COMMENT BOOKS & ARTS
DATA VISUALIZATIO N
Drawing out the meaning
Felice Frankel relishes an inspired handbook on the art and science of picturing data.
1 8 6 | NAT U R E | VO L 4 9 7 | 9 M AY 2 0 1 3
© 2013 Macmillan Publishers Limited. All rights reserved
NATHAN YAU
S
cientists, probably more than most, are
concept or data into a formatted representa­
Data Points:
aware of the ever-increasing presence of
tion — that is, to reduce a visual expression
Visualization That
Means Something
data visualization in newspapers, televi­
to its fundamental information — is one of
NATHAN YAU
sion, online shopping, social media and even
the least discussed but most important ele­
John Wiley & Sons: 2013.
policy-making sales pitches in the US Con­
ments of successful visual representations.
384 pp. $39.99/£26.99
gress. I’ll bet there’s a brand new office build­
I have some quibbles. You might need a
ing in Washington DC devoted to creating
magnifying lens while reading Data Points:
chart-laden poster boards for congressional
a number of the figures need considerable
members. Statistician and visualization
enlargement to make sense on the printed
expert Nathan Yau’s Data Points: Visualispecifically scientific visualizations on offer,
page. And ensure you view online his many
zation That Means Something is a clear and
the creative researcher could find much to
examples of visualizations intended for
passionate exploration of this burgeoning
adapt in these approaches.
online publication, especially those intended
phenomenon.
Yau illustrates and discusses the funda­
for interactivity. I would also have liked to
A detailed handbook, Data Points is espe­
mental components of visualizations and
see a more in-depth discussion on repre­
cially useful for those working on scientific
how small changes can improve readability.
senting uncertainty. We are all aware of the
data visualization, guiding
various levels of impreci­
the reader through fascinat­
sion in our data, and not to
ing examples of data, graph­
communicate that somehow
ics, context, presentation and
in our representations can
analytics. But this is more
be irresponsible — but that
than a mere how-to manual.
exercise probably warrants
Yau reminds us that the real
a separate book. Finally, I
purpose of most visualiza­
found it annoying that many
tion work is to communicate
figures do not appear next to
data to pragmatic ends. He
where they are referenced in
points to the other end of the
the text; I sometimes had to
spectrum: visualizations cre­
turn a page or even two to see
ated by those so seduced by
what Yau was describing.
artful design that meaning
But this remains a master­
becomes inconsequential.
work. I can imagine some
As Yau writes of statistics and
initial eye-rolling in the
design knowledge, “having
visualization community
skills in both provides you
on the first viewing of some
with the luxury — which is
of Yau’s seemingly ‘obvious’
growing into a necessity — to
examples. But there is much
jump back and forth between
to learn from studying what
data exploration and story­
Yau does here. That is, defin­
telling”.
ing and demonstrating good
Yau entices us to look and
visualization with clarity and
think, study, dissect and
precision as “a representation
judge. As a visualizer him­
of data that helps you see
self, he has made intelligent
what you otherwise would
Characteristics abstracted into form: each cell combines two properties of data.
choices of illustrations in
have been blind to if you
this wonderfully varied collection; they are
He teaches us to think graphically. On one
looked only at the naked source … trends,
not there as decoration. Lokesh Dhakar’s
spread (and there are many), he cleverly
patterns, and outliers that tell you about
Coffee Drinks Illustrated (http://go.nature.
uses the ‘visual cues’ described by William
yourself and what surrounds you”.
com/tcxs21), for instance, is accessibly sim­
Cleveland and Robert McGill in their 1985
Thank you, Nathan Yau, for helping us to
ple, yet rich in information. Yau includes
study on graphical perception and meth­
begin. ■
an interactive astronomical visualization
ods (W. S. Cleveland and R. McGill Science
by Santiago Ortiz (http://go.nature.com/
229, 828–833; 1985) in a table that subtly
Felice Frankel is a research scientist in
hwnsdx) as an example of how perspective
encourages the reader to think their way
the Massachusetts Institute of Technology’s
and context are inseparable when zooming
into visual abstraction (section pictured).
Center for Materials Science and Engineering
and rotating the night sky. The visualization
The table shows patterns in data — such as
in Cambridge, Massachusetts, USA. Her
of constellations is both elegant and beau­
an increase or decrease in the population of a
latest book is Visual Strategies: A Practical
tiful, and could become an inspiration for
species — and how these can be represented
Guide to Graphics for Scientists and
scientists creating interactive molecular
graphically through angle, position, area or
Engineers, co-authored with Angela DePace.
configurations. And although there are few
colour saturation. How to visually abstract a
e-mail: [email protected]