08 f 06

SLDS Annual Grantee Meeting November 14, 2008
1
Data visualization for enlightening communication.
Stephen Few, Principal, Perceptual Edge
sfew@perceptualedge.com
(510) 558-7400
© 2008 Stephen Few, Perceptual Edge

SLDS Annual Grantee Meeting 11/15/2008
2
The amount of information that is available to businesses has increased dramatically in the
last few years, but the ability to make use of it has increased little, if any.
Our networks are awash in data. A little of it is information. A smidgen of this shows up
as knowledge. Combined with ideas, some of that is actually useful. Mix in experience,
context, compassion, discipline, humor, tolerance, and humility, and perhaps knowledge
becomes wisdom.
(Turning Numbers into Knowledge, Jonathan G. Koomey, 2001, Analytics Press:
Oakland, CA page 5, quoting Clifford Stoll)
Most of us who are responsible for analyzing data have never been trained to do this.
Knowing how to use Excel or some other software that can be used to analyze data is not
the same as knowing how to analyze it.

3
We live in the so-called information age. We have far more data than we’ve
ever had, but our ability to make good use of it hasn’t caught up. Information
is useless until we understand what it means and can clearly communicate
that meaning to those who need it, those whose decisions affect our world.

4
I work with people and organizations of all types. If they’ve requested my
help, they have one thing in common—they have important quantitative
information to understand and present, and realize they could be doing it
much better.

5
In 1786, a roguish Scot – William Playfair – published a small atlas that
introduced or greatly improved most of the quantitative graphs that we use
today. Prior to this, graphs of quantitative data were little known.

6
Today, 220 years later, partly due to the arrival of the PC, graphs are
commonplace, fully integrated into the fabric of modern communication.
Surprisingly, however, Playfair’s innovative efforts – sprung from meager
precedent – are still superior to most of the graphs produced today.

8
You’ve been invited to another of the many meetings that you’re required to
attend. You’re one of many managers in the Information Technology
department. Like most meetings, this one begins with the light of a projector
suddenly illuminating a screen. Bursting with excitement, the speaker
announces that you and everyone else in the room will now receive a daily
report that will inform you how the network is being utilized, and then the graph
on the next slide appears.

9
You stare at this graph very intently, trying your best to keep any hint of
confusion from crossing your face. From your peripheral vision you can see
that the CIO (Chief Information Officer) is smiling broadly and nodding with
obvious understanding. You and everyone else in the room begins to nod
enthusiastically as well. You feel very dumb. What you don’t realize is that you
are not alone.

10
I wrote the book, Show Me the Numbers: Designing Tables and Graphs
to Enlighten in 2004 to help people like you respond to the challenges that
you face every day when presenting quantitative information.

11
Why? Few people are trained.
Why? Few people recognize the need to be trained.
Why? Few examples of good design exist to expose the problem.

12
In 1954, Darrell Huff wrote his best-selling book about how people often
intentionally use graphs to spread misinformation, especially in favor of their
own products or causes. Today, vastly more misinformation is disseminated
unintentionally because people don’t know how to use charts to
communicate what they intend.

13
When the PC was introduced, software soon made the arduous task of table
and graph creation as easy as 1-2-3 (literally ―Lotus 1-2-3‖, the software that
was the first to legitimize the PC as a viable tool for business). Unfortunately,
this improvement in ease and efficiency was not accompanied by instruction
in visual design for communication. People today think that if they know how
to click with the mouse to create a table or graph, they know how to present
data effectively.
―In the two centuries since [the invention of the first graphs], …charts have
become commonplace. With the advent of modern computer tools, creating
graphs from data involves trivial effort. In fact, it has probably become too
easy. Graphs are often produced without thought for their main purpose: to
enlighten and inform the reader.‖ Jonathan G. Koomey, Turning Numbers
into Knowledge, Analytics Press, 2001
I can talk about this all day, but the best way to make my point convincingly
is to show you.

14
The purpose of this graph is to display how Department G is doing regarding
expenses compared to the other departments. Is the message clear?
Often, when someone creates a graph that appears inadequate somehow, they
try to fix it with sizzle, as in the next slide.

15
Does the addition of 3D improve this pie chart? Definitely not. In fact, it actually
makes it harder to read.

16
Though it lacks flash and dazzle, this simple bar graph tells the story
elegantly.

17
I found this table on the Web site for Bill Moyers’ public television show
―Now‖. I felt that it provided important information that deserved a better form
of presentation. In this case the story could be told much better in visual
form.

