---

Read more:

• Lessons learned from the teacher assessed grades process of summer 2021
• Cause for celebration, not criticism: reflections on the GCSE class of 2021

Dr Chris Yapp, NACE patron

The need for reform of the assessment is system is now being well argued at the national level. It is important to remember that the current assessment framework and exam results are an important part of the accountability framework by which schools are judged. The issue I wish to address here is: how does any new assessment framework that is developed impact on the accountability of schools? Importantly, what issues and problems of the current approach could be addressed by a novel approach?

If I ask you about the ethos of your school, I would probably have no difficulty in achieving consensus that “every child should be able to reach their full potential” would be a core value of teachers and educational leaders near universally. However, if I suggest that 80% of children should achieve their full potential in education by 2025, how would you react?

My experience is that few are comfortable with the target, even though you can’t reach 100% unless you go through 80% at some point. It would be easy to be cynical that teachers may aspire to the vision but react against trying to achieve it. There are numerous reasons why professionals are uncomfortable with this problem. First, how do you measure potential? Importantly, does the assessment framework reflect both achievement and potential?

There is an economic model, Goodhart’s Law, which has a long history of precedent in different fields and is now more widely understood as a general problem. The usual formulation of the law is: “When a measure becomes a target, it ceases to be a good measure.”

Consider the following example: 80% of children should reach Level X by Year Y.

Apart from agreeing how that is measured, there is another big problem to be addressed. Once a measure becomes a target it can be “gamed”.

Consider two schools with similar catchment areas and performance facing this as a target. 

In the first, the leadership team focuses resources on all children and their development and reaches 78% on the timescale set.

In the second, 10% of children are given minimum support and the resources are focused on the remaining 90%. The school achieves 82%.

Which is the better school? Which would you want to work in? Which would you want your children to go to?

In short, targets can distort ethos and with it the morale and self-worth of professionals. It happens with accounting in the private sector, in reward mechanisms and many other walks of life. My experience is that once you understand Goodhart’s Law you start to see it everywhere.

One of my favourite quotes of Sir Claus Moser sums it up well: “If you can measure the same thing in two different ways, you'll get two different answers.” 

The different components of education are heavily interdependent. Teacher development is heavily dependent on curriculum design, which in turn is heavily dependent on assessment. Attempting to reform one without understanding the impact on the other parts is fraught with difficulties.

So, I welcome a focus on reforming assessment in schools. For me, it is long overdue. However, in the context of our 21st century economy and society we need to be more explicit about the ethos of our education system and its individual institutions. I believe in accountability systems, but they must be driven by ethos, not targets. The assessment measures that are developed need to reflect our societal and economic goals for education itself.

If our ethos is to optimise pupil achievement, the wise words of Plato come to mind: “Do not train a child to learn by force or harshness, but direct them to it by what amuses their minds, so that you may be better able to discover with accuracy the peculiar bent of the genius of each.”

---

• Read more from our “rethinking assessment” series
• Share your experiences of remote feedback and assessment: NACE member survey

Sue Riley, NACE CEO

Many of you will have seen the open letter to the Sunday Times from the recently formed Rethinking Assessment group. Born from issues raised this summer, Rethinking Assessment is a broad coalition of state and independent schools, universities, academics, employers and other stakeholders, which aims to value the strengths of every child. At its heart lies four fundamental principles:

1. Many young people find the way our exam system works increasingly stressful and not a true reflection of what they are good at.
2. Many employers complain that exams do not provide them with good enough clues as to who they are employing.
3. Many headteachers feel that high-stakes exams distort priorities and stop them from providing a well-rounded education for their pupils.
4. Many who are passionate about social mobility believe that any system that dooms a third to fail is a system with little sense of social justice.

We want to add our members’ voice and our research to this debate. There are immediate questions to be answered and longer-term opportunities to recalibrate the assessment system so that all learners have their full range of strengths recognised. As a membership organisation we can share and build on the decisions school leaders are taking now and over time provide perspectives that will inform longer-term changes.

Assessment is of course an integral part of learning and teaching. It facilitates daily ongoing review of individual progress and impacts on planning and target-setting. It supports personal learning targets. But we must not let the tail wag the dog. Not everything needs to be assessed, or indeed can be assessed, or needs to be independently assessed. We must consider too the timing of assessment – even more pressing as schools focus on tier 2 rota planning. 

Whilst a decision over summer exams has been made in principle, “fall-back” detail remains unclear and learners are picking up on this, increasingly questioning the reasoning behind assignments, and the part they will play in assessment. All of this detracts from the richness of a subject.  

The Early Career Framework and Teachers’ Standards have done much to support the teaching profession’s development in recent years. We must trust teachers with assessment, but teachers must be clear on what they are assessing and why.

What can we therefore now do in our schools to readdress this balance? One response lies in thinking about what we assess on a day-to-day basis in classrooms, how we build on low-stakes testing, and how we position effective challenge. How effectively do your teachers use retrieval practice for example? Deliberately recalling information forces us to pull knowledge “out” and examine what we know. The “struggle” or challenge to recall information improves memory and learning – by trying to recall information, we exercise or strengthen our memory, and we can also identify gaps in our learning. 

NACE has recently undertaken a literature review of retrieval practice – looking at the theoretical framework and considering emergent related classroom practices and practical amendments and applications for more able learners. To access this review, click here. 

Beyond the here and now of assessment, we need to return to the longer-term focus of the Rethinking Assessment coalition. Against the current backdrop, what could we do to improve the assessment system more broadly? How would we do it differently, allowing us to show non-traditional talents – making assessment more effective for employers, individuals and supporting the practising teacher? Fundamentally, how can we assess the child in front of us?

