Leveraging Jupyter Notebooks in Assessment Development, Completion and Marking to Reduce Cognitive Load and Minimise Errors

Howson, Oliver (2023). Leveraging Jupyter Notebooks in Assessment Development, Completion and Marking to Reduce Cognitive Load and Minimise Errors. In: I-HE2023: Proceedings of the Innovating Higher Education Conference,, 04-06 Oct 2023, Istanbul, Turkey, I-HE, pp. 11–20.

URL: https://zenodo.org/records/10123535

Abstract

When redeveloping a level two Algorithms and Data Structures module the decision was made to centralise the learning material and assessments around Jupyter Notebooks. We shall explore the lessons learnt and advantages gained through an assessment process which now has development, assessment and marking all completed within the Jupyter Notebook infrastructure.

The cognitive load associated with assessment development often stems from the complexity of the task and the risk of introducing errors during various stages of the process. By leveraging Jupyter Notebooks, authors can benefit from a range of features that enhance productivity, promote collaboration, and reduce the likelihood of errors.

This presentation will delve into several key areas where Jupyter Notebooks can significantly contribute to the assessment development process. We will discuss the seamless integration of code, documentation, and visualisation capabilities, which allows developers to write and test their assessment in a single environment. This not only enhances readability and maintainability but also facilitates the identification and rectification of errors through an interactive and iterative development process.

We will explore the collaborative nature of Jupyter Notebooks within the GitHub infrastructure, enabling developers and reviewers to work together. Furthermore we will explore the in-house development of a plugin to reduce the chances of amendments becoming desynchronised between student-facing and assessor facing versions.

Jupyter Notebooks can also help reduce cognitive load for students who are taking assessments. Traditional assessments can often be challenging to follow, with a series of complex instructions, questions, and response formats. Even digitally completed programming assessments may have question documents, answer documents and multiple program code files to be read, changed and submitted. This can lead to cognitive overload, especially for students who may already be feeling stressed or anxious about the assessment.

Jupyter Notebooks can help mitigate this by providing an interactive and visually appealing assessment experience. By presenting questions, code and responses in a single digital format, Jupyter Notebooks allow for more flexible and intuitive navigation, reducing the cognitive load required to understand and respond to questions.

Additionally, Jupyter Notebooks can provide immediate feedback to students, providing a sense of clarity and direction. By offering automated testing and visual cues, such as colour-coded responses, Jupyter Notebooks can help students quickly identify areas to be developed and adjust their approach accordingly.

The final area of utilisation of Jupiter Notebooks is in the assessment process; markers are provided a single document per student to mark, without having to view additional code files. As well as again reducing cognitive load, assessors are saved the potential hassle of having to chase students for additional files which should have been submitted but may be missed. We will look at a second in-house plugin which has been developed to aid and streamline the assessment process.

Plain Language Summary

A review of the use of Jupyter Notebooks and supporting tools to reduce the cognitive load on students studying and colleagues tutoring on M269.

Viewing alternatives

No digital document available to download for this item

Item Actions

Export

About