Though the statistics and methods were all described previously, the reality was this project did not work like it was intended to. I ran into many issues with my data, the processes, and a few other technical issues. My main problem was not having physical access to help from other people due to quarantine, while also overestimating my own abilities. Because of this, I have created speculative results for vulnerability to COVID-19 in Montréal based on the statistics I was able to derive and off of visual analysis from choropleth models of the Census Tract data and locations of elderly homes.
This chart is here for quick definition references:
| PopDen | Population density |
| Pop65 | Percentage of the population over 65 |
| Jewish | Percentage of the population that is Jewish |
| Alone | Percentage of the population living alone |
| LowIncome | Percentage of the population who are low income |
| HC_SA | Percentage of the population working in the health care and social assistance industry |
| Homes_CT | The number of long-term care homes per Census Tract |
The Exploratory Regression results showed that the most important variables are Pop65, Jewish, LowIncome, HC_SA, and Homes_CT. This means that the most vulnerable populations are elderly, Jewish, of low income, health care workers, and those living in or near long term care homes.
Using the identified Exploratory Variables, the OLS Regression produced these statistical results:
| OLS Parameters | AdjR2 Values | AICc Values |
| Pop65 | 6.03 | 1178.47 |
| Jewish | 2.14 | 1169.08 |
| PopDen | 0.96 | 1140.44 |
| LowIncome | 0.55 | 1204.05 |
| Homes_CT | 3.03 | 1137.17 |
While you cannot assume anything geographically about these results because there is no GWR results to compare to, from this global model we can see that populations over 65 are by far the most vulnerable populations, with proximity to or living long-term care home locations and Jewish populations also being significantly impacted.
My GWR analysis failed multiple times. Because of this, I could not perform the Spatially Constrained Multivariate Clustering, but I could make assumptions about what kind of clustering would occur. The clusters I was able to visually identify are (in hierarchical order from most to least vulnerable):
-
-
- Côte-des-Neiges–Notre-Dame-de-Grâce/Westmount region
- Montréal North region
- Ahuntsic–Cartierville/Saint Laurent region
- Dollard-des-Ormeaux/Southwestern Montréal region
- L’Île-Bizard–Sainte-Geneviève region
-
The Côte-des-Neiges–Notre-Dame-de-Grâce/Westmount region by far had the greatest density of elderly folks homes, elderly populations and also Jewish populations. The greatest density of Jewish populations is in Beaconsfield which is adjacent to Côte-des-Neiges–Notre-Dame-de-Grâce, and holds 2 Jewish-specific long term care homes. The Hasidic Jewish population in Montréal is reported to have been the most impacted by COVID-19 after a large wedding for prominent members of the community took place on March 16, 2020 with attendees from New York (Jelowicki, 2020), the biggest hotspot for COVID-19 in the United States(The New York Times, 2020). It seems to be a big shift in culture for most religious communities to halt their weekly services, and this is likely why Jewish populations are so vulnerable. That is to say the fact they are Jewish is not the causation, but more the cultural values and timing revolving around major religious events and celebrations. This region is also home to McGill University, which means that there is probably a lot of foreigners there and many people travelling to and from, so this could have amplified the spread, especially because it is within 10km distance of many long-term care homes.
The Montréal North region has notably become more of a prominent hotspot, with the highest per-capita infection rate within the city during recent times (Rowe, 2020). This has been attributed to dense living conditions and a culture of caring for one another (Bruemmer, 2020), however, it is hard to define this kind of culture and this is based more on speculation rather than fact. This region also holds a large Jewish population, so it is possible that it has COVID-19 spread more up north due to this fact, and would coincide with the culture of caring for one another.
Looking at L’Île-Bizard–Sainte-Geneviève region, it is grouped together because this region has next to no COVID-19 cases. It is the Southwestern part of Montréal, and a majority of this cluster is actually it’s own separate island, so they are very secluded. There are no long-term care homes in this area and population density is very low.
I would recommend that anyone who lives in the top three cluster regions readily assesses their own individual vulnerability before deciding to leave their house for anything including essential services. If possible, travelling through any of these regions even to just go to a big box grocery store should be avoided.
Concludingly, the vulnerability of COVID-19 in Montréal based on the OLS statistics are elderly, Jewish, low income, high population density, and and within proximety or living in a long-term care home. The regions most vulnerable from most to least are Côte-des-Neiges–Notre-Dame-de-Grâce/Westmount region, Montréal North region, Ahuntsic–Cartierville/Saint Laurent region, Dollard-des-Ormeaux/Southwestern Montréal region, and L’Île-Bizard–Sainte-Geneviève region.