The following table (Table 1) shows data used to make a variety of comparisons.
Table 1. Items Categorized by Created Datasets
The first chart (Figure 1) shows
the population density by zone, and the second chart (Figure 2) shows the
distribution of race for those same zones. These charts suggested that
further comparisons might be interesting, particularly within buffer zones
of one mile and one-half mile of the MBTA as well as within the downtown
Figure 1. Population Density for
Each Buffered Zone
Figure 2. Distribution of Race for Each Buffered Zone
Choropleth (thematic or ranged-fill) mapping, using data normalized as indicated in the sequence of maps below, provided visual evidence to create a picture of residential patterns within the subway buffer of the zones identified in the graphs and maps above.
Note on Methodology
In order to determine patterns of
inclusion and exclusion for intersecting polygons that are not nested,
it is important to have a consistent basis for determining inclusion and
exclusion. The next step involved finding differences between three operations
for spatial analyses:
and Intersecting block groups in subway buffer zones. Maps 13 to
15 demonstrate the differences visually.
After creating three different shape
files and datasets, and evaluating (based on field-based knowledge from
having lived in the region) which might be the most accurate study area
for the analyses, it appeared that the shape in Map 14 was the most appropriate.
The region created in Map 13 appeared to have too much spill over outside
the subway zone while that in Map 15 appeared to have insufficient coverage
of the subway zone. Map 16 shows a clearer picture of the overfit
Map 16. Problem in Block Groups
Table 2. Items Categorized by Different Datasets for 1-Mile Buffer Zone
Professor Arlinghaus and Mr. Schlossberg
suggested that there are some different theme editing options, and each
of them makes a different dataset because each uses a different mathematical
algorithm during geoprocessing. Figure 3 shows how to choose "Proportion"
instead of "Copy." This option simply makes a precise, proportional dataset.
Another way involves the use of an ArcView extension called "TwoTheme.avx"
Marc Schlossberg's Web page: TwoTheme.avx "is
an ArcView extension by Kevin O'Malley that allows you to overlay two polygon
themes and extract attribute data proportionally based on area, e.g. overlay
a buffer on census tracts and proportionally extract data underneath buffer."
(See Figure 4.)
Figure 3. Theme Editing Options
Figure 4. The TwoThemes Extension for ArcView
The "Aggregate Data" menu
in the TwoThemes extension enabled the destination shape (1-Mile Subway
Buffer Zone) to collect/count data from a source shape (Boston Block Groups),
and produced a precise result set about the zone. Table 3 shows fixed datasets
acquired after the process.
Table 3. Fixed Datasets for 1-Mile Buffer Zone
In summary, the buffer zones used in this project have been created by the "Clipping" of geoprocessed polygons; their associated datasets for the study areas closely match those obtained from the "Centroid Within" operation. For consistency during the studies, and for ease in replication of results by others, the "Centroid Within" operation was therefore chosen for use throughout this project.