Teaching Data Science Programming Skills to Diverse Student Cohorts

In response to the industry demand for data science skills, universities have created new data science degrees and integrated new data science courses into existing degrees. While data science is now being taught at several universities, there is still limited consensus among instructors on the best way to teach data science. Interviews and surveys with data science instructors revealed that they find it difficult to accommodate diverse student cohorts. Students that enrol in data science courses or degrees have differences in background knowledge, are at various stages of their careers, have various levels of commitment and prefer different learning styles. Although the challenges of teaching data science to diverse student cohorts are often stressed, limited methodologies or guidelines have been developed in response. This paper presents the design of a scaffolding framework developed to teach data science programming skills to a diverse student cohort. The scaffolding framework outlined can be used by instructors to design a project-based data science course that progressively challenges the development of data science programming and self-scaffolding skills.