Implementing Version Control With Git and GitHub as a Learning Objective in Statistics and Data Science Courses

Abstract

A version control system records changes to a file or set of files over time so that changes can be tracked and specific versions of a file can be recalled later. As such, it is an essential element of a reproducible workflow that deserves due consideration among the learning objectives of statistics courses. This paper describes experiences and implementation decisions of four contributing faculty who are teaching different courses at a variety of institutions. Each of these faculty have set version control as a learning objective and successfully integrated one such system (Git) into one or more statistics courses. The various approaches described in the paper span different implementation strategies to suit student background, course type, software choices, and assessment practices. By presenting a wide A c c e p t e d M a n u s c r i p t range of approaches to teaching Git, the paper aims to serve as a resource for statistics and data science instructors teaching courses at any level within an undergraduate or graduate curriculum.

An Interactive Approach to Teaching Git Version Control System

“…More recent studies employed Git as a platform to disseminate course materials and facilitate assignment submissions [5,7,[11][12][13][14][15]. Two studies integrated Git into non-CS degrees.…”

“…Two studies integrated Git into non-CS degrees. Beckman et al [7] employed the tool in four statistics courses, one of them being a graduate-level course, whereas Lawrance et al [15] utilized it in a CS1 course for mechanical and electrical engineering majors.…”

“…Similarly, Isomöttönen and Cochez [5] demonstrated Git concepts, after which students practiced the tool in a single session. Beckman et al [7] prepared a 15-minute tutorial and 30-minute practice sessions for teaching Git to students.…”

“…There are also various kinds of tutorials or courses online for people to benefit from. These include but not limited to, YouTube videos 5 , online courses 6 and games for learning Git 7 . However, such online tutorials lack interactivity and teaching assistants where students can get feedback whenever they are stuck.…”

An educator's perspective of the tidyverse

“…We affirm that students benefit from developing literacy in multiple languages and argue that the "tidy data" (Wickham 2014) approach central to the tidyverse is programming language independent, with notable implementations in the three languages mentioned above as well as domain specific languages like SQL (see Section 4.6). We have chosen to focus our attention on R, in part because there are good models for teaching statistics and data science with reproducible computing practices-even at the introductory level (Baumer et al 2014, Beckman et al 2021.…”