Designing for an Audience - Data Points: Visualization That Means Something - page 597

works and doesn't. It's always a little different for each dataset, so you learn
something new every time.

You might tend to create charts in a rapid fire fashion during the exploration
phase. You want to know what the data is about and if there are any angles
you should explore in depth. Layout and aesthetics is of little concern at this
point, whereas efficiency and speed is what you need to look for. If there's
a lot of data or the dataset is complex, you shouldn't be surprised to spend
most of your time in this phase.

Note: More formal analyses happen during
this time, too. It's an iterative process in which
visualization informs statistical methods and
vice versa. The more you learn about your data
during this phase, the more you have to say to
an audience.

When you know what you want to show, you can figure out
how you want to show it. You can start with paper before
you mess around on the computer, so you need to have a
notebook within arm's reach. Sketch, scribble, and jot down
notes for what you think might be useful, and then try to
translate that through the computer.

Sketching on paper can provide a separation between what you want to do
and the limitations or technical challenges imposed by a computer. It's good
to have limitations in mind, but it's better (and easier) to have a lot of ideas
and then dial back to it within digital limits and time constraints than it is to
restrict work to what you know how to do on a computer.

Note: You tell the computer what to do—not
the other way around.

Use the tools that work best for you. Then learn all you can
about your data through exploration and analysis. Use your
findings to guide design.

A SPECIFIC AUDIENCE

The main challenge of visualization for an audience is that you must make sure
readers can make the leaps from encoding to decoding to understanding the
data. If your audience is already familiar with the background behind your data
or has perhaps even worked with it, the barriers to get to each step are lower.

Note: You have some leeway if your audience
is familiar with your data and analyses, but it's
better to explain too much than too little.

You can take readers through your analysis or the graphics
you created during the exploration phase, and it probably
won't be a stretch that most can follow your logic.

However, because you are presenting data to others, con-
sider how they will examine your work. When you're the only audience, you
design for one person, one distance, and one computer screen or piece of
paper. When there are others in the mix, there is variability. People have