By Jasper Jackson
Data helps journalists paint a more compelling and complete picture – but only if they can interpret and present that data effectively. That was the message from journalists with extensive experience of the benefits, challenges and pitfalls of data and journalism at the Frontline on Wednesday.
If you couldn't make the event, you can watch the whole thing here:
The Guardian's Datablog editor Simon Rogers dismissed accusations that using data isn't proper journalism: “None of the stuff we talk about is particularly new, it's the way we do it and the tools we have that are.”
He described how Guardian readers helped analyse thousands of documents on MPs' expenses and how the paper has gradually developed new ways to visualise data, such as the 92,000 documents on the Iraq and Afganistan wars released by Wikileaks.
Times programmer and editorial developer Julian Burgess showed how computer code can be used to present data as diverse as the body mass index of Playboy playmates and pager messages sent from the World Trade Center on 11 September 2001. He recommended free tools such as Google Docs or MySQL that can be used to work data into stories.
These tools can be used by journalists to tackle data themselves, but programming is still reliant on the quality of the data, he said. "Things are often very easy with programming, or impossible practically."
David McCandless, writer, designer and author of the Information Is Beautiful blog, emphasised the use of design in moulding data to tell a story. Well-designed visualisations can not only present data appealingly, but can also highlight interesting links that might go unnoticed, he said.
As an example, he showed how a simple graph showing peaks in news coverage of violent video games coincides with the anniversary of the Columbine school shootings.
Michael Blastland, a freelance journalist and creator of BBC Radio 4's More or less programme, closed the session on a cautionary note by highlighting the dangers of not interrogating official data or questioning its creation. He cited the use of a single test result - in a sample of 1,500 - as the basis for estimating that three million people in the UK were affected by the Winter vomiting norovirus bug. He also described how valuable sources of data such as the Office for National Statistics often involve huge margins of error.
Finding your data is tough, knowing what kind of data you are finding is even harder. But I think you do need to know what data does, how it behaves and how it misbehaves in order to start making sense of it and start doing some of the wonderful things that my colleagues have been showing us how to do.
This event was part of our On The Media series, in association with the BBC College of Journalism. The next in the series is on 12 October and asks what the future holds for TV journalism in a connected age. More details and ticket booking here.