Ryan Thornburg at NICAR: Why we need more data reporters

The integrity surrounding elections systems has become a hot topic and as a result, experienced data reporters are increasingly in demand. Ryan Thornburg wants to help build spaces where students, reporters and computer scientists can meet to learn from each other and fill that demand. Thornburg joined a panel called “Using data and records to investigate voting” at the Computer-Assisted Reporting conference (CAR) on March 10. Other panelists included Jessica Huseman, a reporter for ProPublica, and Doug Moore, a reporter with the St. Louis Post-Dispatch.

Ryan Thornburg headshot

Ryan Thornburg, director of Reese News Lab and associate professor at UNC School of Media and Journalism

Thornburg is an associate professor in the UNC School of Media and Journalism and the director of the Reese New Lab, where he leads students in innovation and entrepreneurship through experimental projects. He also leads the Carolina Data Desk and NC-Votes, both aimed at using data and public records to expand and strengthen reporting tools for journalists.

The CAR conference is hosted in Chicago by Investigative Reporters & Editors (IRE) and The National Institute for Computer-Assisted Reporting (NICAR). Before the conference began, we caught up with Thornburg to talk about the use of data in journalism, what higher education institutions should be doing to produce better data journalists, and steps for smaller media organizations to use data in their reporting and research.

Your panel’s topic “using data and records to investigate voting” is well-timed with the 2018 midterm elections on the horizon. Have you seen more interest recently from journalists or the public in analyzing voting and elections?

Next to weather reports, coverage of voting and elections was probably the original data journalism. This is one topic in which people have always been interested in the numbers and the details. But I think the thing that has changed this year is all the discussion about the integrity of our elections systems. Some of those claims are utterly baseless, and some have been legitimate. The panel I’m moderating aims to help journalists know how to find the fact from the fiction.
Click here for the information presented and helpful links.

What are people talking about at CAR that’s unusual or special for 2018?

A big topic is going to be the upcoming 2020 census. The census provides so much important context to so many news reports. It’s critical that the census is done well and that journalists understand how to use it. I think there’s some real concern about the quality and focus of the upcoming census. And there is also a new census website coming online that will change the way many journalists have been accustomed to getting their census data.

Information security, data integrity, machine learning and artificial intelligence I think are also some of the interesting topics of conversation this year. The panels at this conference really do a good job diffusing the hype while also getting reporters excited about how they can use these things to find and tell hidden stories that are important for people to know.

Photo by Markus Spiske on Unsplash

Photo by Markus Spiske on Unsplash

How have data collection and reporting with data changed in the last year?

Boy, that’s a hard question. I feel like maybe I’m a frog in a slowly boiling pot of water — maybe I can’t see the changes because the most important ones tend to happen slowly. What hasn’t changed is the incredible demand for more journalists with these skills. The job board at this conference this year is just as big as it was last year. And there are hundreds more newsrooms that need these reporters but can’t afford the salaries they are demanding. If the next year sees more UNC students working in the field, and NC Votes and Carolina Data Desk filling the gap for the communities in which data isn’t playing an important role in igniting the public conversation, then that’s the kind of change I’m going to get excited about.

Do you have a sense that journalism schools are preparing students to understand how to read, source and use data properly in their reporting?

A couple of years ago at this conference, some folks released a well-researched report about the state of data journalism education. What it found was that the best journalism programs had done a great job incorporating interactive and infographic design into their curriculum, but that very, very few provided sufficient opportunity for students to learn how to become world-class data reporters. And not much has changed since that report came out.

Even at UNC, we have a long way to go before we can honestly call ourselves a leader in data reporting curriculum. That said, being at this conference makes me realize how lucky I am to be associated with the school. The top award in the field — which is given at this conference — is named after former UNC journalism professor Phil Meyer. And many of my former students are now leading their own panels here at the conference.

What really makes me proud is that these students are presenting on topics that they never had the chance to learn in class. They did what I tell all my students they will have to do — be lifelong learners. Classes are able to pique students’ interests, but they have to stay hungry to learn more and find newsrooms that give them space to learn. Because guess what? That’s what I did. I’m sitting here teaching a class in data journalism and I never took a college class in it.

That’s why conferences like this are so important for undergraduates. Thanks to my professorship fund from the school, all the students in my class this semester are members of Investigative Reporters and Editors and have access to all the organization’s online tip sheets. And because of the school’s funding from the Triad Foundation and the Don and Barbara Curtis Fund for Extracurricular Activities our students have been able to come to conferences like this one and learn so much, so quickly.

I’m also excited that my classes are starting to draw students from outside the School of Media and Journalism. I’ve had students from statistics, information and library science, public health and computer science find their way to my class because it gives them an opportunity to learn how to apply the skills they are learning in their home departments in a mission they feel is important — the journalistic mission of holding powerful people accountable, shining light in dark places and explaining the increasingly complex and interconnected world we all share.

For local media organizations with limited resources, can you give an example of how they can experiment with data via collaborations with higher education institutions and nonprofits?

Absolutely. I think there are two good ways — first, find stories you’d like to replicate that used data analysis. The most efficient way to use that is to search the curated list of stories on the website of Investigative Reporters and Editors. Bring me that list and I’ll help you get a sense of how you might do versions of those stories locally. Second, would be to bring me a data set in which you think there might be a story. Or a story you’ve done that you think could have benefited from some more data analysis.

And, if you just want to poke around with some great, free, journalism-focused exercises check out the great site by MaryJoWebster.

How have data collection and reporting with data changed in the last year?

Boy, that’s a hard question. I feel like maybe I’m a frog in a slowly boiling pot of water — maybe I can’t see the changes because the most important ones tend to happen slowly. What hasn’t changed is the incredible demand for more journalists with these skills. The job board at this conference this year is just as big as it was last year. And there are hundreds more newsrooms that need these reporters but can’t afford the salaries they are demanding. If the next year sees more UNC students working in the field, and NC Votes and Carolina Data Desk filling the gap for the communities in which data isn’t playing an important role in igniting the public conversation, then that’s the kind of change I’m going to get excited about.

If anyone reading this is interested in strengthening their data skills, fostering transparency at local levels, or learning how to use products like these, what opportunities are available?

Well, right now, folks can check out the #NICAR18 hashtag on Twitter to drink from the firehose of information that this conference presents.

Students at Carolina can get in touch with me so they can join some of the extracurricular opportunities, like the upcoming hackathon we are co-hosting with the Carolina Analytics and Data Science club.

Journalists in the field should keep coming to me with questions they have, or specific data they need help acquiring, cleaning and organizing. The NC-Votes project is going to ramp up as we get closer to midterm elections with a tool to allow reporters to easily draw random samples from the state’s voting registration database so they can ensure they are getting representative voter voices into their stories. And Carolina Data Desk continues to provide support to newsrooms and freelance reporters who don’t have a data expert close at hand.

Follow Ryan Thornburg on Twitter at @rtburg.

Want to receive a monthly roundup of our news and research?