The Story of the Individual: Why Exam Results Days Tell Us Nothing

How compensatory grading masks truth, misleads the public, and fails our pupils

Every August, the exam results ritual follows the same pattern. The media puts out headlines, politicians put out soundbites, and school leaders put out proclamations of success. Meaningless numbers and slogans fly around: “X% Grade 9s!” “Highest ever!” “Biggest drop since records began!”

Why do I call this meaningless?

Well, because these headline figures, for all their polish and prominence, tell us almost nothing. They neither help children grow nor help teachers act. They serve as a kind of national theatre, a performance for the benefit of the policymakers, rather than a moment of reflection that will help our next cohort of pupils. And while the cameras roll and charts flash across screens, the true story of what education is, in my view at least, meant to be about – a human, personal, transformational and meaningful story – is quietly overlooked.

That the national figures mask all nuance and reveal nothing about how learned a cohort is, is bad enough.  That, at the level of the individual school, the headlines are often simply untrue is unforgiveable.

One year’s cohort differs profoundly from the next. The demographics shift. Prior attainment varies. Even the local context – economic pressures, staffing changes, community events – can render cohort comparisons effectively meaningless.

But still, the race is run. One school against another. One year against the last. And all the while, the question that actually matters – what did this individual child come to understand? – is nowhere in the discussion.

What we see is like Goodhart’s law in action: when a measure becomes a target, it ceases to be a good measure. Exam grades and headline figures were designed to tell us something about standards. But once they became the scoreboard for political success and institutional bragging rights, they stopped describing reality and became just a game.

This year has provided a particularly absurd example. Because of the pandemic, Progress 8 figures cannot exist. The baseline data they depend on does not exist. And yet, that has not stopped several Multi-Academy Trusts from announcing dramatic improvements in their Progress 8 scores.

How have they dared to do this? By leaning on mathematical models of such weakness and dubious validity that they would be laughed out of any serious statistical review. These figures are conjured into being not because they tell us something true, but because a certain type of CEO craves the headline. They allow the organisation to claim victory in a race that has not even been run.

It is as though we are keen to create an even purer form of Goodhart’s law: once the measure itself becomes the goal, truth is no longer required. The number matters more than the reality it was meant to describe. And when leaders collude in this fiction, the profession corrodes itself from within.

I think people are generally unaware that England operates a system known as compensatory grading. This is not a standard where “Grade 7 means X level of knowledge” and is preserved by ensuring the content and difficulty of exams remain consistent over time. Rather, compensatory grading means that pupils’ high scores in some areas of the curriculum can compensate for lower scores in other areas. This results in a grade that indicates general aptitude rather than guaranteeing mastery in specific areas of a subject. England's system also allows some fluctuation in exam difficulty because grades are awarded to maintain a statistical distribution of outcomes year on year – the idea is to attempt to create stability in grading over time.  You might think of this as an annual moderation of awarding bodies.

In theory, this preserves “national standards.” But in practice, it creates a fiction. A pupil receiving a Grade 6 in 2025 is not necessarily more or less accomplished than a pupil who received the equivalent grade in 2015 or 1985. The grade is anchored to other candidates, not to any fixed measure of what has been understood.

This is deeply attractive to politicians. It creates the appearance of consistency, which is useful in press releases and elections. But it does little for the individual pupil. Indeed, I would argue, it harms them because it prevents us from acting honestly and planning with intellect for the coming cohorts.

While compensatory grading may maintain the appearance of stability, longitudinal research tells a different story. In mathematics, for example, when pupils today are presented with the same exam question as their counterparts in 1985, performance is significantly lower. There has been a quiet erosion of standards, even as grades have remained outwardly stable.

The system designed to meet policy needs, not educational ones.

And because grades are awarded not for what a pupil knows but for the score the achieve, a pupil can gain a Grade 5 in GCSE English or maths, for example, and yet harbour profound misunderstandings. They may have never grasped the structure of a paragraph, or the meaning of place value. The grade, in this case, becomes a mask – hiding the conceptual gaps that will later cause confusion, disengagement, or failure.  We see this all the time when pupils progress to the next step in education and attempt to continue a subject when their apparent mastery of it is built on sand.

 Here again is Goodhart’s warning. Once the measure (grades) became the goal, it obscured the very thing it was meant to reveal. The number may look reassuring, but it may hide a hollow core. The measure, repurposed as a target, has lost its value.

Education should not about statistics, it should be about people, about individuals and their specific story. About the long, slow journey toward becoming. Toward becoming mathematical, or historical, or articulate, or scientific. Toward developing the habits, knowledge, and ways of seeing that connect them with the world.

That is why I am more or less uninterested in exam results days. I absolutely care about outcomes, but not the statistic that hides the truth.  I care about the outcome for the individual.

You cannot apply a statistic to an individual. What matters is their story. Their progress. Their relationship with the discipline. Their readiness to continue the journey.

I know I have banged on about this forever, but I’ll continue to call for a better way.  It does exist.  And we know it.  It is this: Criterion-referenced assessment.

Criterion-referenced assessments tell us what a pupil has mastered and what they have not. They do not hide behind distributions or comparisons. They are not designed to produce fixed proportions of success or failure. Instead, they are designed to support learning – to show teachers what to revisit, to show pupils what to practise, to illuminate the next steps and to reveal the unfolding story of a subject, its connections and the webs of ideas that are forming.

I appreciate that these assessments are not easy for governments to stomach. They can result in fluctuations in national averages – something few ministers want to see happen on their watch. But they are honest. And they make a better world, because they help each child to become more.

Results day will truly matter it no longer provides another opportunity to spin headlines. Results day will truly matter when it is a celebration of individual journeys – of what each young person has overcome, achieved, and begun to understand. It should be rooted in deep knowledge of the child, not in statistical vanity.

We have built a system around Goodhart’s law, and then acted surprised when it delivered what it always delivers: numbers that look neat, headlines that sound strong, and a truth that grows weaker with every passing year.

When will we stop pretending that aggregate scores are meaningful? When will we stop designing assessment systems for the benefit of technocrats and ministers?