Quantifying Urban Diversity: Multiple Spatial Measures of Physical, Social, and Economic Characteristics - Computational Approaches for Urban Environments - page 172

Tabl e 7. 5 Correlation coefficients between all measures

Dwelling
density

Mean street
seg length

Building year
difference Mixed use UL index

Dwelling density - 0.320 0.114 0.388 0.568

Mean street seg
length

0.320 - 0.139 0.285 0.627

Building year diff 0.114 0.139 - 0.009 0.223

Mixed use 0.388 0.285 0.009 - 0.863

UL index 0.568 0.627 0.223 0.863 -

a result that is close to what the casual observer who is familiar with the city
might expect. Future research should explore ways to build on this methodology
by effectively incorporating a richer mix of uses.
In addition to the correlations presented in Table 7.4 , it is important to examine
the correlations between each of individual parameters. Table 7.5 shows the
correlation coefficients between each of the individual parameters, as well as the
UL index.
Looking at each of the coefficients, it is clear that some parameters exhibit
stronger correlations than others. The mixed-use parameter and dwelling density
parameters appear to exhibit the strongest correlations between each of the other
parameters, while the building year difference parameter displays the weakest
correlations. Although this analysis is not sufficient to draw strong conclusions,
when paired with the strong correlations to the UL index for these two parameters,
it suggests that mixed-use and dwelling density are perhaps the most important of
the four individual parameters in terms of their contribution to the overall livability
of the built environment. Jacobs, as noted earlier, would likely dispute the finding
that any of the four parameters is more influential than the others.
In order to explore this further, some initial within-city scenario testing was
conducted. From the perspective of a city official, it is important to know that
if a targeted investment can be made in only one parameter, which parameter
would provide the greatest improvement for that investment. The scenario tested
is one where the 43 (10 % of the total) bottom performing block groups in a
particular parameter are targeted such that their raw parameter input is increased
(or decreased) to match the raw value of the best performing block group for that
parameter. The subindex is then scaled with these new values and the change in
the raw UL is observed. This was repeated independently for each parameter. For
all block groups outside of the bottom 43 in each parameter, a reduction in the
UL value is observed. This is due to the fact that this test is in effect shifting the
distribution such that the bottom 10 % becomes top performers, making all other
values (aside from the one previous top performing block group) lower, because the
values are scaled relative to one another. Thus, only the effect of this change on the
43 block groups tested is noted below. This testing does provide some challenges
for the mixed-use parameter. This is due to the 94 block groups that contained either
no residential or no commercial. Originally, these block groups were automatically