6 mistakes newspapers make with data journalism


Too much of what is claimed to be “data journalism” in today’s media is really just ego-driven “data porn” — pretty pictures created around numbers with no real reader value, according to an international “data guru” with strong journalism credentials.

Justin Arenstein, the chief strategist and Knight International Fellow at the International Center for Journalists in South Africa, told the World Editors Forum in Bangkok that news companies needed to stop trying to “get into” data journalism.

Instead, they should ask themselves how good storytelling can be aided by data.

“New tools don’t replace traditional journalism,” Arenstein told a captivated audience, as he took them through a three-hour data workshop that felt like three minutes.

Data journalism is “no longer just entertainment and no longer just voyeurism but creating decision-making tools based on news reporting.”

With shoulder-length, curly hair and a black roadie T-shirt with a QR code in the shape of Africa, Arenstein fits the visual stereotype of a hacker-nerd.

And he is. But while most “data journalists” have backgrounds in programming, Arenstein comes with a proud history and a string of awards for investigative journalism, which he achieved by running tiny news teams across Africa.

His passionate belief is that data journalism — and the understanding of the tools that are out there to crunch data, reveal connections, and illustrate it creatively — is one of the greatest tools in the modern journalist’s tool box.

And he now works regularly with Google, Mozilla, and Microsoft to spread the gospel.

This is my summary of the key mistakes that modern companies make — and how to avoid them — as highlighted by Arenstein’s presentation:

Myth No. 1: Data journalism is expensive and requires a lot of resources.

In these days of down-sizing, too many great data journalism ideas fall at the first hurdle of “we’re not allowed to hire anyone to do it.” But Arenstein points out this is just a failure to think creatively.

There are a host of grants and programmes designed to embed data experts into journalism teams internationally run by Google, Mozilla, and Knight International.

Open source code means you can build quite neat deliveries from as little as a few hundred dollars of “real money” and a week of staffing time (including your seconded data cruncher).

Myth No. 2: The best data journalism is around breaking news.

No, it’s not. The rise of data journalism ties into a hunger from readers to not simply be entertained or amused, but to “extract credible and actionable intelligence from the firehose of information” that is the daily media stream.

Arenstein advises to go for long-tail stories, not breaking news.

Ask yourself, “What is the pain that my readers are trying to solve in their lives?” and look at how data can play a part here.

These are invariably areas such as health, education, crime, government, property, etc. Build deep data sets around these topics and allow breaking stories to feed into them.

Myth No. 3: Our newsroom doesn’t have time to learn data journalism.

Nonsense, says Arenstein. Set your staff the task that they need to spend their first hour at work every day for two weeks in some self-managed training. Use free online video tutorials to improve their skills in anything from setting up data in Excel all the way through to using open source code — depending on their skill level.

Check out schoolofdata.org, which provides online learning modules around data.

Myth No. 4: We will need to get IT involved and get a capex to buy the software.

Have you worked it out yet? The biggest impediment to news companies doing better data journalism is thinking like newspaper men, when they should start thinking like hackers.

Check out Knight-Funded Ideas, which invests millions of dollars into funding new experiments around community and engagement, including journalism. Much of the code is open source and can be plug-and-played.

Myth No. 5: We don’t have any data.

Rubbish! So the government has told you they can’t release the figures. And you believe them? Now you need to start thinking like a journalist again — and find another way.

Search on Google for sources such as “open data” and “open institute” to discover what is freely available from data sources in your region.

Turn your articles into a database with tools such as DocumentCloud, which can help you scan data from PDFs. Start geo-tagging your stories and mine names and businesses for connections. Even your own subscription database can be a source of information.

Myth No. 6: We have to do everything ourselves.

“No individual media company needs to take responsibility for reinventing itself,” Arenstein says. “There is a growing international eco-system and community that media companies can start tapping into and a global network of people interested in building solutions and experimenting.”

What are we waiting for?

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.