There is a particular kind of professional loneliness that comes with being a middle leader in a secondary school. You are, simultaneously, a member of the teaching staff and a member of the leadership team. You eat lunch in the staffroom but attend meetings in the conference room. You teach a full or near-full timetable and then conduct observations of the people you share a corridor with. You are expected to be a colleague on Monday and an evaluator on Tuesday and to manage the tension between those roles without any formal training in how to do so.
Middle leadership brings a familiar tension: sitting across from someone you consider a colleague and telling them that their questioning is narrower than they think it is; observing a lesson that is clearly struggling and deciding, in real time, whether what you are watching is a bad day or a pattern; writing feedback that is honest without being hurtful, specific without being prescriptive, and developmental without being patronising — often without formal training in how to do any of it.
The observation problem, for middle leaders, is not primarily a problem of what to look for. Most experienced teachers who step into leadership have a strong intuitive sense of what effective teaching looks like. The problem is structural. It is a problem of tools, training, time, and the relationship between observation evidence and the decisions it is supposed to inform.
The training gap
In Scotland, there is no mandatory training in classroom observation for anyone who is not an inspector. You can become a principal teacher, a depute head, or a headteacher without ever having been formally taught how to observe a lesson, how to record what you see, how to distinguish between what you noticed and what you inferred, or how to give feedback that changes practice rather than merely describing it.
It is a description of a systemic gap. The GTCS Professional Standards for Leadership and Management describe what leaders should be able to do in terms of improving learning and teaching. They do not describe how. The assumption, embedded deep in the structure of professional development for school leaders, is that if you were a good teacher you will be a good observer. This assumption is wrong.
Being a good teacher and being a good observer are different skills. A good teacher knows, instinctively, how to manage the flow of a lesson, how to pitch a question, how to read the room. A good observer knows how to watch someone else do these things and describe them accurately without conflating what they saw with what they would have done. The first is about enactment. The second is about analysis. They are related but they are not the same, and conflating them produces observers who watch a lesson and unconsciously evaluate it against their own teaching style rather than against a research-grounded framework.
I speak from experience. The first observations I conducted as a new faculty head were, I can see now, observations of how closely the teacher's approach resembled my own. I noticed the things I would have noticed in my own teaching. I missed the things that were outside my own repertoire. If the teacher used a questioning technique I did not use, I was less likely to see it as a strength because I did not have a framework for recognising it as such. My observation was filtered through my practice, and my feedback was a reflection of my biases as much as it was a description of the lesson.
It is a description of what happens when you give someone a complex professional task — analysing and evaluating someone else's teaching — without training them in how to do it. The observer defaults to what they know. What they know is their own teaching. And the result is observation feedback that is inconsistent between observers, unreliable across time, and shaped as much by the observer's background as by the quality of the lesson.
The relationship problem
The structural challenge is compounded by the relational one. Middle leaders observe colleagues they work with every day. The observation relationship is nested inside a professional relationship — and often a personal one — that has to survive the observation feedback.
This creates an incentive structure that works against honest feedback. If you are a principal teacher observing a member of your own department, the cost of being critical is not abstract. It is the atmosphere in the department office on Wednesday morning. It is whether the person you observed will volunteer for the school show, or cover your class when you are at a meeting, or support your proposal at the department meeting. The professional and the personal are intertwined in a way that makes honest developmental feedback genuinely risky.
The result, in my experience and in the experience of every middle leader I have spoken to about this, is a gravitational pull toward positive generality. The feedback becomes warmer than the evidence warrants. The development point becomes softer than it should be. The written record, if it is written at all, describes a lesson that is slightly better than the one that was actually observed — not through dishonesty, but through the human instinct to preserve a relationship that matters more than a proforma.
I do not think this can be solved by exhorting middle leaders to be more honest. The incentive structure is real and it is rational. What can be solved is the ambiguity that makes honest feedback so difficult to give. If the observation framework is specific enough — if you are not saying "your questioning could be improved" but pointing to evidence that shows "you asked twelve questions, ten were closed recall directed at volunteers, and wait time averaged under a second" — then the feedback is no longer a personal judgement. It is a description of what happened, grounded in a shared framework that both you and the teacher understand. The relationship can survive a description of evidence far more easily than it can survive an expression of opinion.
