Like many other educators and commentators, I have argued that schools need to redesign the ways in which they assess students’ learning. AI has broken the proxy that traditional methods rely upon: the (mostly written) work submitted by a student can no longer be assumed to reflect their level of understanding. This debate isn’t confined to educators, either: AI in education has become a live topic well beyond the profession itself.
But for school leaders, this will soon become just the latest instance of a familiar type of challenge: making change stick is hard, especially when it runs against the intuition and experience of a subset of the teaching staff. People who have been doing something a particular way for years come to trust that it is the right way, because it has worked for them so far. For some, it will have gone even further and become an ingrained part of their professional identity. These will always be the staunchest holdouts, as any change involves an emotional response as well as a rational one.
Even though they may have been through this many times from both sides, change makers still often underestimate the inertia or, worse, active resistance they will come up against. The reason is simple: they are already convinced that this is the best way forward. The evidence is clear. The PowerPoint is slick. The pilot run with a willing and open-minded volunteer was successful. This is a no-brainer. How could the INSET session possibly go badly?
Fast forward to 5pm on Wednesday, after the INSET introducing the new system, and staff are filing out of the hall in three distinct moods. Some are buzzing, genuinely pleased that the school is finally responding to something that needed to change. A small number are grumbling, already dismissing this as yet another initiative to sit out. And a larger minority are quiet, or have already shifted to talking about their plans for the weekend. Not visibly resistant, but not supportive either. They are unconvinced.
The grumblers were always going to be cynical, no matter the quality of the idea or its presentation. Every team has a few, and there will need to be follow-up discussions to get them on board. Most of the quiet ones, though, might have been winnable but for one problem: to adopt the new approach, they need to let go of the old one. They are not ready to do that, and it is not yet clear why.
Why the new wisdom doesn’t automatically supplant the old
In the early 1980s, science education researchers started to acknowledge something that had no doubt long been clear to teachers. One of the key things that impedes progress is not the raw difficulty of the material. It is the set of misconceptions that students pick up from everyday intuition, television, or wherever else, and bring to the classroom with them.¹ Physics is particularly susceptible, as its principles are not just complex, but also often counterintuitive.
When facts and concepts taught in a lesson contradict the preconceptions that learners arrive with, they do not simply supplant them. The old intuition is suppressed, but carries on interfering: even adults with years of science education are measurably slower and less accurate when a scientific fact contradicts what intuition tells them.²
These days, most educators are well aware that simply teaching the ‘right answer’ may be fine in an environment where students rely on rote learning and regurgitation, but it is not enough to genuinely transform understanding. What works is a refutational approach: naming the wrong belief and its origins explicitly, explaining specifically why it fails, and only then supplying the correct one.³
None of this is exotic. It is why a good science teacher does not just explain that heavier objects fall no faster than light ones in a vacuum. They first acknowledge why this is what almost everyone assumes, then show why the assumption breaks down. And of course the same logic extends to other subjects as well: I have lost count of the number of times I have had to unpick prior thinking around sunk costs in my economics lessons.
Refutation, not exposition
For some reason, we find it all too easy to write this issue off as something that applies to our students and not our teachers. Every staffroom is full of people with long-held beliefs about how their job (and education in general) works. These have accumulated over years of training and professional experience, and sometimes they are simply outdated. If we wish to achieve genuine professional development and align what we do with current evidence, the same rule applies: address and refute the existing belief explicitly or it will continue to be a barrier to change. Research on how adults respond to corrections tells the same story: a correction reduces reliance on what people previously believed, but seldom eliminates it.⁴
Which is where many change initiatives in schools go wrong from the outset. They are rolled out by their champions, who are by definition the people most convinced of their merits. So it is no surprise that they are presented the way anyone presents something they are excited about: here is the new approach, here are its benefits, here are all the case studies of it working elsewhere. That is exposition, not refutation, and leaves the door open to comments along the lines of “if it ain’t broke, don’t fix it” from colleagues.
The problem is often worse when the pitch is handed to an external training provider. A visiting trainer can present the new approach brilliantly, but they usually arrive knowing little to nothing about what this particular staff currently believe, why they believe it, or what has been tried and abandoned before. Without that groundwork done in advance, there is nothing specific to refute, and the session defaults to exposition by necessity.
In short, context is king: the beliefs worth unpicking are part of an institutional memory that is usually opaque to an outsider or a recently appointed leader. They must be identified and refuted before anyone stands up to present the alternative.
Inertia usually reflects personal belief or external pressure, not cynicism
Faced with resistance to a change they are certain is right, most leaders reach for one of two explanations: a problem with the presentation of the case for change, or a problem with the people. Perhaps staff have not really understood the evidence or are unwilling to accept it. We need to give them more, or explain it better. Or perhaps they are simply being difficult, protecting habit and comfort over what is best for students. We need to make our culture more student-centred.
Either way, the conclusion is usually wrong, or at least incomplete. More common is a preconception that the presentation didn’t unpick, a perfectly rational response to a pressure that the supporting case studies ignored entirely, or a mixture of the two. Working out which one, or which combination, you are dealing with has to happen if you wish to overcome the inertia.
Go back to 5pm on Wednesday. The quiet ones filing out of the hall are not being difficult, and most of them have not misunderstood anything either, but they have no intention of doing anything differently tomorrow. Each is holding on to their existing approach for a reason, and those reasons are rarely visible from the front of the room.
Whether in advance or in response to such a reaction, the only reliable way to find out the source of the reluctance is to ask. This is best done in small conversations where people can say honestly what the current approach does for them and what worries them about losing it. The answers will not always be about pedagogy. Two cases, one I was part of and one described to me by a fellow educator, show how differently this plays out depending on whether the barrier is belief, pressure, or both.
