On lesson observations and paracosmic compliance…
Mr Wilson left my lesson smiling, and nodding. This time, I’d hit the jackpot. My annual graded lesson observation was clearly outstanding!
I had taught the fluidity of identities in postmodern sociology in a dramatic way. Actual blocks of ice slowly melted as students listened, enthralled, by my explanation of Bauman’s Liquid Lives. There was a free-flowing (!) discussion afterwards and a palpable buzz in the room.
The door clicked shut behind him.
Emily Johnson waited patiently until he was safely out of earshot.
‘You have to help me Mr White. I have no idea what we’ve been talking about!’
I laughed. The other students laughed. In truth, I was somewhat deflated! For a moment I believed a little too hard in the simple reality of the inference (1*!) which my line manager and I had drawn from the previous thirty minutes of teaching.
For a time, in English schools, there was a particular emphasis upon measuring and then raising the quality of individual lessons. Outstanding lessons became the building blocks with which leaders and policy makers constructed their educational edifices. If you were a teacher in England during this era you will no-doubt recall high stakes graded lesson observations. A senior member of staff, trained in recognising outstanding lessons, would observe and grade lessons from 1-4.
The lesson is a unit or episode in schooling with which we are all familiar. It can be readily brought to mind and animated. Solutions and strategies built with this unit appear substantial, self-evident even. It is easy to imagine a successful lesson and the learning that it produces in a class. The ease with which we can visualise the lesson process leaves us particularly susceptible to the illusion of explanatory depth. Superficial familiarity with a complicated process can induce a false sense of understanding. In reality we know rather less about the mechanics of learning, less still about the mechanics of multiple learners responding in disparate ways to the same ‘teaching’.
During these graded observations, Teachers had two discrete tasks. Firstly, teaching the class, and secondly managing the inferences being drawn by their observer. After the observation, a dance ensued.
The feedback meeting.
So…how do you think that went? Was often the observer’s first step.
From here, a skilled observee may have been able to take the lead. They would steer the process towards an outstanding finale – by framing the lesson in a positive light using the language imposed by whichever grading criteria were in play. Perhaps there was a buzz, perhaps it was evident that all students were making progress, perhaps there were clear hinge-moments in which teaching responded to formative assessment.
Others, with less experience, less gravitas, or greater faith in the judgement itself would be led by the observer. The quality of their teaching explained to them by their line manager on the basis of perhaps a 30 minute episode and select moments, pupil questions, comments, work and other proxies therein.
The observer, tasked with drawing a clear judgement from patchy information was faced with a deceptively simple task. They may have been aware of the limited reliability of the process. They may not. Either way, countenancing this troubling issue would only hamper their capacity to complete it. The lesson observation process discouraged such thinking, and in providing a neat abstraction – a numerical measure of teaching quality – provided clarity around a complex issue.
In ‘Obliquity’, John Kay (an economist!) points out that economists generally question reality if reality fails to correspond to their theoretical models and predictions of it! In the meetings following graded lesson observations, the discussion which followed was a collective, or autocratic attempts to achieve the same end. The lesson was made to fit the model.
Ensuring coherence with the dominant silver-bullet solution often became the end towards both parties work (knowingly or unknowingly). We needed to make the lesson fit in with an imaginary world in which lessons could, and should be consistently ‘outstanding’.
An unsatisfactory explanation:
Despite their initial allure, Graded lesson observations in particular didn’t work properly. Not least because of their unreliability (sommething Prof Rob Coe explores here), and because high stakes further distorted the measure.
But the craze emerged and failed because of a wider dynamic. It was merely one episode in education’s serially monogamous relationship with simplistic solutions. These solutions offer explanations of the problem with teaching which have alluring, but ultimately illusory, depth.
The lesson grading training which some of us went on might have initially produced what psychologist Frank Keil calls an ‘Aha’ moment. There is a rush of understanding when we first see how a particular explanation fits together. Aha, of course, outstanding teaching is the key to ensuring that all students learn successfully. However this intuitive sense of comprehension is misleading. In reality our understanding is less complete (or shallower) than we believe. The explanation was shallow and failed to stipulate the mechanism by which it could actually work.
The explanation of teaching which the process framed was superficial. It had illusory depth. Where proxies of outstanding teaching were identified, there was not a mechanical explanation of how they could induce learning in multiple unique individuals at once. Where proxies of learning were highlighted (everything from detailed answers/performances to a palpable ‘buzz’) there was a similar gap in identifying quite how this could be produced. ‘And then some magic happens.’ This ‘explanation’ of teaching was inadequate. It failed to engage with the inherent complexity of teaching disparate individuals specific ideas within a fixed timescale.
