Data Visualization Exam

Student Name_______________________________________________

(4 points for each answer unless otherwise indicated)

1. The understanding and interpretation of data is an activity of __________________.

a. Software processing
b. Human cognition
c. Statistical analysis
d. Querying from a database using an optimized method

2. Around 150 C.E., Claudius Ptolemy published the Almagest, which was a thorough treatise on

astronomy, solar, lunar, and planetary theory. It contains the earliest preserved example of data

organized into a ________________________.

3. The Scottish engineer ____________________________ was the first to create the time-series

line graph, bar chart, and pie chart. One of his works shows the balance of trade between

England and Denmark & Norway over time.

a. William Playfair
b. Charles Minard
c. Napoleon Bonaparte
d. Edward Tufte

4. The following individuals are well-known contemporary advocates for the practices of effective
data visualization: (Pick two)

a. John Tukey
b. Edward Tufte
c. Stephen Few
d. Charles Minard

5. In 1967, Jacques Bertin published Sémiologie Graphique. In this book, he identified eight pre-
attentive attributes (two of which are related to the planar dimensions of x and y) that our brains

Page 1 of 20
Data Visualization Research Questions
use to perceive data on a typical printed page without conscious effort. List four of these eight
visual variables: (1 point each)

6. In addition to Bertin’s pre-attentive attributes, three additional criteria were discussed that are
impactful based on research in psychology and advances in display technology. Name one of
these: (2 points)

7. Psychologists generally recognize________________ thinking systems that we use to make

sense of the world:

a. Two
b. Three
c. Four

8. When we perceive our world, our System 1 (also known as Bottom Up) process generates
____________________ for System 2.

a. Impressions, intuitions, intentions, and feelings

b. Knowledge, understanding, and depth of perception
c. Quantifiable measurements
d. Complex mathematical solutions

9. In what kinds of situations is our System 2 (Top Down) thinking process engaged?

a. When we need to generate a fight-or-flight response to the logic of System 1

b. When there are complex problems or where System 1 runs into difficulty
c. When we make decisions based on "gut feel"

Page 2 of 20
Data Visualization Research Questions
10. In Exploratory Data Analysis (EDA), John Tukey advocated a concept that has become a
foundational principle in our approach to effective data visualization today. For what did he
advocate?

a. The user should not need instructions in order to operate software applications
b. The main characteristics of a data set should be presented in a visual, easy to
understand form, without requiring the use or understanding of statistical models
c. We should use only use data when it can be completely verified from at least two
sources
d. The use of pre-attentive attributes should be avoided because they require us to “think”
about the data in a multi-dimensional way

11. The acronym “KPI” stands for _______________________________________.

12. True or False: A KPI is a type of performance measurement used to evaluate achievement against
a specific goal.

a. True
b. False

13. In order to properly define a KPI, you must first:

a. Install a compatible database

b. Define your target measure and the ranges of performance that are considered
acceptable and not acceptable
c. Know the difference between causation and correlation
d. Identify the tool you will use to visualize the result

14. Dashboards provide a view of the most important information needed to achieve one or more
objectives. According to Stephen Few’s definition of a dashboard, they should also (pick the
single most important):

a. Be consolidated on a single computer screen so that they can be monitored at a glance

b. Be interactive and “clickable”
c. Have as much information as possible on multiple pages
d. Contain complex diagnostics that the user can use to drill-down into other reports

Page 3 of 20
Data Visualization Research Questions
15. In the Duell Rules for Actionable Visualizations, we discussed that the visualization should be
tailored to the person who will use the information. How does this impact the way we may
design our dashboards and visual analyses?

a. It has little or no impact, because for consistency, everyone should see the exact same
view of the data.
b. It doesn't impact our visualization, but will impact the data underneath.
c. We should have different versions of all visualizations that can be clearly read by users
with each kind of color blindness.
d. We need to consider what kind of information each group of users will need, and then
customize the information, layout, and design to present the data that is most useful.

