195
How Is Data Displayed on a GIS Map?
sizes of each range are based on the total span of values to be mapped. For
instance, in the seasonal home maps, the data is divided into four classes,
and the range of values goes from the state with the lowest seasonal home
percentage (0.6% of the total housing stock in Illinois) to the state with the
highest seasonal home percentage (15.6% in Maine). Equal Interval takes
the complete span of data values (there is 15% separating the lowest and
highest values) and divides it by the number of classes (in this case, four),
and that value (in this case, 3.75%) is used to compute the breaking point
between classes. So the first class represents states that have a seasonal
home value of 3.75% more than the class’ lowest end (for instance, the first
class would have values between 0.6% and 4.35%). Note that this method
simply classifies data based on the range of all values (including the highest
and lowest) but does not take into account clusters of data or how the data
is distributed. As such, only a few states end up in the upper class because
their percentages of seasonal homes were greater than three-fourths of the
total span of values.
The final method is presented in the map in Figure 7.8d, the Standard
Deviation method. A standard deviation is the average distance that a single
data value is away from the mean (the average) of all data values. The breakpoints for each range are based on these statistical values. For instance, the
GIS would calculate the average of all United States seasonal home values
(3.9%) and the standard deviation for them (3.1%). So when it comes to the
percentage of the total housing stock that is seasonal, each state’s percentage
is an average of 3.1% away from the average state’s percentage. These values
for the mean and standard deviation of the values are used to set up the breakpoints. For instance, the breakpoint of the first range is of all states whose seasonal home values are less than half a standard deviation value lower than the
mean—those states with a seasonal home percentage of less than the mean
minus 0.5 times the standard deviation (1.55%), or 2.34%. The fourth range
consists of those counties with a value greater than 1.5 times the standard
deviation away from the mean. The other ranges are similarly defined by the
mean and standard deviation values of the housing data.
Like Figure 7.8 shows, the same data can produce some very different
looking choropleth maps depending on which method is used to classify
the data, with differing messages from the maps. For instance, the map
in Figure 7.8d (Standard Deviation) shows that most states have roughly
an average (or below average) percentages of seasonal homes, while the
other maps show various distinctions between which states are classified
as a higher or lower percentage of homes that are seasonal. Thus, the same
data can result in different maps, depending on the classification method
chosen. When selecting a method, having information about the nature of
the data itself (that is, if it is evenly distributed, skewed toward low or high
numbers, or all very similar values) will aid in ending up with the best kind
of mapping.
Also keep in mind, while different data classification methods can affect
the outcome of the map, the type of data values being mapped can greatly
Standard Deviation
a data classification
method that computes
class break values by
using the mean of the
data values and the
average distance a
value is away from the
mean.
Equal Interval a
data classification
method that selects
class break levels by
taking the total span of
values (from highest to
lowest) and dividing by
the number of desired
classes.
How Is Data Displayed on a GIS Map?
sizes of each range are based on the total span of values to be mapped. For
instance, in the seasonal home maps, the data is divided into four classes,
and the range of values goes from the state with the lowest seasonal home
percentage (0.6% of the total housing stock in Illinois) to the state with the
highest seasonal home percentage (15.6% in Maine). Equal Interval takes
the complete span of data values (there is 15% separating the lowest and
highest values) and divides it by the number of classes (in this case, four),
and that value (in this case, 3.75%) is used to compute the breaking point
between classes. So the first class represents states that have a seasonal
home value of 3.75% more than the class’ lowest end (for instance, the first
class would have values between 0.6% and 4.35%). Note that this method
simply classifies data based on the range of all values (including the highest
and lowest) but does not take into account clusters of data or how the data
is distributed. As such, only a few states end up in the upper class because
their percentages of seasonal homes were greater than three-fourths of the
total span of values.
The final method is presented in the map in Figure 7.8d, the Standard
Deviation method. A standard deviation is the average distance that a single
data value is away from the mean (the average) of all data values. The breakpoints for each range are based on these statistical values. For instance, the
GIS would calculate the average of all United States seasonal home values
(3.9%) and the standard deviation for them (3.1%). So when it comes to the
percentage of the total housing stock that is seasonal, each state’s percentage
is an average of 3.1% away from the average state’s percentage. These values
for the mean and standard deviation of the values are used to set up the breakpoints. For instance, the breakpoint of the first range is of all states whose seasonal home values are less than half a standard deviation value lower than the
mean—those states with a seasonal home percentage of less than the mean
minus 0.5 times the standard deviation (1.55%), or 2.34%. The fourth range
consists of those counties with a value greater than 1.5 times the standard
deviation away from the mean. The other ranges are similarly defined by the
mean and standard deviation values of the housing data.
Like Figure 7.8 shows, the same data can produce some very different
looking choropleth maps depending on which method is used to classify
the data, with differing messages from the maps. For instance, the map
in Figure 7.8d (Standard Deviation) shows that most states have roughly
an average (or below average) percentages of seasonal homes, while the
other maps show various distinctions between which states are classified
as a higher or lower percentage of homes that are seasonal. Thus, the same
data can result in different maps, depending on the classification method
chosen. When selecting a method, having information about the nature of
the data itself (that is, if it is evenly distributed, skewed toward low or high
numbers, or all very similar values) will aid in ending up with the best kind
of mapping.
Also keep in mind, while different data classification methods can affect
the outcome of the map, the type of data values being mapped can greatly
Standard Deviation
a data classification
method that computes
class break values by
using the mean of the
data values and the
average distance a
value is away from the
mean.
Equal Interval a
data classification
method that selects
class break levels by
taking the total span of
values (from highest to
lowest) and dividing by
the number of desired
classes.