I observed our department’s new unqualified teacher yesterday. Goodness knows how tough it must be but he is doing really, really well.
Oops… I have just made a statement that has become controversial. How can I possibly know how well he is doing? The message from twitter and a spate of blogs recently is that student learning can’t be observed; that single lesson observations are remarkably unreliable ways of judging teacher quality.
I do actually pretty much agree with those arguments. Schools across the land need to hear that message and stop judging the quality of someone’s teaching based on one lesson, or even a twenty minute snap shot. It IS crass to do so and often our judgements tell us more about our own teaching preferences than the effectiveness of the teacher we observe. If you are unaware of the research that demonstrates just how unreliable individual lesson observations are for judging teacher quality I urge you to read this. http://www.cem.org/blog/414/
However, and it is a big however, I question the assumption of some that we can’t make any judgements from lessons and thus provide other teachers with guidance.
Our new teacher’s lesson was on the events leading up to WW1. I first taught that lesson twenty two years ago, using some of the same resources as he used yesterday. Subsequently I must have taught hundreds of students of all abilities that very material. As I sat quietly at the back of the classroom I realised I just ‘knew’ what was going on. I could ‘read’ the lesson; see each teacher dilemma, anticipate the pitfalls. I just ‘knew’ how much was being learnt and I just ‘knew’ what some of the class didn’t understand. I might not be able to see inside every child’s mind but I knew the sort of learning that transpires longer term from that teaching. Those are big claims but I don’t know how else to describe how it felt. I had some insight into what was going on because I have done my time in the history classroom, earned those metaphorical leather elbow patches. If you’ve read Gladwell’s Outliers you might remember his observation that the apparent ‘talent’ of the expert who reports they just ‘know’ is actually the product of years of experience. I’m really not claiming infallibility (far from it) but I agree that all my experience does not count for nothing.
Ask yourself who you would rather give you feedback on a lesson. Would you prefer an observation by a teacher of another discipline with a few years solid classroom experience or an observation by a highly experienced practitioner in your own subject with a proven track record of success? Actually if I had neglected that class recently and the lesson was near the end of a series then definitely the former! I would be able to pull the wool over their eyes…
That is the problem with these sweeping pronouncements in education. Apparently research shows homework is not useful, that setting is worse than mixed ability, that class size doesn’t matter and now that lesson observation is unreliable. The problem is that there is some useful truth in all that research. I really don’t think we should dismiss all education research out of hand. I am grateful to the researchers that have demonstrated just how unreliable individual classroom observations can be. Thank goodness brain gym and learning styles have been debunked. However, the more general the conclusion, the less generalisable the finding can be to the classroom.
The usefulness of observation really depends WHO is observing and WHAT the school believes they can judge from the observation.
*YES as an experienced history specialist I can give real constructive help to a new department member but NO I could not judge a French teacher’s (or really any teacher’s) effectiveness for PRP or performance management based on a few lesson observations.
*I feel some confidence that I could identify genuine strengths and weaknesses in an INDIVIDUAL lesson of history when I am familiar with the material covered and especially if I have taught that content often before. Even then any specific grade I might give that lesson would not be reliable.
*A history specialist would be less able to reliably spot the weaknesses in a science lesson though. Their comments on things like behaviour management might be helpful but they just wouldn’t understand enough about what the teacher is trying to achieve and any judgements would inevitably depend on teaching style preferences and weak proxies for apparent learning such as degree of student engagement. I could spot something really bad but my ability to judge well would be limited. I think this is the truth we need to accept from the research.
Let’s not dismiss lesson observation out of hand but let’s also take the research on board. I would argue that WHO is observing and WHAT they are attempting to judge are crucial.
Esse Quam Videri 
I am amazed that no attention is given to the principle of aggregation in these discussions. The best elucidation of the importance of this principle in the social sciences that I am aware of is Rushton, Brainerd, and Pressley’s classic paper from 1983: “Behavioural Development and Construct Validity: the Principle of Aggregation” (available for free at http://pdf.k0nsl.org/R/Rushton%20Materials/Behavioral%20Dev%20and%20Construct%20Validity%2083.pdf).
Every teacher and every social scientist should read this fantastic paper at least 3 times, but the relevant takeaway is that judges’s ratings become more reliable and more highly correlated as you increase the number of judges. It makes a great deal of sense that subject specialists are able to judge teachers of their subject more accurately that specialists in other subjects are, although I don’t know if anyone has tested this. Clearly, however, an easy and vital step to improve the reliability and validity of lesson observations is simply to increase the number of observers. This doesn’t even require multiple observers in that classroom; video analysis can serve much the same function. Error variance specific to one observer gets averaged out and a clearer picture of the underlying construct starts to emerge. Of course, multiple observers observing multiple lessons is better still for exactly the same reason.
