When the coronavirus pandemic prompted the American Statistical Association (ASA) to cancel this year’s annual DataFest, it seemed like another hands-on opportunity that some Virginia Tech students would miss out on this spring.

Everyone understood the reasoning. ASA didn’t want to encourage teams to huddle together to work on data sets in a time of social distancing norms and stay-at-home orders. But still.

“The students were very disappointed it was canceled,” said Christian Lucero, collegiate assistant professor in the Department of Statistics and the Computational Modeling and Data Analytics (CMDA) program, both part of the Virginia Tech College of Science. “Many of them put the ASA competition on their resumes, particularly if they won.”

Before the disappointment with the cancelation could sink in, leaders of a Washington, D.C.-based health care startup named Socially Determined Inc. reached out to Virginia Tech.

Several members of Socially Determined are Hokie alumni, and they have opened up a tech talent supply line, especially among those majoring in CMDA, housed in the college’s Academy of Integrated Science. Socially Determined also sponsored a graduation student in the statistics department this past fall.

Alan Lattimer, Socially Determined’s senior data scientist who earned his Ph.D. in mathematics from Virginia Tech, said the company had begun thinking of how it could help Virginia Tech students while the coronavirus pandemic was disrupting normal events.

“What would be beneficial to both Socially Determined and Virginia Tech?,” Lattimer said he wondered, then thought: “I don’t see why we can’t do this thing online.”

That began an effort to revive a data analysis competition among Virginia Tech students and, not only that, but to hold it on the same weekend originally set by ASA. Lattimer and Lucero would have about two weeks to assemble the whole thing.

Lucero knew they’d need three components: a directive, questions to pose, and a relevant set of data. In years past, ASA had used data sets from corporations such as Ticketmaster and Indeed.com.

It so happened that Socially Determined could provide the framework. Socially Determined is a health care analytics company that focuses on assessing how social determinants — economic distress, food insecurity, housing instability, transportation barriers, and lack of health literacy – can affect a person’s health and the cost of their health care. Data science drives Socially Determined’s work, and they had a data set from the state of Maryland well suited for the competition.

Mark Embree, professor of mathematics, talks with students in his Computational Modeling and Data Analytics capstone project class in this November 2017 photo.

Mark Embree, professor of mathematics, talks with students in his Computational Modeling and Data Analytics capstone project class in this November 2017 photo. Embree is wearing a dress shirt and sports coat as he stands near the students.
Mark Embree, professor of mathematics, talks with students in his Computational Modeling and Data Analytics capstone project class in this November 2017 photo.

Mark Embree, head of the CMDA program and a professor of mathematics, emailed students in the major on April 2 to tell them the good news, that in two weeks a newly fashioned team data competition would be held: “The COVID-19 Social Data Project at VT,” sponsored by Socially Determined. Working together, Virginia Tech — via the CMDA program — and Socially Determined had fashioned a highly relevant, real-world, learn-by-doing experience. All online.

By mid-April, 32 undergraduate teams and six graduate teams had registered, representing more than 100 Virginia Tech students. Under any definition, they’d be tackling what is currently one of the world’s most complex problems.

The data project prompt read, in part: “It has become increasingly apparent that there is a direct correlation between poverty, housing conditions, food insecurity, transportation barriers, and other factors to the health in this country. While improving our overall health has always been important, the COVID-19 pandemic crisis has simply exacerbated the issue.”

The student teams would work with the provided dataset and could import their own data from other sources. Their time limit ranged from noon on Thursday, April 16, until 5 p.m. on Monday, April 20 – a long weekend – to see what they could discover. They would have to present their findings in a five- to seven-minute video, and submit slides. They could seek expert guidance from Virginia Tech faculty and Socially Determined volunteers, who were even available during the weekend.

The teams explored issues such as: where should lawmakers direct resources to have the greatest effect; the impact of a nursing home’s transportation capabilities to get patients to hospitals; which specific counties should be prioritized for economic stimulus to get the most impact?

One team, for instance, focused on a part of Maryland that could be overrun by COVID-19 cases: a three-county area with 29 nursing homes – and just one hospital.

By Friday afternoon, Lucero, Lattimer, and others had finished asking the teams follow-up prompts, gotten answers, and finished the judging. They announced the winners at 5 p.m., when finalists were invited to join a video call. Winners are listed below, along with their team names:

Graduate Level

1st Place: Savage T.Rex and the Exotic Close-Talkers, students of the Data Analysis and Applied Statistics (DAAS) master’s degree program in the Washington, D.C., area: Andrew Knittle, Michael McNamee, Sadaf Rohani, and Adam Wells.

Runner Up: Tri-State Modeling Team (graduate students in the Department of Statistics): Shane Bookhultz, Erica Porter, Stephen Walsh, and Young Ho Yun.

Undergraduate Level

1st Place: Quaranteam 3: Jake French, Jianna George, and Amy Wikiera.

Runner Up: (Tie) R-nought: Allison Desantis, Connor Law, and Klara VanWamelen.

The Data Dudes: Dylan Badger, Porter Carlson, and Jacob Kaminsky

Lucero said that the winning teams’ submissions shared a common quality, an on-going thrust in the CMDA program.

“It’s not just the data science behind it,” he said, “but if you had to make a recommendation, to the governor, or a mayor, or to President [Tim] Sands, what guidance would you give to make it the most valuable?”

Embree said during the video call with the winners that he was impressed by the overall quality of the submissions. “So many of these were so close,” he said, “they needed maybe just an idea or two to push them over the top."

Lucero was pleased that 79 percent of the teams completed the project, more than those who usually finish an ASA-sponsored DataFest – though Virginia Tech and Socially Determined’s event gave the students two additional days.

Tom Woteki, director of the Data Analytics and Applied Science master’s program in the Greater Washington, D.C., metro area, noted that members of the DAAS team that won first among the graduate student teams had many other obligations. “These are people who are working fulltime and completing their studies at night, on the side,” he said.

Socially Determined’s Lattimer said his company, “spends lots and lots and lots of hours” on issues such as the ones the teams tackled, and their insights will be valuable to the company in the coming months.

“We have few opportunities in this lifetime to honestly make a difference in the world, and we should gravitate to them every chance we get,” Lattimer said. “I feel like that’s what we did here.”

Share this story