A rose by any other name? Reading across old and new Ofsted inspection gradings

Twice in the past month, the HMCI Sir Martyn Oliver has made an identical claim about the new Ofsted inspection framework judgements, in speeches to the Livery Education Conference (14^th January 2026) and the Sixth Form Colleges Association (4^th February 2026). That claim is…

“There is no read-across from the old grades to the new.”

In both speeches, he goes on to explain why he thinks that this is important, and I don’t wish to dispute any of these thoughts here. But it is a striking claim and one that is doubted by many in the education sector who suspect that the introduction of the exceptional judgement is like the introduction of the grades 9 in GCSEs and A* in A-Levels (for those of you with long memories).

This blog is an attempt to make sense of the emerging data from reports on inspections conducted between November 2025 and January 2026, to test out the claim that there is no read-across from the four-point grading system of the 2019 framework and the five-point grading system under the 2025 framework. We will be using SSAT’s inspection tracker to do this, as part of our wider work on helping schools and school leaders in their readiness for inspection.

Reading across: Previous grading

One way in which we can test the HMCI’s claim that there is “no read-across” from the 2019 to the 2025 grading systems is to have a look at how well or poorly schools with outstanding, good, requires improvement or inadequate judgements have done in their new inspection gradings.

To do this, we have given a numeric equivalent to each grading from exceptional (5) to urgent improvement (1) with expected standard given a value of 3. Then we have averaged the total of all gradings for all elements of the new framework awarded since November and linked them with the school’s previous overall inspection judgement.

There has been only one previously inadequate school with a new framework inspection report and so we have not included the data here, as it may be far from representative. We have also not included schools without a previous inspection (i.e. new or newly academised schools) but we have included the mean judgements for all schools in the final column.

Finally, we have colour-coded cells with the strongest outcomes for each row in green and the least strong in red, with the middle performance highlighted in yellow.

	Outstanding	Good	RI	All schools
Achievement	4.22	2.79	2.13	2.66
Attendance and Behaviour	4.11	3.03	2.43	2.98
Curriculum and Teaching	4.22	2.81	2.27	2.73
Inclusion	4.33	3.18	2.49	3.09
Leadership and Governance	4.22	3.04	2.38	2.96
Personal Development and Wellbeing	4.22	3.31	2.83	3.25
Mean of all judgements	4.22	3.03	2.42	2.95

What this data tells us is that previously outstanding schools are, on average, performing the best in judgements for all elements of the new inspection framework. Perhaps more to the point in assessing whether there is any read-across from the old inspection framework, these schools are averaging above the strong standard and towards the exceptional judgement.

By contrast, schools formerly graded as being requiring improvement are, on average, performing the worst in judgement for all elements of the inspection framework. They are averaging closer to the needs attention judgement than to the expected standard in all areas, other than personal development and wellbeing. Even by this measure, though, they are not quite meeting the expected standard when their data are aggregated and averaged.

In the middle, then, are the formerly good schools, whose performance against the new inspection framework judgements, sits comfortably in the middle of that shown by the formerly outstanding and formerly RI schools. When all the gradings for these schools are averaged out, they sit almost exactly on the expected standard score.

The smallest gap between the previously outstanding and previously RI schools are for personal development and wellbeing (1.39 grades difference), which is easily the least challenging of the seven judgement areas based on the average for all schools.

The biggest gaps between the formerly outstanding and the formerly RI schools are related to achievement (2.09 grades difference) and curriculum and teaching (1.95 grades difference). These are the most challenging judgements according to the mean for all schools.

So, by this analysis, what we can see are that:

• Formerly outstanding schools are averaging at the strong standard and towards the exceptional judgement
• Previously good schools are coming in at almost exactly the expected standard
• Schools that required improvement at their last inspection are averaging at the needs attention level of the new framework.

In short, then, there clearly is a read-across between the old judgements and the new ones, except for the exceptional judgement (more on this below).

Reading across: Indicative overall gradings

A second way of testing the claim that there is no read-across, is to have a look at the outcomes for schools with a previous overall judgement grade and their average performance across the new grading system. This is not unproblematic as it involves our ascribing an indicative overall grading to each school. To do this, we average out each school’s grading for each component of the new framework and use it to award an indicative overall grading. So, for example, we would identify a school with a mean grade between 2.51 and 3.5 as being of expected standard.

I appreciate that this methodology runs against the spirit of the changes made since September 2024. We do not publish these indicative overall grades anywhere and only use the data generated to provide helpful analysis for schools and their leaders. Our intention is to help.

The table below is an example of this system-focused data analysis. Across the top are the overall effectiveness gradings of schools inspected under the new Ofsted framework. Down the side are the indicative overall gradings generated using the methodology described above. To see what happened to schools with previous judgements under the new framework, we have colour-coded certain squares dark, medium or light green to signify the frequency of new gradings. The darker the colour, the stronger the frequency.