18
This series of related graphs tells the story in vivid terms and brings facts to
light that might not ever be noticed in the table.

20
Paul Grice was a 20th century philosopher whose work ventured into the
realm of linguistics. He is well known for his conversational maxims, which
attempt the describe the characteristics of polite conversation.
Every one of these maxims of conversation apply equally well to the
communication of quantitative information in the workplace. We’ll strive in
this workshop to translate these maxims into effective and polite
communication via tables and graphs.

25
Often, the simplest form of display is the most powerful.

26
Dressing things up is appropriate for advertising, because the illusion
pleases and sells. When you’re responsible for discovering the truth and
understanding it, makeup only gets in the way.

27
But besides the data, what else is there? According to Edward Tufte, tables
and graphs consist of two types of ink: data ink and non-data ink. He
introduced the concept of the ―data-ink ratio‖ in his 1983 classic The Visual
Display of Quantitative Data. He argued that the ratio of ink used to display
data to the total ink should be high. In other words, ink that is used to display
anything that isn’t data should be reduced to a minimum.

28
―In anything at all, perfection is finally attained not when there is no longer
anything to add, but when there is no longer anything to take away.‖ Antoine
de St. Exupery
John Maeda, in The Laws of Simplicity, offers a maxim about design
simplicity, which I have massaged into the following statement:
Simplicity is about eliminating the obvious (and everything else that
doesn’t support your purpose), and enhancing the meaningful.

© 2008 Stephen Few, Perceptual Edge 29
This is the kind of graph that software products, including Excel, encourage
us to create. They give us an infinite selection of poorly-designed graphs
from which to choose. What we really need, however, is a small selection of
graphs that really work.
Using this graph, try to see the pattern of change across the months in
actual expenses. Try to determine one of the actual values. Try to compare
actual expenses to the budget across time.
Let’s transform this graph into one that communicates.

We have now removed the useless 3-D effects and angle, which makes the
data easier to read.

We have now removed the background fill color.

We have now replaced the silly cones with regular bars.

We have now removed the tick marks, which aren’t necessary. Tick marks
are not needed to separate the months along the X-axis and because
horizontal grid lines are being displayed, there is no need for tick marks on
the Y-axis either.

We have now enlarged the text, making it easier to read.

We have now removed the unnecessary decimal places in the dollar
amounts along the Y-axis.

We have now removed the redundant dollar signs and labeled the unit of
measure (U.S. $) clearly.

We have now reoriented the Y-axis label to the horizontal and placed it
above the axis to make it easier to read.

We have now reoriented and repositioned the legend to make it easier to
associate it with the data bars.

We have now changed the color of the Budget bar to be more visually
pleasing in relation to the blue Actual bars. Changing from the color red also
removed the possibility people interpreting the data as something bad or a
warning, which red is often used to represent.

We have now reduced the visual salience of the Budget values, because
they are less important that the Actual values, and have done so in a way
that reduced clutter.

We have now made it much easier to see the pattern of change through time
by using lines rather than bars to represent the data.

We have now made it much easier to examine the differences between
actual expenses and the budget by spreading them across more space.

We have now labeled the lines directly, removing the need for a legend.

We have now changed the lines to two shades of gray to guaranty that even
if the graph is printed on a black-and-white printer or photocopier, they will
still look distinctly different from one another.

We have now represented the variance of actual expenses from the budget
directly, as a single line.

As our final step, we have expressed variance as a percentage, to provide a
better measure of performance.

Our final solution, which we produced in sixteen steps, could have easily
been our original solution. It usually takes no longer to design effective
graphs than those that communicate poorly, if at all.

48
1. You begin by determining the best medium for your data and the
message you wish to emphasize. Does it require a table or a graph?
Which kind of table or graph?
2. Once you’ve decided, you must then design the individual components of
that display to present the data and your message as clearly and
efficiently as possible.

50
The saying, ―A picture is worth a thousand words,‖ applies quite literally to
quantitative graphs. By displaying quantitative information in visual form, graphs
efficiently reveal information that would otherwise require a thousand words or
more to adequately describe.
Joseph Berkson once stated what happens quite powerfully: When we visualize
the data effectively and suddenly, we experience ―interocular traumatic impact‖—a
conclusion that hits us between the eyes.

