DataFest 2015: An Insider’s Perspective
What is DataFest? This was the first question we asked ourselves when we heard of the event last fall. Turns out DataFest is a 48-hour-long event, where teams of undergraduate students work together and compete against one another in order to produce insightful presentations from a large group of data. This year, we were given the chance to work with Edmunds.com data. After listening to the Edmunds.com representatives and reviewing the given data, we were confident with our initial idea…or so we thought.
We all rushed from the auditorium to the room where we’d be working for the next couple of days to get our first look at the data. Before diving in, we made a beeline for the snack table to grab all of the essentials: coffee, Red Vines, brownie crisps, and anything else we could get our hands on. Being locked in a confined space for an extended amount of time brings out the sweet tooth in everyone!
After grabbing a hefty handful of sugary snacks to fuel us for the coming 48 hours, our team reconvened. Before the competition, we planned to split our team of five into different strategic roles, but quickly realized that this plan wasn’t going to be as productive as embracing the collaborative spirit. During the idea stage on day 1 we all realized that we needed to be flexible and open to compromise with each other. A quick note to future ASA DataFest competitors: don’t get closed-minded and stuck on your first “bright” idea, sometimes the better idea will come along later!
Our original idea required very specific outside data, but we ran into a snag when we couldn’t access what we needed to further investigate our idea. Until midnight came around we were reluctant to give up our initial idea, but eventually realized we needed to settle on a new path. The time we spent attempting to make our original idea work could’ve been better spent cleaning the dataset and planning a new strategy.
From our experience, teams with clear working ideas from the start had better luck with stress and avoided losing valuable hours of sleep, which is definitely useful when you need to be innovative, productive, and efficient.
Throughout the event, many UCLA graduates and professors, along with consultants from statistically oriented careers helped us develop our ideas and debug our code. From the three judging categories: best recommendation, best visual, and best use of outside data, we decided to go for the best recommendation category because we felt like our team was producing strong, creative solutions.
Since we only had five minutes to give our presentation, we thought that having only one of our team members speak would allow for the smoothest presentation. This turned out to be a unique opinion as most teams split up their presentations evenly. Regardless of our performance in the competition this was a great learning experience because we were able to hear the perspectives of future colleagues and network with others. Our suggestion to future DataFesters is to attempt to develop a clear objective as soon as possible; if you’re fortunate to have your plan well outlined early on you will better your chances of being recognized during the awards period.
Overall, we really enjoyed our experience at DataFest and can’t wait for next year! Participating in DataFest gave us a peek into our future careers and the type of work we might be completing after finishing our undergraduate degrees. We recommend this experience to anyone who wants to improve their communication skills, learn how to better collaborate with others, and produce creative and effective solutions for industry.