Much of the data used in this analysis were obtained from the Spatial Information Systems lab at Simon Fraser University. Several shapefiles for Vancouver were used and include enumeration areas, parks, street network, and shoreline themes. These data were obtained from the City of Vancouver or digitally created by Peter Schaub, a former Geography 354 student. This data were projected as UTM 10 and has a NAD83 datum. Other data obtained from the SIS lab includes shapefiles for bus stops and the SkyTrain lines in the GVRD. This data were originally obtained from TransLink and is also projected as UTM 10 with a NAD83 datum. Using my own knowledge of Vancouver as well as printed maps of the city, I believe this data to be reasonably accurate.

Using these shapefiles in ArcView, I created two additional themes. Using the shoreline theme and a recent map of Vancouver, I selected only those line segments of the shoreline that are co-existant with the seawall and created a new shapefile.

To create an amenities theme, I used the existing street network theme, a 1994 Vancouver land use map published by the City of Vancouver, and my own knowledge to identify the portions of streets that have shops and services that are likely to be used by senior citizens. While this approach is highly subjective, I think it is more suitable than using a land use theme that identifies all commercial and office districts as these districts in their entirity are not considered desirable and may even skew the results innapropriately.

Using the SkyTrain route shapefile, I digitized the locations of the existing SkyTrain stations and created a new shapefile.

To create a shapefile that showed the locations of existing, assisted living homes for seniors that are potential competitors for a new development, I used the phone book and internet to derive a table of their locations. With that table and the street network theme for Vancouver, I was able to geocode their locations and create a shapefile.

1996 StatsCan census population data were downloaded from the Research Data Library at SFU and were manipulated in Microsoft Excel and ArcView to produce data that showed the percentages of the different age groups relative to the total population for each enumeration area. This population data is based on a 100 percent sample rate (though in reality it is likely less than that) and is considered to be reliable. Although the census data is seven years old, it is the most current available at this time for enumeration areas. In hindsight, I should have placed more importance on the 60-69 cohort than I did in my weighting for the multi-criteria evaluation as most of this group has likely already reached the age of my target market - the 70 - 84 age group.

Average dwelling value data were also derived from the census data though at a 20 percent sample rate. I have my doubts about the quality of this data as their were several enumeration areas that I know to contain residential uses for which no values were recorded. Furthermore, this data shows some enumeration areas in the Downtown Eastside to have average dwelling values between $500,000 and $1,000,000 though I think this is unlikely. In absence of a land use theme, I am relying on census data pertaining to average dwelling value to eliminate all industrial and other non-residential areas however, it indicates average dwelling values in areas that I know to be used for industrial purposes. The questionable quality of this data compromises the quality of the analysis results.

After these vector data were inspected and edited in ArcView, they were imported into IDRISI and then converted to an IDRISI vector format. Finally, the vector files were converted to raster files for analysis in IDRISI. The cartographic model provides a detailed illustration of the data manipulation processes carried out in IDRISI.