Researchers have found that even experienced test graders can have trouble distinguishing answers generated by artificial intelligence (AI).
The research was conducted at the University of Reading in the UK, where university leaders are working to identify potential risks and opportunities from AI in research, teaching, learning and assessment, and its findings have already led to updated advice for staff and students.
The researchers call on the global education community to follow the example of Reading and other universities which are developing new policies and guidelines and do more to address this emerging issue.
In a rigorous blind test of a real university exam system, published today (June 26) in PLOS ONE, ChatGPT-generated exam answers submitted to several undergraduate psychology modules went undetected in 94% of cases and earned higher grades on average than real student submissions.
This was the largest and most powerful blinded study to date challenging human educators to detect AI-generated content.
Associate Professor Peter Scarfe and Professor Etienne Roche, from the University of Reading's School of Psychology and Clinical Language Sciences, who led the research, said the findings should serve as a “wake-up call” for educators around the world. A recent UNESCO survey of 450 schools and universities found that less than 10% had policies or guidelines for the use of generative AI.
Dr Scarfe said: “Many educational institutions have moved away from traditional examinations in order to make assessments more comprehensive. Our research shows that it is internationally important to understand how AI will affect the integrity of educational assessment.”
“We won't necessarily go back entirely to handwritten exams, but the global education sector will need to evolve in the face of AI.”
“That we have rigorously scrutinised ourselves to take a lead on this matter is testament to our commitment to straightforward academic rigour and research integrity at the University of Reading.”
Prof Roche said: “As an industry, we need to agree how we expect students to use AI in their work and recognise its role, as well as the wider use of AI in other areas of life, to prevent a crisis of trust across society.”
“Our study highlights our responsibility as producers and consumers of information. We need to redouble our commitment to academic and research integrity.”
Professor Elizabeth McCram, Vice-Chancellor for Teaching and Student Experience at the University of Reading, said: “It's clear that AI will transform many aspects of our lives, including how students are taught and how their learning is assessed.”
“At Reading University we have embarked on a major programme of engagement to look at all aspects of education, including making greater use of technology to improve the student experience and increase the employability of our graduates.”
“Solutions include moving away from outdated assessment methods and towards ones that are more aligned with the skills students need in the workplace, including the use of AI. Sharing alternative approaches that enable students to demonstrate their knowledge and skills with colleagues across disciplines is crucial.”
“Through a detailed review of all the courses the University of Reading has already established, we are confident that we are in a strong position to help current and future students learn about and benefit from the rapid developments in AI.”
/Public Release. This material from the originating organization/author may be out of date and has been edited for clarity, style and length. Mirage.News does not take any organizational stance or position and all views, positions and conclusions expressed here are solely those of the authors. Read the full article here.