This is really not rocket science, but it is frustratingly still too-often ignored in the social sciences. The tools and the resources are there to vastly improve on current practices of lesson observation. I cannot recommend the Rushton et al paper enough for how clearly it explains why and how to do so.
Thank you for your comment. Sounds like something I really should read.
Anyone who assumes “we can’t make any judgements from lessons and thus provide other teachers with guidance” is a fool. I for one would never make sucha claim 🙂
I hope I didn’t imply you did. I have heard it said. 🙂
We use lesson observations at my site as part of a collaborative project to improve practice. We select our own partners, choosing someone who we believe might be able to offer insights because of their special skills as a teacher. At its most simplest, classroom observations are another pair of eyes seeing things that the teacher delivering the lesson might not. I asked one of my partners to observe me demonstrating a complex task to a class with the brief to note student engagement, clarity of my instructions and the effectiveness of me doing the task myself in front of the students before asking them to do the same. He was able to draw a map, detailing where the students who lost interest stood in relation to me and what contributed to their lack of focus. He also included notes on my questioning techniques and the clarity of my words (use of language) along with noting when and how often I myself noticed that attention needed to be recalled. Using his observation notes along with the data I myself collected on how the students performed when they began the task themselves, I gained valuable insights into my effectiveness in communicating knowledge to my students. This helps me to make adjustments to my delivery and management of my classroom to ensure better results. Observation helps me to inform my practice. Very useful and effective in my opinion.
It also builds relationships with colleagues.
When it was my turn to observe my colleague he asked me to observe a lesson where he used a strategy he had just heard about and wanted to try to help his students get the most out of reading a text (reading for meaning). My observation notes helped him determine if it was effective and he made adjustments from this feedback to suit his cohort. Another pair of eyes sees things the teacher delivering the lesson misses. We also collaborated on the assessment of students work with the brief to discuss if his feedback at the drafting stage was effective in improving student outcomes.
He is a Drama and English teacher, I am Visual Arts. We teach different subjects, but we helped each other and in the process also learned something new and learned to appreciate each other’s perspective and skills as a teacher.
Clearly I see many benefits regardless of the research.
Enjoyed your post Heather, thank you 🙂
This is brilliant.
I’d agree that the research evidence shows that grading lessons is not a reliable process and that we are probably not able to predict examination results by observing one (or even several) lessons.
Yet clearly observers can say meaningful things and you highlight several here.
The question which leaps out at me (and which you ask) is *why* you can do these things in an observation and someone else can not.
My answer would be that you have the following:
(1) a good knowledge of history and what ‘good history’ looks like
(2) a good knowledge of how children learn history (common misconceptions, common errors, common assumptions) gained from years of experience and professional discussions
(3) a good knowledge of history curricula (how historical knowledge is structured, the different ways this is possible, how parts relate to wholes, etc.)
Now of course in any one observation (or series of observations) you might get it wrong – this is why judging someone’s career on observations is a bad idea. But in terms of providing useful, expert feedback to a new teacher, your knowledge base is what they lack and your expertise in (1), (2) and (3) is exactly what they need.
Love it.
Those are precisely the points I wished I had made in more detail once I had pressed ‘publish’. Thank you very much for making them!
We have a problem here: because we are too vague about what outcomes we expect from our pupils (merely understanding why people acted as they did in bygone times is clearly inadequate to modern History teaching), Ofsted and hence SLTs focus on process. Thinking back to my school days, which are so far in the past that teachers were seldom if ever observed, all teachers used quizzes and tests on a weekly basis. Teachers were actually trusted because their tests were transparent. Teachers varied greatly in terms of how they taught; some of my best teachers were rigidly organised, and others relied more upon charisma. They all understood the vital importance of questioning, and in STEM subjects worked examples were nigh-well universal. Otherwise, the only unifying factor was the focus on ensuring that what was taught had actually been learnt.
The currciulum was stable, which made a huge difference: you could tell how long someone had been teaching by how yellowed and grubby their notes were. Even though there was no universal system of testing back then (this was in the US), weak teachers would be found out straight away when their pupils went up to the next grade.
I suppose we will never return to that Arcadia (which admittedly was far from perfect), but at least we could learn from it. The starting point would be to disband Ofsted. It is oppressive in concept and performance, and it is depressing to hear intelligent teachers defend it. Then we would have the difficult task of deciding exactly what it is that our children should be learning, and how it can be objectively assessed.