Is your chart a detective story or a police report?

Wired? - I did not expect this to be article to be good, and it turns out to be very worthwhile. The article is by Andrew Gelman, an authority on statistical modeling, and Jessica Hullman, an authority on statistical visualization. In my words, they make the point that a data visualization is a kind of a model, and as such the foremost job of the viz is to frame a judgment (a specific comparison) by the viewer.

In their words: "This connects back to our idea of graphs as comparisons. The viewing, and hence the construction, of any graph is enhanced by an understanding of the comparison that it represents. As scientists, we can make better graphs if we go beyond the idea of “displaying data” to consider the model we’d like to present, including the reference to which we want the data to be implicitly compared. As consumers of scientific information, we can better read an infographic by being aware that its function is to tell a story through comparisons, which unfold through our own active participation."