An R Autograder for PrairieLearn

TL;DR

This paper presents an R autograder integrated with PrairieLearn, utilizing Docker containers for customization and an extension for context-aware unit testing to improve automated grading accuracy.

Contribution

It introduces a customizable R autograder for PrairieLearn with a novel extension for context-dependent unit testing, enhancing automated assessment capabilities.

Findings

Effective integration of R autograder with PrairieLearn

Custom Docker containers enable course-specific adaptations

Extension improves feedback quality through context-aware testing

Abstract

We describe how we both use and extend the PrarieLearn framework by taking advantage of its built-in support for external auto-graders. By using a custom Docker container, we can match our course requirements perfectly. Moreover, by relying on the flexibility of the interface we can customize our Docker container. A specific extension for unit testing is described which creates context-dependent difference between student answers and reference solution providing a more comprehensive response at test time.