“Data + Other data + People = Readable story” - Himanshu Ojha at Hacks/Hackers London
Hacks/Hackers London returned this week, with an evening of talks about financial and data journalism. I’ve already posted notes from Marianne Bouchart’s and Sam Arnold-Forster’s talks. Next up was Himanshu Ojha, who like Sam Arnold-Forster, works at Thomson Reuters. He was talking about the story behind “The Unequal State of America”.
“Data + Other data + People = Readable story” - Himanshu Ojha
Himanshu described himself as “a journalist who had come to data”, rather than someone who had always done data journalism. His major project for the last few months had been a Reuters series entitled “The Unequal State of America”. Himanshu said he worked on it for several months “up until it was released.”
I rather liked that “released” metaphor for features, almost like “releasing” an album or DVD. It also echoed in my mind the way Bobbie Johnson talked about Matter at news:rewired, treating each story like an individual project.
Himanshu Ojha said that framing the “Unequal State” series had been hard — they knew they wanted it to be about inequality and data-driven, but trying to discover new angles had taken some time. He spent a little time talking about the behind-the-scenes tools he had used, which was actually the less-than-glamorous workhorse of Excel. “Lots of large Excel spreadsheets” as Himanshu put it. In the end, with over 50 million rows of data, they also employed SQL Server and SPSS, although Himanshu was quite clear that the coding part of the project wasn’t his — he was the story-teller.
He had a rather lovely equation for data journalism. One piece of data makes a dinner party anecdote. Two data sources combined can make a story. But it is “Data + Other data + People = Readable story”. I do worry that in the rush to have trendy infographics and “me too” data journalism, people are tempted to make visualisations out of any data they can get their hands on, rather than focus on telling a good story.
Several years ago at one of our London IA meet-ups, Max Gadney talked about how designers can easily fall into this trap too. Often having to do work at the bidding of someone else’s business requirements, when left to their own devices, he argued, designers could latch on to data as something they could “design with”, without building a compelling narrative.
There were three key things that Himanshu wanted people to take away from the talk. He argued that “getting lost in the weeds” where you end up obsessing over the minutiae of the detail of a data journalism story is a necessary part of the process, but only for a short period of time. Treat it like “a holiday fling” — brief and intense. He said that during this project he had even been dreaming about “margins of error”.
Secondly, he suggested befriending experts in the field you are telling the story about, in his case sociologists and academics. Show them your work before you publish it. You will almost certainly have made mistakes, since in a data-driven story “there are so many moving parts.”
Finally, allow more time than you think you’ll need for checking, and re-checking. The “dirty secret” of data journalism, he said, is that “if you want to do a good long piece, you’ve got to spend a long time on it.”
Another organisation being represented at Hacks/Hacks London was the FT. Next up I’ll have my notes from the talk by Emily Cadman & Martin Stabe…