The "message" or the "story" of an infographic is an often hotly debated topic in the world of information design. There are those who believe every chart should tell a story, and then there are those who are strictly against the idea of defining a narrative for the audience. I can certainly see the value in both, the former can guide the viewer through a complex set of data by highlighting key points, and the latter allows freer exploration. Both sides however will no doubt agree that no matter what form the message takes, the data always comes first.
Putting the Blinkers On
But what about starting with the message first and working backwards? One designer interviewed over on Visual Loop discusses his method for creating infographics:
A worrisome thought process indeed. Putting the blinkers on and and only finding evidence to support your predefined message or narrative is a sure-fire way to completely mislead your audience (even if you have good intentions).
Take this wonderful example from io9. If my predefined message or hypothesis was that "an increase in organic food sales causes autism" I could very much create a chart that appeared to show this. Now nothing on the chart below is technically incorrect, indeed, both organic food sales and cases of autism have increased over the past decade but correlation and proximity should not necessarily imply causation.
Yet by cherry-picking data and presenting it in this way I can illustrate my intended message to an audience and as if by magic I appear to have proven my own hypothesis. As io9 puts it:
It's very easy as an information designer to mislead the audience via visual means, and there unfortunately is a startling number of examples out there (I'm looking at you Fox News).
By starting with a predetermined hypothesis, you're misleading your audience, whether intentionally or not. Infographics should be about conveying the truth and finding the meaning within a set of data, and not about picking and choosing the data that simply supports your message or claim. I'm all for using data from different sources and combining them into one graphic to build up a narrative, to provide some back story and additional context. It's dishonest however, to leave out those pieces of data that don't quite fit with your intended message. Unfortunately this cherry-picking is all too common.
I'll sign off with Alberto Cairo's thoughts on the matter: