Visualizing Spatial Relationships - Data Points: Visualization That Means Something - page 272

What to Look For
You read maps much the same way that you read statistical graphics.
When you look at specific locations on a map, you still look for cluster-
ing in specific regions or for example, compare one region to the rest of
a country. The difference is that instead of x- and y-coordinates, you deal
with latitude and longitude. The coordinates on a map actually relate to
each other in the same way that one city relates to another. Point A and
Point B are a specific number of miles away, and it takes an estimated
time to get there. In contrast, the distance on a dot plot is abstract and
(usually) has no units.
This difference brings with it a lot of subtleties to maps and cartography.
There's a reason The New York Times has a group of people in its graphics
department who exclusively design maps. You need to make sure all your
locations are placed correctly, colors make sense, labels don't obscure
locations, and that the right projection is used.
This chapter covers only a handful of the basics. These can actually take
you pretty far in terms of finding stories in your data, but keep in mind
there's a whole other level of awesome that you can strive for.
Things can get especially interesting when you introduce time. One map
represents a slice in time, but you can represent multiple slices in time
with several maps. You can also animate changes to, say, watch growth (or
decline) of a business across a geographic region. Bursts in specific areas
become obvious, and if the map is interactive, readers can easily focus
in on their area to see how things have changed. You don't get the same
effect with bar graphs or dot plots, but with maps, the data can become
instantly personal.
Specific Locations
A list of locations is the easiest type of spatial data you'll come across. You
have the latitude and longitude for a bunch places, and you want to map
them. Maybe you want to show where events, such as crime, occurred, or
you want to find areas where points are clustered. This is straightforward
to do, and there are a lot of ways to do it.