Why the f**k do we see so many bar charts?

With respect to data visualization, humans are really good at some things and really poor at others. Good data visualization should take advantage of what people do well and avoid what they do poorly.

Consider this example that shows two ways to present the same data.

Which one is more “interesting”? Most people would say the collection of colorful circles on the right.

But which one quickly answers questions such as “which category is second from the top? Fourth from the bottom?”

I want you to give something a try to drive home how good most people are at comparing length from a common baseline (and how bad at judging the size of circles).

Please visit https://bigpic.me/estimate. You'll be able to see the results there after you submit your response.

Don't feel too bad if you missed the circles. While almost 80% of people who participated estimated the size of the bar correctly, only 40% were able to estimate the circle. People are much better at comparing the length of bars from a common baseline than they are at comparing the area of circles.

And that's one reason we see so many bar charts.

Please do not come away with this exercise with the notion of "bars are good, circles are bad." Circles in data visualization can be very helpful (especially on a map); they just aren't very good if you want to make accurate comparisons.

Steve Wexler is the author of the new book The Big Picture: How to Use Data Visualization to Make Better Decisions--Faster. You can download a free sample of the book from https://bigpic.me.