Department of Statistics hosts 10th annual DataFest

An error occurred while preparing your download

Department of Statistics hosts 10th annual DataFest

Students tackle real-world data in 2-day international competition

One dataset. 48 hours. 100 students. Thousands of megabytes of data. For data analytics students and statistics faculty at The Ohio State University, these are the ingredients for one of the biggest events of the year: DataFest. On April 10-12, the Department of Statistics hosted the 10th iteration of the event in Cockins Hall.

"DataFest is an intensive weekend competition where teams of three to five students tackle a large, real-world data set to uncover insights under time pressure," said Asuman Turkmen, a statistics professor who serves on the planning committee for the event. "It blends competition, collaboration and creativity at the same time, and it is not only about data analysis. Students also learn how to think critically about data, how to communicate their findings and how to make data-driven decisions."

First hosted at the University of California, Los Angeles, in 2011, DataFest is now sponsored by the American Statistical Association (ASA) and has spread to over 60 universities around the world. From March to May, more than 2,000 undergraduates gather for their institution's weekend-long iteration of the event, all analyzing the same real-world data set and then presenting their findings to compete for university-level awards like "best visualization" and "best insight."

Although the source and details of the data are kept secret until all DataFests worldwide are complete, past sets have come from major organizations like Ticketmaster, Expedia and the Canadian women's rugby team.

What makes DataFest unique is that this data is all the students receive. There are no questions or requirements that guide how students should analyze the data set or what form their final five-minute presentation should take. For instance, one team might analyze how travel destination affects booking rates while another might create a 3D map visualizing how website users jump between pages while browsing.

"The students do not have a roadmap. Typically, when you do homework, you have data, and you have some questions, like 'estimate the parameters,'" Turkmen said. "That is very different from the messy, real-life industrial data students will work with after they graduate, so they get to see this kind of experience through DataFest."

Fortunately, the students aren't completely on their own. Throughout the event, faculty and industry mentors are available to answer questions or provide guidance.

"The mentors are typically grad students or recent graduates, so they've just been in our shoes and can share advice from that perspective," said Tony Stabile, a junior microbiology major and two-time DataFest participant. "They didn't just answer questions but gave me advice about classes, skills to prioritize and how to interact with people."

In the weeks leading up to the event, registrants can also attend virtual workshops for quick lessons on using data analytics software and attend team formation nights to meet other registered students.

"Any Ohio State student can be a part of DataFest, and when you make teams, maybe you have someone very good at computation, someone very good at biology, someone very good at art," Turkmen said. "Gathering an interdisciplinary team is the way to go in this era that we live in, because nobody can know everything, and having a diverse team gives you different points of view."

Sometimes, the event also inspires students to become more interdisciplinary in their own studies.

"Last year, I had zero experience with data science, and I mostly listened to my teammates do the analysis. But I thought, 'Wait, this is cool. I should learn this.' So I picked up a statistics minor," Stabile said. "This year, I was able to use programming software to manipulate the data. It was great to see that progress."

Executing such a large event is no small feat, and the planning committee begins preparing for DataFest up to six months in advance. Statistics department faculty, graduate and undergraduate students and university staff work together to book event space and catering, secure sponsors, purchase prizes and recruit mentors and judges.

Turkmen believes the months of planning are worth it, though, when the committee sees how much the students grow throughout the course of the weekend.

"I strongly encourage any Ohio State student, regardless of knowing statistics or not, to get involved with the event," Turkmen said. "It's a great opportunity to work with real data, learn some new skills and connect with peers, faculty and professionals in the industry. Most importantly, it's a very fun and supportive environment, and you always gain a lot by being a part of it."

More Ohio State News