Descriptions of the features of outstanding lessons were essentially a ‘design stance’ explanations masquerading as a mechanical ones. A design stance (or teleological) explanation describes a process by emphasising the ends towards which individual components are directed. I.e. ‘The teacher differentiates appropriately’ ‘Explanations ensure progress in learning for all.’ These are design oriented explanations or descriptors. They indicate an intended end. They are not mechanical explanations – a mechanical explanation would instead emphasise the specific actions a teacher takes and the effect this has on individual learners. (Keil explores the implications of these Dennett’s ‘stances’ here.
Why were the limitations not immediately obvious?
The self-referential nature of the dialogues (or judgements) which teachers and leaders were involved in discouraged analysis of lesson grading itself as a means of monitoring and improving teaching quality. This was in part due to the complex and, arguably, imbalanced relationships in our educational system. The onus on making an alluring solution ‘fit’ pushes down. There is not, a commensurate feedback mechanism for modifying the solution to match reality pushing back up.
If a teacher’s lesson was not ‘outstanding’; or if a school recorded a troubling number of unsatisfactory observations, or if Ofsted visited and concluded the same; then a problem was immediately apparent. The pressure was on to align with the solution. However, if the act of compliance (in this case to graded lesson criteria) were having no impact, or worse a latent negative impact, upon lessons this would not be apparent for some time. Possibly never.
The teacher’s role in this process was not one which invited critical engagement. Whether or not they agreed with the premise, its validity was generally off-the table in the follow-up meeting and judgement of their teaching. The professional closest to the complex reality of schooling, often had to work hard (consciously or unconsciously) to ensure that this messy world aligned, with a simplistic conjecture of outstanding teaching.
Some dismissed it as a minor annual irritant. A hoop through which to jump before returning to business as usual. One teacher friend rolled the same lesson out each year, regardless of year-group or time of year. (It worked). Others gave it rather more credence than it should have been afforded.
The latter position is more common we might first think. Inexperience, the isolated nature of the teaching experience and a deference to the knowledge and seniority of ‘superiors’ (school leaders or those who create the latest criteria) can all contribute to uncritical credence in the validity of the inferences which flawed processes actually afford. This was me for a while. I laboured through my lesson plans trying to spark the magic which my line-managers would be looking for. Eventually, as I became more experienced (and gained other sources of positive professional identify reinforcement) I came to treat the process more sceptically. But at this point, I probably jumped too quickly to dismiss senior leaders who continued to carry these high-stakes gradings out.
They were just doing their job. They may have had reservations about the process, or may not. Either way, a role premised upon tracking and improving the quality of teaching across a school necessitated clear judgement about the same. Lesson grading afforded a feasible route by which this could be carried out. Furthermore, Ofsted, also leaned on this data, cross-referenced against their own judgements when grading an entire school.
The focus of the meeting become managing the imagined school. Immediately, the imagined school as conceived by the observer. Later in relation to the collective imagination of the wider leadership team, perhaps the teacher who was observed, perhaps other members of staff. Of course, accountability dynamic further distorted this measure. Ofsted would also grade lessons and it was important that a schools’ internal records of teaching quality aligned with the ratios harvested during an inspection.
My line-manager’s job largely entailed constructing and maintaining a ‘paracosm‘ an imaginary world (with links to the real one). This world was one where simplistic solutions could and should ‘work’ as this ‘evidenced’ the quality of our school. That ‘1’ that he recorded indicated, that for one year I was one of the school’s ‘outstanding’ teachers.
Have we learned from this episode?
The underlying problem is not really to do with lesson observation as such. It is driven by the paradox at the heart of the school improvement challenge. We do not know as much about learning as we would like. We know less about how to facilitate learning in groups of people simultaneously. Learning, even defined narrowly as long-term change to the brain – is often hidden, and only becomes apparent in the long-term. Schools, from a design stance, may be seen as places where students learn. But their particular form, the activities which happens within them, and the diverse lived experiences of students, teachers, parents, leaders, policy makers represent a messier reality.
Simultaneously, the school system is populated with roles which demand action and clarity despite our fundamental unknowledge of exactly how they should (and do) work. From within these roles, it is perhaps possible to look back, with hindsight, at the inadequacy of earlier failed solutions. For example,many of us are as now as sceptical about the data wave, as we are about lesson grading. However, we are not exempt from the challenge. The way in which we imagine our role, the concepts and patterns we recognise, the specific demands of our jobs, and the dominant ideas imbibed from training courses, accountability measures and the wider educational discourse appear rather more solid than are actually the case. We need to act, and need a wider context which informs (or at least rationalises) our actions.
