By Kara McCormack (UNC - Chapel Hill)
Information
Cloud computing and GitHub Classroom can make teaching R in a statistics course much easier, but instructors may not have the time to learn these infrastructure tools before starting their first class. For new instructors, particularly those teaching R in an undergraduate regression course for the first time, navigating these technologies while designing course materials can feel overwhelming. This presentation provides a structured template for effectively teaching R and regression without requiring cloud-based tools or GitHub Classroom, serving as a stepping stone for instructors who may want to build toward these resources in the future.
The course structure consists of lectures, application exercises (short in-class coding activities), lab assignments (coding & scientific communication practice), and homework assignments (blending conceptual and applied components). All instructional materials are developed reproducibly using Quarto.
This template was implemented via Canvas in an undergraduate biostatistics classroom of 37 students (a mix of juniors and seniors) at an R1 institution, many of whom had no prior experience with R. The approach allowed students to progressively build coding confidence and scientific communication skills while ensuring instructors and TAs could efficiently manage course materials without additional software training. Success of the template was measured with a mid-semester feedback survey.
This template provides concrete examples of each assignment type and practical strategies for structuring a course using only local R installations. This resource empowers junior instructors and faculty to confidently teach R and regression in both undergraduate and graduate courses, even without access to/training in cloud computing or GitHub classroom.
Take Aways:
- Learn a practical framework for designing a regression course that integrates R coding without requiring cloud-based tools or GitHub Classroom.
- This template not only provides a reproducible framework for course materials using Quarto but also serves as a structured way to teach useful models—such as linear and multiple regression—by integrating hands-on coding exercises that reinforce statistical thinking.
- Gain insights into common hurdles when teaching R without cloud computing, along with practical solutions to support students and streamline instruction.
- Access to a GitHub repository containing sample assignments, course materials, and implementation tips, providing a jumpstart for designing one's own regression course.