How To Boost Data Literacy in Journalism
This year’s MIT-Knight Center for Civic Media Conference, “The Story & the Algorithm,” showcased how data innovations are changing storytelling, burgeoning models for seeding that innovation and, of course, a whole lot of apps. In addition to exposing me to fascinating ideas for engaging new audiences in data, it also gave me, as deputy director of the Knight International Journalism Fellowships, a chance to connect with this extraordinary community of data thinkers, doers and disrupters that the Knight Foundation is building.
It didn’t take long for me to tweet: “Gotta nerd out for a second: I love that journalists are having these conversations about data. Too little of this in journ. #civicmedia.” [This Storify aggregates all of my tweets from the conference.]
A lot of people might think “data journalism” is the flavor of the week or that wrestling with data is not new for journalism. Based on what I’ve seen going on around the world – and heard from our Knight International Fellows – there is much room to boost the data literacy of journalists to improve the flow of quality information. Far too many journalists are at the mercy of statistical analysis by “experts” who may have an agenda. As a result of low skill levels with data, journalists also may inadvertently mislead the public or simply miss the real story. Furthermore, we are still just scratching the surface of what is possible in data-tech innovation for everyday users, if this conference is any indication. I can’t help but be excited that journalism conversations are increasingly dominated by talk about data.
The panel, "Turning Data into Narrative," illustrated this, raising many of the important concerns for journalists using data. For instance, Jonathan Stray, who leads the Knight News Challenge-winning Overview Project, asked what can be done with analytical results that don’t make seem to make sense. He cited an example in which his students found data suggesting that bicycle theft decreases as bicycle commuting increases. Possible next steps could include examining the data over a longer time period, comparing the data to other data sets and looking at the methodology behind the data. He was concerned that it took his students ten minutes to think of actually asking the people who had collected the data. Meanwhile, Dan O'Neill of Smart Chicago Collaborative noted the example that airports that do a better job of collecting data on bird strikes on aircraft were misconstrued by media as having more incidents of bird strikes – simply because they were tracking that problem. In fact, other airports that were not collecting data may have had more strikes.
For most people at this conference, data-literacy concepts like Stray’s and O’Neill’s are probably basic. Certainly, most know that data and statistics can be subjective, that depending on how it is analyzed data can tell the wrong story and that sampling populations means making assumptions. However, these issues are seldom explained clearly alongside numbers and infographics to ensure that every day audiences understand them.
Had this been my conference to organize, I would have emphasized mainstreaming data literacy among journalists and audiences. If data journalism is to serve the public interest, we need to keep having these conversations as a way to mainstream data literacy. This is what is truly needed to push data to the forefront of journalism and drive audience demand for data-driven storytelling.
In fact, something Dan Sinker, who heads the Knight-Mozilla OpenNews project, said took my thinking about this issue one step further. He pointed out at the end of the Ignite Talks, a project show-and-tell, that conferences can be a waste of talented people’s time. “Instead, what we should be doing is plucking those [people] out of the conference, giving them a time and a place to make something amazing, which is exactly what we did this weekend with the Knight-Mozilla-MIT ‘Story and Algorithm’ Hack Weekend.”
With all of these brilliant data innovators in the same room, it might have also made sense to host smaller, unconference-style ideations for attendees to interact and brainstorm. Imagine the innovation we could accomplish from that, especially in driving audience demand for data-driven news and taking data literacy to the far reaches of the globe.