For the potential of getting a clearer trend, Manhattan, which has the highest amount of and most spread-out public Wi-Fi hotspots, was chosen as a case study to see if there is any pattern existing between income levels and Wi-Fi hotspot density. Dataset that contains median household income (MHI) by census tract level in Manhattan was downloaded from American Census Survey data, and the two variables were plotted side by side.
The comparison of the plots does not seem to show a consistent correlation between MHI and density of Wi-Fi hotspot, as census tracts with lower or higher MHI can be seen to have both higher and lower density of Wi-Fi hotspots. For example, some areas with low MHI, such as the Lower East Side and East Harlem, also have low density of Wi-Fi hotspots, but areas with high MHI, such as the Upper East Side and Chelsea, have low density of Wi-Fi Hotspots.
This map shows MHI and density of public Wi-Fi hotspots for each census tracts in Manhattan. The density of public Wi-Fi hotspots are demonstrated by the circles with darker-green representing higher density. By overlaying the two variables together, this map provides a clearer visualization for the relationship discussed in the earlier section between the two variables.
From the bivariate map, it is obvious that the Lower East Side area has low MHI, as shown by the blue polygons. At the same time, census tracts in this area also tend to have very low density of Wi-Fi hotspots, as indicated by the dots. Most of their density is at the lowest interval. However, there are also a few census tracts in this area that has low MHI but high public Wi-Fi density. These suggest the limitation from identifying trends from simply observing the plots and the need to quantify data to provide more information.
In order to provide more information and a quantitative understanding of density of public Wi-Fi hotspots and median household income, a scatter plot with a linear line of best fit is shown on the left. From the result of this exploratory correlation analysis, there seems to be a weak positive correlation between wifi density and median household income. The r is only about 0.19, which suggests that the correlation between these two variables is very weak, and there is need for further research and analysis into other variables.