All entries for Monday 18 March 2024

AI Marking Criteria

Rob Liu-Preece is the Academic Technologist for PAIS, Sociology and Philosophy at the University of Warwick. He has also been an IELTS marker for Writing and Speaking for 20 years and previously taught Academic Skills to international students both in the UK and overseas for 20 years.

This is the first of two posts written by Rob about AI and the ANTF Project:

The sudden explosion in the availability and use of generative AI technology, especially by university students has left education professionals in a position of playing catch-up. With ChatGPT gaining 1 million users in just 5 days and 100 million in 2 months, I feel like educationalists have just arrived at the point of coming up for air. As part of that process, I’ve written marking criteria aimed at marking the use of AI by students completing assignments.

I think in a learning environment characterised by uncertainty and disruption, students will benefit from an explicit expression of how the university wants them to use AI. Applying a marking framework like this could also lessen the need for tutors to follow a punitive/academic integrity route for dealing with misuse of AI. It could achieve this by opening up and defining ‘poor academic practice’ more closely aligned to AI as an alternative. I also hope this type of approach will help steer the development of pedagogy and AI, providing a structure for on-going debate and discussion. Lastly, having a set of criteria like this enables reverse engineering of training and coaching on AI for both students and tutors.

To address these issues, I’ve written a set of marking criteria based on the existing Politics and International Studies assignment marking criteria for undergraduate students. I would anticipate students including a short report to their written assignments covering their use of AI. The framework is based around 2 main categories, namely appropriacy of use and awareness of key issues. The criteria is by no means a finished piece of work, is not necessarily fit for purpose and hasn’t undergone any road testing or standardisation. Rather it is designed to signal a possible route forward for those of us concerned and interested in shaping the take-up of AI in education. It does raise some thoughts in my own mind about whether such an approach is the right way to go. Should we be setting or defining an orthodoxy in quite tight terms for AI use, like this? Is a literacy model approach implied here the correct one, or would a better way be to focus on conscious use of AI by students?

Please note I used Google Gemini to help with the overall structure of this blog and for the statistics in the first paragraph.

Abigail Ball : 18 Mar 2024 09:00 |  Tags: AI Marking Criteria Rob Liu-Preece |  Comments (0) |  Close comments |  Report a problem