Jonathan Barnbrook and I probably have a lot in common. His studio recently produced The Little Book of Shocking Global Facts, a collection of graphic illustrations based on the topics of world trade, the environment, human rights, arms trade and other issues of our time. For Barnbrook, it’s clear that the powerful statistics he’s selected are a condensed vehicle for strong statements in his graphic experiments. In many ways this is what attracted me to the field of information visualisation. The world of data presents a designer with rich fodder with which to write, edit and chart, but bears an added responsibility of authorship.
The Little Book of Shocking Facts eagerly adopts the visual language of information graphics but eliminates much of the procedural rigour that it demands. While the book has received favourable press, the professionals who work with data narratives have felt compelled to point out the shortcomings and dangers of such a casual approach to our profession. In fact, the Visualising Data blog (visualisingdata.com) wonders in a recent post whether a pie chart included in the book is the “worst graph design ever”.
Even if you haven’t been paying attention, the upsurge in output of information graphics over the past few years is evident. From graphjam.com’s pie charts of rap lyrics to the Nike+ visualisations of running data, these methods are being applied all around us to communicate an abundance of numerical stories. As with physical products, our expectation leads us to believe that the craft of the object is an extension of the quality of the materials. With information visualisation this translates into a linear relationship between visual polish and data rigour. Between the extremes of the index card charts (like those at thisisindexed.com) and the refined interactive displays of The New York Times, stretches a spectrum of assorted quality of content and execution that can be difficult to evaluate.
This scares me
In isolation, Barnbrook’s World Bank pie chart does not concern me. Its disdain for legibility makes it almost impossible to glean any information from the graphic, and any message is undermined by this lack of concern with communication. A treatment like this is completely at odds with the goals of information design and the needs of the content it contains; stylistic noise scrambling the signal.
The page that scares me is a map that appears elsewhere in the book, titled Top 20 Carbon Emitting Countries. This map bears all the trappings of rational information design: clear typography, good contrast and a geometric minimalism of forms. To the casual observer, it communicates a damning image of the US and China producing what seems to be around 80% of the carbon emissions from the top 20 countries. Had the quantities of emissions been included in the graphic, a viewer might be able to scrutinise the visualisation, but without data presented alongside the image one must turn to the index of sources at the back of the book for a link to the International Energy Annual’s list of carbon dioxide emissions by country (1980-2006). While the US and China emit an inordinate amount of CO2, their actual share is closer to 52% of that of the top 20 countries.
What went wrong?
Where did this graphic go wrong and why has it sparked another firestorm of criticism? At its heart is either a misuse or a misunderstanding of how scaled shapes are used to represent quantitative information. When geometric primitives like triangles, squares or circles are used to convey quantitative information, the area of the shape is used to represent relative amounts. Rather than showing ‘20’ by creating a circle with a diameter of 20, one needs to do a few calculations to determine that a circle with an area of 20 has a diameter of 5.05. I fear that this error may be repeated elsewhere in Barnbrook’s book alongside more subtle breaches of information graphics standard practice.
I don’t worry too much about the US and China being made CO2 scapegoats, what does concern me is the proliferation of lax visualisations and limp visual fact checking. Unlike an image on the internet, materials, production and transportation costs all increase the verisimilitude of a book like this and stand in the way of its correction. If obvious errors like this are making their way past the team responsible for publishing the Shocking Facts book, what is the hope for any credibility amongst the myriad jpeg information graphics in circulation today? Without more astute creators and consumers, these abuses are destined to continue.
I see this concern mirrored and addressed in a recent post by Nathan Yau of flowingdata.com titled 7 Basic Rules for Making Charts and Graphs. His fifth point “Keep your geometry in check” and sixth point “Include your sources” are particularly relevant advice for the makers of infovisuals. I’ve been encouraged recently by an increase in attribution amongst the casual infographics afloat on the web, and encourage readers to not mistake aesthetic polish for information legitimacy as more and more of the world is translated into data visualisations.
Fortunately, I am finding signs of progress. There is a special species of internet-only information graphic that for lack of a better name I term ‘tall infographics’. These images are characterised by spanning several screens in length. Recent examples I’ve noticed include The Facts About Bottled Water, Sextoys and Cyborgs on the Screen. Most of this breed are missing the visual refinements we equate with data rigour, making them less convincing and serious, but in the last few months I’ve noticed one new improvement. Each of the graphics mentioned now cites their sources with footnotes.
Nicholas Felton is a New York-based information designer. See feltron.com
In answer to Felton’s criticisms, Jon Abbott of Barnbrook Design responds: “The graphic may have been created using diameter rather than area, which has been heavily criticised in some information graphic circles. However, the circles are colour coded and the URL to download the figures is in the book. We understand the criticism directed at this graphic but are disappointed that it has received such disproportionate attention. The graphic is part of a book which is not, and never attempted to be, a compen-dium of information graphics, but is a brightly illustrated book which includes illustration, photography and playful typography.”