Visualization: The Medium - Data Points: Visualization That Means Something - page 386

Visualization has been around for centuries, but it is relatively new as a field
of study, and experts in the area haven't even decided what exactly visual-
ization is yet. Should it be used for only analysis? Is visualization specifically
for quantitative insights, or can you use it to evoke emotions? At what point
does visualization—a field deeply rooted in, well, visual things—become art?
The answers to these questions vary between who you ask. The questions
have created heated debates within and in between subject areas, and this
is just amongst the academics and practitioners.
I was on a consulting gig at a large, data-centric organization because it wanted
to inject more visualization into its work. It wanted the public to know what
it was doing and wanted to improve its existing work in reports and data
summaries, along with tools within the organization.
So I was in a meeting with about 40 people, which was a diverse group of
marketers, developers, and statisticians. The group worked on a variety of
projects, from quick, made-for-blog graphics to interactive data exploration
tools. We were discussing an online application, and part of the group felt
there should be more editorial content on what the data was about, whereas
another part insisted that any interpretation should be left to the users. A few
others leaned toward graphics that looked like abstract paintings. The ideas
for visualization were all over the place, and a long argument ensued.
They were all right. Everyone argued in support of visualization for a specific
purpose and insisted that others' visualization had to fit the same criteria, even
if the others designed applications for different reasons and with a different
audience in mind. They approached visualization as if it were a monolithic
thing that had a defined set of rules. This might have been true a century ago
(or not), but visualization has grown into more than just a tool. Visualization is
a medium: a way to explore, present, and express meaning in data.
Rather than disjoint categories that work independently from others, you can
think of visualization as a continuous spectrum that stretches from statistical
graphics to data art. There is visualization that is clearly one or the other,
but there are many works that are a blend of both and can't be put in a bin.
Where statistics, design, and aesthetics find a balance is where a lot of the
best work comes from.
This is not to say that the blend is always best, nor are statistical graphics better
than data art, or vice versa. They all serve their own purposes and should be
judged by how well they achieve their goals. You don't critique a documentary