little things can mean a lot
No small number of you took on our small multiples challenge this month, with 74 entries making it in by the deadline. As always, you covered a variety of topics and aesthetic approaches, from the simplicity of lines or bars in monochrome, to choropleth maps, to 650 pie charts, in two distinct formations. (We did say “never say never” to forty-five pie charts last year, but Neil is really pushing the boundaries of “never.”)
Many contributors found ways to use the inherent power of the small multiple format: effectively showing changes over time across multiple dimensions. Fred used multiple tilemaps of the U.S., repeated across 12 years of monthly data, to show state-by-state unemployment rates. This layout made it easy to see a year-over-year view (by reading horizontally), a single year’s change (reading vertically), and an overall trend. Claire, on the other hand, used a single tilemap, but within each tile was a stacked area chart showing the composition of energy usage in the state over time. Claire’s use of blue for renewable energies vs. gray for coal and other nonrenewables made the overall message easy to see from a distance, but also detailed enough at the individual panel level to be insightful. Simon chose to use one London borough per panel, and within each, used an area chart with a color ramp to show the rise in property prices over a 25-year period. The color highlights the boroughs that have increased the most, while the area charts reinforce the reality that the price rises were not confined to any one borough.
Others did a great job of using the visual appeal of the small multiple chart to draw the users’ attention in, and then told a story well within their chart. In applying spider charts to a striking palette, Max created an eye-catching visual that conveyed the message of the visualization—the feelings of discrimination experienced by LGBT youth vs. straight, cisgender youth—clearly and strongly. Kate’s creative approach used stacked bar charts across Harry Potter books to show how the cover palettes varied between the American and the U.K. published versions. This is a case where it may not be important to know the exact percentages of each color in each cover (although that is the data encoded in the chart), because the value is in comparing the books and the publishers at a higher level, getting an overall aesthetic sense of the differences.
Will used satellite imagery of his own farm, and plotted treemaps of the soil composition over individual sampled locations. He also included a larger, inset treemap that showed the overall composition of his farm, and that inset treemap was labeled in a way that it served both as an additional information chart and as a legend for the rest of the visual. This is a clever solution to a common challenge in small multiple charts, which is how to include legends or axes or labels so that each panel can be read clearly, without adding too much repetitive clutter.
You can see all 74 submissions in the SWD community. As you browse entries, we encourage you to participate by leaving comments and adding datapoints to work you enjoy.
There are not a lot of tools out there that let you make small multiple charts easily out of the box, so for everyone who took the time to build one, we commend you for your efforts, we thank you for your contributions, and we look forward to your participation in future challenges. Why not start with February's challenge, which is all about the simple bar chart?
