This three-part series moderated by Brant Houston of the University of Illinois walks through a data-driven story, from interviewing the data to proving the story to your editor to telling it to your audience.
Using a database (available from the NICAR data library) of government-backed loans from the Small Business Administration, a group of reporters and editors around the country worked together remotely in the days leading up to NICAR. In addition to some insights and promising questions about the data itself, they pulled together lessons and documented their process for reporting out a data-driven story.
Part 1: The approach
Speakers: Jennifer LaFleur, ProPublica and David Donald, The Center for Public Integrity
Begin with analysis of the database, treating it like a story tip from any other source. You will interview the database, but before you get there, do your homework: Study the tipsheets and previous stories available in the IRE Resource Center; talk to experts inside and outside newsroom, check out inspector general and government accountability reports for leads.
If you’re getting the data from the NICAR database library, some of this will already be part of the data package; you will want to pay special attention to the “known problems with data” — something to think of less as a weakness than an inevitability.
The Power Point that will be available for this session gives a great checklist of “interviewing questions” to ask of the data and methodology for analysis. But the important point to take away is that no data is perfect. “Bring as much healthy skepticism to data interviewing as you would to any other source,” Donald said.
Part 2: From where your editor sits
Speakers: Maud Beelman, the Dallas Morning News
Your editor may be psycho, Houston admitted, but it’s less likely a function of personality than job description. The important thing is to understand what is important to your editor and why. The more you anticipate this, the more successful your pitch will be.
One of the most important parts of your editor’s job, Maud underscored, is to allocate resources smartly. For this reason, it is essential for you to get a handle on the minimum and maximum story you can get out of your data — and the timeline and resources required to achieve it.
Maud also explained that a lot of editors may seem resistant to data-driven stories, but it’s likely because they don’t know how to bulletproof data reporting as well as they know how to bulletproof standard reporting. She shared her list of questions for reporters, which she has developed over time with the help of other data editors:
As taxing as it is, she said, she requires reporters to re-import their data from scratch and re-run the entire analysis before publication. It is also essential to share your results with your targets for them to put it through the ringer.
Part 3: Big/Medium/Large
Cheryl Phillips, Data Enterprise Editor at Seattle Times, Tony DeBarros, formerly USA Today database editor, now Gannett Digital
Depending on the results of your data analysis, you may have to decide whether to make a big, medium or little splash with your reporting. A rule of thumb about the types of visual tools most suitable for the different sizes is:
Small: employing templates and tools — usable by many
Medium: more advanced tools — some coding, design skills required
Large: You’ll need a developer and a designer
But before you make plans on how to go from reporting to publication, you need to really size up what you’ve got:
Once you’ve determined the size of your story, you’ll find a wealth of tools (which also come in big, medium and little sizes and costs) discussed in other sessions during this conference.