As part of a personal project, exploring novel ways to visualise pandemic-related data, I produced this gradient chart. It shows that although there was uncertainty, there was strong evidence that people from minority ethnic backgrounds had an increased risk of dying from Covid-19 (compared to people from a white background).
As the Covid-19 pandemic unfolded, the relative risks faced by people from different ethnic backgrounds were often discussed in an overly simplistic way. For example, ‘people who are black are twice as likely as people as who are white to die from Covid-19’. I wanted to find a way to visualise the uncertainty around statements like this one. Without detracting from the core message conveyed by the data. There is a higher risk.
My approach to showing this uncertainty was inspired by an example in the excellent Better Data Visualizations by Jonathan Schwabish. I used R (Tidyverse packages) to clean the data and create a basic outline for the visualisation (ggplot2). The ggdist packages was really helpful when creating the colour gradients which show the uncertainty. The final figure was put together and edited it in graphic design software (Affinity Designer and Affinity Publisher).
The gradient chart allows the viewer to see the uncertainty around the relative risks from Covid-19 for different groups. The annotations explain to a general audience the tricky statistical idea which allows use to estimate uncertainty (confidence intervals). The chart attracted significant engagement and interest on Twitter.
