I was alerted today by Alberto Cairo to a piece by Kaiser Fung of Junkcharts talking about (the state of) datavis teaching.
Kaiser cites an example that a reader sent in, from a recent course she had attended, on data visualization for academics.

This example was held out as an example of good work and the quote that stood out for junkcharts and indeed me, was…
“always try to find a graphic that relates to your subject, like the bullets here representing military spending, and use it in the chart.”
I have spoken to academics and researchers about graphics for the scientist and non-scientist, from my perspective, and have spoken about giving the reader something tangible to relate to. In this case I feel the data has been lost in the execution.
So, is this really a data visualization or just some pictures with numbers printed next to them?
Even if the figures were to be represented by the height of the bullets, it is still portrayed incorrectly. I quickly charted the data to see what it looked like…

so quite different from the original. Junkcharts went even further and looked at some more meaningful metric. A scatterplot plotting GPD per head against Military spending per head.

What this shows is the importance of knowing your data and knowing what you are wanting to show – and more importantly don’t forget the data, its more important than the imagery.
I’m looking forward to seeing what else was sent in