Shared Goals and Contrasting Objectives: Understanding DataFest Through Dual Perspectives

Jessica Karch, Jennifer Noll, James K. L. Hammerman, and Traci Higgins

DataFest has emerged as a prominent informal data science education event, offering participants a unique opportunity to engage with real-world data challenges. This extracurricular event is less structured than a formal university course, takes place over a weekend in the spring, and develops teamwork and communication skills while engaging students in an open-ended challenge with large, complex, authentic data. The core elements of DataFest are similar across sites but the local context and multifaceted motivations, goals, and expectations of both organizers and participants creates opportunities and challenges statistics educators at each site will benefit from understanding. 

This work is part of an Improving Undergraduate STEM Education project funded by the National Science Foundation. The research team spent two years at six DataFest sites conducting surveys and interviews with organizers and participants to better understand the goals and motivations for hosting and taking part in DataFest. 

Five DataFest organizers, who are primarily statistics professors and lecturers, participated in a focus group interview related to their goals and expectations for DataFest at their sites. Organizers consistently articulated a clear set of two goals for the event. The most fundamental goal organizers described was their aim to provide students with more authentic, out-of-class experiences with real data. This core aim manifests in several key areas. A primary driver for organizers is to facilitate hands-on learning that complements and extends classroom instruction. They envision DataFest as a space for students to apply classroom skills to a real-life data set, learn how to work with messy data, and sharpen presentation skills and soft skills. For many, it’s about giving students a tangible sense of what majoring and working in data science feels like. 

Organizer Goals

Organizers like John Tukey (a pseudonym) from a private suburban research institution emphasize exposing students to “working with real data outside the classroom / in an unconstrained setting,” while Margaret Hamilton (a pseudonym) at a large public research university sees DataFest as a direct application of learned skills. The event is also viewed as a significant curriculum vita item, providing participants with a valuable credential for their résumés and an experience they can discuss in future interviews, thus linking directly to career development and widening perspectives on professional opportunities.

Beyond individual skill enhancement, a second goal for organizers is to foster community. Organizers actively look to bring a sense of community to the stats department by bringing together alumni, faculty, grad students, and industry mentors and even creating regional connections by inviting satellite sites. Hamilton notes the goal of keeping alumni “involved and engaged in an authentic way,” while others aim to rebuild communal, collaborative culture lost during COVID. This goal highlights DataFest not just as a learning event, but as crucial for building networking and shared experiences, where students can connect with professionals and peers. At some larger university sites, DataFest is seen as a capstone event for the related majors, while at others (that may not have a data science major) the event is used to get students excited about learning quantitative analysis skills.

Design and Recruitment Impact

Organizer goals directly influence design elements and recruitment strategies that span sites. To support skill development, structures include inviting mentors from industry and, at many sites, offering workshops for hard and soft skills. The 48-hour structure is seen as a way to manage data complexity while providing a challenging experience. The time constraints require participants to prioritize and make decisions quickly. 

For community building, organizers aim to involve a wide range of stakeholders, from alumni to local schools. While recruitment strategies vary across sites, they generally target students with some level of ability/comfort with data and often invite students from different majors to encourage interdisciplinary teamwork. Some organizers, like Ada Lovelace (a pseudonym) from a primarily undergraduate institution, also try to “attract newbie students who can get excited about data science,” even if the event tends to be dominated by more experienced students. 

At other sites, DataFest is seen more as a culminating event for some of their students who are majoring in statistics and data science. In addition, some of these sites have more students who want to participate than the site can accommodate and teams are turned away. Because DataFest tends to attract primarily statistics, data science, and computer science majors, organizers mention the tensions between wanting to broaden participation at the event and the space constraints of their sites.

