By Emma White, Founder and Managing Director at Mark My Papers Ltd
The galloping pace of unfettered tech dominates much of our lives. From doom scrolling to abdicating our thought processes to ChatGPT, the very essence of being human is fighting for survival in the midst of it all whilst society insists on delegating the most mundane of tasks to neuro plasticity. Mock exams are not alone, for some teachers the option of machine led marking is a temptation regardless of all the legal and GDPR responsibilities for which schools are accountable. We are regularly approached to work with AI firms in delivering up what we do in machine learning format. We always say no.
As an early adopter of many tech related life improvements, I am however, a believer that some processes are best left manual and human. I’ll give you a good example. Last year my husband purchased a new car which came with invisible door handles. It looks fabulous, but in its almost servile willingness to pre-empt human thought it actually brings irritating complications to the most basic of tasks. Opening doors remotely from the house means they are locked and handles sequestered just as I try to climb in. I could go on and regale against the touch screen heater than is impossible to operate without a pair of eyes as well as fingers, whilst driving, but my point is this; the car is simply too clever for its own good. In frustration my default has been to repurpose our 11 year old faithful Jeep that does what it says on the tin – from door handles to demisting windows it allows us to decide who wears the pants and seems to understand that when it comes to the simple matter of operating a vehicle, it is a servant to us and not in charge. It is the same with AI in assessment.
There is an embarrassment in being left behind or being seen as a technophobe. A luddite in the face of progress. Three years ago, I was proud to embrace Synthesia and worked hard at polishing scripts and garnishing our website with the trappings of the modern world. Only when I looked at the viewing figures did I realise that it had all been in vain. Nobody watches them. Why? Well probably because in the feeding frenzy of social media, our attention span has apparently been reduced with computer and phone use showing an average focus time on a single screen drop from 2.5 minutes in 2004 to roughly 47 seconds today. Faster we may be but more impatient and less satiated too!
The desire to embed AI in education is all about us enabling youngsters to utilise AI wisely and productively. It is very much on the agenda for modern day educationalists, in a bid to enable these children to be able to face head on the world of employment, as we believe today how it may look tomorrow. But the truth is, none of us, not even the wisest of sages, have a clue how the world of employment will look in five years. Change happens at such a remarkable pace that skills learnt today may well be outmoded in a few months, let alone once our current cohort of Year 11s have reached the age of graduation.
Perhaps a wiser policy would be to push tech to one side in education and focus on fostering a willingness to learn. Ultimately going backwards to go forwards. This seems obvious but much of modern teaching appears to be about the end game – results – rather than furnishing children with the skills to self-learn and adapt. One thing we can be sure of, whatever we teach youngsters today, it will be obsolete before they reach the workplace and so the benefits of engendering the understanding and patience they will need to keep learning and adapting to change is crucial. It is worth remembering too that every advancement and piece of tech that bolsters our life today was invented by people who were educated in an age before such science was even dreamt of – but they clearly had the capacity to learn and the confidence to believe they could enact change.
The astronauts who landed on the moon in 1969 were not schooled in ‘how to land on the moon’ when they were teenagers.
I found myself in a most unusual place when I came to home educate my two younger sons in 2015. For much of the eight years our home education journey took we lived in rural France, with very little internet, tech or printer. Hence, our sole resource for learning was text books – text books, note taking and conversation. Lots of it. And endless reading – we had books a plenty! Clinging rigidly to the national curriculum and navigating our way through 18 GCSES (between two children) and eight A Levels (also between two children) we reached the end without a laptop or a smart phone but with both boys winning places to Cambridge University – one for History and Modern Languages (Christ’s College) and one for Natural Sciences (Pembroke College). Bright boys you may say but the younger did fail his 11 plus and went on to be accepted at Cambridge at the age of 16. Remarkable yes but truly a case of the tortoise and the hare when old fashioned learning – books, pen and paper – accelerates a child who fails his 11 plus through to Cambridge, without even touching the newfangled world of tech.
Many would say having had no exposure to tech would have limited my son’s opportunities that may otherwise have been available to him, but interestingly, when he completed his Masters this summer, his (soon to be published) thesis on ‘Assessing Hippocampal Sclerosis in Drug-Resistant Temporal Lobe Epilepsy Patients Using Automated Hippocampal Subfield Segmentation Across 3T and 7T MRI’ involved working extensively with AI in order to make the process of MRI scanning for epilepsy patients less invasive. Clearly, a love of learning and a self-belief enabled him to assimilate quickly what was needed for a totally new challenge. This is a young person who had never had an IT lesson in any school environment but whose teaching had been focussed on embracing and enjoying a challenge and developing the skill of learning. For him there were no short cuts, no quick fix learning remedies just the understanding that knowledge and skills take time to accumulate and education should be taken at leisure, rather than a rush for the gate marked exit.
This summer has been an eye opener for me. Annually we ask our examiner team to provide us with evidence of their official marking for the main public exam boards. It is a core tenet of what we offer that we only work with official examiners who have been standardised across the nuances of each particular exam paper. Increasingly I am receiving requests from markers who only want to receive hard copy scripts. This request may surprise those educationalists who want to see all exams go on line and hunger for the days where schools will have several hundred plug sockets in one room, state of the art laptops by the thousand and broadband capacity that enables 250 plus students to complete an exam paper and upload it to the grid without calamity – and with a government who sanctions the budget to pay for it all too.
As a company which receives many thousands of scanned papers annually, I await the cherished moment when a large cohort succeeds in submitting their scripts without some sort of phantom catastrophe interfering with page numbering and sequencing to the point where very many candidates have to be reprocessed and delivered again.
It might just be that hard copy is easier because much as we are socially conditioned to embrace tech, some manual processes are a better fit for the homo sapien brain. However much we would rather believe otherwise, our neural processes are a good fit for hard copy. We remember text better on paper because physical pages give the brain spatial anchors, touch cues, and a fixed layout that reduces mental fatigue. Screens often encourage fast skimming, cause cognitive overload, and lack the physical checkpoints needed for deep memory encoding. I know from my own experience that books I have read on my kindle fade into oblivion far quicker than those which I have read as hard copy texts.
When it comes to assessment delivered by AI we are doing our students a disservice by delegating a key role of human judgement to a machine. If we are expecting young people to pick up their pens and their thoughts and put them on paper, then we surely should have the common decency to read their responses and engage our minds on that particular individual’s efforts as a duty of care. Teaching children to think and write by hand is a very different challenge to asking them to collect their thoughts and then process them on a laptop, bypassing the key skill of thinking and planning subconsciously from one paragraph to the next, at a gallop which is a crucial factor in completing an exam in the allowed timeframe.
Nursing an AI programme through a rubric allows us to bend the grading a little to how we perceive we want AI to interpret a mark scheme, but if we are a little ‘out’ in our judgement, that inaccuracy pervades all students and there is little gain and much to lose. If one is wrong, they will all be wrong. At MMP we insist that our examiners never resort to AI but use their vast experience to treat each child as an individual and hopefully, improve that child’s chances of a better grade in the process and more accurate results for the school.
Introducing AI increasingly into the world of education only serves to make us a little more lightweight intellectually. Tempting though it is for time-poor teachers, we really need to pause and think. Increasingly we grab and snatch at morsels tech throws our way as we march briskly through our learning journey, but very little is seeded for long term germination because the current end game is to project a glossy veneer of knowledge rather than work on hardwiring the skills we need for long term learning.
That hardwiring will enable young people to face off any change or challenge this world of tech many bring. We will be doing them a disservice otherwise.