Data Visualization and

Dashboard Generation
Data is everywhere
• Many varied sources of data

• Data and information overload

• How to effectively use this data to make decisions?

Data Visualization
• Graphical representation of data and information using charts,
graphs, and maps

Examples
• Weather map
• Seating chart of an airplane
• Car dashboard
• Fitness tracker
• Chart displaying performance of stock market index
Why visualize data?
Why visualize data?
• Exploring data [Exploration]

• Communicating / explaining a message or an insight [Explanation]

DV for exploration
• Visual data exploration is a critical aspect of descriptive analytics

• By visualizing data, we can

• Identify patterns easily
• Recognize anomalies / irregularities in data
• Better understand relationships between variables

• DV can help monitor KPIs (key performance indicators) – measure

how well an organization is doing relative to its goals → Data
Dashboard
Identifying patterns

25,000
Column Chart of Zoo Attendance by Month

20,000

15,000

10,000

5,000

0
Jan Feb Mar Apr May Jun July Aug Sept Oct Nov Dec
Month

Camm, Cochran, Fry & Ohlmann, Data Visualization - Exploring and Explaining with Data, 1st Edition. © 2021
Cengage. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible
website, in whole or in part.
Anscombe’s quartet
Learnings from Anscombe’s quartet
• Plotting data before analyzing it can be a very valuable activity

• Outliers can cause problems when included in summary statistics

DV for Explanation
• Explaining relationships found in the data
• Explaining the results of predictive and prescriptive models

• Communicating effectively with the audience – ensure that your

intended audience understands your intended message
Data Visualization for Explanation
Data visualization is also important for explaining relationships found in
data and for explaining the results of predictive and prescriptive models.

Camm, Cochran, Fry & Ohlmann, Data Visualization - Exploring and Explaining with Data, 1st Edition. © 2021
Cengage. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible
website, in whole or in part.
Data Visualization
• Part art and part science

• Two key concerns

• Accurately conveying the data → do not mislead or distort
• Be aesthetically pleasing → to enhance and not distract from the message
being conveyed
Ugly, Bad, and Wrong Visualizations
Types of data
Overview
• Quantitative and Categorical data

• Cross-sectional and Time-series data

• Discrete and Continuous data

• Why is the type of data important for visualization?

• It influences the type of graph / chart we should use to convey the message
Quantitative and Categorical data
• Quantitative data – numerical values denote magnitude
• Arithmetic operations can be performed on these data
• Ex – volume, length, share price

• Categorical data – labels or names identify categories of like / similar

items
• Arithmetic operations cannot be performed
• Frequency of occurrence can be calculated
• Ex – company, symbol, industry
Cross-Sectional and Time-Series data
• Cross-sectional data – collected from several entities over the same
time frame

• Time series data – collected over several time periods

• Frequently found in business and economic publications
• Can help understand what happened in the past, identify trends over time,
and project future levels for the time series
Discrete and Continuous data
• Continuous data – values for which arbitrarily fine intermediaries
exist (Ex: time duration)

• Discrete data – values for which arbitrarily fine intermediaries do not

exist (Ex: number of people in a room)
Typical variables in data visualization
Type of variable Examples Appropriate scale Description
Quant / numerical 1.3, 5.7, 83 Continuous Arbitrary numerical values – these
continuous can be integers, rational numbers, or
real numbers
Quant / numerical 1, 2, 3, 4 Discrete Numbers in discrete units – most
discrete commonly but not necessarily
integers
Qual / categorical Dog, cat, fish Discrete Categories without order – these are
unordered discrete and unique categories that
have no inherent order
Qual / categorical Good, fair, poor Discrete Categories with order
ordered
Date / time Jan 5 2018, 8:03 AM Continuous or discrete Specific days and/or times
Text My name is Tom None, or discrete Free from text – can be treated as
categorical
Aesthetics / Visual Elements
Overview
• Describe every aspect of a
given graphical element

• Some represent both

continuous and discrete data
→ position, size, line width,
color

• Some only represent discrete

data → shape, line type
Advantages and
disadvantages
Advantages of data visualization
• Helps an audience quickly see trends and outliers
• Easily share information
• Interactively explore opportunities
• Visualize patterns and relationships
Disadvantages of data visualization
• Making an inaccurate assumption while viewing a visual with many
data points
• Wrongly designed visual → biased or confusing
• Biased or inaccurate information
• Correlation does not always mean causation
• Core messages can get lost in translation
Data Visualization is a
double-edged sword
Ineffective and misleading visuals
• Visualization is just a tool that needs to be used effectively

• Relatively easy to create misleading visuals, meant to massage the

data and convey incorrect insights

• Not all graphs are intentionally misleading

• Further, some are just ineffective in conveying the intended message

Types of Charts
Introduction
• Many different types of charts are available

• Each chart is designed for a specific purpose and / or context

• Hence, we need to understand

• What are the different types of charts available?
• Why are some charts more appropriate for a certain purpose?
Same data – people <= min wage per state
Type of data influences chart selection
• Summarizing data about categories – bar / column chart (ex –
students’ grades in a course)

• Relationship between two quantitative variables – scatter chart

• Time series data – bar charts, scatter charts, line charts with
horizontal axis

• Data with spatial component – geographic map

Six common chart categories
• Comparison / Ranking • Distribution

• Trends • Hierarchy

• Correlation • Composition
Bar Chart
• Most used
• One number and one category

• Horizontal bars v/s Vertical bars?

Pie chart
• Very commonly used
• Used to represent parts of
whole

• Recommended for few (4-5)

categories
• Otherwise difficult to interpret
Scatter plot
• Helps show relationship between
two numbers

• Correlation → does change in

one lead to change in another?

• Is sales correlated with profit?

Scatter plot
and correlation
Histogram
• Frequency distribution

• Useful to show data

centrality and dispersion

How is it different from a bar

chart?
Box-and-Whisker Plot
• Another way to show
centrality and dispersion
• Here we are concerned with
percentiles
• Helps to easily visualize and
identify outliers