The time problem
Middle leaders in secondary schools teach. Most teach a substantial timetable. The time available for observation, feedback, and follow-up is carved from the margins of an already compressed day — the free period that is also the only time to complete the tracking, respond to the emails, meet the parent, prepare the materials for tomorrow's lesson, and deal with whatever pastoral or behavioural issue has surfaced since this morning.
In this context, the aspiration that every observation will be followed by a prompt, detailed, developmental feedback conversation is exactly that — an aspiration. The reality is that observations happen, feedback is delayed, the moment passes, and the evidence — such as it was — degrades in the observer's memory until it is no longer useful even if the conversation eventually takes place. I have been guilty of this myself. Not through negligence but through the simple arithmetic of having more responsibilities than hours.
The implication is that the observation system must be designed to produce useful evidence quickly — not evidence that requires a forty-minute write-up after the event, but evidence that is captured in real time, structured in a format that makes it immediately shareable, and specific enough that the feedback conversation can happen in ten minutes rather than thirty. The observation tool has to do some of the analytical work that the middle leader does not have time to do. If it does not, the middle leader will default to the fastest route available: a brief, general, positive note that satisfies the compliance requirement and contributes nothing to professional development.
What middle leaders need
I have thought about this for years, and I think the answer is simpler than most CPD programmes suggest.
Middle leaders need a shared framework for what they are looking at. Not a generic proforma with open text fields, but a structured set of observable practices — grounded in research, specific enough to be useful, consistent enough to be fair — that makes one observation comparable to another regardless of who conducted it. The framework does the calibration work that training alone cannot achieve. If two middle leaders are both looking for the same practices, described in the same terms, and recording the same qualitative detail, their observations will converge — not because they have been trained to think identically, but because the framework constrains the variation that would otherwise make their observations incomparable.
Middle leaders need a tool that captures evidence in real time without demanding their full attention. The observer should be watching the lesson, not filling in a form. The recording mechanism should be fast enough that the observer can glance down, tap what they see, and look up — an instrument played by feel, not a document composed by thought. If the tool demands more than a few seconds of attention per data point, the observer will stop using it and revert to the notebook. And the notebook, however comfortable it feels, produces anecdotes rather than evidence.
Middle leaders need the feedback conversation to be structured by the evidence rather than by memory. If the observation record shows that specific practices were observed at specific points in the lesson, described in specific terms, the feedback writes itself — or rather, it is already written. The conversation becomes "here is what I saw, here is what the research says about it, here is what you might try" rather than "I thought the lesson was good but your questioning could be stronger." The former is a professional exchange between colleagues. The latter is an opinion dressed as feedback.
And middle leaders need to see the aggregate picture — not just the individual lesson, but the pattern across their department, their faculty, their stage. The observation of a single lesson tells you almost nothing reliable. Coe's research is unambiguous on this point. But ten observations across a department, captured with the same framework, analysed for the same practices, reveal patterns that no single observation could: that questioning is narrower in the lower school than the upper school, that feedback is predominantly generic praise, that metacognition is mentioned in every department development plan but observed in almost none. These patterns are what drive improvement. They are invisible to the observer in the moment. They become visible only when the evidence is structured enough to aggregate.
I built Learning Lens, in significant part, because I was the middle leader who needed these things and could not find them. The framework, the real-time capture, the evidence-structured feedback, the aggregate picture — all of this came from the experience of doing the job without the tools the job required. In my current school, the approach we have taken to learning visits — focused, evidence-based, conducted with a shared framework and followed by professional dialogue — has changed how observation feels for the people doing it and the people receiving it. Not because anyone became a better observer overnight, but because the system became better. And a better system produces better evidence, which produces better conversations, which produces better teaching.
That is the cycle that middle leaders are supposed to drive. Most are driving it without a map.
Jamie Scobie writes from extensive experience in Scottish secondary education, including pastoral care, data for improvement, and school self-evaluation. This blog is an independent publication: he writes in a personal capacity as the creator of Learning Lens, writing about classroom observation, teaching evidence, and education policy. He speaks here only for himself and for Learning Lens, not for any employer or other organisation. He holds an MSt from Cambridge (Distinction) and a Masters from Stirling.