Designing for an Audience - Data Points: Visualization That Means Something - page 610

PUTTING IT TOGETHER
Put everything together—from understanding data, to exploration, clarity, and
adapting to an audience—and you get a general process for how to make data
graphics. The amount of time and effort you spend in each phase varies by the
data and what you want to do with it. For example, if there's not much data
to look at, you spend less time exploring, but if you have a lot of data, you
most likely spend more time in exploration and iterate between the phases.
Now try this with an actual dataset. The Centers for Disease Control and
Prevention (CDC) keeps track of obesity rates at the county level in the United
States. You can download data from the site, which includes obesity rates from
2004 through 2009 for each county. There are approximately 3,100 counties.
How have rates changed over the years? You probably know rates have
increased, but by how much and with what variation in between years? Start
with box plots to see the changing distributions, as shown in Figure 6-25.
There's clearly an annual increase (and a good bit of variation).
FIGURE 6-25 Comparing box
plots to see changes over time