Journalists, scientists, academics and data experts held a conference in UCD on July 6th and 7th to discuss data journalism, sharing success stories, research and new tools. Here are some of the takeaways from the two-day event, which invited stiff competition, accepting just 35% of applications.
Journalism needs data, journalists need data skills
Colin Scott, professor at UCD, said journalists need "significant" skills in handling data, especially where official sources - with their own agendas - are concerned.
Results from the Global Data Journalism Survey showed 86% of respondents considered themselves to be data journalists - but only 18% rated themselves as experts. The survey asked 206 journalists working with data in 43 countries to rate their own skills and education and give their thoughts on data.
90% of respondents said data adds rigor to journalism, and 70% said they couldn't do their work without it.
"Just Data Isn’t the Story" -Tomas Petricek from the Turing Institute
Tomas demoed The Gamma, an open-source tool which he hopes can bring transparency to data stories in a world that increasingly distrusts statistics. The tool would allow journalists to give their audience the full data set to play with along with a story: doing the heavy lifting in sorting the data for the audience, while letting them draw their own associations and conclusions with simple coding.
For @thegamma_net demos I showed at #datajconf see carbon emissions viz: https://t.co/MBjBLOfCuw and playground: https://t.co/ksDoRlss0g
— Tomas Petricek (@tomaspetricek) July 6, 2017
One journalist wondered whether that would remove all storytelling and emotion from the piece - but Petricek insisted his tool was a supplement to, not a replacement for story.
"That’s just the background, that every data story should have" he said.
Throughout the day, speakers emphasized that data can't just be numbers - it has to connect with people.
Great quote in presentation by @EvaConstantaras at #datajconf re. a starting point for #data stories. pic.twitter.com/2U41zNNjIT
— Kathryn Torney (@Kath_TheDetail) July 6, 2017
"Data is not the goal" – Conor Ryan, RTE Investigations Unit
Conor said journalists should look for leads in data, but not consider it the final product. Journalists must verify, follow leads, and investigate before publishing - particularly in Ireland, where defamation and data protection laws are strict.
Data journalism requires understanding law, policy, and deep understanding of data structures (don't settle for PDFs). #datajconf #OpenData
— Kalpana Shankar (@Dublin_kshankar) July 6, 2017
He also advised the audience to know the law, be persistent, and try and think of the bigger picture when making requests for data and try to get access for everyone.
Good data does not come easy - @conor_w_ryan #datajconf pic.twitter.com/ohl6jWED0a
— Martin Chorley 💾 (@martinjc) July 6, 2017
Google Trends "basically represents what they are wondering about, without any filter or agenda" - Clara Guibourg, Google News Lab
Clara showed how journalists can use Google Trends, "one of the largest publicly available data sets", to show people’s thoughts in real time.
"When you compare it with social data, and that’s really great... but you might be embarrassed typing (a question) into Twitter" she suggested, pointing out that one of the most popular questions around Teresa May was a slightly out-of-the-loop: "When was Teresa May elected?"
John Lennon once said The Beatles were bigger than Jesus. Google Trends says not #datajconf pic.twitter.com/EXPyGrjZDd
— Glyn Mottershead (@glynmottershead) July 6, 2017
The tool allows anyone to compare multiple searches over time - you can search for politicians, cities, search terms, even new slang words. For instance, the word "selfie" began to rise in 2013.
Google don’t release the search count - it’s indexed to 100, 100 being the highest point in the period you are searching within, 0 meaning the lowest point, which could be hundreds or thousands. So, it’s a tool for comparison and to gauge changes and trends in interest, and not quantity.
Clara says the data is more meaningful to viewers this way, and gives media a tool to reach out to the public on issues they care about.
"Be wary of new technologies" - Jennifer Start, University of Maryland
Jennifer audited Google's home page to look for bias in how top images are selected for display in searches. The lab-coat-scientist, as she describes herself, was cautious about coming to any conclusions, but said people should be careful- search engines and other tools are guided by underlying instructions we can't see.
"Unless you establish yourself as a non partisan mass media, you’re probably just preaching to the choir" – Eva Constantaras, Internews
Eva said there was a shortage of good data journalism during the US election - but pointed out some people were doing excellent work. Unfortunately, they weren't reaching across the party divide.
The NPR's fact checks of debates were great, said Eva, but what's the point if only liberals read them?
Next year's Data and Computational Journalism Conference will be held in Cardiff. More tweets and research from the event can be found on Twitter, #datajconf or on datajconf.com
