Partitioning of Data - Spatial Mathematics: Theory and Practice through Mapping - page 145

Figure 6.18 Histogram, associated with map, made using Esri software.
Expressed in “standard” rather than in “exponential” notation, note the pat-
tern of distribution of the mean and standard deviation of the population of
world countries—Mean: 2.58 E 7 = 25871761.475806452, or about 25.8 million
people; Standard Deviation: 1.11 E 8 = 111792975.28287828, or about 111.7
million people. Most of the countries on this map will fall into the lowest cat-
egory with the exception of China, India, and the USA. Maps such as this are
not very helpful except for examining outliers.
Use “Classify” and change to Quantile, again noting the histogram. Is this quan-
tile map more helpful than the Equal Interval map? Change the Symbols to
Manual. What manual class breaks make the most useful map? How many
classes do you believe are optimal? Again, you are the map producer and
map reader, so you have the power to make the choices. But, most studies
have shown that people can easily process between six and nine classes,
after which it becomes much more difficult to distinguish the classes. Of
Equal Interval, Quantile, or Manual, which classification method was most
useful for this population data? Quantile, as explained in this chapter, is not
as useful to use at this scale and with population data, because most of the
small countries at a global scale will not be visible.
Quantile places the same number of observations into each category (by
“Quantity”). Here, with 248 countries and five classes, about 49 or 50 countries
fall into each category. In the above activity, the Quantile map did a better job of