“Risk of student burnout” as concerns over AI cause number of assessments set by universities to rise

Universities risk ‘over-assessing’ students in response to rising AI usage in higher education, some academics have warned, as the number of exams faced by students increase.

The Student Academic Experience Survey (SAES), which polls over 10,000 undergraduate students annually, has found that, since 2020, the frequency of summative assessments has risen from an average of five per term to 5.8.

Similarly, formative assessments are now assigned 4.1 times per term, up from 2.5 in 2020.

The supposed goal of this increase in assessments is to counteract the risk that AI poses on universities, by improving confidence in the assessment results.

Josh Freeman, co-author of the SAES report, also suggested that an increase in assessments may lead to students having to sacrifice social activities, sports participation, and other ‘extra activities’

Prof. Michael Draper of Swansea University has suggested, however, that this increase could “run the risk of student burnout” by placing too much pressure on students’ time.

Josh Freeman, co-author of the SAES report, also suggested that an increase in assessments may lead to students having to sacrifice social activities, sports participation, and other “extra activities”, other factors which could also increase the risk of widespread burnout.

Aside from the increasing number of assessments, the SAES survey also found that 68% of students in the UK now work during term-time, a record-breaking percentage.

11% of students also said they would choose a path outside of higher education if they could decide again, while only 37% of students said they perceived their course as good value for money – a decrease from 39% in 2024.

One positive factor that the SAES report identified though was that “the quality and speed of assessment feedback has improved significantly” as the frequency of assessments have risen.

Jason Lodge, an Education Professor at the University of Queensland, believes that ‘systemic overhaul’ is needed to avoid AI-facilitated cheating in universities

This suggests – as Thomas Lancaster, a Teaching Fellow at Imperial College London, notes – that assessments are being “split into smaller stages”, and feedback is being given continuously, benefitting the students’ end results.

However, Prof. Draper has claimed that even this may place too heavy a burden on students, with their learning potentially suffering from more frequent assessments.

He claimed that this approach takes away from valuable reflection time for students between assessments.

Jason Lodge, an Education Professor at the University of Queensland, believes that ‘systemic overhaul’ is needed to avoid AI-facilitated cheating in universities, instead of approaches like increasing the number of assessments.

Lodge suggested moving away from detecting cheating with AI, and instead looking to detect ‘actual learning’.

Rather than banning AI, which can in fact be a crutch for students who speak English as a second language, Jemma Skeet of Deakin University suggests that it should be permitted in certain stages of the assessment process, such as planning, but a clear line should be drawn on its limitations.