Methods
In order to address our research questions, we used esri ArcGIS (10.3) to: 1) output descriptive statistics of West Point Grey and 2) perform an experimental road network analysis. Our data sources are as outlined in Table 1 below.
Table 1: Data Sources for Variables
First, we displayed Walk Score data by each dissemination area in West Point Grey (Figure 1). A dissemination area (DA) is the smallest geographic area that has all census variables available. They consist of 400 to 700 individuals and can be the size of one or more city blocks. Second, we explored the geographic dispersion of vulnerable populations residing within West Point Grey. We overlaid five population variables from 2011 Census data (under the age of 19, over the age of 65, unemployment rate, rental status, and median income) with Walk Score, using a proportional dot method, to visually compare which vulnerable group is most likely to reside within a low Walk Score dissemination area. Although all variables were dispersed relatively evenly, we noticed that a higher proportion of older adults fell within areas of low Walk Score. We used pie charts to symbolize the total population of adults over the age of 65 in the 2011 Census and displayed this over the Walk Score data(Figure 1). From this analysis, we selected four dissemination areas in the northern section of West Point Grey that had both higher populations of older adults and lower Walk Scores as our focus for further analysis. While the dissemination area in the north-west corner of West Point Grey was considered “Car Dependent”, we decided not to include it in our analysis given the lower proportion of older adults.
Figure 1: Population Age Distribution and Walk Score in West Point Grey
From the map we were able to identify some trends: Walk Score is generally higher throughout middle of the neighbourhood (running in the east/west direction), while scores tend to decrease as you move north or south of this region. Dissemination areas with the lowest scores fall along the northernmost edge of the neighbourhood. This pattern is likely explained by the geographic isolation of the northernmost dissemination areas from amenities. Not only are these areas surrounded by parks and ocean, an investigation in Google Maps revealed that the majority of amenities within West Point Grey are located along West 10th Avenue, which runs east/west through the center of the neighbourhood.
Within the Walk Score data, we examined the characteristics of our four focus dissemination areas. These dissemination areas have relatively low Walk Scores (61, 42, 37 and 43) compared to the the overall Walk Score of West Point Grey (74) (Walk Score, 2016). For the purpose of this study, we chose to focus on food services. Previous research on the travel behaviours of older adults living in highly walkable neighbourhoods in Vancouver supports this decision. Winters et al. (2015) found that the most frequent destinations visited by study participants (adults over the age of 60) were grocery stores, followed by restaurants. Access to food services appear to be an issue in these areas of West Point Grey, as Walk Score data indicates that the ratings for coffee shops, restaurants, and grocery stores are relatively low (Table 2).
Analysis
The purpose of our analysis was to reveal the distance to various food services from the average residential location of the residents within our four target dissemination areas, and the impact that a new proposed location would have on these distances.
To prepare for our analysis, we located population weighted centroids. Population centroids are more representative of where individuals actually live, compared to geographic centroids which lie in the middle of a geographic region; furthermore, given the large area of land covered by parks in our target dissemination areas, geographic centroids would not accurately capture where the target populations reside. We extracted the longitude and latitude of each centroid from the Walk Score data and plotted them on our map. We then used the geoprocessing buffer tool in ArcGIS to draw 400 meter and 800 meter buffers around each population weighted centroid. We selected these distances based on the Walk Score methodology of awarding the highest scores to amenities within a five minute walk (or approximately 400 m) with less points awarded as the distance increases. The 400 meter buffer can be used to visually assess which amenity locations fall within a five minute walking distance from each population weighted centroid. This method also aligns with previous research, as Nathan et al. (2012) employed 400 and 800 meter distances to assess older adult mobility in relation to neighbourhood amenities.
In order to perform a road network analysis, we geocoded all amenities (using the BC Physical Address Geocoder) that provided food services within the 800 meter buffer around each population weighted centroid by searching for address listings on Google Maps. We operationally defined food amenities by two categories: 1) businesses with the main purpose of selling food or alcohol products (ie. stores such as a butchers, grocery stores, or liquor stores) and 2) businesses with the main purpose of selling prepared food and drinks (ie. restaurants or cafes). We grouped these businesses by functionality, as restaurants and cafes provide a similar service - a place to enjoy a meal and socialize with others. On the other hand, food and alcohol stores, may provide less opportunities for social interaction, as purchased food and drinks are often prepared and consumed within the home. We opted to omit certain locations from this analysis: Shell Circle K and 7-Eleven (due to the limited fresh food options sold in these stores) and beach concession stands (as concession stands typically have limited hours or are closed during the Winter).
After geocoding the locations of amenities in West Point Grey, we searched Google Maps for potential areas within the neighbourhood to redevelop to increase access to food services. Through this process we identified a location that appeared to be underutilized (Figure 2). Using the VanMap tool provided by City of Vancouver’s geographic information system department, we were able to determine that this 3878 meters squared piece of land is already owned by the City of Vancouver. We deemed this area appropriate for small scale development for multiple reasons: it is accessible to walkers within the low walk score dissemination areas as well as possible beach goers, the land is currently unused, yet it is not designated as a park (Figure 2) which means the city can save money by not having to purchase the land from homeowners, property site lines will not be affected by such a development as the lot is bordered by parks on three sides, and finally, parallel street parking already exists in this location (City of Vancouver GIS Department, 2014).
Figure 2: Satellite Images of West Point Grey
Finally, we attempted to quantifiably measure if the addition of amenities to our selected location would have an impact on walkability within West Point Grey. We performed road network analyses in ArcMap using BC Roads to find the nearest grocery/liquor store and cafe/restaurant to each population weighted centroid. We performed the same analysis after the hypothetical addition of one cafe and one grocery store to the new proposed location. From this we were able to calculate the change in distance to the nearest food amenity for each population weighted centroid.
