"I Read Where I Am. Exploring New Information Cultures" - читать интересную книгу автора (Various)

9. Highway Drugs and Data Visualization – Catalogtree

Experts do not need visualization to be able to read data. A musician can hear the music as he reads a score. The rest is at the mercy of graphics and audio.

A few years ago we were working on a project about the motorway between Arnhem and Nijmegen , the A325. We used a series of silk-screen prints to show various aspects of the motorway. Where did the most accidents occur? Where did people drive fastest? Which approach and exit roads were used most? We made use of data that the province collected 24 hours a day at various measurement points along the road. Three civil servants worked in this measuring department: one was responsible for traffic counts, using a computer system dating from 1977, one was responsible for speed measurements on a system dating from 1992 (not compatible with the former), and a third was the coordinator of the other two.

We kept in touch with them by telephone and e-mail and asked for interesting data. They were pleased to help. They went out in a mobile measuring unit to take measurements that they then sent us by mail. We received an enormous e-mail with tens of thousands of figures, separated by commas. A completely abstract jumble of figures that was explained in the subject line of the mail: ‘Look, a monster tailback!’

A frequent misconception in data visualization is that treating data subjectively is tantamount to lying. It is often said that data must be presented as objectively as possible, but to our minds that leads to an exact reproduction of the original data set. That is perhaps the best method if you’re dealing with the top ten best selling books, but if the visualization represents tens of thousands of pieces of data, that makes a good interpretation impossible. To be able to tell a story means that it is necessary to place the emphasis somewhere, to make an editorial choice, so that the reader understands what is being told without having to be an expert.

Another misconception is that a visualization must be clear at the first glance, otherwise it has failed. We once drew a map about drug transports between South and North America . It had to show which drug cartels were active (around twenty such cartels), what they were transporting (various types of drugs, weapons, money, people), how they were transporting them (various means by land, by sea, and through the air), which routes they took, and which alliances they had formed in order to work efficiently. A fairly complicated map of course and one which required some effort from the reader if it were to be fully fathomed. A data visualization can easily have a higher information density than the written word. Yet nobody glancing at a page covered with words expects that the content is immediately clear. You have to read it first.

Although those provincial civil servants didn’t need our visualizations to know what was happening on their roads, they nevertheless derived considerable pleasure from our silk-screen prints. They quickly replaced the art in the corridors which had been purchased by the province. ‘Well’, they said, ‘this is art that actually means something!’

Joris Maltha and Daniel Gross run the multidisciplinary design studio 'Catalogtree', specialized in making information visualizations.