Database Reference
In-Depth Information
However, as with many of the chart types developed prior to this, there are
failings. Although the pattern of growth and decline is easy to see, the abso-
lute values of the deaths are not easily discernible. Being able to quickly get
to an absolute value that can be compared to other charts and data sources
is a key goal of data visualization.
Visualizations such
as these use color
as a fundamental: it
must be noted that
this will not work
as effectively for
color-blind people,
or when figures are
printed in gray scale.
In addition, a clearer
indication of what
each color indicates
would be useful—the
descriptions are
embedded in the
paragraph without a
color key.
To summarize, the possible goals of visualization are
uu
To present more data than otherwise possible
uu
To illustrate patterns that are not immediately apparent
uu
To answer questions posed by a viewer
uu
To compare values
uu
To show changes over time
uu
To easily extract the underlying data points used
uu
To draw a viewer toward a visualization
uu
To create a quick mechanism to view a value
HuMAn PeRCePTuAl ABIlITIeS
Human visual perception is not as clear-cut as you might think. The percep-
tual difference in the size of a full moon just above the horizon versus the full
moon directly overhead is the most commonly known instance of how an
optical illusion can trick you. The moon is in fact the exact same size in terms
of angular diameter, or what fraction of your visual field it takes up. Knowing
and taking into account perceptual differences are key to creating visualiza-
tions that communicate the intended message.
You have already discovered how the use of 3D imagery in visualizations can
be misleading—partly due to the technologies used to represent it and partly
due to our perceptual abilities—but there are additional pitfalls you need to
be aware of.
The most important pitfall is context—the shapes around a visualization may
distort the message of the visualization. An extreme example is shown in
Figure 2-2.
