Fig. 3

Box Plot Left: Boroughs divided by high (n=15) and low (n=11) incidence rates and their associated Φ-values. It is possible to notice a clear difference between the two sets medians (bands inside the boxes). Moreover, high rates boroughs exhibit a considerably higher interquartile range, meaning that cumulative contact rates for boroughs with ILI incidence rates higher than 10 tend to exhibit a significantly higher range variability respect to boroughs with lower ILI incidence rates. This distance, which represents the middle 50% of data, touches considerably higher Φ-values for high ILI-rates boroughs respect to the values covered by low ILI-rates boroughs. A Mann-Whitney U-test was run with MATLAB with the null hypothesis of both sets having an equal median and the test rejected the hypothesis with a p-value of 0.0293.Right: Boroughs divided by high (n=15) and low (n=11) incidence rates and their associated children demographic rates. It is possible to notice that the difference between the two sets medians (bands inside the boxes) is almost non existent. Moreover, even if high rates boroughs exhibit higher interquartile range, this distance, which represents the middle 50% of data, touches lower children demographic rates for high ILI-rates boroughs respect to the values covered by low ILI-rates boroughs