Evergreen Data

Does your org have a style guide? Chances are the answer is yes (and if you didn't know that, you're probably at least part of the reason the comms team cries themselves to sleep at night). This is the document where you’ll find the exact colors you should be using in your charts. If you’re lucky, you’ll also find details about font sizes. This takes out the guess work and brings you cohesion and consistency to your data viz. But, chances are good that your style guide needs some things added to it for data viz purposes. Check with your comms team about adding things like a condensed font type, chart types to use for common data sets, and specific formatting for those chart types. You can even include examples to help your team understand how to format their data viz, like this example. So, do you have a style guide? What is it missing for data viz?

Even a new kid on the block (my favorite was?Donny) knows that statistics can lie. And charts can aid and abet. To do our audience justice, we have to visualize data responsibly. I sound like I’m on some moral high ground but even after being in the data world for 20 years I still fall into some traps and only recognize it when I start feeling the ick. Here’s a round up:

What color is Math? If you've been following me for a while, you might remember me asking this question a few months back. Soooo many people responded here and on my Instagram. I've got the results for you. It'll come as no surprise that people have associations between colors and *things*. So let's use that to our advantage when making charts so cognition is even easier for our audience. I'll walk you through how to do this step-by-step.

Your Tableau graphs don’t work for slideshows. Unless you design with that destination in mind. Let me show you a SMDH example and provide some clear advice on how to make this work.

You should avoid using red and green together in your charts, especially red to mean “bad” and green to mean “good.” Because the most common form of colorblindness is red-green. Those two colors appear as a muddy green-brown and aren’t distinguishable to people who are colorblind. You can always check how your colors look using the Data Visualization Checklist. When you get to the colorblind guideline it’ll show you what your graph looks like so you can determine whether you’re accessible in this way. (I have a class on the Checklist in a few hours - check out my other posts.) Making data visualization that’s accessible also goes beyond attending to colorblindness. I have 9 other totally do-able tips in my latest article.

When you use a legend in your chart, people have to do mental gymnastics to connect each legend entry to the associated part of the graph. And if it looks like it’s going to take mental gymnastics, people are gonna quit. As much as possible, stop using legends in your chart. What should you do instead? I’ve totally got you. #dataviz

Where should your graph’s scale begin? Where should it end? If you’re in the camp that thinks it should always start at 0% and go to 100%, that’s outdated thinking that’s likely skewing an accurate read of your data. Jump in and get a more sophisticated strategy for choosing the right scale for your chart. #dataviz

You see the constellation where your audience only sees a random smattering of stars. What makes sense to us (who have been so steeped in the data we're dreaming about it) will not be readily obvious to an outside viewer (even if it's someone who cares quite a lot about the results). We can help our audience see the connections between our data points by assigning a color system to our graphs. What’s a color system? Let me show you two different ways you can make this happen.

The only movie I ever watched on repeat as a youth is My Cousin Vinny. I realize I’m dating myself with this reference. With the hindsight of today, the movie surely has its problematic moments. It also has so many golden scenes that have made it a cult classic. Like the one where the lawyer is questioning whether a witness was able to get a clear view of the defendants, when there’s a lot of obstruction between the witness’s house and the Sac-O-Suds where the murder occurred. We, too, put obstruction in our data visualization. All the unnecessary lines in our graphs are the dirty window, crud-covered screen, the trees with all the leaves, and the seven bushes standing between your audience and your graph. They get in the way of being able to see clearly. This doesn’t mean *every* line in your graph has to go. Don’t take this too far! Let me help you discern which lines to keep and which to ditch. #dataviz

When we think about #dataviz, we usually focus on the graph. The visual part. The picture. Makes sense. But don't neglect the text. That's half your story! You only get a little bit of text in your graph. Here’s how to make sure you get the most mileage out of it.