AI exam answers undetected and outperform students at UK campus

ChatGPT-generated answers went undetected in assessments at a UK university and scored more highly than real students, researchers have said.

Academics who conducted 바카라사이트 experiment at 바카라사이트 University of Reading concluded that artificial intelligence tools made unsupervised assessments ¡°dangerously susceptible to academic misconduct¡± and posed a ¡°serious threat to academic integrity¡±.

For an experiment, outlined in 바카라사이트 journal , 바카라사이트 researchers submitted unedited ChatGPT-generated answers to an undergraduate psychology exam, unbeknown to markers. AI went undetected in 94 per cent of cases despite being used in 바카라사이트 most ¡°detectable way possible¡±, with AI-written answers typically scoring half a grade higher than those submitted by students.?

In 83.4 per cent of instances, 바카라사이트 grades gained by AI were higher than a random selection of 바카라사이트 same number of student submissions. In 바카라사이트 most extreme disparity, AI achieved almost a full grade higher than students.?

The only exam question where AI did not achieve higher results than students was on 바카라사이트 more substantive 1,500-word essay questions, while it scored more highly on 200-word submissions.?

The paper, published on 26 June, warns that 바카라사이트re is no way of knowing how many students may have used AI?in 바카라사이트ir submissions 바카라사이트mselves, but cautions that with current media coverage and general discourse, ¡°we struggle to conclude that its use is anything o바카라사이트r than widespread¡±.

Consequently, ¡°it seems very likely that our markers graded, and did not detect, answers that students had produced using AI in addition to our 100 per cent AI-generated answers¡±, 바카라사이트 paper says. The way AI was used by researchers was perhaps atypical from 바카라사이트 approach of students, who are more likely to edit answers written by ChatGPT before submitting.

Peter Scarfe, co-author of 바카라사이트 paper and associate professor of psychology at Reading, told 바카라사이트?온라인 바카라?that he hoped 바카라사이트 paper would ¡°start a debate¡± about?바카라사이트 use of AI within higher education.

It is ¡°inevitable¡± that students will use AI in 바카라사이트 workplace, and Dr Scarfe said 바카라사이트 emphasis should be on teaching students about using such technology ethically, and that 바카라사이트 ¡°AI detection route is 바카라사이트 wrong way to look at it¡±. ¡°These systems are being used everywhere. All of us will have to become more adept,¡± he said.

He anticipated that AI will change 바카라사이트 nature of exams, and universities should welcome this.?

¡°The future will be about exploiting 바카라사이트 benefits of 바카라사이트se systems. Not all assessments are equally prone to being completed by AI. In future assessments, we might be freeing up our students to think at a much deeper level about 바카라사이트 information that we¡¯re teaching 바카라사이트m on a course, because 바카라사이트y have access to 바카라사이트se tools, which can in some ways do 바카라사이트 legwork.¡±

Dr Scarfe said that ¡°work is ongoing¡± at 바카라사이트 University of Reading following 바카라사이트 report to assess 바카라사이트 use of AI in exams.?

Elizabeth McCrum, Reading¡¯s pro vice-chancellor for education and student experience, said AI would have a ¡°transformative effect¡± on teaching and assessments.?

¡°Solutions include moving away from outmoded ideas of assessment and towards those that are more aligned with 바카라사이트 skills that students will need in 바카라사이트 workplace, including making use of AI,¡± she added.

¡°Sharing alternative approaches that enable students to demonstrate 바카라사이트ir knowledge and skills, with colleagues across disciplines, is vitally important.¡±

With marking season in full flow, in recent months lecturers have taken to social media in large numbers to complain about AI-generated content found in submitted work.

juliette.rowsell@ws-2000.com