16. When it comes to exploring data visually, there are two general modes of discovery that directly
relate to the needs of the user. Pick the two modes and their primary application (Pick two):

a. Exploratory discovery, which allows the user to find new patterns and relationships
b. Requirements discovery, which allows the user to isolate needs for a project
c. Synthetic discovery, which combines multiple data models to arrive at a single view
d. Directed discovery, which allows the user to find answers to pre-known questions

17. What is one data modeling approach that supports faster reporting and analysis, and is good for
“data out?”

a. Online transaction processing model

b. Relational model
c. Analytical models used in OLAP systems
d. OLTP normalized data
e. Replicated model

18. What are the characteristics of this approach that make it desirable for data analysis?

a. Data redundancy makes it possible to conduct a rich, highly attributed analysis

b. Reduction of redundant data makes complex queries operate more efficiently
c. Normalized modeling helps the database operate within “normal” parameters
d. None of the above; it depends on the amount of memory allocated to the server

Page 4 of 20
Data Visualization Research Questions
19. In an analytical data model, attributes or dimensions are _______________________ and facts
or measures are ____________________.

a. Not countable, countable

b. Countable, not countable
c. Both are not countable
d. Text, units of measure

20. Based on the data set below, identify which columns contain measures and which contain
dimensions: (6 points)

List the Dimensions: List the Measures:

21. In the chart above, there are three measurement components. Fill in the blanks in each
question below using the text target value, actual value or performance levels. (2 points each)

The long black horizontal line shows the:

The short black vertical bar shows the:

The shaded areas show the:

Page 5 of 20
Data Visualization Research Questions
22. The shimmering or jittering caused by the overlapping lines is an example of the
________________________ effect.

23. The unnatural colors below would be used to ______________ the reader.

24. Most fonts can be grouped into one of two categories, based on whether they do or do not have
“the little decorative feet” shown on the letter H below at left. Please name these two
categories: (2 points)

h __________________ h _________________

Page 6 of 20
Data Visualization Research Questions
25. Who created the chart shown above? _____________________________ (2 points)

26. Even without reading the detailed text, we can intuitively understand some aspects of this
conflict. Which dimension of Napoleon’s march on Russia does this visualization help us to
understand? (2 points)

a. The mix of weaponry used in the assault

b. The number of Napoleon’s troops that started the assault vs. the number that returned
c. The amount of time involved in making the march and the retreat
d. The defense strategy of the Russian troops

27. Edward Tufte described the chart above as the “best statistical graphic ever drawn.” Others
disagree. Based on what you have learned in class, which argument can reasonably be made
against this approach to data visualization for making daily business decisions? (2 points)

a. Busy decision makers may not want to study this visualization to understand its message
– they will want the meaning and message to be clear
b. There are not many colors – it needs to be more “exciting”
c. No one knows how to read maps anymore, so they would be confused
d. There is no argument – this is a perfectly reasonable approach

Page 7 of 20
Data Visualization Research Questions
28. Which chart type would be the best choice to use as a redesign of this pie chart?

a. A bullet graph
b. A bar chart and a stacked bar
c. A box plot showing the distribution by company
d. A line chart showing Company B has the highest percent

29. What is this chart type and how would you describe it?

a. A line chart that is poorly designed because it does not have any axis labels or data
labels to guide the reader.
b. A candlestick chart used to describe price movements of a security, derivative, or
currency.
c. A bullet graph showing the actual sales versus the target sales
d. A sparkline with a performance band that is data-intense, design-simple and a word-
size graphic.

34. What is the name of this chart type and who invented it?

Page 8 of 20
Data Visualization Research Questions
a. A sparkbar invented by Edward Tufte
b. A sparkbar invented by Stephen Few
c. A bullet graph invented by Stephen Few
d. A thermometer chart invented by Galileo Galilei

35. Breaking the axis on a chart can be problematic, but it can also be useful in some cases. Which
chart type should never have a broken axis?

a. Bar Chart
b. Line Chart
c. Dot Plot
d. Box Plot

36. Humans are very good at quickly and accurately estimating:

a. Size of Circles
b. Angles of a pie chart
c. Arcs of a donut chart
d. Length and height of bars on a bar chart

37. Small multiples could be a useful tool as all of the following except for:

a. a series of maps showing data changing over time

b. a trellis chart showing comparisons across many categories
c. a series of pie charts showing multiple comparisons
d. a series of line charts showing rate over time by category

38. Which of the following is a good design practice for placing charts on a dashboard in a quadrant
according to emphasis of importance?

a.

