Team Data Science

In collaboration with Steve Marron, I am developing an approach to teaching Team Data Science. Based on my experience, most data scientists will spend much of their careers working on teams. Effective team members increase their impact on their organization, science, and society. So how does one go from an academic background in statistics or other data science related specialty to a valued member of a team?

Here is a list of skills that are important for team members:

• Learning how to use collaboration tools such as GitHub and Slack.
• Learning how to organize your code so that others can understand it and assess its correctness and functionality.
• Learning how to create readable, eproducible analyses.
• Developing a systematic approach to problem solving that minimizes errors and takes advantage of inputs from team members.
• Understanding the different roles team members play and team dynamics such as learning how to promote your own ideas, accept suggestions and criticisms, accommodate different viewpoints, and manage conflicts.
• Understanding the role of the team leader both from the perspective of the leader and a team member.
• Negotiating and bringing to fruition the final product of the analysis.

In the approach that Professor Marron and I are developing, we assign a group of 3-6 students to a real data science problem. We teach them the basic tools and skills as necessary. Then we guide them through the experience with as much independence as possible so that they get a first hand knowledge of how a team works. In addition to solving the actual data science problem, we hope to produce data scientists who are ready for successful employment, and we aim to prepare students to take leadership roles in subsequent team projects.