The Texas Tribune’s data visuals team’s goal this year was to use geography and readers’ locations in Texas to customize the content of of our graphics and help them know more about politics in policy right where they live. For these stories, we partnered with the Tribune’s audience team and reporters to use location to contextualize news about school district ratings, elections and housing costs for our readers.
These projects fill readers’ need for detailed, local data with context. For instance, there is nowhere else Texans could go to see what congressional and legislative districts they could vote in for the 2018 midterm election. The state’s district lookup page doesn’t show who is on the ballot, and it locates Texans based on their zip codes, which is imprecise. Our customized data stories have no peer in the state — national outlets don’t know the communities we report in like we do, and large metro papers don’t cover Texas’ breadth and depth in the same way.
What makes this project innovative?
We know that maps aren’t the only way to add geographic context to our stories. We set out to see how many ways we could use technology to capture readers’ locations and customize their news experience. This means that these stories are all about innovation. We used web development tools — including writing custom code and designing creative data visualizations — to help localize the stories for all parts of Texas. Using geocoding APIs, we built a shape-finder that can take an address, pinpoint its latitude and longitude and return the shape around it — a school district, political district or Zip code — for drawing a map or finding information. Then, we built maps and charts to tell a story with that data.
What was the impact of your project? How did you measure it?
These projects were designed as team efforts between the Tribune’s audience and data visuals teams, specifically the housing graphic and the ballot guide. Our missions are the same — to use storytelling technology to deeply engage Texans on custom stories about their state. These graphics and interactive stories received nearly a million page views, mostly in October and November, in the heat of the election cycle. Nearly half of the views were to the Tribune’s election results page, which gave readers the option to search their address and see local outcomes. The voting guide page also benefited from frequent updates as registration and early voting deadlines passed — and readers told us it was an invaluable resource that they kept turning to again and again.
Source and methodology
These projects all combine data with creative web development to help readers learn more about Texas. For each one, we obtained public data and parsed it to make sure we understood the story it was telling. Then, we used web development tools to build a narrative around the data — making easy points of entry for readers both in and out of Texas. Readers in the state can opt to let some of the pages geolocate them or enter an address for more detailed information. The data wasn’t perfect — for example, when creating our early voting tracker, which showed the number of early ballots cast each day in the 30 counties where 78 percent of Texans live, we discovered errors in the Secretary of State’s early voting data for previous elections. We eventually called all of the counties’ election offices to make sure we were using the right data — and told the state about their mistakes.
All of the data visuals team's project start with our in-house graphics development environment, which is built in Node.js. This is the framework that makes our innovative work possible. Our graphics are built with D3. For the elections results, we also used Python scripts and a Django application format and prepare the results to update continuously on election night.
Ryan Murphy, Chris Essig, Darla Cameron, Elbert Wang, Emily Yount, Naema Ahmed