A reader named Jess has asked my opinion on a set of charts in China vs United States: A Visual Comparison. I’ve captured part of one chart below. I found these visualizations interesting and fairly self-explanatory.
Some commenters seemed confused by the flow of the charts, as directed by the green lines and arrows. I though that the flow was apparent, wth arrows leading from a broader statistic to a more granular breadown.
Many commentators disliked the lack of scale or data labels. My first impression was that the charts are showing qualitative or semi-quantitative comparisons, and that the tallest bar in a logical grouping of bars represents the largest value, and values of the other bars are proportional to their heights relative to this tallest bar.
A little quantitative information never hurts, of course. Including a simple data label at the top of each bar would add an increment of value to these displays.
In any case, I think this is more effective than a lot of the chaff being passed off nowadays as information visualization.
What do you think?
Mike Woodhouse says
I don’t like the top progression: is the last one growth per capita, which to me is implied from the arrow leading from “growth rate” or is it gdp per capita, in which case the arrow should come from the left. From knowing that China has 3-4 times the population of the USA, I’d say the latter makes more sense.
So I’d say it’s a nice idea, very badly executed.
Ken Puls says
I think it’s an interesting (set of) chart(s), now that your comment broke down what it’s meant to convey. At first glance I didn’t see that it was meant as a drill down. (Long time since I looked at GDP!)
While interesting, I’m not sure it really meets the purpose of why I use charts, and that is to be able to quickly interpret the information. I really had to look at this one to make sense of it. I would have probably broken it into four separate charts… the top 3 on their own and the bottom two combined in one. More importantly, a title for each that clearly states what it is. (Scale would be nice as well.)
One thing it does do, that separate charts may not, is clearly define that there IS a relationship between the sets of data though. I do like that.
Ken Puls says
Honestly, I haven’t dealt with GDP in so long that I didn’t realize the charts were trying to convey a relationship between the data. Knowing that, it’s kind of a cool thing, as it does add a layer of information that you might not see if you split them apart.
Big question to me is that, even with the extra info present, do they fail at conveying the information quickly? I had to really look at them to draw those relationships back and forth and figure out what each piece was.
Personally, I would probably have opted for four charts. (The top 3 as individual and combine the bottom 2 into one chart.) I’d also have clearly labeled them with titles and scales as I couldn’t tell you for a certainty which of these is graduated in percentages and which is dollars. (I’m about 95% confident, but that does leave some doubt.)
Ken Puls says
Oops… sorry for the duplicated comment. I thought it got eaten (and re-typed the thing).
jhmorse says
The need for numbers depends on actual context. Tough to tell viewing this on my G1, but this looks like a magazine article, in which case, I don’t need numbers. The objective is to show comparisons, and these charts do that well. Only assumption is that there is no skewing of numbers here through non-zero based axes or logarithmic axes.
Parts I don’t like are using the red “China” and white “United States” as the legend. I had to look around and figure out what I was looking at. Also, those two colors together make me think of Japan (picture the flag).The split of the countries at the “agriculture/industry/services” layer is also a break in pattern, which slows down the comparison.