News beyond numbers: Seeing stories in an age of data

By Inyoung Choi

In a world of “TL; DR” (too long, didn’t read), even the most captivating news stories must actively engage with readers to grab and maintain their attention. For instance, the Wall Street Journal’s coverage on the Greek economic downturn enabled readers to “scroll” through the recession through an embedded, adjustable timeline.

Stanford’s Computation + Journalism Symposium featured a panel, “Finding Story Ideas in Large Datasets,” where New York University Assistant Professor Meredith Broussard, Atlanta Journal-Constitution Deputy Managing Editor Shawn McIntosh and Stanford Communications Assistant Professor Jennifer Pan shared their respective experiences on looking beyond numbers to discover news.

Because even with all the cool interactive data — it’s still all about the stories.

Broussard, who created an online journalism tool to track political campaign finance data, believes numbers provide journalists the insight to what may be a good story.

“Every time you see something that’s weird, and think ‘what’s going on here,’ that’s an opportunity for a story,” Broussard said. “This is a lede. It’s going to give all kinds of potential for investigative news. It parses out the information in a way that is easy to understand and it triggers the reporter’s natural creativity. Reporters are really good at coming up with story ideas, so story discovery engines will help them be better.”

Data is not only a source for new stories — it enables journalists to better engage with readers in innovative ways, McIntosh and the other panelists suggested.

McIntosh supervised an investigative journalism project on doctors who sexually abused their patients. The AJC team used data so readers could compare allegations in their state to those of others. Her team charted 2,400 cases of alleged sexual abuse by developing a machine model based on public documents, keywords, and domain variables.

The Atlanta Journal-Constitution used machine learning to investigate doctors who have abused patients.

“Once we developed the tools, then we could really use it to find the specific story that would engage the readers,” McIntosh said in an interview. “I think what was really most engaging to readers is that we used the data to grade (patient sexual abuse in) each state, so readers can look up their state and see how it is.”

That’s not to say, however, that all aspiring journalists must study computer science or statistics. The key question lies in how reporters can collaborate with researchers in other disciplines to discover stories, Angèle Christin said in an interview. Christin, an assistant professor of communications at Stanford University, moderated the panel.

“Academics need to engage with a broader audience, and journalists sometimes need academics for technical skills, such as analyzing data, and for questions that academics care about that journalists wouldn’t see otherwise,” Christin said. “I really see a dialogue that’s going on in these quantitative projects, and that’s what I’m really excited about.”

Inyoung Choi is a student reporter at Stanford University.

.@Stanford University's Journalism Program and Stanford Computational Journalism Lab focuses on multimedia storytelling and data journalism.

.@Stanford University's Journalism Program and Stanford Computational Journalism Lab focuses on multimedia storytelling and data journalism.