Open Government Data policies promote transparency and accountability by making government data available to the public. Despite the fact that publicly available data can be difficult to decipher without additional cleaning and analysis, four students from the University of Virginia (UVA) were able to find success analyzing public data from the US Department of Veterans Affairs. With their Tableau for Students licenses and data skills, the UVA team created a viz for a local data hackathon that helped improve access to healthcare benefits for US veterans—ultimately bringing a better quality of life to this community.

It began with Brian Harris, a major in the Army Reserves and a student at the University of Virginia’s Master of Science in Business Analytics program (MSBA), who jumped at the chance to participate in the hackathon framed around a real-world issue: helping veterans see if they qualify for healthcare benefits by analyzing open government data. As an active member of the Washington, DC veterans community, working on solving this problem was especially meaningful to Brian.

The hackathon was inspired by current public policy on healthcare access for veterans. A piece of legislation had been introduced to extend healthcare benefits to service members who were exposed to the toxic chemical, Agent Orange, while aboard aircraft carriers and other ships. To be eligible for additional benefits or healthcare, veterans and their family members had to manually search through handwritten deck logs to find proof of eligibility. This process could take days or weeks to find a single record, making it difficult for veterans to prove that they could qualify.

How the UVA team did it

Step 1: Build a winning team

To tackle this problem, Brian put out a call in the UVA MSBA Slack channel seeking out other students with Tableau for Students licenses to help. He partnered with Alexi Himarios, Matthew Jacobs, and Jackson Sutherland, who are also students in the Master of Science in Business Analytics program, for the hackathon team. Once assembled, they made a promise to themselves that "[regardless of what] happened with politics or budget, potential beneficiaries should not be denied their medical care simply because determining eligibility was deemed ‘too hard’ of a task or was ‘not technically feasible.'"

At this point, the UVA team set out to build a prototype with the goal of creating a way for veterans and their family members to more easily find proof of eligibility. But before they could build the prototype, they needed to take the information from the handwritten logs and create digital records to analyze. Jackson Sutherland described how they approached the time-consuming task:

"We first needed data for a number of ships to demonstrate how [it] could function. After the four of us had burned through the majority of a Saturday, with only two ships digitized and 754 ships remaining, we realized that we either needed a lot more beer or a lot more help. Luckily, we settled on the latter and established a digitization network where volunteers could each sign up to digitize a small segment of data on a given ship. We now have 24 ships digitized and a couple of veterans groups are even spending their own money to pay for digitization of remaining data."

Step 2: Choosing sources and features wisely

With easier access to the ship logs, the team was ready to dive in. They needed to uncover which ship coordinates met the eligibility requirements for benefits. Service members who traveled within 12 nautical miles seaward of a defined boundary off the coast of Vietnam could be eligible for certain benefits. After playing around with the data, they found an answer by joining two data sets in Tableau. The group looked at ship coordinates, and the exposure zone boundary, then applied a minimum distance calculation to figure out which of the ships came within the 12 nautical miles of the exposure zone.

Once this was done, Alexi, Jackson, Brian, and Matthew continued to refine their dashboard with the end user in mind, making it easy for anyone to use on the dashboard—regardless of whether they are a veteran, family member, or Veterans Affairs analyst. Jackson outlined how someone might use the viz:

"For example, a veteran who has served on the USS Ajax can select the specific ship from a drop-down, click on any of the waypoints that are color-coded for the condition of ‘12 miles seaward of the exposure zone’ and follow a hyperlink to the national archives record of that ship during that time to get the information they need prove eligibility for healthcare benefits."

Step 3: Presentation is everything

The UVA students presented the data visualization to the panel of judges, which included members from the Veterans Affairs Benefits office, Tableau leaders, and industry professionals at the end of the hackathon. The team highlighted how their prototype allows anyone to find answers in a matter of minutes, compared to what could take a single analyst several weeks combing through paperwork to complete. The potential impact their viz could have is significant, and resulted in the team capturing first place in the competition.

Here is the UVA team’s winning viz:

The possibilities for working with open data sources are endless, and anyone can use that data to find new insights, solve real-world problems, and have a lasting effect on people’s lives. Brian, Jackson, Matthew, and Alexi recognized they were all passionate about the same subject, had data skills with Tableau for Students licenses, and wanted to make a difference. Even when things were difficult, their commitment did not waver. Instead, they enlisted the veteran community to help by digitizing the handwritten records—crowd-sourcing a task that would have taken the team weeks to complete. With the visibility of winning the data viz hackathon, they showed that all a student needs to make a difference are data skills, open data sources, and passion to make an impact.

Start your own data journey with a free student license.

You might also be interested in...

Subscribe to our blog