DATA ACQUISITION AND Conversion
The data for COVID-19 cases is from April 14, 2020, meaning that this analysis is immediately out of date, and thus, is more speculative than concrete. Additionally, the precise locations of COVID-19 cases has not been mapped and is almost impossible to really know, so the locations are based on their proximity to green spaces and elderly folks homes within the Census Tracts. Although this is the best way of interpolating the locations, it also introduces a source of error because COVID-19 exposure could have very easily occurred elsewhere, even outside of the Census Tract. Additionally, the cases are recorded in aggregate based on borough, not individual points, and administrative boundaries do not play any role in the spread of COVID-19. This also introduces a source of error because I had to transform the spatial units from boroughs to Census Tracts, as explained in Methodologies.
ASSUMPTIONS
Because it has been established that long-term care homes are the centre of the COVID-19 outbreak, due to low pay, a heavy workload, and lack of resources (Kestler-D’Amours, 2020; Lowrie, 2020), I made assumptions that the people who work in these homes and live adjacent or in proximity of one are more at risk. This is not necessarily true, and could present a major source of error because if population density around a long-term care home is quite low, or the Census Tract average income is quite high, this may not necessarily be true.