The educational embracers of AI appear outnumbered by the machine-learning sceptics and there seem way fewer of both than the majority who are bewildered by how students will embrace the technology and what it means for teaching and learning practise, and especially assessment.
The Tertiary Education and Quality Standards Agency is here to help, commissioning views from a range of researchers to try to set a context and offer advice on assessment now that students use of AI will be inevitable and ubiquitous.
AI Trainer and Consultant and Founder of AI Her Way, Dr Nici Sweaney, said that every university would need to be considering the types of issues flagged by TEQSA in the new year – with the pace and scale of change meaning that inaction was no longer an option.
“TEQSA are rightly flagging that a tremendous amount of work is going to be required at universities and TAFEs across the country to adjust to the arrival of AI in student backpacks across the nation,” Dr Sweaney said.
“Reverting to closed book exams and squeezing your eyes shut to avoid seeing the tsunami of AI change approaching just isn’t going to cut it past December 2023.
“There are tremendous opportunities to improve learning outcomes with AI, but assessment and teaching will have to change. We need to recognise that AI must be made part of learning in 2024, and in doing so, eliminate AI as a pathway to cheating.”
Dr Sweaney said the TEQSA-commissioned paper was useful to start the ball rolling on consideration of AI in assessment, proposing:
- Appropriate, authentic engagement with AI
If assessment tasks in meaningful ways are set then students will regard it as an essential part of their learning, rather than a supplementary component
- Align assessment with disciplines/qualifications
“(it) then becomes a matter of educational design, allowing for multiple methods, integrated tasks, and meaningful feedback/dialogue between educators and students to support judgements about progress and attainment”
- Process of learning
“where it is appropriate for assessable products to be created by both AI and students, the assessment design should provide clear opportunities to gather evidence where learners critically engage with the use of AI, demonstrate judgement in how to best use AI and reflect on the learning process”
- Students working with each other and AI
set out the boundaries on using AI and incorporate it in curricula inclusive access/capabilities
- Set points in a programme to assess what students are capable of themselves
“likely to be related to programme-level learning outcomes and are important either to the student’s journey through the course, or to judging program completion”
