As you may have seen, Ofsted have published a report which lays the ground work on how they might start observing lessons once more: Six models of lesson observation: an international perspective.

Most people will probably accept that if Ofsted are going to inspect schools then should almost certainly observe lessons as part of the inspection process. And, as someone who spends a fair bit of time visiting schools around the country, it’s clear that you can learn a lot about a school from seeing how lessons unfold. But when I observe lessons, I do so informally. I’m not attempting to make a judgement. Sometimes I’m trying to see if there are any themes evident, and anywhere I might be able to help. If I think I’ve spotted anything interesting or seen evidence of a problem I might have some insight into solving, I’ll talk, in general terms, to someone on the leadership team and ask them what they think. Other times I’m observing in some sort of ‘coaching’ capacity where the goal is to help a teacher improve some aspect of their practice. In this case I’ll write down a list of questions as they occur to me and then, after the lesson, I’ll use them as a means for discussing what should come next. I try really hard to withhold any kind of judgement and assume – until proved otherwise – that the teacher is aware of everything I’m aware of and is at least as knowledgable and thoughtful as I am.

Both of these approaches have some merit, and neither have anywhere near enough reliability or validity to make any kind of high stakes decision about a teacher or a school. Happily, this is a battle that appears to have been won. Ofsted have stopped judging lessons, and, slowly, this is starting to trickle through the system with more and more school leaders beginning to understand that judging lessons leads to perverse incentives.

But, as anyone who’s been through an inspection in the last few years knows, even though Ofsted may no longer be judging individual lessons, they still observe them. And they still judge the quality of teaching and learning in a school. In the current inspection framework, it’s clear that there is nothing that “Oftsed wants” and that schools are free to decide their own priorities and preferences, as long as results are good enough. Part of what inspectors currently do when observing lessons is to check the convergence between what a headteacher says is happening in her school, and what teachers are actually doing. If the head says that all teachers meet and greet students at the classroom door, inspectors will check to see whether they do. If inspectors are told that all teachers mark in purple pen and write down ‘two stars and a wish’ after every piece of written work, they will have a look to see it this happens. If inspectors discover evidence that policies are not being followed, then schools might find their ‘quality of leadership and management’ takes a hit in the report.

This is a bit of a weakness of the current inspection regime. It assumes that one approach to teaching is as good as any other, and it encourages school leaders to look for compliance in their teachers. One wonders what would happen if a school went about enforcing a patently foolish policy and got complete compliance? Wouldn’t it be better if inspectors used their power to guide school leaders towards certain practices and away from others?

The research supporting teacher-led instruction, and against enquiry learning is increasingly well-known and, whether you agree or not, you’ll no doubt be aware that there’s a growing consensus that some pedagogical approaches more likely to be effective than others. This being the case, should inspectors be tasked with making sure that teachers understand and are using, say, Rosenshine’s Principles of Instruction?

Even though I wholeheartedly support these principles, such an idea fills me with horror.

No matter how you go about collecting data, and no matter how carefully you try to aggregate what you see across a whole school, getting inspectors to look for ‘good’ and ‘bad’ things will lead to perverse incentives. In the bad old days, Ofsted was the ‘child-centred inquisition’ burning teachers who talked for too long at the stake. Group work, ‘active’ and ‘self-directed’ learning were held up as unquestioned good things and any dissent was crushed. We can look back with a shudder and say that obviously, this was a silly thing to have done. Not only did it result in lots of pointless group work, it ended the careers of some excellent teacher raconteurs. But maybe, just maybe, Christine Gilbert and her advisors thought they had the right of it? Maybe they didn’t know they were the baddies?

Even if the research really did show that enquiry learning was unequivocally bad all the time (it doesn’t) forcing teachers to step into line stops them thinking. If we really want teachers to be their best we must hold them to account intelligently. The three sacred principles of any well-run accountability process are as follows:

1. School leaders and teachers know they will be accountable to an audience
2. the audience’s views must be unknown
3. the audience is well-informed and interested in accuracy.

When we decide we know better than a classroom teacher how they ought to teach their classes, no matter how pure our intentions might be, we inevitably end up doing something foolish. Instead, we should always ask teachers to talk about the reasons for the decisions they’ve made, and then listen. Teachers should be asked what they think needs to be done and what support they need to make these things happen, and then hold then to account for doing whatever they’ve said they should do.

Unless we believe those holding us to account are interested in accuracy rather than simply having their preferences met, we tend to become fearful, dishonest and risk averse. If those in authority – whether it’s school leaders or Ofsted – pre-define what good looks like and hold us to account for meeting a set of standards, we will give the appearance of meeting those standards. This results in two equally undesirable outcomes:

  1. Compliance – some people will just do whatever they are told to do. Some will do it well, others will struggle. They will assume managers know best and try hard to please them.
  2. Pretence – some people will feel they know better and assume managers are foolish or corrupt. They will sometimes give the appearance of playing the game, but will, as far as possible, ignore the accountability process.

Some of those who are successfully compliant will feel pretty good about being able to meet inspectors’ demands but everyone else will experience a combination of guilt, fear and anger. None of these emotions are particularly useful for improving teaching. The bitterest irony though is that even when these accountability systems appear to be successful they promote a lack of curiosity and blind adherence to a set of partially understood principles. We lose the ability to make considered professional judgments and embark on Cargo Cult teaching – following the forms and structures of instruction but without understanding the underpinning theory or science.

And that’s the crux of it – teachers should be encouraged to be professionally sceptical and to ask “Why?” If inspectors want lesson observation to help them calibrate the effectiveness of a school they must resist the temptation to think they know best and to instead ask teachers and school leaders why they are doing what they’re doing. And then listen.

Each of the proposed lesson observation frameworks considered in the new report share the same basic weakness: using any kind of generic observation check list is “not even wrong.” Even if you have iron clad evidence that you’re looking for the right things, you’re missing the point. The point is this: how teachers teach is only the tip of a very large iceberg; the question of real importance is whether how they’re teaching is matched to what they’re teaching. If we’re genuinely interested in improving the system then we need to stop messing about at the point of pedagogy and focus on the quality of the curriculum. It’s not that how teachers teach is irrelevant, it’s just that what they teach is both more important and this importance is less well understood.

Ofsted inspectors should definitely observe what goes on in lessons, but equally, they should resist the temptation to judge.