Abstract

Graduates in the data sciences are expected to work with large, messy data in teams and to explain their findings to non-specialists. These demands are hard to reproduce in a classroom of clean datasets and well-posed problems. This presentation shares experiences of the ASA DataFests hosted in South Africa since 2025, where students, mostly from Statistics, analysed authentic, unfamiliar datasets under time pressure and presented their findings to a panel of judges, as they would to clients or decision makers. The format calls for competencies rarely practised in class: collaboration, solving ill-defined problems and data storytelling.
Drawing on Kolb's experiential learning cycle and the notion of pre-professional identity, the presentation argues that DataFest bridges classroom statistics and professional practice. It highlights the collaboration with partner institutions in the UK, built on an exchange of ideas on how to run events that benefit students. The latest of the events was held on the same weekend, and participants have connected through recorded greetings and live exchanges, inviting them to see themselves as part of an international community of data practitioners rather than a local cohort alone.
The presentation connects these practices to practical outputs by sharing participant feedback on how DataFest shaped their thinking about data, including from those now in industry or postgraduate study. It also reflects on the limitations of this evidence, the challenges of funding and scheduling, and lessons for institutions using hackathons to build employability and global citizenship.