The secondary measures
Attainment 8 is a points score from a pupil's best eight qualifications, arranged in slots: English and maths, both double-weighted, three further subjects from the English Baccalaureate list, and three others including approved non-GCSE qualifications. A slot left empty scores zero. It is a raw attainment measure and nothing more.
Progress 8 is the one worth attending to. Each pupil's Attainment 8 is compared with the national average for pupils who had the same key stage 2 results in English and maths, and the difference is scaled down by ten. A school's figure is the average across its pupils. Zero is the national average; a score of +0.5 means pupils achieved around half a grade more per subject than similar pupils nationally.
EBacc measures entry to and achievement in a specified combination — English, maths, the sciences, a language, and history or geography. It describes curriculum breadth rather than quality, and schools can raise it simply by changing what they enter pupils for.
Grade 4 and grade 5 figures give the proportion achieving a standard pass and a strong pass in English and maths. Under the 9 to 1 grading, a 4 is broadly where the old C sat.
The confidence interval, which almost nobody looks at
Progress 8 scores are published with a confidence interval, and there is a single rule for using it: only where the whole interval sits above or below zero can anyone be reasonably confident that a school genuinely differs from the average.
This matters more than it sounds. Small schools have wide intervals, so a small school with an apparently dramatic score — in either direction — frequently has a result that is statistically indistinguishable from average. A great many confident local opinions about schools rest on differences that the data does not support.
Two further cautions follow from the same place. Progress 8 is meaningful for comparing groups of schools, not as a verdict on one school and certainly not as a prediction for one child. And when the measure was introduced it produced far more movement in rankings than attainment did, which is itself a clue about how much of any single year's figure is noise.
At primary
The headline figure is the proportion of pupils reaching the expected standard in reading, writing and maths combined at the end of key stage 2. Progress measures also exist, though they depend on a baseline whose arrangements have changed in recent years, so they are less straightforward to compare across time than they appear.
The size of the cohort matters enormously here and is routinely forgotten. In a one-form entry primary, a single child is worth three or four percentage points, so year-to-year swings that look like a school improving or collapsing are often four children having a good or bad year.
What the data cannot tell anybody
The largest single determinant of raw attainment is intake. A school serving an affluent catchment — which in practice means admissions selected by house price — will post strong attainment on unremarkable teaching, and a school serving a poor one can teach superbly and publish modest figures. Progress measures are an attempt to correct for this, and they do it partially rather than completely.
Beyond that, nothing in the tables captures any of the following: whether a particular child will be happy there, what happens when a child starts struggling, the quality of pastoral care, special educational needs provision, what behaviour is actually like in a corridor, how many staff left last year, or how the school responds when something goes wrong. Those are the things parents cite afterwards as having mattered, and none of them is measured.
One practical note on comparability: the years from 2020 to 2022 are not comparable with what came before or after, for obvious reasons, and any trend drawn across that period should be treated as broken rather than informative.
Ofsted, which changed in November 2025
A new inspection framework replaced the 2019 one for state-funded schools, early years settings and further education from 10 November 2025, and for non-association independent schools and initial teacher education from January 2026.
The single overall effectiveness grade has been removed entirely. In its place, a report card gives a grade for each evaluation area with a written explanation, a met or not met judgement on safeguarding, and published data alongside.
The evaluation areas for state schools are safeguarding, inclusion, curriculum and teaching, achievement, attendance and behaviour, personal development and wellbeing, and leadership and governance — plus early years and sixth form provision where a school has them. A not met on safeguarding automatically places a school in a category of concern.
The five-point scale, from highest to lowest, is:
- Exceptional
- Strong standard
- Expected standard
- Needs attention
- Urgent improvement
Inspectors work from the expected-standard criteria for each area as their starting point, which is a different exercise from deciding whether a school is good or not.
On frequency: all routine inspections are now full graded inspections on a four-year cycle, and the old ungraded inspections ended in November 2025. A school with any area at needs attention or urgent improvement also receives monitoring visits — five over eighteen months where significant improvement is required, six over twenty-four months in special measures. From September 2026 Ofsted updates its inspection toolkits annually, with changes published in advance.
The practical gain for a parent is considerable, and it is worth stating. The old single word concealed the most useful information, because a school can be strong on curriculum and weak on inclusion, or excellent academically and poor at handling a child who is struggling. A report card shows that split instead of averaging it away — which means the right question is no longer how a school was graded, but how it was graded on the area that matters for a particular child.
What to actually do with all this
- Read the report card area by area and weight the areas according to the child. Inclusion and the written explanation beneath it, where special needs are in play. Attendance and behaviour, where that is the worry.
- Look at progress rather than attainment, and check whether the confidence interval clears zero before concluding anything.
- Look at three years, not one. Single-year figures move for reasons that have nothing to do with teaching.
- Then visit, and ask what the data cannot answer: what happens when a child falls behind and who notices first, how many teachers left last year and why, and what happened the last time there was a bullying problem — not whether there is a policy, but what was actually done.
- Talk to parents of children already there, and preferably to one whose child has had a difficulty, since that is the only account that reveals how the school behaves under pressure.
The figures themselves are published through the government's find and compare schools service, free and without registration.
Where results suggest a child needs additional help, the evidence on what actually works is under hiring a tutor — and the short version is that the school should be asked first.