As our thinking about what works in education develops, the concept of formal lesson observation conducted by school leaders (and visiting consultants and inspectors) seems to me to be increasingly flawed – even preposterous. And yet, it is deeply embedded in our systems and culture that in doing their duty in relation to securing high standards of teaching and learning, leaders should observe lessons, evaluate ‘quality’ and then provide a cocktail of judgements and feedback. Teachers expect this to happen even if unions set out policies about the maximum level of observation that is acceptable.

But what if ‘formal observation’ of teachers by leaders in these one-off set-piece episodes that happen a few times a year or happen on the fly – with emailed feedback appearing in people’s inboxes – is just a giant waste of time? What if it’s just people playing out the process they think they’re meant to be doing in the (false) hope and (false) belief that these rare intense events and exchanges drive standards and improvement – doing it because we always have rather than because it’s the best thing to do.

I think we need to stop. Instead the whole business should be reconfigured around instructional coaching. That should be the standard; the habit; the given; the default setting: all teachers engaged in an ongoing process of professional development (and, yes, quality assurance) through multiple coaching cycles.

What’s wrong with top-down observation?

I’ve explored this in a few posts in the past:

---

In that post I highlight just how little impact observers have on the actual quality of teaching relative to the self-directed actions that teachers decide and enact for themselves. And yet all across the land, SLTs devise comprehensive systems for making sure all teachers are observed and that they receive feedback. This has many forms including:

• Graded formal observations: This really is the pits; something of a stain on our system that grading of one-off lessons persists. I can barely be bothered to talk to people who justify the whole delusion of this nonsense – but there it is.
• Formal lesson observations twice year supported by detailed observation forms, filled in in detail by the observer and kept on file, accompanied by a giant post observation debrief. Twice a year.
• Fly-by lesson observations where leaders go on learning walks and then issue rapid-fire feedback – sometimes in person but often via email or a pigeonhole slip. These are entirely legitimate, important fact finding missions in themselves – leaders should see their school in action as much as they can. But to use these snapshots to issue fly-by feedback about what a teacher could do better is something I increasingly struggle to justify.

The problem with all these systems is that they are built on presumptions that are often just wrong:

• that the observer has insights of value by dint of their status in the hierarchy
• that their judgements and feedback are accurate, correctly identifying the factors that could be addressed or reinforced
• that, if conveyed to the teacher directly, the feedback will be absorbed and thus will improve quality in the long run
• that the judgements and feedback have some kind of status – power, meaning, weight – that a teachers’ self-evaluation doesn’t have.
• the frequency and depth of the top-down fly-by analysis and feedback is optimal for driving improvement.

None of these things is necessarily true. Think about it. An experienced, effective teacher has a visit from an observer that lasts 10-15 mins. They later receive feedback (verbally or in writing) telling them their lesson was ok but could have been better for stated reasons. What? There’s the teacher – in that classroom week in, week out, knowing the curriculum and the students, and you’re going to drop in and then presume to judge? To evaluate? To fill out all those boxes on your neat little form and give advice? WITHOUT ASKING THEM FIRST?? Get out of here. It’s so obviously wrong -surely?

What should happen instead?

In the work I’ve been doing with Oliver Caviglioli and various schools and colleges I support, the scenario we envisage as the ideal is depicted here. The observer and teacher, side by side, on a level, talking through the problems and actions associated with teaching the class in question. I would say that, as a minimum, all feedback from lesson observations should be communicated face to face with the fundamental principle that that it is the teacher’s meeting, for them, about their teaching and their students. They own it; they drive it – and, ideally, they record it. The feedback isn’t given and received – it is co-constructed and agreed. The nature of these feedback conversations is superbly captured by Bambrick-Santoyo. I summarised his ideas in this post:

---

In this context, observers suspend judgement. However ‘good’ or ‘poor’ they think the lesson may be in their private subjective view, they look for specific ways in which they can help the teacher so that students could be supported to learn more. They solve the learning problems. That’s all that matters.

---

At a basic planning level this means that lesson observations need to be linked automatically to feedback conversations. Time constraints now come into play so we have to talk about short observations and short feedback meetings. And that’s ok. The bloated dissection of a one-off observation is pointless and ineffective relative to a series of short interactions spread out over time, focusing on specific elements of practice.

Recently I was asked my thoughts about designing a lesson observation form. Although I’ve done this plenty of times in the past, now I wouldn’t. I would say to an observer – sketch notes of what you see in anyway you like to inform your conversation but keep them to yourself. The only things we need to record are the concrete actions steps that a teacher agrees on as part of their feedback conversations. No judgements; no loaded ‘quality’ language with all its inherent biases. Just the specific actions that will follow. Who writes them down? The teacher – via whatever recording tool is found to be helpful: lean, accessible, no fuss.

Why is instructional coaching even better?

The optimal model is where every teacher works with someone with the required level of expertise to support an instructional coaching process. Here, the assumption from the beginning is that the coach will engage with the teacher on a sustained basis over time – weeks, months, terms – seeing them teach regularly and meeting to discuss their progress regularly, each time helping them to identify problems and associated action steps; helping them to sustain practice on specific aspects of their craft. The log of action steps serves to inform the process as it moves through various cycles of review and improvement. The frequency varies according to need.

At a logistical level, schools need to map out a coaching tree of some kind. How many people can any one person coach? Three? Five? Eight? It depends on their loadings in general. How many teachers are there? So, how many coaches need to be identified – including SLT, middle leaders and other expert teachers. ? Once you map this out it seems doable. If I have my allocation of people to coach, I set about getting to know them, meeting them, observing them, supporting and challenging them.. week to week, month to month. This is the engine of real sustained improvement in quality.

Of course, in addition, there will be general learning walks and themed quality reviews etc. But these serve the purpose of giving leaders insights that inform decisions about CPD, assessment, curriculum. They are not used to give ad hoc fly-by feedback to individuals. That’s just not the deal. We learn to acknowledge the inherent poverty of those feedback snippets and reject them. I won’t give you feedback every time I walk through your lesson – because I’m only there to see what’s going on; not to judge.

Imagine that as a shift. Instead of SLT taking time to design their observation form; their observation regime – they map out how to engage all teachers in cycles of instructional coaching all year long. They talk about how to build the capacity in the school for delivering effective coaching sessions. They talk about how teachers can record their action steps using the shared language that is part of the school’s framework for teaching. That’s the way to go.

One day, the fact we once graded lessons will be laughed at for sure – but so will the very idea that leaders were presumptuous enough to fly into lessons and pass judgement; that the quality of individual lessons as entities could be judged at all. One day, the default expectation of teachers arriving at new school will be : so, when do I meet my coach?