On Grading Lesson Observations

In Debates and Polemics, Evidence in Education, Research Evidence by Alex Quigley11 Comments

20140118-230656.jpg

“Are you ready for your annual lesson observation Mrs Crane?”

It is clear that the topic of lesson observations is an emotive one for all teachers. Few topics arouse fear and unease like the prospect of being judged on your ability to teach in a singular lesson. Still, too often, the norm is that this fear is exacerbated by the fact that such observations are a rare, all-eggs-in-one-basket occurrence. They take on disproportionate importance – necessitating disproportionate planning – with the result becoming a disproportionate and often ill-judged measure of what we do in the classroom.

Chuck ’em out is the bold and brave call. Evidence from respected academics like Professor Robert Coe, has ploughed the furrow of extensive research to unveil the catalogue of bias that attends these lesson observations. The Gates Foundation has undertaken vast research that opens a can of worms regarding their validity.

Yet, the nagging doubt – no, the immovable obstacle – is the stranglehold of accountability that attends our school system. It is a brave – or foolish- Headteacher who disbands graded lesson observations if our inspectorate is breathing down their neck with an imminent visit to judge whether they are ‘Requiring Improvement‘ or not. The inconsistency of such inspection visits makes any school leader pause for thought. If results are beyond reproach perhaps, but for the vast majority of schools that position of safety is not the lived experience.

And yet…and yet…

Our experience, and the growing body of evidence, would appear to prove that such judgements are unreliable. Considering they can also have a corrosive effect on staff morale, as well as not proving useful in actually developing teacher practice in the classroom, we must question using them at all.

We must have lesson observations as part of our performance development process in my view. How can we improve our teaching without some feedback on our performance in the classroom? Indeed, ‘lesson study’, peer observations and ungraded, formative observations appear to be a much better model to promote the improvement of pedagogy – and therefore student attainment – in our schools.

I feel that in my position as a performance development reviewer, I can give useful formative feedback about pedagogical knowledge that promotes a useful dialogue about teaching and learning. I also think that I can observe whether behaviour is positive and that fertile conditions for learning are established.

I am much less confident about judging whether students are making genuine progress in their attainment and their learning in any given hour of time. Learning indeed is often invisible, which encourages a ‘show’ of performance. It can encourage style over substance – the classic one-off showpiece lesson. I know whether any learning is happening, but estimating whether achievement, and indeed teaching, is ‘Good‘ or ‘Outstanding‘ is too often a dodgy judgement call.

In light of my experience, albeit limited I should add, I would propose that lesson observations are conducted, without graded judgements, as part of a raft of approaches to judge teacher effectiveness. Work scrutiny and internal and external assessments are a much better indicator of student achievement. Observing a lesson can simply be a place to talk to students about their learning, bolstering and unpicking the analysis of a work scrutiny.

If we consider that we can establish an evidence base of work scrutiny, student attainment, 360 degree reviews, evidence of planning and schemes of learning etc. then surely the potentially corrosive impact of graded lesson observations should be removed. Each school can create a sound, rounded model to judge teacher effectiveness, where no one indicator becomes disproportionately important.

I think the difference between ‘Outstanding‘ and ‘Good‘ is particularly arbitrary. Does either label help us become a better teacher? Perhaps as much as damning a student with the label of being ‘4a’ does. If in our experience of lesson observations the behaviour of students is patently an issue, or the subject knowledge is lacking, surely these can be more reliably observed. We should therefore abandon the OFSTED model and define our own model of formative observations that do have a mechanism for recognising a significant ’cause for concern’.

As a parent of young children I would be happy for this system to be in place, knowing what I know about how we improve as a professional and the limits of our judgement. I would be happy for school leaders to apply their wisdom and judgement.

Related posts:

This topic is gaining real traction with people thinking very deeply about teaching and learning. Here are some very useful links on the topic:

Professor Coe presents the evidence base here.
David Didau composes a cogent argument for learning over performance here and his practical post about how observations should be here.
Joe Kirby presents an impassioned rallying call to be rid of lesson judgements here.
Stephen Tierney presents a Head’s perspective on improving the system here.
Ross McGill presents a really useful collation of the views and resources related to the topic of lesson observation here.
An excellent article by Mary Myatt on solutions beyond graded lesson obs here.

