One of the main conclusions of last year’s Accountability Commission report, Improving School Accountability, was that an obsession with pupil test results was narrowing the curriculum and failing schools serving the most deprived communities.

Credit, then, to Ofsted’s chief inspector Amanda Spielman for attempting to tackle this head-on through changes to school inspection. Ofsted’s proposed new framework aims to extend the focus of inspection beyond test performance to look more broadly at the substance of education.

The ambition is sound: few would argue that a better balance is long overdue.

Early feedback suggests that there has indeed been a shift of emphasis in the pilot inspections taking place this term. In the last few weeks, I have heard reports from some headteachers, inspectors and those that have observed these practice inspections that the quality of discussion on curriculum has been good. School leaders want to talk to inspectors about what they are doing and why they are doing it. Inspectors are enjoying talking about things that actually matter, and observers have noted a positive professional dialogue between the two.

Perhaps we shouldn’t be too surprised. Having a low-stakes professional discussion with an experienced HMI is generally a very good thing. A number of years ago I led the subject inspection team at Ofsted. These were one-day inspection visits to a sample of schools, designed primarily to gather evidence on how the system was performing overall rather than make a judgement on individual school performance.

These were, on the whole, positive experiences all-round. Once school leaders realised the terms of the visit - that they were not going to receive a judgement or a possible public shaming at the end of it - conversations generally became more open and honest regarding strengths, challenges and vulnerabilities. These inspections were led by genuine experts in their field: HMI national advisers. On more than one occasion these visits were described to me by heads as ‘the best free consultancy I’ve ever had’.

They worked for two reasons - firstly, there was not an overall judgement at the end of the visit and secondly, the inspectors that conducted these visits were truly expert in their field.

Which is where current proposals for scrutinising curriculum become somewhat problematic.

Discussing the choices that a school makes in design and implementation of the curriculum is one thing, but it is a completely different order of expectation to be able to make a reliable and consistent judgement about how good those decisions were, with little time on-site, without simply defaulting to pupil outcomes.

As drafted, the evaluation criteria that inspectors will use to inform their judgements is of little help - it is imprecise, vague and open to wide interpretation, and there is nothing to say how inspectors should weigh and balance the many aspects that make up the “quality of education” judgement.

In the right hands, ambiguity could actually be a good thing. I might trust the very best HMI to apply flexibility to good effect, drawing on their insight gained through hundreds of inspections to make, on balance, the right judgement call about whether a school is good or outstanding.

But ask yourselves this: do you want an inspection system that is quite so subjective?

Would you be happy for an award of ‘outstanding’ to be the educational equivalent of a Paul Hollywood handshake on ‘Bake-off’?

Because, under this system, an enormous amount rides on how good that inspector really is. And the big obstacle remains that the majority of inspectors are not HMI - they are serving leaders who do, on average, nine days inspection activity a year.

So for every Paul, Prue and Mary, there is a Sandi, a Noel, a Mel or a Sue - all brilliant at what they do but not necessarily well placed to judge a baking competition.

It is asking a lot of school leaders, even with the very best training, to make consistent judgements with so much that is open to interpretation in the new framework. This is doubly true for those that will be asked to inspect out of subject or phase.

We should, therefore, be very cautious of rushing too quickly to embrace curriculum inspection as currently designed. If we are to retain high stakes judgements following inspection, first and foremost we must ensure that those judgements are fair, reliable, and consistent. And with so much at stake for individual schools, we should be wary of creating an inspection system where judgements are overly influenced by personal taste.

Nick Brook is deputy general secretary of the NAHT headteachers’ union, and chair of NAHT’s accountability commission