|
Abstract:
|
Most education and pedagogical conversations involve "hard" content topics like data visualization or modeling. Indeed, it's challenging to discuss teaching many data science ideas because they do not make up an entire course themselves or do not relate strongly to an existing concept. However, many of these topics are just as necessary in data science education as basic statistics. Reproducibility, documentation, ethics, team dynamics, version control, choice of tools, and communication are just some of the topics that could be considered part of "workflow" as they contribute to many parts of the data science pipeline of work no matter what the project actually is. While each of these may not get its own dedicated course, it is important to thoughtfully include discussions of them across multiple data science courses. In this presentation, we discuss essential considerations for including these workflow ideas into existing data science coursework.
|
