The power of the puzzlegraph

January 12, 2013

(This article was originally published at Statistical Modeling, Causal Inference, and Social Science, and syndicated at StatsBlogs.)

The Organisation for Economic Co-operation and Development reports that the following project from Krisztina Szucs and Mate Cziner has won their visualization challenge, “launched in September 2012 to solicit visualisations based on the OECD’s data-rich Education at a Glance report”:


(The graph is interactive. Click on the above image and click again to see the full version.)

From the press release:

Entries from around the world focused on data related to the economic costs and return on investment in education . . . [The winning entry] takes a detailed look at public vs. private and men vs. women for selected countries . . .

The judges were particularly impressed by the angled slope format of the visualisation, which encourages comparison between the upper-secondary and tertiary benefits of education. Szucs and Cziner were also lauded for their striking visual design, which draws users into exploring their piece [emphasis added].

I used boldface to highlight a point that Antony Unwin and I have been making recently, which is that the obscurity and difficulty of certain infographs can, paradoxically, make them appealing and powerful by first sucking you (the reader) in and then giving you a puzzle to chew on: what exactly do those numbers mean? Once you’ve figured that out, you’re deep inside the subject.

To be frank, I don’t think many people will actually learn much at all about education or education statistics from playing with the Szucs and Cziner visualization, but it does a good job of selling the topic in the sense of making the numbers look potentially interesting and surprising.

More here from Patrick Love at the OECD, who writes:

They chose Krisztina and Maté’s graph from over 30 entries because it “does a great job of breaking down the complex interplay between costs and returns into a form that is easy to compare”. And also because “instead of the many-country approach used by most entries, the project takes a detailed look at public vs. private and men vs. women for three selected countries (which you can change)”.

Huh—only three selected countries??? That I can’t see the benefit of, given that the graph is interactive.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science