	Outstanding	Good	RI	All schools
Exceptional	22.2%	0.8%	0.0%	0.9%
Strong	77.8%	15.0%	4.3%	14.2%
Expected	0.0%	65.3%	43.0%	61.4%
Needs Attention	0.0%	18.9%	32.3%	20.0%
Urgent Improvement	0.0%	0.0%	20.4%	3.5%

Over three-quarters of previously outstanding schools have an indicative strong standard, when all their gradings have been aggregated and averaged. The remaining schools – over one in five – that were formerly outstanding had an average grading of exceptional.

Almost two-thirds of formerly good schools averaged out to the expected standard, with almost all the remainder fairly evenly split between the strong standard and needs attention. Less than 1%of this cohort achieved a set of judgements that averaged at exceptional.

Although the expected standard was the most common indicative overall judgement for schools that were RI, there was more chance that the schools would fall below the expected standard and only a very small chance (fewer than one in twenty) of a previously RI school achieving an average indicative of a strong standard.

Using this extrapolated data, the table provides more nuanced findings than the one above. The key insight is that the better the school’s previous grading, the stronger the correlation with the new grading system. Previously outstanding schools can be more confident, based on this data, of a strong standard or better than good schools can be of an expected standard or better. And previously good schools, in their turn, can be more confident than previously RI schools can be, in self-evaluating at expected standard or needs attention.

This table, then, would further support the thesis that there is a read-across from the previous inspection grades to the current ones but also suggests that the higher the previous inspection grade, the more confidence schools can have in that read-across.

Reading across: Overall gradings over time

One final way of assessing the claim that “there is no read-across from the old grades to the new” is to use the data to explore patterns in the awarding of grades over time. The table below shows the proportion of improvement needs we have curated since September 2023, linked to the overall effectiveness grade or – since September 2024 when these were no longer used – our indicative overall grades (see above for an explanation of these). The colour-coding here simply relates to the 2019 (peach) and 2025 (lilac) frameworks being utilised at the time.

	Exceptional	Strong Standard or Outstanding	Expected Standard or Good	Needs Attention or Requires Improvement	Urgent Improvement or Inadequate
Sep 23 to Jul 24*	N/A	8.5%	68.4%	18.9%	4.2%
Sep 24 to Jul 25	N/A	8.6%	69.1%	17.8%	4.5%
Nov 25	1.8%	24.4%	64.1%	5.5%	4.1%
Dec 25 to Jan 26	0.3%	7.8%	59.5%	29.4%	3.0%

* Indicates actual overall effectiveness grades. All other rows are based on indicative overall grades.

One line that stands out is the data from November 2025, when the schools inspected were all volunteers piloting the new inspection framework. A far greater proportion of these schools achieved indicative overall gradings of strong standard or exceptional than achieved the same gradings after that time, or outstanding gradings in the two academic years before this. This is not surprising given that these schools were likely to ready for and confident about inspection.

This table also shows that the proportion of outstanding gradings from the 2019 framework and strong or exceptional gradings from the 2025 framework, beyond the volunteer cohort of schools, is very similar at 8.5%, 8.6% and 8.1%. This suggests a read-across from the old gradings to the new ones at the very highest levels, albeit that they seem to be a little tougher to achieve.

The biggest change between the old Ofsted framework and the new one is that the expected standard appears to be noticeably more challenging to achieve than the previous good judgement.

Consequently, the new needs attention grading is more common than the previous requires improvement judgement was for schools. There has been a notable reduction in the proportion of urgent improvement gradings compared to the prior inadequate grading, but the data in this area are too small at present for a reasonable level of confidence.

Indeed, the data from the new framework inspections are still too few to reliably draw concrete conclusions from, but they are pointing us in a direction that will be interesting to follow in the months to come, as more of the new reports drop. And that direction appears to be that there is a strong correlation (or read-across) for the new exceptional or strong standards with the outstanding judgement, and a weaker, but still noticeable, correlation between the expected standard and the good judgement, and between the needs attention and requires improvement judgements.

This table, then, supports the hypothesis that there is a read-across from the old grading system to the new one, but that the boundaries have shifted downwards in the early stages of the application of the new framework. This downward shift is more pronounced for schools fighting for the expected standard over the needs improvement judgement.

Conclusions

Although we are still in the early stages for the analysis of new gradings for schools, there does appear to be a read-across from the old inspection judgements and the new ones, despite the claims of the HMCI.

Previously outstanding schools are routinely securing the strong standard and are much more likely to achieve exceptional judgements than schools with other previous gradings. The pattern is more nuanced for previously good schools in achieving the new expected standard and for schools that were formerly requires improvement in achieving the new needs attention grade. The read-across, or correlation, is weaker for these schools than for those with high prior inspection outcomes, but it is there, nonetheless.

These findings are bolstered when we look at data on the gradings awarded under the old and new frameworks, regardless of schools’ prior gradings. Fewer schools are hitting the expected standard than used to achieve the good judgement and more schools need attention than required improvement in the past. This may be evidence that there is no read-across from the old gradings to the new ones. It may also be evidence that schools close to the expected standard are falling short because of new and as-yet-unfamiliar inspection processes.

Time, and further analysis of the data, will tell.