Schools have rightly moved away from lesson observation gradings, a damaging system created by Ofsted and whose legacy persists. As Chris Watkins rightly notes, we want lesson observations to be about learning and improving, not proving and performance:
The legacy of lesson gradings is the tendency of teachers to try to ‘perform’ when being observed. And when teachers are ‘performing’ for the observer, one of two things will happen: either the observer will witness an all-singing, all-dancing lesson that is in no way reflective of that teacher’s day to day practice, meaning that any feedback the observer gives is not useful in terms of improving regular class teaching. Or the teacher’s attempt to teach an all-singing, all-dancing lesson means they lose focus on the learning they are trying to achieve and again, the lesson is therefore a disaster and the feedback the observer can provide is not relevant to the teacher’s daily practice.
In fact, even if the observer does manage to see a lesson that is broadly reflective of the teacher’s actual practice, there is no real evidence that the feedback the observer gives makes a difference to classroom practice. A Sutton Trust exploration of what supports teacher development stated that the evidence of impact of lesson observation on student outcomes is ‘generally limited’.
But even if it doesn’t work, there is surely an obligation on senior leaders to observe teachers in their schools in order to have a handle on teacher quality? Here again however, the evidence is dubious.
‘Hill, Charalambous and Kraft (2012) estimated that using observations of practice to produce ratings of teacher quality with a reliability of 0.9 would require seeing a teacher teaching five different classes and having each lesson observed by six independent observers.’ RSA, Licensed to Create
So what should we look at? The same Sutton Trust report evaluates some familiar sets of data. Progress measures are noted to be notoriously unreliable since even tiny changes in measures can dramatically change judgments, the effects of the ‘best’ teachers are long and not short-term (i.e next year’s teacher will benefit), there is a strong chance of measurement error, and a ‘cohort effect’. Reviewing work outcomes or planning is reliable only when raters follow a specific protocol for evaluating them. There is one measure that does provide a reliable evaluation: ‘Student ratings are valid, reliable, cost-effective, related to future achievement, valuable for teacher formative feedback and require minimal training.’ The report states that the only disadvantages is the need to consider how best to manage the feedback process in relation to the pupils’ ages.
Given that none of these ways of judging teaching can claim to be fully accurate, how should a school leader proceed? The Sutton Trust advises triangulation: look across more than one set of data before coming to conclusions about a teacher’s performance.
And yet it is this advice that has given us punitive performance management systems which have significantly contributed to teacher workload and stress. Many schools conduct frequent book looks, half termly assessment weeks in which children are tested, termly lesson observations and more regular learning walks or drop ins. Are these systems really built around the needs of the majority of teachers in our schools? Or are they built around the need to ‘catch out’ the weaker members of staff, whilst maintaining a veneer of all teachers being treated equally?
We would never build our class teaching around the needs of the lowest achievers; instead, we take extra care to monitor their progress and put in place an additional intervention to support their learning.
What would a performance management system look like, if it was built around the learning needs of the majority of teaching staff, rather than the minority? Might it encourage self-assessment of strengths and needs, and self-identification of suitable professional development. Surely it would involve evidence, but this might be evidence that is used to corroborate a self-assessment, rather than to judge. A group of experienced headteachers I met in March explained how the removal of regular lesson observations had had a surprising impact: teachers started asking senior leaders to observe them as they focused on developing a particular aspect of their practice, and often repeatedly.
It is possible to develop performance management systems which enable teachers to focus on learning and improving, instead of proving and performance. And such systems are predicated on a belief not only that the majority of teachers do know what they need to do to improve learning in their own classrooms, but that they are also motivated to make this happen, without the need for constant checking and feedback from senior leaders.