GRS51803 Data Visualization: See the study handbook for learning outcomes and further course details

• Masters elective
• Number of students: around 15
• Prior knowledge and skills in GenAI: when the course ran last time, GenAI was fairly new, but it was already clear that some students were using it actively.

Currently there is a disparity between students who are able to communicate their ideas clearly (and do so in English) and students who are less able to or who have a lower level of English, and between students with differing levels of experience in coding. GenAI can be used to level the playing field between students with differing communication, English, and coding levels. In this course specifically, GenAI was used for coding.

In this course, students work on a data visualisation project of their choosing. In this project, they are allowed (but not required) to use GenAI as a tool to help them with coding. (At the start of the course, they are not given much guidance or tips as to how; this is because when the course ran, ChatGPT had only just been released.) Near the end of the course was a reflection session on the use of GenAI. This presentation focused on the ethical issues that come with using GenAI, possible ways in which students could have used it, and strengths and weaknesses of AI usage. It became clear that GenAI is a powerful tool for coding, but not a complete problem solver. It can provide inspiration in smaller pieces, but is as of yet unable to create full finished products from scratch. Thus, using it as a coding tool requires an iterative process and an inherent understanding of what the code is doing.

The setup of the course worked very well. The students evaluated the assignment positively. A majority had used AI in their projects – notably, the people with less coding experience seemed to have used it more. It was clear that instructions at the start of the course on which tools to use or how were not needed: in contrast, the students even gave tips to the lecturer on how to improve GenAI tools’ output.

A tip to lecturers looking to implement something similar in their courses is to be open to it. As a colleague said, “if we embrace AI, it’s going to be messy, but if we reject it, it’s going to be messy too”. Secondly, listen to the students: they often know more than you do. Let them lead. A last tip is to set clear boundaries on which usage is allowed: that way, students know AI is supposed to be a support tool, and do not have to worry about plagiarism accusations.

Jessica Wreyford (course coordinator) and [email protected]