Breaking the filter bubble with data journalism on marginalized communities

By Kelsie Schrader

For many, data journalism is a complex and daunting task. It requires time, skill and access to data and sources. Data stories on hard-to-access, marginalized communities, then, can often seem unapproachable.

The perceived difficulties of reporting on marginalized communities have resulted in a lack of data stories about and for non-white, non-elite communities. Three journalists discussed this issue at the CAR Conference and offered tips for how to develop a data story on a marginalized community.

Adriana Gallardo, an engagement reporter for ProPublica, and Anjeanette Damon, a government watchdog reporter for the Reno Gazette-Journal, have both reported on sensitive, undercovered populations. Gallardo covered maternal mortality in the U.S., which has the highest rates of maternal deaths of any developed country. Damon reported on prison deaths in Reno, which increased significantly after a new sheriff took over. Both stories came with challenges such as difficulty accessing data and sources, but positively affected marginalized communities and inspired conversation and change on local and national scales.

Eva Constantaras, a data journalist at Internews, used these stories as examples of quality data stories on marginalized communities. She offered three main steps to help journalists begin developing data stories on similarly marginalized populations.

Start with the background

Before jumping into a story, see what’s already been reported in other outlets. Stories on marginalized communities often come from breaking news stories. Use headlines in other papers as an opportunity to reveal the systemic issues underlying the breaking news of the day. Analyze what is and isn’t being covered. See what data is available on the subject. Discover what you can add to the conversation.

Form a hypothesis

After you’ve analyzed what already exists on the topic, identified what’s missing and examined available data, write your hypothesis. What do you think your story is about? What could be the causes of inequality and discrimination? Be careful to avoid common mistakes with hypotheses, such as statements that are too simple, too broad, too narrow or unable to be proved with data. 

Develop questions that will prove whether your hypothesis is true

Questions can fall into four categories: problem questions, impact questions, cause questions and solution questions.

• Problem questions look at how big the problem is. For example: How many people does it affect? How expensive is it?
• Impact questions examine who is affected by the problem. For example: Are some people or groups more affected than others?
• Cause questions analyze the causes of the problem. For example: What has caused the problem to worsen?
• Solution questions seek to discover what will solve the problem. For example: What is the solution, and how can you measure the effectiveness of those solutions?

Ultimately, Constantaras advocated for stories that have an impact. Stories should have information that marginalized communities can act on and use to engage policy makers. “These problems are everywhere,” she said. “Governments will brag about policies they say are helping people, and it our job to check it.”

Kelsie Schrader is a journalism student at the University of Missouri.