Designing and Choosing symbols 165
Another consideration in choosing colors is the nature of the data. As we have
seen in Chapter 4, a variety of color schemes are possible
and one should match
sequential data with sequential schemes and double-ended (diverging)
data with
diverging schemes.
Legends for choropleth maps normally show the various categories and the val-
ues associated. Since the lines have no value, it is best to put the category values next
to the pattern or shade representing the value, as in Figure 8.25. There is some dis-
agreement over whether the highest values should be placed at the top or the bottom
of the legend, but as long as the legend is clear, high to low or low to high is a matter
of preference.
Data analysis
for Choropleth Maps
A variety of techniques both mathematical and graphical can be used to analyze the
data and aid in choosing an appropriate series of categories. Although category types
can be selected from the software, the methods of selection are described here to aid
in understanding the categories and choosing appropriate categories for a map.
exaMining THe DaTa
A first step in the analysis of data is to place them in rank order on a spreadsheet.
That is, the data are sorted from highest to lowest. An inspection of the ordered data
can often give an idea of the nature of the distribution, but creating a histogram or
number line will graphically show its nature. A
histogram (Figure 8.26) is a type of
bar graph for which the area of the bars is proportional to the frequency of the obser-
fIgURe 8.24.
A computer-generated “unclassed choropleth” map. Each state has a different
pattern.
166 sYMBOLizaTiOn
vation. The histogram can be examined for obvious breaks in the distribution that
are used as a basis for the natural breaks method or to determine if the distribution
is normal, equally frequent, or resembles an arithmetic or geometric progression. A
number line is a line on which all places are plotted as points (Figure 8.27). It can
represent
data values,
natural breaks, and divisions between classes.
An unclassed choropleth (Figure 8.24) can be used to see the spatial arrange-
ment of the data and aid in choosing categories; however, most often categories are
41-50
31-40
21-30
11-20
0-10
0
10
20
30
40
50
Choropleth
Isarithm
