The question of how school leaders can make "sound judgements" about colleagues' performance is bound up in the difficulty of defining "effective" teaching.
We don't want judgements to be based wholly on test scores, because great teaching is longer lasting and more multidimensional than that. And how we teach is just as important as what we teach. So why shouldn't performance be judged largely on direct observation of the teacher in action?
Yet this approach has been criticised on the grounds of validity and reliability. Even allowing for the "Hawthorne effect" - where the observed improve their performance because they are being observed - how can a few isolated observations represent a teacher's total activity, opponents say?
At best, these objections focus on methodological problems; at worst, they make it sound as if teachers have something to hide. A stronger argument for being cautious about observations is that the lesson is not the atom of education.
Treating a lesson as the fundamental unit of delivery, and therefore of analysis and judgement, leads us into all sorts of difficulties - including the temptation for teachers to shoehorn every conceivable approach into one observation to ensure they are demonstrating a sense of variety, motion and a wide-ranging repertoire.
It isn't just observations that encourage this tendency. The modular nature of many programmes of study, and the orthodoxy that lessons should follow a template with a clear start and ending, both encourage the idea that lessons are microcosms of the whole.
Behind the scenes
Yet we know that much of what makes for effective teaching has already been put in place before the pupil has sat down. German education expert Anna-Katharina Praetorius has shown that effective teaching can be identified principally in terms of classroom management, expectations, relationships with students and personal qualities such as fairness and respect. These are not created de novo at the start of each lesson; they are established over time, as a relationship of trust develops between the teacher and students. These are the structures of classroom management that hardly vary for an individual teacher from lesson to lesson.
We know, too, that the way a teacher teaches depends on where she is in the course. Researcher Steen Beck showed that teachers adopt roles that relate to the task in hand. The "teacher as presenter" might be prominent at the outset of a unit of study; over time, this might give way to the "teacher as instructor", working with students to handle simple problems. Student-led learning then takes over, with the teacher acting either as a consultant or a participant. Any one classroom snapshot will capture only one stage of the course. But an individual lesson cannot - and should not - try to be a version of the whole.
Specific teaching and learning styles will depend on what has gone before and what is still to come in the programme of study. The tools chosen will vary according to need and circumstance. Observations that focus primarily on the toolbox, the teaching style and pupil activities are at risk of producing unreliable judgements because they're looking at the wrong elements of the lesson.
This is supported by a recent Sutton Trust report, in which the don'ts of effective teaching were found to relate to surface characteristics - such as grouping by ability, a focus on learning styles and indiscriminate praise - and the dos pointed towards the deeper structures of effective teaching like content knowledge, classroom climate and management.
Let the story unfold
A well-planned lesson has much in common with a television drama series, in which each episode has a self-contained story but is part of a narrative arc that spans the entire season. Some things are clarified quickly, but others take time and aren't resolved until the very end. Teaching is a long game and education has a narrative arc.
Classroom observation should form part, but only part, of the portfolio available to evaluate the effectiveness of teaching (alongside pupil progress data, scrutiny of planning and the work produced, peer review and pupil feedback). A few isolated observations cannot be expected to bear the burden of summary judgement. Observation should explore aspects of practice, following up on issues identified by other evidence. It should pay more attention to deep structures around planning, climate and management, and focus less on "surface" performance, techniques and activities.
Dr Kevin Stannard is director of innovation and learning at the Girls' Day School Trust
- Praetorius, A K, Lenske, G and Helmke, A (2012) "Observer ratings of instructional quality: do they fulfil what they promise?" Learning and Instruction, 226: 387-400
- Beck, S (2008) "The teacher's role and approaches in a knowledge society", Cambridge Journal of Education, 384: 465-81
- Coe, R et al (2014) "What Makes Great Teaching?", Sutton Trust