Comments

  1. Alex,

    After (over a period of 5 or so years) first using grades, then removing them, then reinstating them, then finally removing them last week, our SLT here at KEGS Chelmsford have at last officially allowed formative dialogue without numbers to be at the heart of what we do as regards lesson observation so – hopefully peer observation, possibly using the lesson study model which headguruteacher is leading on here will now become the norm. Hopefully, we will get away from formal Vertical observations altogether …

    Why did it take us so long! Why did we flip-flop so much … Why have so many schools been in thrall to numbers – I blame the usual data fixated suspects, including too many Senior Leaders …. As I have often said to our school leaders, misquoting the apostle, “the number killeth, but the dialogue giveth life.” A plague on all number crunchers – they have sometimes made worms’ meat of teachers’ morale, as Mercutio would say …

    Most of us Middle leaders have always played down the grading anyway … so, hopefully no more counterproductive mock Ofsted internal inspections either!

    1. Author

      Great to see KEGs leave the graded obs behind. There is so much scope for a more intelligent system. The addiction to data, for good or ill, as a proxy for improvement is indeed rooted in the system. We need outliers proving it can be done better. Peer obs and lesson study offers so much more. I am looking on intently – hopefully we won’t be far behind.

  2. Bravo! Keep talking sense, SLT across the country need persuading that they CAN evaluate T & L reliably and show inspectors that they have all without a 30 min random number generator. It’s almost as if people had never heard of comment only marking!

  3. Alex, thanks for your writing as always.
    A couple of comments:
    ‘We must have lesson observations as part of our performance development process in my view. How can we improve our teaching without some feedback on our performance in the classroom? Indeed, ‘lesson study’, peer observations and ungraded, formative observations appear to be a much better model to promote the improvement of pedagogy – and therefore student attainment – in our schools.’
    Feedback, study, observation all involve the collection of evidence and it’s communication. Coe’s argument ( http://www.cem.org/blog/414/ ) is centred on the meaning of evidence and the amount of trust we can put in it – reliability – in making judgements based on it. In his article he refers to the the process of trying to convert know-how into know-what: knowledge into information through ‘a scientific development and validation process, ….training, and a test they are required to pass.’ Given all this, assuming the ‘worst case’ based on Coe’s figures, only about 20% of observers would give a teacher the same assessment, using OFSTED criteria. So that’s the level of reliability you might expect when the must-have lesson observations are carries out. Whether or not the intended purpose of the observation is formative or summative is irrelevant from this point of view, because the evidence on which future improvement can be made, by assessing potential, is as suspect as it is of a producer of hard data on which to base performance assessment.

    ‘I feel that in my position as a performance development reviewer, I can give useful formative feedback about pedagogical knowledge that promotes a useful dialogue about teaching and learning. I also think that I can observe whether behaviour is positive and that fertile conditions for learning are established.’
    I’ve been writing a bit about my understanding of what ‘knowledge’ means that you might be interested in reading. (see my website)
    If you take ‘pedagogical knowledge’ to mean pedagogical know-what, information/data that the teacher holds about their teaching in terms of its pedagogical type, the intended learning outcomes associated with the pedagogical choice, how long any teaching component will take in the classroom in minutes, etc. etc. then you could give useful feedback. Because the teacher can tell you the facts.
    But the evidence that is worrying Coe and others is when ‘pedagogical knowledge’ means pedagogical know-how. Know-how is difficult for the teacher to communicate, is produced in the context in which the teaching happens, includes aesthetic and other values, cannot be precisely reproduced. Know-how is knowledge produced by the teacher in the moment of teaching, the observer can interpret it according to their own understanings, but it cannot be directly accessed and transmitted as can the raw information/data of know-what.

    What to do about it? Ask the expert – the teacher. What went well in that lesson? What else? What else? How can you tell? What’s your best hope for your work with year x over the next month? What might change a bit for things to go better? When might that happen? It’s a plan, it’s subjective (because there’s a paradigm shift). Because it’s difficult to communicate the conversation needs a structure – and you’re right, it’s necessary and important to do it.

    I’m very interested to see where you go next with your own knowledge. Best wishes. Geoff

  4. Pingback: On Grading Lesson Observations | HuntingEnglish | The Echo Chamber

  5. Pingback: ORRsome blog posts from the week that was! Week 3 | high heels and high notes

  6. Pingback: The great grading debate | M J Bromley's Blog

  7. Great article! I agree that judging an educator’s ability to teach in one lesson (for me that’s only a 40 minute period) is not an accurate way to produce better teachers. Peer review and informal feedback would make the process less nerve wrecking for teachers and could also give teachers new ideas for their lessons. Having multiple teachers view your lessons, even teachers from different disciplines, can help open your eyes to new ideas and different teaching methods.

  8. Pingback: NUT Conference 2014 #NUT2014 (Shock Expose!) | emmaannhardy

  9. Pingback: Bloggers lead the campaign to reform Ofsted

  10. Pingback: Lesson observations for ‘Genuine improvement’ | @mrocallaghan_edu

Leave a Reply