Accessibility of Jupyter Notebooks on M269

• Project leader(s): Alexis LansburySharon Dawes
• Theme: Access, Participation and Success
• Faculty: STEM
• Status: Archived
• Dates: April 2021 to March 2025

When M269 (Algorithms, data structures and computability) was rewritten in 2021, all learning, practical work and assessment content was presented by means of Jupyter notebooks rather than having a mixture of theory on the VLE, two printed textbooks and practical work using an IDE.

The project aimed to examine the usability and the accessibility (Web Content Accessibility Guidelines) of this learning approach from the perspective of tutors and students of the rewritten module.

We recruited 6 students and 6 tutors from the 21J presentation and asked them to answer a short questionnaire in the form of diary entries at 3 key points in the presentation. We followed this up with 2 focus groups for the tutors, the student volunteers having dropped out of the study before the second point. We analysed our findings using NVivo.

Our key findings were that while some tutors had adapted well to the new module style, others struggled to find information in the notebooks to help students. The way of working with notebooks, in marking student assignments, was resulting in much longer marking times than before for all tutors. Having to be online to answer student queries was also resulting in much longer response times.

Students were generally positive about the ability to run code examples easily alongside reading of theory, but they struggled to find information from earlier sections, and some found it tiring working on a screen the whole time rather being able to read a textbook offline.

Our recommendations included that content and index pages be provided for the notebooks and that ways of reducing the time to mark the assignments for tutors be explored.

We followed the study up 3 years later and found that the module team had made significant improvements to the marking guide and processes for marking as well as adding automated testing and making software available on the Open Computing Lab to improve accessibility.