Want to play a game with The Dashboard Spy and Information Visualization expert Stephen Few? In his blog post, The Billion Pound-o-Gram Redesigned, he takes a stab at redesigning a pretty well known chart by David McCandless.
Take a look at the original chart here:
Here is how it appeared in Guardian.co.uk’s Information is Beautiful Friday:
289 billion spent on this. 400 billion spent on that. When money reaches this level it literally becomes mind- boggling.
Yet these figures are regularly issued by the government – and the media – as if they are self-evident facts that everyone understands.
Frustrated by this, I created The Billion Pound-O-Gram. It’s a cousin of the Billion Dollar-O-Gram.
In this version, I’ve mixed up of 2008/09 figures from the Treasury and the Guardian. Visualising the numbers like this puts them in visual context, making them easier to relate to.
I was pretty shocked by the size of the UK budget deficit – essentially the country’s overdraft. It’s more than an entire year’s worth of income tax.
So, now for the game. Stephen Few has redesigned the graph. Take a look at his version:
Here is what he had to say:
All of these comparisons are incredibly simple to make using the bar graph below. Take a minute to notice how easy it is to see the relationships between these values from largest to smallest and to compare them. Notice especially how easy it is to compare each of the values with the budget deficit, which appears as the vertical black reference line.
In the bar graph, I stuck with the colors that McCandless chose to make it easy to compare his chart with mine, except that I tweaked a few colors a bit to resolve minor problems. In McCandless’ chart, some colors stand out more than others, but they should be equal in salience unless there’s a reason to feature some items over others. Also, for some unknown reason McCandless sometimes altered a single color from rectangle to rectangle, which serves no purposes and creates potential confusion. For example, notice that some of the green rectangles are lighter than others, yet they all represent “Earning.”
I can’t imagine anyone seriously arguing that McCandless’ chart communicates this information as well as the alternative above, but is his chart more engaging? Some folks might find it more engaging purely on the level of entertainment, but not in a way that encourages or supports meaningful consideration of the information, resulting in optimal understanding. Journalism should tell the story truthfully and clearly.
So, which version do you like and why?
Hubert Lee
The Dashboard Spy
Hubert,
Great article. I can answer your question with some authority, as I actually presented a similar McCandliss chart (A Billion-O-Gram geared more to US citizens) to a group in the context of “telling stories with data.” What I can tell you is that it was *very* effective in engaging the group. (BTW, I’m generally a fan of Few’s work.)
However, most of the comparisons were already very carefully crafted. For example, $Spent in Iraq compared to the $Predicted before the war was set up go make the reader think “wow”, as the “predicted” rectangle was embedded inside of the $Spend rectangle. The colors differed, because of the embedded rectangle. Comparisons across categories were much more difficult, but that was hardly the point.
Few’s display certainly appears as more “ordinary” to most people.
I actually tested engagement and insight with the group. I showed a bar graph very similar to Few’s after the “wow” factor of McCandliss’ chart. With some reluctance the group that got the bar chart went against the group that got McCandliss’ graphic. The general consensus of the entire group was that the well-constructed bar chart was far better as an analysis tool (creating more insight through easy, more precise comparisons), but that they were all far more excited by the Billion-O-Gram.
After some discussion, they agreed that McCandliss’ chart was directing the reader toward a more pre-fabricated conclusion. This isn’t bad of course. All stories (with data or not) have a point of view. But in designing something that will be able to let the reader create his own insight the simple bar graph was much better.
McCandliss designed the display well, as it gets his points across. But it’s better as an exclamation at the end of the sentence, while a well-designed bar graph can be the heart of prose.
From a comparison point of view. The simple bar chart was much easier to quickly gauge the relative values. The Pound-O-Gram required a more difficult volume analysis. In both, I found the scattered categories required scanning all the data looking for like or comparitive values. I would use the bar chart, but organize them by putting the categorits together and then the comparitive categories above or below.