51
The stories contained in numbers all involve relationships. Numbers that measure
public health, in fact, involve six fundamental types of relationships.
If you know the relationship you’re trying to communicate graphically and you know
the best way to graphically encode that relationship, you possess a simple
vocabulary for telling quantitative stories.
Allow me to introduce the six relationships that you should get to know.

52
This graph features the relationships between values as they change through time,
which is perhaps the most common quantitative relationship that you need to
communicate.

53
Here’s an attempt to display a time-series relationship regarding actual vs. budgeted
sales, which works, but the trend and patterns could be much more clearly
displayed.

54
Here’s the same exact data presented in two ways: to top graph uses bars and the
bottom graph uses a line. Which displays the shape of change through time more
clearly?
Bars work well for comparing individual values to one another, but lines show the
shape of change through time much more clearly.

55
We’ve been talking about time-series relationships, but this graph features a
different relationship between the values. The values are arranged in order of size,
in this case from big to small. Arranging values sequentially makes them easier to
compare to one another and directly communicates the relationship of rank.

56
In the display of trauma registry injuries by county on the left, notice how difficult it is
to compare the values and to get a sense of rank when they aren’t sequenced by
size.
The same information is displayed on the right, this time with the counties arranged
by the number of injuries. If the purpose of the display is to look up individual values,
which is the only thing that alphabetical order supports, a table would work much
better. The ranking display on the right, however, tells a useful story.

57
This graph features another relationship that is commonly displayed. Notice that if
you add the values they total 100%. This is what I call a part-to-whole relationship,
which shows how the individual values that make up some whole relate to one
another and to the whole.

58
Pie charts use 2-D areas and the angles formed by slices to encode
quantitative values. Unfortunately, our perception of 2-D areas and angles as
measures of quantity is poor.
Since all graphs have one or more axes with scales, there must be one on a
pie chart, but where is it? The circumference of the circle is where its
quantitative scale would appear, but it is rarely shown.
Try using either one of the pie graphs to put the slices in order by size. Can’t
do it, can you? Now see how easy this is to do when the same data is encoded
in a bar graph.
Coda Hale once expressed his opinion of pie charts quite colorfully:
Pie charts are the information visualization equivalent of a roofing
hammer to the frontal lobe…[Piecharts] have no place in the world of
grownups, and occupy the same semiotic space as short pants, a runny
nose, and chocolate smeared on one’s face. They are as professional
as a pair of assless chaps. Anyone who suggests their use should be
instinctively slapped.

59
This graph features the next common relationship between values. This approach
is used when you want to feature how one set of values differs from another
reference set of values. I call this a deviation relationship.

60
When people primarily need to see the differences between things, show them the
difference directly, rather than showing them the two sets of values and forcing them
to construct a new picture in their heads of how they differ.

61
This graph directly expresses how actual sales differed from the budget through the
year.

62
The relationship that this graph features the spread of a set of values from lowest
to highest and the shape of their distribution across that range.

Here’s a graph that attempts to show the distribution of overweight children by grade
separately for boys and girls, but doing it in this way results in clutter that makes the
patterns difficult to segregate and compare.

This pair of histograms—one for boys and one for girls—however, are arranged in a
way that makes the patterns of each easy to see, yet still easy to compare.
Even better, by using lines rather than bars, the separate patterns can be shown in
the same graph in a way that features the shape of the patterns and how they differ.

65
This graph illustrates the last of the six relationships. Graphs such as this feature
co-relations between two paired sets of values, so this relationship is called a
correlation.
Correlations show whether two paired sets of measures, such as these purely
fictional sets of people’s heights and salaries, vary in relation to one another, and if
so, in which direction (positive or negative) and to what degree (strong or weak).

66
This example, based on WHO data, explores the correlation between adult literacy
and fertility rate by country. A correlation clearly exists: higher literacy corresponds
to lower rates of fertility. It is also clear from this display that the highest rates of
fertility all occur in Africa (the blue circles), which the one exception of Yemen (the
one green circle at the high end of fertility).

67
Almost every quantitative story involves one or more of these fundamental
relationships, these comparisons between numbers. The skills required to tell
these stories clearly are easy to learn.

68
Adding a third dimension of depth to the bars on the right without adding a
corresponding third variable is not only meaningless, it makes it more difficult
to decode the data.

69
Can you determine which of the lines in the graph on the right represents the
East region? Are you sure?
A third dimension with a corresponding variable is too hard to read.

This chart of Escher’s changing popularity through time was created by B.
Brucker. I found it at www.GraphJam.com.