Leadership roles can all to easily drift into an exercise in paracosmic compliance. We work hard ensuring that the imaginary world which we create and document, aligns with dominant ideas of what a good school should be. This process is clearer than the more complex challenge of actual school improvement. As a collective endeavour it can gain a faux solidity. Improving lesson gradings can seem substantial, as could improving the accuracy of exam predictions, or ensuring consistency across curriculum statements.
The monopoly afforded to dominant solutions – in no small part amplified by accountability measures – is expedient in some ways. It allows for clear judgements, clear solutions, and clear narratives to be told explaining the problem with school and how to solve it. But it comes at a cost, the strain across a system where we attempt to shoehorn reality into the form we have been told it should take – if we’re doing the job properly.
The problem, in as much as there is one, is not the existence of imagined worlds. We need internal models, abstractions, or maps to help us make sense of and act within a complex system. The issue is that we often forget (or never learn) that this is what we are doing. As a result contrasting models; information which doesn’t fit how we feel things should be; and the disparate experiences (an imagined worlds) of others can all appear as a threat to our efficacy rather than merely being reminders of the complexity of schooling.
The imagined worlds which matter most are those constructed by our students. They will be diverse and sometimes surprising of course. Regardless of whether I imagine myself to be teaching an outstanding lesson, targeting students for intervention, or carefully sequencing my curriculum, it is likely that some of my students will love learning psychology. Some will take pride from their learning. Some will be bored in my lessons and most will forget most of what I teach them (regardless of how powerful I deem it to be). All will learn in unique and surprising ways. All will continue to learn and grow after the 100 or so lessons they spend with me fade in their memories. But, we understand that that’s kind of the point. School provides a controlled environment in which students are exposed to and construct new understandings (abstractions) of the world. Teaching, as Mary Kennedy points out is a human improvement profession we can expect the process to be somewhat messy and unpredictable.
Emily Johnson helped remind me of the complexity of teaching. I discussed the incident with friends afterwards. But in my lesson observation meeting remained quiet on the matter, and let the ‘1’ stand! The context did not leave room for such discussion. I am skeptical of explanations of, solutions to, schools’ ills which fail to leave space for messiness. Stories and solutions premised on lesson gradings, data tracking and curriculum mapping can easily fall foul of this problem. I would like to think that it’s possible to bring more nuanced and reflective sense-making into professional conversations.
The how is complex.
I suspect that there’s much more to consider in terms of ego development. Just as students need different models and instruction at different stages of development, the same is true of adults too. Neil Gilbride’s work in this area is definitely worth exploring further. There are times when teachers need clarity and simplicity in being told or shown what to do. Especially early in careers. For example, if in my lesson observation pupil behaviour had been ‘off the wall’ I probably would have benefited from practical, direct instruction on how to improve in that area. Nevertheless, in the main, professional dialogue can benefit from being more a collaborative sense-making endeavour rather than an exercise in paracosmic compliance.
There are also likely lessons to learn for leaders in relation to line-management relationships. How do we ensure that our support and management genuinely are supportive? It is easy to get distracted by imaginary school maintenance rather than more meaningful work. Once this happens, our interventions are likely to be viewed as irritating and largely irrelevant demands. Role out a buzzy lesson. Change some of those predicted grades. Write a curriculum statement.
There are further implications for accountability measures. The dictate the rules of a game which schools often feel compelled to play. The lesson observation game was largely shaped by a need for schools to record good and outstanding teaching if they were to fare well in inspection. The data game did the same for internal data. The Covid19 delayed curriculum wave may risks doing the same for curriculum sequencing and statements.
I am in the final few weeks of my current teaching role. After Easter I start in an AHT position, with a view to eventually becoming a head-teacher. One lesson I will hopefully take into the role is not to take myself, or the imagined school I construct too seriously. In ten years, dominant ideas will change and I’ll likely look back at much of this era just as I now do at the lesson grading moment above.
However, the students who pass through the school every few years only get one cycle. And the staff embarking on their teaching careers only do so once. Fostering an environment which supports them, and helps them to grow will always be important. This does require some imaginary school building – bringing structure, familiarity, routine and purpose to their daily experiences. But humility and collaboration are also important. Perhaps, being forewarned will help. We need to be wary of perceived or real pressure to practise paracosmic compliance!