All Charts Have The Potential To Suck
Monday, 11th July 2011
I have been known to give pie charts a bad time because in a vast majority of cases, they are the wrong choice for displaying the kinds of data they are used for. It isn't just pie charts that are utterly rubbish though. Any chart has the potential to suck in many ways.
The problem with charts is that the people who make them forget why people are using them. A chart doesn't just show a piece of information in an attractive way, the entire point of any chart is to show values in comparison with other values. You either want to see how this month compares to last month, or you want to see if green is more popular than red - and by how much.
Because of this, there are simple visual hints you can give on a chart that make it easier to read. In a chronological chart, we are almost certain to put time on the horizontal axis - to put it on the vertical axis would be crazy! But what if we aren't dealing with time? What help can we give people. Here is a genuine example that I have re-created from information published recently on an excellent website. As usual, I mention no names - especially as the website is truly great.
As you can see, there has been a survey about where people work and lots of people have answered - the problem is that while you can tell that "at home" is the most popular answer, and "open office" the clear second, it gets harder and harder to rank the answers as you work your way through third, fourth, fifth and so on. With one very simple change, the whole chart is easier to judge.
All the hard work has been done for us in this chart. We can tell exactly what is going on and we can easily asses that if you add "open office" and "cubicle" together, that there are more office workers than home workers, even before we add the "office w/door" category.
So next time you put together a chart, why not spend an extra five minutes doing the first thing each viewer of the chart will have to do in their head. They will really appreciate it.