71
Lots of bright color is great is you’re a preschooler. For adults, frequent use
of bright colors accosts visual perception. Pastels and earth tones are much
easier to look at. Use bright colors only to make particular data stand out
above the rest.

72
The top graph varies the colors of the bars unnecessarily. We already know
that the individual bars represent different countries. Varying the colors
visually separates the bars by making them look different from one another,
but we want them to look alike to encourage people to compare them and to
see the ranking pattern that they form as a whole.

73
Notice that, despite the softness of the colors in the example of natural
colors, they still do the job of separating the sections of these pies just as
well as the other examples, but do so in a manner that is much more
pleasant to look at.

74
Dashboards are extremely popular today. Everyone assumes that
dashboards are useful. Like all aspects of business intelligence, however,
they can be done well or they can be done poorly.

75
A dashboard is a visual display of the most important
information needed to achieve one or more objectives;
consolidated and arranged on a single screen so the
information can be monitored at a glance.
This definition first appeared in the March 20, 2004 issue of Intelligent
Enterprise magazine in an article written by Stephen Few entitled
―Dashboard Confusion.‖

76
Dashboards like this are typical. You can find them on the websites of most
business intelligence vendors. In their attempt to dazzle us visually, they fail
entirely to present information is a way that can be understood at a glance.

77
It is because dashboards are rarely designed effectively that I wrote the
book Information Dashboard Design. The principles are not difficult to learn,
but they aren’t obvious until someone points them out.

78
The fundamental challenge of dashboard design is to display all the required
information on a single screen:
• clearly and without distraction
• in a manner that can be quickly examined and understood
Think about the cockpit of a commercial jet. Years of effort went into its
design to enable the pilot to see what’s going on at a glance, even though
there is much information to monitor. Every time I board a plane, I’m grateful
that knowledgeable designers worked hard to present this information
effectively. Similar care is needed for the design of our dashboards. This is a
science that few of those responsible for creating dashboards have studied.

79
The process of visual monitoring involves a series of sequential steps that
the dashboard should be designed to support. The user should begin be
getting an overview of what’s going on and quickly identifying what needs
attention. Next, the user should look more closely at each of those areas that
need attention to be able to understand them well enough to determine if
something should be done about them. Lastly, if additional details are
needed to complete the user’s understanding before deciding how to
respond, the dashboard should serve as a seamless launch pad to that
information, and perhaps even provide the means to initiate automated
responses, such as sending emails to those who should take action.

80
This failure is not rooted in technology; it is a failure of communication,
rooted in poor design. The main problem that plagues dashboards today is
that they don’t say enough and what they do say they don’t say very well. In
other words, they are not very informative. When designed properly, a
dashboard provides an overview of what’s going on, clearly and rapidly.

81
Dashboards are not for show. No amount of cuteness and technical wizardry
can substitute for clear communication. My purpose today is to present the
design practices you cannot ignore if you wish to build dashboards that set
the stage for informed response.

82
Measures of what’s going on in the business rarely do well as solo acts; they
need a good supporting cast to get their message across. To state that
quarter-to-date sales total $736,502 without any context means little.
Compared to what? Is this good or bad? How good or bad? Are we on track?
Is this better than before? The right context for the key measures makes the
difference between numbers that just sit there on the screen and those that
enlighten and inspire action.
The gauges above could have easily incorporated useful context, but they
fall short of their potential. The center gauge tells us only that 7,822 units
have sold year to date and that this number is good, indicated by the green
arrow.
Quantitative scales on a graphic, such as those suggested by the tick marks
around these gauges, are meant to help us interpret the measures, but they
can only do so when scales are labeled with numbers, which these gauges
lack. A great deal of the space that is used by these gauges tells us nothing
whatsoever.
(Source of gauge image: Website of Informmersion Incorporated.)

83
Many dashboards—perhaps most—are data poor. They report numbers, but
don’t express them meaningfully enough for people to gain the level of
understanding needed to respond. To be effective, dashboards must be data
rich. This is achieved by comparing measures of what’s going on to related
measures (for example, to a target). When monitoring what’s going on,
people usually need answers to these questions:
• Are we doing well or poorly?
• How well or how poorly?
• What has led to what’s happening today?

