Massifying Data Science Education through Immersive Datathons

As the demand for data-driven decision making continues to rise, data science education (DSE) is becoming more and more crucial. The primary aim of the study was to understand how DSE may be scaled to non-science participants, particularly using immersive DSE training sessions that culminated in a datathon (data science hackathon). The immersion programme was based on a project-based pedagogy informed by the CRoss Industry Standard Process for Data Mining (CRISP-DM) to integrate data science's multidisciplinary nature. Quantitative research was used with 107 datathon participants. Using various statistical measures including Exploratory Factor Analysis, the key results revealed that non-science participants who completed the immersions and datathon were sufficiently knowledgeable in all the CRISP-DM components as well as those who attended the full DSE programme. This means data science skills can be attained by non-STEM individuals by spending focused time in immersions on a specific data science concept. In this case, immersion sped up learning as the participants picked up new data science concepts while applying them simultaneously. The study recommends immersions through datathons to encourage transdisciplinary collaborative learning to massify data science skills.