What Ofsted’s science comments show that results day doesn’t

Ask a Head of Science where the attainment gap “shows up,” and most will point at results week.

Fair enough. That is when the numbers go public, the trust dashboard lights up, and the narrative writes itself: some schools pull ahead in GCSE science; others do not.

But results day is the end of a longer story. On 18 September, SchoolDash published a Gatsby-commissioned analysis of Ofsted secondary reports — reading what inspectors have been saying about science education. Not only final grades: teaching, staffing, enrichment, pathways and assessment, and how those comments land differently by intake.

The gap is visible on results day. Ofsted’s science comments suggest it has been forming earlier.

We already know outcomes are uneven. We have written before about what the widening GCSE regional gap means for secondary assessment strategy. That piece stays on the outcomes side of the story. This one does not remake it.

SchoolDash’s point is more uncomfortable for anyone who thinks exam results are the whole equity problem. Inspectors have spent years writing qualitative remarks about science lessons, labs, pathways and staffing. Their method turns that archive into something closer to a comparable picture — with an important caveat we will come back to.

They processed 40,641 secondary Ofsted reports published as PDFs on or before 27 June 2026, then focused on 31,748 where science mentions were substantial enough to analyse — coding themes such as attainment, teaching, staffing, curriculum, assessment and enrichment, plus sentiment.

The headline finding is blunt.

Schools serving higher proportions of pupils from poorer families, or higher proportions of lower-attaining pupils, received consistently less positive inspector comment on their science provision. The gap is largest for academic attainment — which matches what we already see in exam outcomes — but teaching quality, enrichment, assessment and curriculum also show clear disparities.

That is the upstream picture. Not only “these schools got worse grades.” Also: inspectors were less positive about the conditions that produce grades.

When inspectors were negative about science teaching, the same themes kept resurfacing: staffing instability and a lack of specialists; inconsistent quality and low expectations; passive methods; weak assessment, marking and feedback; and a lack of practical enquiry.

Read that list as a science department, and it stops sounding abstract. Non-specialists covering classes. Feedback that does not actually correct the misconception. Practicals squeezed out because the room, the technician, or the time is not there. Assessment that records a mark but does not change tomorrow’s lesson.

That last one matters. A mock only earns its keep if the information comes back in time for someone to act — and if the feedback is specific enough to change what a student does next. We have written about that trap in The Feedback Graveyard: marking labour that never becomes student learning is just exhaustion with a paper trail.

Through Ofsted’s own words, weak feedback and misconception-handling show up repeatedly in the negative teaching comments — and that picture looks uneven by intake.

Curriculum comments tell a related story. Inspectors flagged inadequate planning and sequencing, timetabling pressure, narrowed or inappropriate pathways, thin practical work, and resource or staffing constraints. Ability-appropriate routes cut both ways: lower-attaining pupils steered into options that limit future routes; higher-attaining pupils denied access to triple science.

That is not only an inspection anecdote. The same analysis shows large disparities in GCSE double and triple science entry rates by geography and, even more starkly, by school deprivation. Triple science entry is higher in more affluent intakes; in the highest-deprivation schools, a meaningful share of pupils are not sitting science GCSEs in the same pattern at all.

If your department has already lived through the triple science entitlement debate, none of this will feel theoretical. Access to a demanding science route is part of the equity story long before a grade appears on results day.

SchoolDash is careful about causation, and so should we be. The method uses large language models to code themes and sentiment. They note that worse sentiment could mean worse provision, or harsher judging of schools serving poorer communities. The available data cannot distinguish those two possibilities cleanly.

That caveat does not make the piece ignorable. If provision really is weaker upstream, results-day gaps are partly a delayed read-out of earlier inequality. If inspectors are simply tougher on disadvantaged schools, that is a different equity problem — and still a serious one. Either reading should make science leaders uneasy about treating GCSE outcomes as the only dashboard that matters.

National datasets will not timetable your physics cover or conjure triple-science specialists out of thin air. They can still sharpen what you refuse to outsource.

If Ofsted’s science comments keep landing on weak feedback, misconception-handling and inflexible pathways, the departmental question is not “how do we look better on results day?” It is: where, in Years 9–11, does assessment actually diagnose and act — and where does it only produce a number after the moment to reteach has gone?

That sits in mock design, question-level review, how quickly scripts turn into teaching decisions, and whether pathway conversations are backed by evidence or by habit.

At Excelas, we spend a lot of time looking at science and maths scripts after the papers have been sat — the layer where marks become information teachers can still use. We are not an awarding organisation. What the Ofsted archive underlines is simpler. Science equity is not only a recruitment story, and it is not only a results-day story. If inspector comments keep pointing upstream — to teaching quality, feedback that corrects misconceptions, and pathways that stretch or narrow — then departments that treat assessment as timely diagnosis, rather than delayed confirmation, are working on the part of the problem they can still touch.