Excel skill + graphic design ≠ Good data visualization (unfortunately)
One of the unique lessons of my time in Europe. I think I realized it when I was told one day that I ”have to think” (about the data graphics I was creating). Have I not been thinking? I thought I was smart – I began to question that notion. Sitting at a cluttered desk, staring at the mess of charts and graphs in my spreadsheet, I really started to rethink my professional self. Am I smart? Is this an identity crisis? No… At least I hope I passed that stage years ago.
In the one place at work where I could be alone and think (not outside for a smoke – the other not-so-glamorous solitary space in offices), I had a good sit-down to do what I was told and think. Why was I having trouble coming up with good ways to visualize data?
- Good performance in mathematics courses? (check.)
- Good at graphic design? (check.)
- Understanding of the subject + audience? (BINGO. Time to flush and get out of here…)
Data by itself is a tricky thing. Presenting it in different ways can distort the impressions and decisions based upon it, no matter how objective the data itself is. Adding the creative freedom that comes with graphic design to the mix creates an ambitious and complex job for the unfortunate colleague who ends up with the task of making data pretty and simultaneously idiot proof. Without a good understanding of the audience, context, and content itself, there are essentially two options that come to mind for visualizing a set of data:
- Bar chart
- Pie chart - how impressive…
It’s a shot to my ego thinking that the bar and pie charts I’m creating could be made by a 10-year-old kid these days (with the right supervision and guidance). But although bars and pies are easy to create, they aren’t always easy to read.
As information and the relationships it conveys get increasingly complex (human development indicators, event performance measurement, qualitative surveys – just some of the things I’ve been wrapping my head around in various visual forms) bar and pie charts get harder and harder to understand. No more pizzas, no more simple addition… it’s no longer a question of how many of this or what percentage of that, but how is inflation rate related to corruption in politics? How can we track how this new contact becomes a new sales lead? Is there a way we can visualize that with these numbers… can you make it idiot proof?
Despite the simplicity of the adjective “idiot-proof,” presenting complex and specialized data in this way can get pretty complex. After a while, I began to realize that getting good at it is a matter of 1 step:
1. Exposure to the creative data visualization community
This is most important. Learning how other specialized and visually-creative minds have simplified complex relationships into colourful lines and circles and blocks is a way to expand your understanding the infinite possible ways there are to understand a set of data. I saw a youtube video that turned the simple bar chart into a funny-looking 30-second moving visualization of the progression of popular technologies over time. So simple, but it was amusing to see Apple products move rapidly upwards as a modest yellow bubble on a screen.
Some favourites
