10
Lesson 1
Data
Visualization
11
12
13
14
16
Principles of Data Visualization
1. Simplest: Everything should be made as simple as
possible, but no simpler
If you take something out and the visualization can be understandable,
so it should not be there!
2. Clear: Clearly communicate information you need to
present
3. Precise: Precisely present data (label, legend, scale, etc.)
4. Efficient: Audience can understand it as fast as possible
(within 10 seconds!)
17
Keys of
Data
Visualization
WHO?
20
Executives Investor
Manager Customers
Operations Experts
21
Audiences
What they should know?
What they want to What did they know? What they don’t
know? (Existing knowledge) want to know?
22
10 Seconds Test
• Sit in Group
• Show your visualizations
for 10 seconds
• Ask them
“What did they learn
from it and what did
they don’t understand?”
23
WHAT?
25
26
Storytelling Techniques
Comparison Prediction Composition Distribution
Tell differences Show future Show ratio of all Show data
of data trend composition distribution
HOW?
Visualization
with Graphs
Bar Chart
Story: Comparison
X-axis: Discrete data
Notes:
• Set ‘Zero Baseline’
• Slightly change matters
• Standard Error Bars
• Arrange data
• Use vertical bar chart with limited
horizontal space
• Box-plot is bar chart with quartile
34
35
Line Chart
Story: Comparison, Prediction
X-axis: Mostly with continuous data
Notes:
• Set ‘Zero Baseline’
• Slightly change matters
• Arrange data timeline from past to
present
• Annotation in line chart makes
audience understand better
• Line chart can be used with Bar
chart
38
39
40
41
42
Pie Chart
Story: Composition
Notes:
• Total value must = 100%
• Must show value
• Arrange value clockwise
• Should not use 3D Pie Chart
• If you have many group of data, you
should use Stacked Nar Chart
instead of Pie Chart
43
44
45
46
47
Scatterplot
Story: Prediction
X-axis: Continuous data
Notes:
• Show future trends
• Bubble Chart is Scatterplot with
different size of dot which indicates
differences of data
48
49
50
51
52
53
54
55
56
57
58
Common
Mistakes
5
9
60
What do you think?
Credit: Yanofsky D & Fernholz T.
Credit: Hutchins Center
Great Visualization begins with
understanding basic concepts
6
0
61
1 Objectives of Creating Data Visualization
Function over Aesthetics
Simplest
Clear
Precision
• No x-Axis info.
• Different range of time
Efficient
62
1 Objectives of Creating Data Visualization
Function over Aesthetics
Simplest
Clear
• Compare or Predict?
Precision
• X, Y Axis?
• X is continuous value?
Efficient
• Difficult to understand
• Meaning of color
• Need Annotation
63
2 Understand Data Characteristics
Continuous Data Discrete Data
64
3 Respect Scale
Simplest
• Color may not be necessary
Clear
• Clear comparison
Precision
• Is IQ105 smarter than IQ93??
• Y axis should start with 0
Efficient
• Rearrange data
65
Respect Scale
matters
66
4 Hierarchy
Simplest
• Color tone
Clear
• Clear comparison
Precision
Efficient
• Which point should be read
first? Deaths or Rate?
• People will read from left to
right
67
Analyze this Chart
68
Function over Aesthetics
Do not put beauty before functionality
Correct data for X and Y axis
Conclusions Do not overuse Colors
Do not lead audience for misperception
Reference
• How to visualize data like a pro by 24Slides.
• Data Analytics for Upskilling by AIT
