This static visualization showcases the inverse correlation between the financial impact of various types of disasters and the number of people killed by disasters over the century.
I used Tableau and Sketch for this assignment. I loaded the data into Tableau and filtered out rows of data that had null values for ‘Cost in Dollars’. Initially, I had separated the circles into rows that each represented a unique type of disaster. However, the inverse correlation between mortality and cost over time was downplayed in that iteration because there were a few rows that were largely empty.
For the final design, I decided to aggregate all of the instances of disasters into one row to further emphasize that contrast (the specific subcategories of disasters became less relevant to the purpose of the goal of this visualization).
I reduced the opacity of each circle to 50% in order to increase the visibility of each disaster instance, and to make the clustering more apparent. The categories of disasters were not as important, so the individual types were not color-coded. I used red throughout this visualization because it is an impactful color that is often linked with danger and trauma. I wanted the visualization to have a powerful impact on the viewer as quickly as possible.
After I assembled the graphs on Tableau, I screen-shotted them and imported the images into Sketch, a design application. I added custom labels and titles to better fit with the visual theme and emotional message.
The primary visual channels are size of the circles and x-axis distribution. Each circle on a timeline represents a disaster. There are two separate timelines: the first one has circles whose sizes are proportional to the number of people killed in disasters, and the second has circles whose sizes represent the cost in dollars of those disasters.
My goal was to immediately draw attention to the polarizing clusters in this visualization. On the ‘Mortality’ timeline, there is a collection of larger circles between the 1900–1950 tick marks, indicating that more people were killed in disasters during those years. Meanwhile, on the ‘Cost in $’ timeline, the larger circles start clustering between the 1960s-2000s. This implies that fewer people are getting killed, but disasters are becoming more financially-straining as we approach the present day.
I wanted the message of this visualization to be simple, straightforward, and powerful. In doing so, one of the main trade-offs was abstracting away from smaller details in the data. It is not as easy to view more details about or differentiate between individual disasters/categories of them. Some of the smallest circles on the timelines overlap, so there may be more disasters plotted than what is immediately evident.