Technical innovation in assessment and the use of AI

As part of the wider support for the 25 schools in the GDST, the Trust has specialist teams offering training and advice on areas as diverse as health and safety and educational trips and visits. As you might expect, I work closely with the Innovation and Learning team directed by Dr Kevin Stannard, whose work (including ‘Why (and how) girls thrive in girls-only schools’) may be known to some of you already. 

You might consider this part of the GDST to be the ‘academic’ directorate, which would be correct, as they do maintain the overview of educational provision and achievement across our schools. However, the choice of words in the title ‘innovation and learning’, to me, speaks volumes about the priorities and values of the team working under Dr Stannard.

This week I attended a conference of other academic deputy heads and colleagues with responsibility for innovation and educational technology (edtech). There were two areas of focus; developments in public examinations incorporating digital technology, and artificial intelligence (AI) in schools, but with a focus on assessment specifically. We were joined by speakers from the main examination boards as well as experts from the wider assessment sector. 

You may have read that a number of exam boards are trialling digital assessments currently, with the AQA board initially planning to introduce a limited number of modern language GCSEs using on-screen technology in the next academic year. AQA has also developed an adaptive assessment for Maths that can help teachers save time when diagnosing learning gaps from earlier years. The OCR board too has come on board with this concept and has plans in place to make its Computer Science qualifications available online from 2025. Interestingly, though, some of these developments appear to have been pushed back by at least a year as the boards seek Ofqual approval for the changes.

The case for on-screen assessments has perhaps become clearer since the pandemic, but moves towards this as a principle go back many years, with exam boards initially setting up working parties in the 2000s. The benefits of the approach have been widely understood in the world of work and there are effective digital assessments taking place in many areas of professional life already. The technology is tried and tested in this respect. However, concerns about the viability of running digital assessments still prevail within the education sector.

This is in spite of the positive views students have of the potential for digital assessments to improve their experience of testing. AQA has conducted research specifically into this, leading to a report which points to 68% of students agreeing that increasing digital learning and assessment would be a beneficial move. Young people mention reasons such as this being truer to the digital world they are growing up in, the reduced risk of examiner bias based on handwriting ability, improvements in accessibility for SEND pupils and, tellingly, the environmental benefits of reducing paper use and avoiding the mass transportation of exams.

The role of Artificial Intelligence (AI) is also a hot topic in the world of assessment, as it is in many other areas of education. Currently exam boards put a huge amount of time and energy into the production of examinations for the range of subjects offered, and also place a high premium on the security and integrity of these papers. This is because the production of the papers is enormously time consuming and costly on a human resources level. However, AI can effectively create multiple versions of exams with minimal time and resources expended, so this becomes less of an issue and papers can be replaced as needed. We see this already in place in many areas, with the driving theory test, for example, existing in ever-changing versions so that it cannot effectively be rote learned. 

A further benefit of this approach is that exams can be administered in a more adaptive manner. Students do not need to take the paper at exactly the same time and the logistics of finding suitable spaces and resources to be provided simultaneously is less of an issue. Notably, SEND pupils can be given assessments that not only match their learning styles but operate in a way that supports their personal needs more effectively too.

While most examinations will still be paper based for the next few years at least, there is no doubt that a change will come. Ofqual has very reasonable demands of the assessment authorities to ensure that pupils are not put at a disadvantage and that the validity of examinations is not undermined by any changes. They are certainly not rushing into this. However, while this approach is understandable and honourable, it is equally important not to underestimate the value that a change could offer to many students. Our world will only become more digitally focused and young people will not thank us for holding them back.

Mr Rickman
Deputy Head Academic