA REDUNDANCY
30
Example 1

CHINA LEADS IN FEMALE LABOR FORCE PARTICIPATION WHEREAS INDIA
LAGS SIGNIFICANTLY BEHIND AMONG OTHER ASIAN-PACIFIC COUNTRIES IN
2013.
(%)
RATIO OF FEMALE TO MALE 100%
75% 82%
70% 69%
50%
25% 34%
0%

DATA & MESSAGE REDUNDANCY
31
Example 1
G ENDER EQUALITY IN LABOR FORCE PARTICIPATION G ENDER EQUALITY IN LABOR FORCE PARTICIPATION
CHINA LEADS IN FEMALE LABOR FORCE PARTICIPATION WHEREAS INDIA
100% LAGS SIGNIFICANTLY BEHIND AMONG OTHER ASIAN-PACIFIC COUNTRIES IN
LABOR FORCE PARTICIPATION (%)
2013.
RATIO OF FEMALE TO MALE
(%)
75% 100%
RATIO OF FEMALE TO MALE

75% 82%
50% 70% 69%
50%
25%
25% 34%
0%
0%

Source: Gender Statistics 2013, World Bank Source: Gender Statistics 2013, World Bank
ORIGINAL DATA & MESSAGE REDUNDANCY
32
Example 2
LIFE EXPECTANCY VS ECONOMIC GROWTH
100
75 KOREA JAPAN
LIFE EXPECTANCY CHINA
INDIA
50
25
0
$0 $13,750 $27,500 $41,250 $55,000
Gross National Income per capita (US$)
Source: World Bank national accounts data, and

OECD National Accounts data files (2013).
ORIGINAL
33
Example 2
100
75
LIFE EXPECTANCY CHINA KOREA JAPAN
INDIA
50
25
0
$0 $13,750 $27,500 $41,250 $55,000

DATA REDUNDANCY
34
Example 2

RELATION BETWEEN LIFE EXPECTANCY AND GROSS NATIONAL INCOME
OF ASIAN-PACIFIC COUNTRIES IN 2013
100
LIFE EXPECTANCY 75
CHINA KOREA JAPAN
INDIA
50
Life expectancy increases
with greater income
25
0
$0 $13,750 $27,500 $41,250 $55,000

DATA & MESSAGE REDUNDANCY
35
Example 2
LIFE EXPECTANCY VS ECONOMIC GROWTH LIFE EXPECTANCY VS ECONOMIC GROWTH

RELATION BETWEEN LIFE EXPECTANCY AND GROSS NATIONAL INCOME
100 OF ASIAN-PACIFIC COUNTRIES IN 2013
100
75 KOREA JAPAN
LIFE EXPECTANCY
CHINA
75
LIFE EXPECTANCY
INDIA CHINA KOREA JAPAN
50
INDIA
50
Life expectancy increases
25 with greater income
25
0
$0 $13,750 $27,500 $41,250 $55,000
0
Gross National Income per capita (US$) $0 $13,750 $27,500 $41,250 $55,000
Source: World Bank national accounts data, and Source: World Bank national accounts data, and
OECD National Accounts data files (2013). OECD National Accounts data files (2013).
ORIGINAL DATA & MESSAGE REDUNDANCY
36
Example 3
VERY FEW W OMEN A RE STUDYING C OMPUTER SCIENCE

% of bachelor degrees awarded for women
COMPUTER S CIENCE 18%
MATHEMATICS 42%
S OCIAL S CIENCES 53%
P SYCHOLOGY 72%
0% 25% 50% 75% 100%
Source: National Student Clearinghouse Research Center (2014).
WITHOUT PICTOGRAM
37
Example 3
VERY FEW W OMEN A RE STUDYING C OMPUTER SCIENCE VERY FEW W OMEN A RE STUDYING C OMPUTER SCIENCE
% of bachelor degrees awarded for women % of bachelor degrees awarded for women
COMPUTER S CIENCE 18% COMPUTER S CIENCE 18%
MATHEMATICS 42% MATHEMATICS 42%

P SYCHOLOGY 72%
P SYCHOLOGY 72%
0% 25% 50% 75% 100%
Source: National Student Clearinghouse Research Center (2014). Source: National Student Clearinghouse Research Center (2014).
WITHOUT PICTOGRAM WITH PICTOGRAM

38
Eliminate clutter

39
Example: Lack of visual order

40
Improve the placement and formatting of elements

41
Example: Weighted performance index across categories
• Shows the weighted performance index across

categories for your company and five
competitors

42
Use contrast more strategically

43
Eliminate distractions

44
Focus your audience’s attention

45
Recall preattentive attributes
• Preattentive attributes are visual properties that we notice

without using conscious effort to do so.
• use preattentive attributes strategically, they can help us enable
our audience to see what we want them to see before they even
know they’re seeing it!

46
Example: Leverage color to draw attention

47
Example: Create a visual hierarchy of information
• Modified focus and text to lead our audience

from the macro to the micro parts of the story.

48
Use color sparingly

49
Aesthetic design

50
Add explainations

51
Text in data visualization
• Useful to annotate important or interesting

points directly on a graph.

