By Glynn A. Hill
Four journalists discussed tips for reporting out a data story at the 2015 CAR Conference in Atlanta. Steve Myers, visiting professor at Texas Christian University and special projects editor at The Lens, facilitated the panel. Speakers included James Ball, a special projects editor at The Guardian; Andy Lehren, a reporter for The New York Times; and Kendall Taggart, a reporter on the investigative team at BuzzFeed News.
So you have that data set. Now what? The next step, they said, is understanding what you have, which demands the following:
Lehren talked about the potential pitfalls of data. He suggested determining its internal and external validity. The former involves looking inside the data set and asking what makes sense. Do codes match? Is the basic information accurate?
The latter involves asking whether the Government Accountability Office or top researchers have ever looked at the data set. He said you should think outside the box and find experts, reports, and documents that support the information. It’s important to cross-check the data.
Lehren suggested looking to any of the following for corroboration:
Furthermore, you should check the data against other public sources. Look for rates in a data set and figure out how the math was done, keeping in mind the world at large and how things look more broadly. To corroborate data against published sources, you should examine the following:
Glynn A. Hill is a 2015 CAR Conference Knight Scholar and budding multimedia journalist from Howard University in Washington, D.C. He's worked with The New York Times Student Journalism Institute and interned for the USA TODAY, and most recently, The Washington Informer.