How is it that 91% of the US didn't vote for either Hillary or Trump in the primary, but yet they're still the final two candidates in the presidential election? Let's break it down with a simple graph!
I recently saw a really cool slideshow on the nytimes website, that answered this question very nicely. They used an 18x18 grid of squares, with each square representing 1 million people, to represent the 324 million people in the US (18x18=324) . Each slide labeled a subset of the grid, representing various groups of people who voted (or didn't vote). In the end, the red and blue squares showed that just 9% of the population voted for Hillary and Trump in the presidential primaries (ie, 91% didn't vote for them ... most didn't vote for anyone). Here is one of the final slides.
I really liked their slideshow, but it was missing a graph that showed all the groups labeled at the same time. So I decided to create one myself using SAS software.
This might surprise you, but I used some code I had recently written to create the following Pokémon Pikachu picture, to create my voter grid graph:
This Pikachu is a grid of data-driven colored squares, therefore it only required minimal changes to repurpose the SAS code to create the voter grid graphic. I changed the dimensions of the grid, added some code to annotate borders & labels around certain groups, and in almost no time I had my own graph. Try clicking the image of my graph below, to see the interactive version with mouse-over text that provides more details about each group:
Do you think the grid graph is appropriate for analyzing this type of data? What are the advantages and disadvantages? What other kinds of graphs might be better, for answering specific questions?