In October, Haley Wixom ’20 and Lucas Ewing ’19, two of the co-heads of Concord Academy’s Big Data Club, spent five days in Las Vegas, where they competed in the 2018 Teradata University Network (TUN) Analytics Challenge. From a pool of more than 60 submissions from around the world, their team was one of 16 finalists — 11 of them in their division. They were the first high school team to take part in this global, university-level competition.
At the Teradata Analytics Universe conference, attended by around 6,000, the two CA students presented their analysis that showed improving mental health support can reduce gun violence in schools. And they had a chance to speak with educators and bag data executives. “It was so inspiring that all of these professors and professionals were so enthusiastic about our work,” Haley says. The team took home the People’s Choice Award for best presentation.
Because the TUN challenge was designed for college teams, the members of CA’s Big Data Club hadn’t expected to be selected. After they learned they were finalists, Haley and Lucas spend the summer working on their posters and practicing their presentation. It was a continuation of the intensive effort that had gone into shaping their project.
One of the days the Big Data Club met last March to consider topics coincided with the national student walkout following the mass shooting at Marjory Stoneman Douglas High School in Parkland, Fla. With that devastating event fresh in their minds, and knowing they wanted to explore some sort of solution to an open-ended topic, the team set its sights high.
Determining the angle from which they could approach gun violence with statistically significant findings and recommendations was a complex task. They considered an exhaustive array of potentially influential factors, from regulatory gun laws to suicide rates, from community income levels to weather patterns. Each week, the 10 to 15 members of the Big Data Club would split up and independently create visualizations, then regroup to discuss them. “We would analyze our correlations and trend lines and talk about the logic behind our conclusions,” Lucas says.
The limits of publicly available data meant that even though they suspected some factors were contributing to gun violence, they often didn’t have datasets of sufficient size or detail from which to generalize. “We ran into so many blocks in terms of what data was available,” Haley says. When they eventually started sifting through a Kaggle dataset on U.S. mass shootings in K-12 and college school settings, which provided information about shooters’ mental health backgrounds, they decided their project would examine links been mental illness and gun violence.
The correlations they drew were striking. As they reported in their presentation, from 1966 to 2017, school shooters were 23 percent more likely to have had a history of mental illness. Furthermore, school shooters with a history of mental illness affected on average more than double the number of victims compared to school shooters with no documented history of mental illness. Finding that states with better mental health rankings had fewer firearm deaths, they suggested that states’ investments in mental health resources could protect their citizens.
The majority of the team’s recommendations — what they call their “three Es” — were aimed at schools: educating school communities about gun violence, establishing resources and support, and evolving practices and attitudes to reduce the stigma of mental illness and help students seek the support they need. Their goal was to identify deliberate actions that schools could take independently, without having to wait helplessly for national movement on gun reform legislation.
Lucas Ewing ’19 and Haley Wixom ’20 presenting on behalf of the CA Big Data Club at the Teradata Analytics Universe conference.
The team’s recommendations for schools:
1. educating school communities about gun violence,
2. establishing resources and support, and
3. evolving practices and attitudes to reduce the stigma of mental illness and help students seek the support they need.
“We were very aware of stereotypes linking mental illness to violence, and we didn’t want to contribute to that stigma,” Haley says. That’s why they spent hour upon hour carefully distilling down their 30-second elevator pitch to include a specific focus on reducing stigma as a central element of the approach.
A recent addition to CA’s student groups, the Big Data Club as founded in 2015 by Ken Lin ’18 and Tim Lyu ’17. Lucas says he didn’t know anything about data analytics when Lin recruited him, but he was game to find out. The students involved spent a good deal of time locating resources and teaching themselves. The story was different for Haley, who grew up in a household where data analytics were a part of life. Her mother, Barbara Wixom, is a principal research scientist at the MIT Sloan Center for Information Systems Research, where she studies how and why companies around the world use big data and analytics. Haley joined the club the year after Lucas did. The club has been growing, and its members have recently had opportunities to learn from major corporations such as Microsoft and individual influencers, including sports analytics thought leader Dave Schrader.
One of the reasons they entered the TUN competition was a hope to connect with other high school data analytics teams. “We’re looking for other high schoolers who are interested in data analytics, and we want other high schools to start data analytics clubs and apply for this challenge,” Lucas says. “Hopefully next year our competition will include teams from other high schools as well as universities.”
Since those teams have been hard to come by, Haley has taken her involvement with TUN a step further, joining their advisory board. As high school editor, she is tasked with supporting high school students interested in pursuing data analytics. With her mother, she traveled in early November to Cincinnati to launch the new initiative, delivering a keynote address at the city’s Women in IT conference.
The Big Data Club members hope to help other high schools that are looking to form clubs gain experience with data analytics tools and connect with TUN. This year they’re focused on learning the programming language R and gaining familiarity with the analytics program Tableau, about which they will develop a presentation they can share with fellow teens. They plan to compete at the TUN Analytics Challenge again next year, but for now they’re taking their award in stride and keeping the club’s biweekly meetings fun by pushing each other to keep learning in a lighthearted way: Whoever completes a challenge first, or brings in the best project, earns the right to wear a crown for a day.