52
Example 4
PERCENTAGE OF U NINSURED A MERICANS

% of uninsured people in the U.S.
20%
17%
16.0% 15.1% 14.7% 14.4%
14%
11.5%
11%
8%
5%
2010 2011 2012 2013 2014 2015
Source: CDC/NCHS, National Health Interview Survey, 2010–2015
BAD TITLE
53
Good title
PERCENTAGE OF U NINSURED A MERICANS A MERICA’S U NINSURED RATE D IPS BELOW 10%

% of uninsured people in the U.S. % of uninsured people in the U.S.
20% 20%
17% 17%
16.0% 15.1% 16.0% 15.1%
14.7% 14.4% 14.7% 14.4%
14% 14%
11.5% 11.5%
11% 11%
9.2% 9.2%
8% 8%
5% 5%
2010 2011 2012 2013 2014 2015 2010 2011 2012 2013 2014 2015
Source: CDC/NCHS, National Health Interview Survey, 2010–2015 Source: CDC/NCHS, National Health Interview Survey, 2010–2015
BAD TITLE GOOD TITLE

54
Tell stories with data

55
Bach et al. CG&A’17
56
Why telling a story with data
• A single chart may not be enough

• To explain
• To engage an audience in decoding a message
• How do we engage an audience?
• How do we break down complexity?
• How do we guide the audience?

57
Stories
Stories are the most powerful delivery tool for information, more
powerful and enduring than any other art form
58
The best stats you've ever seen
• https://www.youtube.com/watch?v=hVimVzgt
D6w
Use of elements from storytelling

to convey compelling data stories
59
Data Storytelling Process: transforming data into visual stories
[B. Lee et al 2015]

60
Iterative exploration:
Traditional visualization process
[B. Lee et al 2015]

61
[B. Lee et al 2015]

62
[B. Lee et al 2015]

63
7 Key Elements of Compelling Data Storytelling
• Clear and coherent narrative structure

• Captivating introduction, logical flow, connecting
data points and insights, compelling climax,
memorable ending
• Characterisation of data points
• Relevance, impact, and alignment with the narrative
• Engaging visuals and data representations
• Contextualisation and relatability
• Emotional resonance
• Simplicity and clarity
• Adaptation to the audience
64
Narration Styles
• Author-Driven • Reader-Driven
• Strict ordering of • No prescribed
scenes ordering
• Heavy messaging • No messaging
• No interactivity • Free interactivity

65
https://www.youtube.com/watch?feature=player_detailpage&v=hVimVzgtD6w#t=159s

66
http://playground.tensorflow.org/
https://archive.nytimes.com/www.nytimes.com/interactive/2010/02/02/us/politics/2010
0201-budget-porcupine-graphic.html?_r=0
68
https://guns.periscopic.com/
69
http://www.fallen.io/
70
Visual narrative design

71
Employ narrative structure
Bach et al 2017

72
73
E. Segel
74
E. Segel
75
E. Segel
76
Genres of data stories

77
Use Multiple Charts
• Break down the complexity of the story and

progressively reveal different facets of the
data.
Bach et al 2017
78
Storytelling, double-edged Sword?

79
[South China Morning Post
2011]
80
[South China Morning Post Flipped upside down….
2011]
81
Same Data, Different Stories
[South China Morning Post Flipped upside down….
2011]

82

Storytelling With Data - v2

Uploaded by

Document Informationclick to expand document information17. Storytelling with Data_v2

Document Informationclick to expand document information

Copyright:

Available Formats

Storytelling With Data - v2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Storytelling With Data - v2

Uploaded by

Copyright:

Available Formats

Credit: prof. Nam Wook Kim 1

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• Domain Experts • General Audience

1. Understand the context

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• The point of communicating our analysis to our

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• A few things to think about and be extremely

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• When you have just a number or two to share,

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• A way to visualize data in tabular format

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• Focus primarily on those cases where cost per

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

• Show relative increases and

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

G ENDER EQUALITY IN LABOR FORCE PARTICIPATION

Source: Gender Statistics 2013, World Bank

G ENDER EQUALITY IN LABOR FORCE PARTICIPATION

Source: Gender Statistics 2013, World Bank

G ENDER EQUALITY IN LABOR FORCE PARTICIPATION

CHINA KOREA JAPAN INDIA

DATA & MESSAGE REDUNDANCY

RATIO OF FEMALE TO MALE

CHINA KOREA JAPAN INDIA

ORIGINAL DATA & MESSAGE REDUNDANCY

LIFE EXPECTANCY VS ECONOMIC GROWTH

Source: World Bank national accounts data, and

LIFE EXPECTANCY VS ECONOMIC GROWTH

Source: World Bank national accounts data, and

LIFE EXPECTANCY VS ECONOMIC GROWTH

Source: World Bank national accounts data, and

DATA & MESSAGE REDUNDANCY

LIFE EXPECTANCY VS ECONOMIC GROWTH LIFE EXPECTANCY VS ECONOMIC GROWTH

ORIGINAL DATA & MESSAGE REDUNDANCY

VERY FEW W OMEN A RE STUDYING C OMPUTER SCIENCE

COMPUTER S CIENCE 18%

S OCIAL S CIENCES 53%

0% 25% 50% 75% 100%

Source: National Student Clearinghouse Research Center (2014).

COMPUTER S CIENCE 18% COMPUTER S CIENCE 18%

MATHEMATICS 42% MATHEMATICS 42%

S OCIAL S CIENCES 53%

WITHOUT PICTOGRAM WITH PICTOGRAM

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG

TRƯỜNG CÔNG NGHỆ THÔNG TIN VÀ TRUYỀN THÔNG