“Data journalism is rather like sex at university: lots of people are talking about it but few are doing it, and fewer are doing it well.” So said Neil McIntosh, deputy editor of the Wall Street Journal, at Wednesday’s (May 16, 2012) AOP forum on data journalism, held in London.

McIntosh, who chaired the event which brought together both those already immersed in data journalism and people eager to learn more about it, added: “I suspect many of us still think there’s some mystery about data journalism.”

Among the speakers at the AOP forum was Alex Graul, interactive developer at the Guardian who is working on its infographic Miso project.

Graul outlined this structure of how data journalism should work, from start to finish:

  • Find the data
  • Clean the data
  • Visualise the data
  • Define the stories
  • Visualise the stories
McIntosh interjected at this point to say he thought that in Graul’s time frame, “defining stories is quite late in the process”, to which Graul replied that sometimes after delving into the data, it turned out that there were no stories there after all, just like in traditional journalism – this was just the way things go.
Next up was Kevin Anderson, digital strategist at the Media Development Loan Fund, who said that among the main drivers behind the “data journalism revolution” were better and easier tools and being able to crowdsource data.

He went onto outline immensely useful tools for visualising and sharing the data, such as Google Fusion Tables, and good sources of official data – his presentation is now available online if you want to find out more.

Another speaker, interactive producer Martin Stabe from the Financial Times, mentioned in his talk how approaches to Freedom of Information (FOI) requests have evolved.
Early FOIs, he said, might have focused on how much a department spent on biscuits, for example. But looking at it from more of a data journalism perspective, the new approach would be to ask for all the data sets, rather than individual records – in other words, if you have all the data around a certain topic, it gives you the opportunity to find other stories in it which you wouldn’t be able to discover if you’d just asked for one or two figures.
However, he warned that cleaning the data – making it usable by putting it into the right format, for example, or sniffing out data that had originally been filled in incorrectly – was a big part of the process.
Also mentioned at the forum was the free Data Journalism Handbook, an excellent source of information, and there are more links to key examples discussed on the day in the AOP’s event summary.
I thought the AOP forum was full of valuable information and good examples of what may be achieved once you realise that data, far from being an impenetrable mass of numbers, can be another useful tool in the journalist’s arsenal.