Two grades, three categories, central data analysis and no aggregation. Go on Amanda Spielman, urges Matt Hood, what do you say to that?

We have a new Ofsted chief inspector agreed by Number 10 and awaiting sign-off. Hurray. But what should Amanda Spielman do on day one? Ofsted has already made big improvements in the quality and consistency of their inspection teams. But there’s more to do.

If it was me, what would be the first order of business?

1. Get rid of four grades and replace them with two – compliant and not compliant

It’s not Ofsted’s job to provide headteachers with “outstanding” logos to fly above their schools. Nor do I think Ofsted can reliably distinguish between good and outstanding schools.

By continuing with the divisive good/outstanding approach Ofsted will damage behaviour and hamper innovation. I want people to worry about being compliant but I don’t want them to stop innovating because they worry about not being “outstanding”. Ofsted should think food hygiene inspector not Michelin star-awarding food critic.

2. Replace the existing four categories in Ofsted judgments with three new ones –strategy, implementation and safety –and inspect each differently

The current categories don’t capture the two pillars of good leadership: having a good plan and making it happen. By looking at themed categories it’s not always clear what aspect of leadership you’re looking at.

In a perfect world the first part of an inspection – booked a week in advance – would focus on strategy and implementation.

Strategy would involve a select committee-style submission of evidence (the school improvement plan) followed by a Q&A with the leadership team. The question would be simple: “does the school know itself and have a credible long-term plan to continue to improve?” If the plan is credible, the school is compliant.

To test implementation, inspectors would explore lines of enquiry from the school strategy (maybe six – half chosen by the school and half chosen by the inspectors). The question would change to: “does the inspection team have confidence that the school can implement its plan? Is the school actually doing what it says it is doing?” If it can demonstrate the capacity to implement then the school is compliant.

Safety, meanwhile, should be separated from strategy and inspected on a no-notice basis.

3. Separate data judgments and give them to a national team of specialists, and judge them as compliant/non-compliant

Understanding school data is difficult. I regularly see school leaders and inspectors making statements about progress and attainment that lack even a basic understanding of statistical significance, correlation, causation or standard deviation. Training every inspector to be an analyst is neither possible nor necessary.

Taking data judgments out of the hands of inspectors and giving each school a compliant/non complaint rating on their data linked to contextual floor targets (that is, a standard below which schools should never fall relative to its intake) makes more sense.

4. Make judgments on strategy, implementation, safety and data distinct; do not give a single overall grade

By keeping the four judgments separate and discussing them separately, you’d immediately jump to a more nuanced discussion about what a multi-academy trust/local authority (MAT/LA) might
need to do next.

It would also give regional school commissioners (RSCs) a clear idea of what support the MAT/LA should be providing, and clarify exactly under what circumstances a change in school sponsor would be required.

For example, if a school had compliant data and strategy, but was struggling with implementation, an MAT/LA should provide implementation support. Only where a school was non-compliant across multiple categories over time would a change of sponsor be required.

Safety is different: any school found non-compliant in this would risk being taken over unless rapid improvements were made, regardless of other factors.

Under this approach, improving schools are given space if they are compliant with strategy, implementation and safety –even if their data is not (yet) compliant. Coasting schools would be challenged on their lack of strategy even if their data and safety was compliant.

And poor leaderships who talk the talk on strategy but fail miserably when it comes to implementing plans would be called out.