|
In this page some of the technical
components of the project are discussed, such as: the source
of the spatial data and how we transformed it, the sources for
the attribute data and how we transformed and classified it,
and choices made with respect to the reporting periods used
in our analyses.
The base maps used in producing the maps presented in the Immigration
and Greater Vancouver Atlas were obtained from Statistics
Canada under the Data
Liberation Initiative. Using Arc/Info we imported the cartographic
boundary files and extracted only those Census Tracts which
fell within the CMAs of Abbotsford and Vancouver. We then projected
the data from Geographic coordinates (longitudes and latitudes)
to UTM coordinates.
It was necessary to project the data since, at Vancouver's latitude,
a degree of longitude is approximately 65% that of a degree
of latitude, and a map produced using longitudes and latitudes
(e.g., a Platte Carre map) appears severely distorted. The UTM
projection is the most commonly used projection in Canada and,
while it is a conformal projection, the amount of areal distortion
is minimal when examining a region as small as the Vancouver
region.
The attribute values (i.e., the census data) were also obtained
from Statistics Canada under the Data Liberation Initiative, and
through the Metropolis Project itself. For details on the attribute
data, refer to the Discussion and
the Definitions pages. In producing
the maps Census Tract (CT) data was used rather than Dissemination
Area (DA) data for several reasons, including graphic constraints
(using DAs would make the online maps far too busy, and most of
the detail would not be decipherable anyway), confidentiality
issues and data suppression
by Statistics Canada.
We mapped the attribute data using location
quotients (LQ). They are a widely used index which summarizes
how concentrated, or not, a group or activity is within an area.
The location quotient for a specific group i in an area is the
ratio of the percentage of the total population in group i in
the CT under consideration to the percentage of the total population
(of the CMA) who belong to group
i. Location quotients above 1 indicate that you are more likely
to find members of a group in the CT than you would expect on
average, whereas values below 1 indicate that you are less likely
to find members of a group in the CT than you would expect on
average. For example, if the percentage of people born in Hong
Kong is 4.8% for the entire CMA, then a census track which has
9.6% of its inhabitants born in Hong Kong will have an LQ of
2.0 while a census track with only 2.4% of its inhabitants born
in Hong Kong will have an LQ of 0.5. The LQ is a dimensionless
index, so the LQ values are directly comparable across the 21
maps.
The legend categories were selected in order to highlight those
census tracts that are either above or below the CMA average.
The central division was designed to straddle the expected value
of 1.0 (+- 0.15). The adjacent classes used fixed spans unless
the data itself has a smaller range, in which case the upper
or lower class limits were set explicitly to the data value.
Census tracks colored white are either tracts for which no data
exists, or tracts in which no members of the highlighted group
are found.
Throughout the Atlas we have used standard Statistics Canada
census reporting periods (e.g., 1981, 1986, 1991, 1996 and 2001)
and classification schemes. If you have any questions concerning
the definition of a variable, please refer to our Definitions
page first. If you can't find what you are looking for there,
please look at the information Statistics Canada provides--links
to the most relevant pages can be found at UBC's Data
Library Census of Canada Documentation
pages (If any questions still remain, please contact us.)
|