Contribute to the debate:

• What approach is your school taking? What would you like to see happen, this term/year and in the longer term? Contact NACE CEO Sue Riley to share your views. 
• The Chartered College of Teaching is hosting a webinar discussion with Rethinking Assessment at 6-7pm on 5 November. Find out more and book a free ticket.

Read more:

• Information sheet: recall and retrieval practice
• Making the grade...time for change? 
• The future of education and data: 3 key questions 
• Lockdown learning: the good, the bad, and the shape of things to come? 
• Beyond the “recovery curriculum”: rebuild, reconnect, reignite 

Ahead of his keynote speech at this year’s NACE National Conference, NACE patron Dr Chris Yapp offers an aspirational yet realistic vision of how artificial intelligence (AI) could help solve current challenges in education.

It is easy to be overwhelmed with both the scope and pace of change in the modern world. Faced with educating children for a world that does not yet exist, for jobs that we don’t yet understand, it is easy to close one’s ears, put heads down and hope it will all go away. Be it as a government minister, as parents, school leaders, teachers or governors, the lack of clarity on the world we are preparing children and young adults for can cause us to become risk-averse and confine our efforts to the areas we feel most confident about.

My experience over a lifetime working in the public, private and third sectors is that this is a mistake. I have seen sectors and companies fail because they clung to a world view that was past its sell-by date.

Adapting to a fast-changing world

When I first became engaged with educational reform nearly 30 years ago, it was common to hear the argument that teaching was inherently a conservative profession and resistant to change.

On the contrary, in my work I have found health professionals, lawyers, accountants, various built environment professions to face the same challenges in shifting professional practice to adapt to a changing world. In all sectors, including education, there are people who thrive on change and others who prefer a comfortable silo.

I have often argued that it is our attitude to change that needs to be rethought, not change itself.

My experience is that in general people do not resist change, they resist being changed. If people feel that change is being thrust on them without their acceptance, engagement and understanding, they will resist. Yet if they feel engaged in and responsible for delivering change, the same changes can feel far less stressful, indeed liberating.

During my time at school and university, the best teachers embraced change and encouraged that in me. The teachers I meet who I find inspiring are constant innovators.

That said, I would argue that society and the economy are facing levels of change over the next few decades – from climate change, technological change and societal and demographic challenges – that require systemic responses that cannot be delivered by individual teachers, educational institutions, advisers and consultants simply “doing their best”.

An aspirational but realistic approach to AI

In this post, I want to illustrate just one strand, the impact of artificial intelligence (AI). Each week I see yet another claim around AI developments in health, education, law, autonomous vehicles and so on. There is a lot of hype around, as ever, but the potential is real, and the possibilities will grow over next few decades.

So, instead of throwing AI tools and techniques at teachers, schools and colleges and hoping something will stick, can we enable students and teachers to become masters, not victims, of AI-enabled change?

I’d like to commend a recent report from Nesta on AI in education which is grounded in the real world yet is aspirational about a more effective, human system of education.

Let’s start by hitting the hype on the head. AI will not replace teachers. While progress in a variety of technologies under the banner of AI is genuinely impressive, we are a long way from the full AI vision. Artificial general intelligence (AGI) is a long-term goal for many in AI research and industry. AGI is about building systems and machines that have a broad range of capabilities that match the breadth of human skillsets.

What we can do now and for the foreseeable future is use AI on specific tasks. The narrower the task, the easier it is to build an AI system that can match a human at that task. We can build AI systems that beat champions at Go, but also outperform doctors in diagnostics.

What existing challenges can AI be used to address?

So, let’s put AI in its proper place, starting with the problems and challenges currently faced by schools:

1. What are the biggest impediments to improving learning at individual, class, institution and national levels?
2. How can we free up teachers’ time to enable them to develop the skills to become masters of change?
3. What tasks can AI perform to support education at all levels from individual to the UK and beyond?

Too often, we have tried to answer the third question without being clear on the first two. Of course, more funding would always help, but it is not a panacea if it acts as a sticking plaster to old ways.

This leads to the next important step…

4. How can education stakeholders engage with developments in AI to shape the capability to address the real challenges and opportunities that AI presents?

Practical and ethical implications

Let me illustrate this with two examples.

The first is in assessment. Imagine you have two pupils who are “B” standard at the end of the year. Based on your experience, Pupil 1 is strong in X and weak in Y, and Pupil 2 the reverse. Understanding next year’s curriculum, you might believe that Pupil 1 might get an A but 2 might move towards a C.

That insight would be very helpful to next year’s teacher but your workload to give far more detailed assessments on each pupil would be unacceptable.

This is precisely the type of task AI is well suited to. Reform of assessment is a perennial bugbear in my experience. We have the technology now to do what many have aspired to for decades, which is to build a more learning-focused, rich assessment framework that supports teachers and learners without adding to the burden and stress on teachers. That is easier to say than do, of course. If we follow the model of building technology and imposing it on schools the potential will be missed, and we will all feel the loss.

The second example is around the ethical implications of developments in AI. We already have examples of AI developments where the systems can be shown to discriminate on gender, race and other grounds. This has profound importance for curriculum developments in teaching AI within schools, but also for the ethical frameworks within which teachers and schools operate. AI, like all technologies, does not exist in a vacuum.

Indeed, I would argue that climate change is an ethical issue as much as it is a scientific discourse. I’ll leave this as an open question here: “Is a grounding in ethics part of the 21st century core curriculum?”

Throughout my years engaged with education at all levels, there has been an aspiration for education to become more evidence-based and research-led. I think we will only deliver on that goal if education professionals are developed and recognised as masters of change. Only then can we thrive on change rather than implement stress-reducing coping strategies. If we are to prepare this generation of young people for the world to come, we owe it to them and ourselves to work together and build an ambitious vision of the teaching workforce.