Let’s ditch exams and tests


Rose Luckin, Professor of Learner Centred Design, UCL 

I learned how to do well in exams and tests when I was at school, but for many people they were and still are a cause of great anxiety. They are blunt, outdated instruments that are meant to assess how well someone is progressing in their studies. However, in addition to causing a great deal of anxiety, exams and tests are inequitable, inaccurate, and expensive. They are no longer fit for purpose and learners and teachers deserve something much, much better. To make matters worse they focus assessment on the knowledge that it is the most amenable to automation in the workplace.

Artificial Intelligence (AI) offers us a realistic alternative to exams and tests. AI is part of everyday life for many people across the globe, so let’s use big data and AI to replace the exams and tests in our education systems. Using AI, we can continuously assess each person throughout their learning life and give them (or their parents) access to this constant stream of information about their progress.

Let’s help everyone: teachers, students, parents and employers, for example, to understand what these continuous learning streams are telling us about a person’s learning. This learning might be in a particular subject, a key skill, a characteristic like resilience or the student’s wellbeing. AI can present evidence about the sort of learning that will enable learners to perform well in the changing, automated workplace. The use of AI for continuous assessment does not require students to be sat at a screen day in and day out. The data that the AI algorithms would process can be captured in many ways, such as video, social media interactions, self-report, teacher input or eye-tracking. We have the technological capability to both capture the data we need and develop the algorithms to process it intelligently, so that we can report back on the detail as well as the headline changes in a student’s learning and wellbeing over time. Yes, we need to we need to discuss the ethical, security and privacy concerns associated with this idea, so let’s do that, what is stopping us?

Rose Luckin is Professor of Learner Centred Design at UCL Knowledge Lab in London. She has a particular interest in how AI can be used to identify learners’ cognitive, social, emotional and metacognitive progression. She has taught in schools, FE and HE. Rose is a ufi charity Trustee, and a Governor and Trustee at St Paul’s school in London and the Self-Managed Learning College in Brighton.

Check out more of the 100 ideas