84
The visual orientation of dashboards is important due to the speed of
perception that’s usually required to monitor information. The faster you must
assess what’s going on, the more you should rely on graphical means to
display the information.
Text must be read, which involves a relatively slow, serial process.
Certain visual properties, however, can be perceived at a glance, without
conscious thought. With the graphical display on the right, it’s quick and
easy to see which bars exceed target, marked by the short vertical line, and
which fall short.

85
People tend to overuse color on dashboards, assuming that the brighter and
more colorful they are, the better they work, but the opposite is true. People
who understand color and how it can be used to display information use it
sparingly.
Too much color is visually overwhelming; it tires our eyes. Also, if you use
color gratuitously, you undermine its ability to be used to make things stand
out. Notice how the red alerts clearly stand out in the bottom display in
contrast to the neutral grays and blacks that been used elsewhere, rather
than being lost in the meaninglessly colorful display above.

86
To clearly present everything on a single screen, even the slightest lack of
organization will result in a confusing mess. You must condense the
information, you must include only what you absolutely need, and you must
use display media that can be easily read and understood even when they
are small, which is often necessary.
Elegance in communication can be achieved through simplicity of design.
Too often we smear a thick layer of gaudy makeup on top of the data in an
effort to impress, rather than to communicate the truth in the clearest
possible way.
―Simplify, simplify, simplify.‖
Henry David Thoreau

88
Direct dynamic interaction with the visualized data allows you to see something in the data
visualization and interact with it directly to filter out what you don’t need, drill into details,
combine multiple variables for comparison, etc., in a way that promotes a smooth flow
between seeing something, thinking about it, and manipulating it, with no distracting lags in
between.

90
Visual data analysis is a process that consists of many steps and many paths to get us from
where we start, knowing little, to where we need to be, understanding much. Some ways of
navigating your path from step to step are more effective than others.
Data analysis, like experimentation, must be considered as an open-minded, highly
interactive, iterative process, whose actual steps are selected segments of a stubbily
branching, tree-like pattern of possible actions.
(The Collected Works of John W. Tukey, John W. Tukey, Wadsworth, Inc.: Belmont, CA,
1988, pages 5 and 6)

91
When new recruits by intelligence organizations are trained in spy craft, they are taught a
method of observation that begins by getting an overview of the scene around them while
being sensitive to things that appear abnormal, not quite right, which they should then focus
in on for close observation and analysis.
A visual information-seeking mantra for designers: ‘Overview first, zoom and filter, then
details-on-demand.’
(Readings in Information Visualization: Using Vision to Think, Stuart K. Card, Jock D.
Mackinlay, and Ben Shneiderman, Academic Press, San Diego, California, 1999, page
625)
Having an overview is very important. It reduces search, allows the detection of overall
patterns, and aids the user in choosing the next move. A general heuristic of
visualization design, therefore, is to start with an overview. But it is also necessary for
the user to access details rapidly. One solution is overview + detail: to provide multiple
views, an overview for orientation, and a detailed view for further work.
(Ibid., page 285)
Users often try to make a ‘good’ choice by deciding first what they do not want, i.e. they
first try to reduce the data set to a smaller, more manageable size. After some iterations,
it is easier to make the final selection(s) from the reduced data set. This iterative
refinement or progressive querying of data sets is sometimes known as hierarchical
decision-making.
(Ibid., page 295)

92
Shneiderman’s technique begins with an overview of the data – the big picture. Let your
eyes search for particular points of interest in the whole.

93
When you see a particular point of interest, then zoom in on it.

94
Once you’ve zoomed in on it, you can examine it more closely and in greater detail.

95
Often you must remove data that is extraneous to your investigation to better focus on the
relevant data.

96
Filtering out extraneous data removes distractions from the data under investigation.

97
Visual data analysis relies mostly on the shape of the data to provide needed insights, but
there are still times when you need to see the details behind the shape of the data. Having a
means to easily see the details when you need them, without having them in the way when
you don’t works best.

98
Once you have the basics down, you can begin to tell more complex stories using
more advanced techniques, such as this example from GapMinder.org, which uses
an animated display to tell the story of how the correlation between the number of
births per woman and mortality among young children throughout the world has
changed through time.

99
The good news is, although the skills required to present data effectively are
not all intuitive, they are easy to learn. The resources are available, such as
my books Show Me the Numbers and Information Dashboard Design, but it
won’t happen unless you recognize the seriousness of the problem and
commit yourself to solving it. It is up to you.

08 f 06

More Related Content

Similar to 08 f 06 (20)

More from Elsa von Licy (20)

Recently uploaded (20)

08 f 06