Student Motivations and Expectations Met

As detailed in “Spending the Weekend with Data: The Appeal of DataFest for Students,” most participants are drawn to DataFest for many of the same reasons organizers promote it: skill development and career enhancement. These goals are by and large met during their DataFest experience. Participants also look for real experience to discuss in interviews, a chance to apply classroom skills, and the opportunity to learn to work in [interdisciplinary] teams. The networking aspect, allowing them to connect with industry professionals and meet alumni, faculty, and other students, is also a significant draw. The competitive element, while sometimes a source of stress, also motivates many to sharpen their presentation skills and learn to tell a data story. 

Student Expectations Unmet

There were four areas in which we saw variability in terms of participants’ expectations being met or unmet based on post survey reflections. First, while some participants expressed great satisfaction with their team experience and learning from one another, teamwork and team dynamics were a challenge for others. Second, while many participants found their mentors a major source of support, a few did not. Third, though many participants found the networking and presence of industry professionals engaging and helpful, a few did not think there were enough structured networking opportunities. Finally, some participants expressed expectations around the judging process and overall desire for more feedback. 

Generally, participants who did express an unmet expectation about the judging fell into one of three categories: (1) a lack of transparency in the judging process; (2) no established rubric for judges; and (3) a lack of feedback about the details of their work. A few participants also seemed to expect more structure and guidance throughout the entire event. 

In reflecting on our conversations with organizers and participants, we think some of the participant expectations around more feedback, structure, and guidance at DataFest may be grounded in their formal school experiences. There may be an implicit distinction between how organizers and students understand what counts as a learning experience in that students may expect feedback as part of the learning experience of DataFest, while organizers may be more focused on the practice of working through complex, authentic data as the learning experience. 

Recommendations

Guided mentorship is crucial, and some sites already have some mentor training. Prior mentor training can equip mentors with strategies for effectively balancing time among groups, proactively engaging with teams, and providing guidance on active listening and constructive criticism approaches. Mentor training also empowers industry professionals who may be shy or less comfortable working with students. 

Expecting judges to have familiarized themselves with the data set in advance and provide feedback on the technical aspects of each team’s work during each event is a lot to require for judges who are volunteering their time over a weekend. However, each site may be able to provide a general rubric judges would use for all participants at the start of the event to help clarify expectations. In addition, organizers could implement a peer review/judging in which teams in the final round could be judged by participants and there could be a peer award given based on peers’ perceptions of the final set of presentations. Engaging in community voting in this way may also keep participants who did not make it to the final round present and engaged. It is also an opportunity for those most familiar with the data (the students) to give their peers constructive feedback on their analyses. Peer feedback has the potential to develop soft skills around giving and receiving feedback, an important skill for working on teams in industry settings.

Finally, team dynamics could be addressed through pre-event meet ups, guided team-building icebreakers, or mechanisms for organizers to help teams address their goals and expectations for DataFest so each person on the team is better prepared to be a contributing member. Setting shared expectations about teamwork ahead of time could help address conflicts before they appear. Some DataFest sites already have a pre-event, and organizers from those sites might have more suggestions for how to make the most of these meet ups. 

DataFest is a powerful vehicle for data science education that complements participants’ formal educational experiences. By explicitly acknowledging the nuanced motivations and expectations of both organizers and participants and trying to create diverse structures that support participants with varied goals and expectations, statistics educators can ensure DataFest continues to be a transformative and enriching experience. The goal is not to host an event, but to cultivate a vibrant, supportive, and truly educational environment in which every student—regardless of their background, current skill level, or goals—feels empowered to engage with the world of data. 

Insights gleaned from organizer and participant expectations and goals offer a valuable roadmap for the broader landscape of informal undergraduate STEM education events. Informal events such as DataFest offer an environment for undergraduate students to gain important experiences working with authentic data in ways many students do not have access to in their formal academic settings. Yet, there can be tensions between student expectations rooted in formal settings. Better understanding these tensions and studying different approaches might allow for best practices for fostering important informal STEM learning experiences.

Editor’s Note: This material is based on work supported by the National Science Foundation under Grant No. DUE 2216023. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.