People keep asking what the difference is between data visualization and infographics. Since I’m not completely satisfied with the available answers I thought I could return to the subject and write my own.

First, you have to recognize  that you can’t compare them because they are not at the same level. Things get much simpler if we assume that data visualization is an umbrella concept that means some kind of visual transformation of some underlying data. This transformation takes advantage of the human perception to reduce cognitive overload. That’s a fact, not a goal.

You can’t say much more. The goals, the tools, the design, the visual objects, the interaction model, are all irrelevant to define what data visualization is. But, like all umbrella concepts, it groups several other concepts where these dimensions become relevant: infographics, business visualization, visual statistics, data art. And those are the concepts we should compare.

Let’s look at one of the dimensions that more clearly define each group: the conscious use of design and aesthetics.

datavis-authors

On the left there is a group of visual statisticians like Cleveland and Tukey, and I’ll add Bertin to the mix (yes, I know he was a cartographer). They are not aware, don’t care or simply don’t pay much attention to graphic design. Their designs reflect that.

When you walk towards the center you’ll find a large group of functionalists. They are aware of the functional aspects of design and they want to use it that way (for example, they love pre-attentive processing). It is a heterogeneous group, and the more you walk to the right, the more important  aesthetics become. The first author you’ll encounter will be Stephen Few and the last one will be Alberto Cairo. Tufte is also here, you’ll see him before meeting A. Cairo. You’ll find me here too, between Few and Tufte.

Say goodbye to Alberto Cairo and say hello to the designers, where function becomes decoration (that’s not necessarily bad). They like to play with the data but they don’t take it too seriously.  This means that they’ll happily sacrifice some degree of effectiveness to original, aesthetically pleasing designs, but they will not call them “art”. They will also try to convince you that data visualization is what they do. You’ll find David Mccandless here, obviously, and a multitude of smaller characters, including all those default-loving Excel users.

Please don’t get me wrong: David Mccandless and the Excel users look very similar from this point of view because this is a single dimension. It’s obvious that if we add a second dimension (“originality”) he’ll move away from them (and they will remain in the “canned effects” group).

And then on the far right you’ll find the last group, the artists (I like several of them, can’t choose one). Decoration becomes art and effectiveness doesn’t make sense. Data is just another object the artist plays with to find Beauty. Flowing Data has a good list of artistic visualization projects.

Stephen Few wrote about this and he says:

… the goal that data be visualized in a way that leads to understanding. Whatever else it does, it must inform. If we accept this as fundamental to the definition of data visualization…

He also wrote in a different post:

Therefore, a data visualization should only be beautiful when beauty can promote understanding in some way without undermining it in another. Is beauty sometimes useful? Certainly. Is beauty always useful? Certainly not.

So, Few believes that if, and only if, a visualization  is made to inform and help understanding then it is data visualization. I agree with that if, and only if, we define data visualization by its purpose.

I prefer a definition that focus on the “what” and lets subgroups of users to define their own “what for”. Data visualization is a federation, not an empire.

What about you? Where can I find you in the diagram above?

 PS: Designers, please make up your mind.