Visualizing Hierarchies - Spatial Mathematics: Theory and Practice through Mapping - page 164

how the counties nest within states. What pattern of change do you notice
by county, and what are the range and the mean? The county level pattern
is much more varied in pattern and also in magnitude, as evidenced by the
new range of − 5.5% to 9.0%, with a mean of 0.5%. Zoom in to the census tract
level, choosing an area in the county that you are interested in. You should
see how the census tracts nest within counties. Again, indicate the pattern of
change, the range, and the mean. Your answer will vary depending on the
area that you are studying, but the range should be greater than the range
for the counties. Zoom in and repeat for a larger scale (that of block groups).
You should see how the block groups nest within census tracts. In the map
legend, the census geography you are examining (in this case, block groups)
is listed at the top of the classes. Indicate the pattern, the range, and the
mean. Again, your answer will vary, but the range should be even greater
at the block group level. Why does the range increase as the scale becomes
larger? The range increases because the more detailed the level of geography,
the more variability occurs, due to the factors of age, employment, percep-
tion, and others. Why does the mean change depending on the scale? The
mean changes because the mean reflects the range of its particular hierarchy
of geography. Can you zoom to the block level? The reason you cannot do so
is because most of these Census data items are not disseminated at the block
level because of privacy issues.
Next, choose another variable in the list, that of median age, and compare pat-
terns, ranges, and means as you change the scale. What does it signify to have
a “mean” of the variable “median age?” The mean here is simply the average
of the median ages in the particular level of geography that you are examin-
ing. Next, choose another variable in the list, such as the median home value
or household income. Which variables seem to be positively correlated, and
why? Median home value and household income seem positively correlated,
as those households with more income tend to purchase and live in more
expensive homes. Median age and household income also seem to be posi-
tively correlated. Which variables seem to be negatively correlated? Obviously,
the median age and the population aged younger than 18 are negatively cor-
related, as are the percentage younger than 18 and the percentage over 64.
Other variables, such as the population density and the change in population
may be positively or negatively related to home value and household income,
depending on the area examined. Which variables seem to have a similar spa-
tial pattern? Again, it depends on the region examined, but home value and
household income have similar spatial patterns. Make the population density
layer semi-transparent and examine the topography underneath it. Population
density is strongly influenced by terrain; densely settled areas tend to be flat
or only moderately hilly. San Francisco is a notable exception! Finally, exam-
ine the tapestry theme and do some research on what tapestry segmentation
indicates. Why can you not compare the numbers on the tapestry data in the
same way as those in the other data layers? The tapestry number is a value
indicating the type of lifestyle and consumer spending habits that the typical