Behaviour management: a false dichotomy
In one school I worked in, the pastoral team led a shift from a largely punitive behaviour system to one built around restorative practice. Many staff (particularly those who had worked with similar systems before) were immediately on board, but there was a significant minority on the other side of the fence. They were concerned that when a student did something wrong, they would no longer be able to simply hand out a detention, or some other punishment that they had resorted to for years and remembered from their own schooldays.
Amongst some colleagues not in the leadership team, there was an almost reflexive dismissal of that concern as people wanting to keep the quick option and minimise their input. Since restorative practice is widely seen as time-consuming and demands far more conversations than a detention slip, a few proponents effectively accused their colleagues of caring more about their workload than their students. As is so often the case, those comments happened in the staffroom and corridors, not the meeting room.
But that reading missed what sat underneath: a genuine belief that a consequence, applied consistently, deters the behaviour that earned it and teaches the student something about responsibility in the process. That belief deserves to be taken seriously rather than dismissed, because it describes how consequences work in plenty of contexts, other schools included. Further discussion exposed the true misconception behind the concern: that moving to a restorative focus meant entirely abandoning the sanctions they had always relied on. What they needed was not more convincing of the benefits of restorative practice, but assurance that action would still be taken when things went badly wrong.
The response from the pastoral team was not a better pitch for restorative practice. It was a clearer account of what would actually change: there would still be consequences, but a restorative response would become the default, with more punitive sanctions reserved for repeated or severe incidents, and sometimes used alongside a restorative conversation rather than instead of one. This had always been the plan, and those making the initial presentation had simply taken it as read. They had not accounted for the misconceptions held by staff who were unfamiliar with the restorative approach.
Scepticism began to fade, the interaction between restorative practice and punitive sanctions was set out more clearly in the policy and guidance documents, and resistance to the change reduced significantly. That is the refutational structure in miniature: name the misconception (in this case, a false dichotomy between restorative practice and sanctions), show specifically why it is wrong, and only then present the alternative.
Marking: an unspoken pressure
An educator I spoke with recently described a change that, on paper, looked like a no-brainer: moving staff away from covering student work in red pen. Intensive written marking is one of the biggest drains on teachers’ time, and there is little robust evidence that it justifies the hours it consumes.⁵ Lighter, more focused feedback promised to give teachers their evenings back and to give students something they would actually act on. The school moved from permission, to encouragement, to an explicit direction to reduce the quantity and density of written feedback, and the practice persisted regardless.
What struck me was not that it persisted, but why, when staff were asked directly. Some of the resistance was a genuine belief, concentrated among longer-serving staff trained at a time when visible, heavy correction was simply what good marking looked like. But most of the answers had nothing to do with pedagogy. Many staff said parents expected to see the red pen. The school had a history of very active parents challenging teachers, and there was a real fear that this could return if lighter marking gave the impression that teachers were doing less for their children.
No amount of research on marking will reassure a teacher who is bracing for the next complaint from a parent. Directing staff to change, however firmly, does nothing to that fear. It just adds a competing pressure, and pressure from two directions with no route through either is a reasonable description of why the practice never actually went away. Ultimately, what was needed was not another direction to staff, but a different kind of communication that could only come from leadership: a clear message to parents about the change and the reasons for it, and an equally clear message to staff that they would be backed if a parent challenged them.
What frustrated the educator most, looking back, was how long it had taken to get to the bottom of why intelligent, well-intentioned staff kept doing something they had been told, with evidence, not to do. Part of the answer lies in the difference between “please do this” and “why aren’t we doing this?” Permission, encouragement and direction are all versions of the first; only the second surfaces what is actually going on.
It is surprising how often a stalled change rests on an assumption that a simple, direct question could have overturned, provided it was asked of the right people in the right setting. Few teachers will voice concerns about parents in a meeting with leadership sitting at the front of the room. Many will be happy to do so over coffee in the staffroom. Diagnosis, in other words, is about knowing who to ask, when, and where.
Laying the groundwork
Which brings us back to AI. Over the next few years, most schools will ask their staff to rethink how they assess learning, and too many of those conversations will begin with the benefits of the new approach. For the champions, that is the easy and enjoyable part. The harder work has to come first: finding out where resistance and inertia are likely to come from in your particular school and staff body. There will be unnamed preconceptions and unacknowledged external pressures, and neither will yield to a better slide deck.
If colleagues believe the old way ain’t broke, the first job is to show them, specifically and respectfully, where it is. Only then will they be receptive to a fix.
¹ Rosalind Driver and Gaalen Erickson, “Theories-in-action: Some theoretical and empirical issues in the study of students’ conceptual frameworks in science,” Studies in Science Education, 10(1), 1983, pp. 37–60: https://doi.org/10.1080/03057268308559904. A foundational paper in establishing the study of students’ misconceptions in school science.
² Andrew Shtulman and Joshua Valcarcel, “Scientific knowledge suppresses but does not supplant earlier intuitions,” Cognition, 124(2), 2012, pp. 209–215: https://doi.org/10.1016/j.cognition.2012.04.005
³ Christine D. Tippett, “Refutation text in science education: A review of two decades of research,” International Journal of Science and Mathematics Education, 8(6), 2010, pp. 951–970: https://doi.org/10.1007/s10763-010-9203-x
⁴ Ullrich K. H. Ecker, Stephan Lewandowsky, John Cook et al., “The psychological drivers of misinformation belief and its resistance to correction,” Nature Reviews Psychology, 1(1), 2022, pp. 13–29: https://doi.org/10.1038/s44159-021-00006-y
⁵ Victoria Elliott, Jo-Anne Baird, Therese N. Hopfenbeck et al., “A Marked Improvement? A Review of the Evidence on Written Marking,” Education Endowment Foundation, 2016.