Data visualisation: an aid to understanding?

The subject of graphic design rarely makes the national media in the UK, so it was with some surprise that we noted a segment on the BBC’s Newsnight programme last night. Particularly as its subject was the previously arcane world of data visualisation.

After a filmed report on the emerging influence of data-driven imagery – both practical and artistic – presenter Kirsty Wark quizzed Information is Beautiful‘s David McCandless and ‘legendary designer’ Neville Brody on the potential and pitfalls of attempting to extract meaning from data by creating arresting visuals (watch the debate here, starting at 26.05 minutes in).

That Newsnight was interested at all in the subject marks something of a high point for an area of practice that has become increasingly prevalent and controversial. As the debate between McCandless and Brody revealed, data visualisation has both its advocates and its sceptics.

 

 

The advocates maintain that we are swamped by data. New methods, they argue, are required to make sense of that data. By using them, we can tell compelling stories about our changing world that will engage people more powerfully than words alone and that, therefore, may act as effective calls to action. Which all sounds reasonable enough. But some don’t stop there. Data visualisation, it is also sometimes contended, will actually aid our understanding of an ever more complex world.

 

 

 

Detail and full image of Left Vs Right by David McCandless and Stefanie Posavec (more here)

 

Here, things are confused slightly by the presence of a splinter group of artists using data as a source to create generative works. For example, Stockspace by Marius Watz, which was displayed as part of the Decode show at the V&A last year, drew on activity on the German stockmarket to create abstract geometric forms. Beautiful, certainly. Informative, less so.

 

 

The sceptics, as represented last night by Brody, argue that what in fact data visualisation is doing is creating a lot of pretty pictures that do nothing more than please the eye. At one point in the discussion Brody dismisses one of McCandless’s pieces as something he might like to hang on his wall – the implication being that this is all it is good for.

 

 

A beautiful geometric pattern may well be visually seductive, but is it any more successful in getting over the facts or arguing a case than bald, unadorned figures? Data visualisation’s adherents argue that a certain amount of visual flair is necessary in order not just to explain complex data but also to encourage engagement with it. The question is at what point does the pursuit of the latter undermine the former?

 

An image from The Little Book Of Shocking Global Facts, described as “possibly the worst graph I have ever seen” by information designer Andy Kirk on his Visualising Data site

 

Further tension surfaced recently in debate over Barnbrook Design’s Little Book of Shocking Global Facts. Information designers accused Barnbrook of ignoring the basic rules of the craft to produce visually arresting imagery that was confusing at best and deceptive at worst in the way that it interpreted data visually (Nicholas Felton will address this in a piece for the September issue of CR, see also here and here). A counter argument, as put forward by MLA in a comment on our original piece on the book, is that “while the pages do use some aspects of information design, these aspects are primarily employed as graphic devices and not as objective, statistical referents that convey quantifiable information. In that sense they are much closer to the subjective quality of illustrations, and in particular the satirical cartoon, that conveys facts as politically charged struggles.” In other words, the book is not a neutral presentation of data but a polemic that employs some of the language of data visualisation to make its highly-illustrated arguments. (See this interview with Barnbrook’s Jon Abbott on the ideas behind the book).

This I think approaches the core of the issues surrounding data visualisation. By deploying the visual language of science it suggests a cool, rational neutrality. Here are the facts, it says, QED.

A core of information designers believes that the authority of data visualisation rests on its rigour in adhering to the accurate presentation of data.

But it’s not as simple as that. As Brody points out in the Newsnight interview, information is most often deployed in a political cause. Just as statistics are routinely manipulated in order to fit the argument, so inevitably will be charts, graphs and their ilk. The Barnbrook book takes the language of data visualisation and runs riot with it. It is presented as ‘fact’. The sources for those ‘facts’ are listed extensively in the back. And yet the way in which those ‘facts’ are given visual form is highly subjective, in some cases twisting them in the service of its agenda.

 

Graphic from The Little Book of Shocking Global Facts purporting to show relative carbom emmissions by country. Felton and others have accused it of representing data in a misleading way

 

In the way that it is employed, data visualisation is no more inherently neutral than any other form of statistical analysis. It should come as a surprise to no-one that its tools will be used to make politically-charged and politically-biased arguments. Graphics that purport to make a statement of fact should be approached with as much caution as a government press release.

This, then, is where the great hope of some data visualisers runs into the buffers of reality. Yes, graphical invention can be used to explain complex ideas and present detailed data in digestible form in the cause of an argument or political position, but this will not necessarily aid understanding. As newspapers have known for decades, a graph is just another way of telling a story. But whose story?

 

No Responses to “Data visualisation: an aid to understanding?”

Post a Comment