3MW (Two alternatives to paired bar charts)

Guten Tag!

Many greetings from Ulm, Germany. Let's dive right into this week's issue.

Check out this paired bar plot. It compares the life expectancies of selected countries in 1952 and 2007. The data is courtesy of the Gapminder foundation.

You can find such a plot almost everywhere. I think that's because paired bar charts are easy to make. But I'm not a big fan of them. I find them hard to read and it annoys the crap out of me to move my eyes back and forth to make comparisons.

So today, I'm going to show you two alternatives that can replace paired bar plots.

Dot plots aka dumbbell plots

Let's transform the bars into points. Then, we can draw both the 1952 and the 2007 value on the same line. I like to draw a line between two points so that it's easy to see the connection. Consequently, each line looks a little bit like a dumbbell. Hence the name.

Did that impress you? I hope you like this new plot. I like to think that it makes comparisons waaaay easier (even if it's not ordered yet. We'll come to that).

But in case you are not convinced yet, let me add a little extra. The thing is: Paired bars are crappy at scaling. Have a look at the following monstrosity when I compare more than five countries.

Hideous, isn't it? Check out how the dumbbell plot fares for the exact same data.

Notice how Zimbabwe actually decreased its life expectancy over time. That's both sad and something we have to watch out for. With this kind of plot it's important to order the dumbbells. Make sure that countries where the green dot is left of the orange dot are grouped together (and vice versa). That's why I've plotted Zimbabwe at the top.

While you're taking care of the ordering, why not order the bars to emphasize something else as well? Here, I've ordered the dumbbells by life expectancy in 2007. But you could also order by the 1952 value or by the amount of change. The choice is yours.

Slope charts

I know that I promised you another alternative. So, here goes. This one is called a slope chart. It can be put to good use when you have a lot of data.

Here, I've compared the life expectancies of even more countries. For each country I drew a straight line from the life expectancy of 1952 to that of 2007. Of course, not every line can be colored. Otherwise, you'll get a mess. But that gives you the chance to highlight what you think is important. It really depends on what you want to communicate.

Recommended Resources

The code by me: I'm writing a full blog post on how you can create the previous plots. But it's not ready yet. But I won't leave you empty-handed. For now, you can use my preliminary code on GitHub.

One chart at at time by Jonathan Schwabish: Seasoned readers of this newsletter know that I love Jon Schwabish's book on data visualization. Imagine my delight to find out that Jon also hosts a YT video series on chart types. So many data viz specialists chimed in and you can learn a lot in a short amount of time. Definitely check it out.

That's a wrap! As always, I'm happy to hear your feedback about this week's issue. Feel free to write me on Twitter or reply to this mail.

Enjoy the rest of your day!Albert