Most Important Least

Important
Neutral Neutral

Page 9 of 20
Data Visualization Research Questions
b.

Most Important Least

Important
Neutral Neutral

c.

Most Important Least

Important
Neutral Neutral

d)

Most Important Least

Important
Neutral Neutral

39. Which color palette would be a better choice for this data visualization and why?

A. B.

a. The Red-Green because red is universally understood as bad and green as good.
b. The Red-Green because it’s a color-blind friendly palette.
c. The Blue-Orange because it’s a color-blind friendly palette.
d. Neither color palette is good because we should use a categorical color scheme for this
purpose.

Page 10 of 20
Data Visualization Research Questions
40. Name the following color scheme used in the legends:

a. Sequential
b. Categorical
c. Diverging
d. Exponential

41. Name the color scheme used in the following visualization:

a) Sequential
b) Categorical

Page 11 of 20
Data Visualization Research Questions
c) Diverging
d) Exponential

42. Name the color scheme used in the following visualization:

a. Sequential
b. Categorical
c. Diverging
d. Exponential

43. The best way to show a comparison of crime by state on a U.S. map would be:

a. Plot the total number of crimes in each state.

b. Drop California, Florida, New York and Texas because they will be outliers.
c. Plot a crime rate by using total number of crimes and state population.
d. Plot the total number of crimes in each state divided by the entire population of the
United States.

44. Which of the following chart would be best choice for showing the most precise quantitative
comparison?

Page 12 of 20
Data Visualization Research Questions
a. Bubble Chart
b. Bar Graph
c. Choropleth Map (shaded Map)
d. Treemap

45. Which of the charts below is the best way to show profit over time?

Page 13 of 20
Data Visualization Research Questions
a. A line chart going from left to right, with the most recent date on the right.
b. A line chart going from right to left, with the most recent date on the left.
c. A bar chart with the most recent date at the bottom.
d. A bar chart with the most recent date at the top.

46. Which of the following bar charts has the best design?

Page 14 of 20
Data Visualization Research Questions
a. A bar chart in alphabetical order using x-axis labels and chart borders.
b. A bar chart in ranked order using both data labels and x-axis labels and chart borders.
c. A bar chart in ranked order using data labels and no chart borders.
d. A bar chart in alphabetical order using no labels and no chart borders.

47. Which dashboard has the best design considering data visualization best practice?

Page 15 of 20
Data Visualization Research Questions
a. Human Capital Management Dashboard
b. Company Income Dashboard
c. Information Technology Dashboard
d. Vendor Dashboard

Page 16 of 20
Data Visualization Research Questions
48. Please explain what is going on in the visual illusions below. Please indicate specifically what
illusion is occurring (what you’re seeing that is not actually there or true), and what causes it. (16
points – 4 points each)

Page 17 of 20
Data Visualization Research Questions
OPTIONAL
These questions can be used as replacements or for alternate versions of the test (ex. make-up tests).

Data Presentation (6 points each)

Create a graph that represents the following statements (please include axis labels).

a) Using a histogram - The largest number of b) In September, the turnover rates for the
employees earn between $30,000 and six divisions were about the same.
$35,000.

c) There is a strong correlation between d) Using a scatter plot - The size of company
profitability and compensation. pay increases is not related to tenure.

e) The sales manager spends only 15% of his f) The typical salary range for Programmer
time in the field (don’t use a pie chart). II is approx. 10% higher than Programmer
I, but the top and bottom range limits are
about the same.

Page 18 of 20
Data Visualization Research Questions
30. On the following page (landscape if you choose), design a dashboard with the following 4
components. The order of the components indicates their importance, which should be
reflected in the emphasis they are given on the page: (30 points)

1.) Actual vs. Projected revenue for 4 regions of the country

a. North – actual revenue is 105% of projected revenue
b. South – actual revenue is 95% of projected revenue
c. East – actual revenue is 135% of projected revenue
d. West – actual revenue is 65% of projected revenue

2.) The company’s earnings per share has been declining, but Sales have been increasing and
are forecasted to increase over the next 3 years.

3.) In August, two plants outproduced the other six by a wide margin.

4.) Of the top 10 states ranked by productivity, Ohio ranks last when compared to other states

Page 19 of 20
Data Visualization Research Questions
Page 20 of 20
Data Visualization Research Questions