Abstract
In three and a half years, the University of Minnesota Carpentries Initiative has taught over 180 hours of synchronous workshops covering research computing skills to over 400 students. During this time, learners from every college in the university have been exposed to best practices and hands-on guidance in the use of programming tools to streamline a range of research activities, from cleaning data to conducting analysis to creating visualizations. This cross-campus initiative has allowed departments and individuals to expand their networks and skillsets, creating opportunities for professional growth through a scalable, sustainable service model. This chapter describes lessons learned in recruiting, maintaining, and growing a multi-departmental team of librarians, technology specialists, and graduate students to deliver data science education.
TopBackground
At the core of the Carpentries are a series of public and open-source online lessons, resembling teaching scripts, that are used to deliver workshops on a variety of specific computational tools, methods, and best practices. Carpentries workshops generally combine a variety of these open-educational modules to cover fundamental computing tools and data science topics for researchers from a variety of disciplines. Before the onset of the COVID-19 pandemic, workshops were usually offered as two-day, in-person events, but in 2020 and 2021 were often taught in shorter online segments. A “full” Carpentries workshop encompasses about sixteen hours of synchronous instruction, drawing on the public online curricula from any of the three branches of the Carpentries: Software Carpentry, Data Carpentry, and Library Carpentry. Typically, full workshops are offered under the title of one of the three branches. For example, an institution may offer a “Software Carpentry workshop” that consists of lessons on the Unix shell, version control with Git, and programming with Python. Software Carpentry lessons are also available on R, databases and SQL, MATLAB, and the automation tool Make (Software Carpentry, n.d.). Software Carpentry is intended for anyone who wants to learn basic research computing skills, regardless of their previous programming experience.
Key Terms in this Chapter
Open Science: An approach to research in which both the products and the processes of research are transparent, accessible, and reproducible to the greatest extent possible.
R: A programming language commonly used for statistical computing, and often used in concert with RStudio, which provides a graphical user interface for working with R.
Carpentries: A global organization dedicated to teaching data and open science skills and tools for researchers. Includes three primary branches: Data, Software, and Library Carpentry.
Open Educational Resources: Openly accessible and freely reusable materials for teaching and learning purposes.
Unix Shell: A command line text-based interface that helps users automate repetitive tasks. Bash is one of the most popular shells for Unix.
Python: A general-purpose programming language that is popular for data science applications in academic research.
Research Computing: Refers to the wide variety of programmatic tools and methods used in academic research contexts.
Jupyter: Jupyter Notebooks and JupyterLab provide interactive computing environments for working with Python, R, Julia and other programming languages. Jupyter is helpful for teaching and learning contexts and popular for academic computing projects.
GIT: A free and open-source tool used for version control in software development (and other computational) projects. GitHub is a popular hosting service for sharing and collaborating on Git repositories.