All the ways you’re making bad data viz choices
There’s a reason why the pie chart is universally hated: The iconic visualization actually makes it harder to understand the data. But avoiding pie charts isn’t the only rule for making good data viz. A handy guide called From Data to Viz unearthed by Flowing Data lays out 37 different charting “caveats,” from the controversial cutting of the Y-axis to unreadable “spaghetti graphs,” making it something like a Wikipedia for data viz don’ts.
While the site’s main purpose is to help designers figure out what kinds of visualizations would suit their data best, the don’ts are just as useful. In a world awash with data, where people use graphics to make the numbers say what they want, it’s crucial to avoid anything misleading. A few tips? Avoid using the full rainbow of colors as your palette–in fact, if you have to use more than seven colors to communicate your point, you should probably try using another kind of chart altogether. Don’t plot too many points because it’ll make your graph unreadable–one quick fix is to make the size of each data point smaller, so hidden patterns can emerge. There’s no need to get fancy and change some of the most followed conventions in data viz–for instance, don’t put your Y-axis horizontally. It’ll just mislead people.
And above all: Avoid the pie chart. As From Data to Viz creator Yan Holtz writes, “it’s bad by definition.”
Check out the rest of the caveats here.
(27)
