Tag Archives: Professor Robert Coe

Schrödinger’s homework: the problem with takeaway menus

One of the signifying mantras of progressive education’s child-centred approach is the idea of giving pupils elective choice in what or how they study.

One example of this choice in action is the phenomenon of ‘takeaway homework’.

Screen Shot 2016-07-12 at 16.54.28

For the uninitiated, this is where pupils are given a menu (usually emblazoned with the branding of a high street fast food chain) from which they get to choose to complete one (or some) from a range of homework tasks.

Whilst I’ll concede that it isn’t completely at odds with it, this idea does seem to sit uncomfortably with another of progressive education’s bogeymen: the marketisation of education (you can also add ‘Poundland pedagogy‘ as another bedfellow in this conflicted ménage à trois).

But that isn’t my main concern with takeaway homework. Neither is it the stealthy promotion of junk food that these menus might seem to endorse. It isn’t even, as Chris Moyse suggests, the excessive workload that takeaway homework creates.

No. The concern I have with takeaway homework is that, whilst it claims to be promoting valid homework, it’s actually doing the opposite. And that’s because it’s doing both.

You see, I think that takeaway homework can be seen as a thought experiment, similar to the paradox of Schrödinger’s cat.

This is what I mean. Take a look at this schematic figure of takeaway homework.


Pupils are offered 4 tasks to choose from. We tell them that these tasks are all valid and useful. Pupil One chooses the circle task. This means that they do not complete the other three tasks. We are now saying that, if we are happy for Pupil One not to complete these, they can’t be important. On the other hand, as they are completing the circle task, this must have validity.

Pupil Two chooses the square task. This means that the circle task that Pupil One completed does not have importance or validity. We are happy for Pupil Two not to complete that task; they can miss out on the learning from this task. We must, therefore, also be happy for Pupil One not to complete that task, even though they chose it.

Pupil Three chooses the pentagon task. This means that nobody chooses to complete the triangle task. We are happy for nobody to complete that task, so it must be unimportant. The learning provided by the triangle task can be bypassed by all pupils.

Put simply, the tasks on this menu are both valid and invalid at the same time. By organising homework in this way, we are suggesting that each task is simultaneously important and unimportant; useful and useless; they have both a learning outcome that we think pupils need and no learning outcome at all.

And the crux of all this is: if we are saying that some of those tasks are unnecessary but it doesn’t matter which, then we are actually saying that all of them are.

This is the problem of takeaway homework.

I think that homework needs to be directed, with a clear intention and learning outcome to be effective. Woolly, ‘anything goes’ approaches like takeaway homework is the opposite of this. It seems to hinge all of its claimed ‘effectiveness’ on things like motivation and engagement, which, as Professor Robert Coe tells us, are actually poor proxies for learning:Poor proxies fro learning

Where Hattie has thrown some doubt over the effectiveness of homework as an intervention, wouldn’t it be better to, as Tom Sherrington says, “be more specific and precise” in the tasks we set?

Even its advocates must agree that takeaway homework is far from specific and precise. And with that in mind, I’m personally hoping that we soon see yesterday’s takeaway homework menus as tomorrow’s fish and chip paper.


Blights of the round table: the damage of poor proxies for learning

(This blog post was originally posted on Staffrm)

Look at the image below and answer this question: how many empty spaces are there at the table?

Slide1Now look at this image and answer the same question: how many empty spaces are there at this table?Slide2Actually, the question I should ask is: which of the two people above appear to be less alone?

It seems a strange question. But not for a particular Seattle-based, world-dominating coffee peddler. According to author Karen Blumenthal, the belief that “people look less alone while seated at a round table” is the reason why you’ll rarely see a square table in Starbucks stores.

The company conducted research by interviewing hundreds of customers and studied the psychology behind what makes them tick. The idea behind the round table is that it doesn’t have any clear ’empty spaces’, unlike a square table. When you looked at the images above, according to Starbucks, the person at the round table should have looked less ‘alone’ than the person sat at the square table. Even if you didn’t register this consciously, you may very well have registered it subconsciously.

But the fact is that the two people in the different images are both as alone as each other. Even though one seems less alone, it isn’t true. They are both solo coffee drinkers.

The problem is that we are often easily fooled by what we glimpse, and we don’t often unpick the underlying truths to the meanings we’ve inferred. And no place is that more conspicuous than in lesson observation. We see things happening in lessons and automatically infer that learning has taken place. Often, we are very wrong.

According to Professor Robert Coe of CEM, we “readily accept poor proxies for learning, rather than seeking direct and valid evidence of true learning”. Whilst he concedes that it is “much harder” to do the latter, it doesn’t excuse the fact that we often judge, and are judged on these “poor proxies” – things that we assume show learning, but actually don’t:

Poor proxies fro learning

Much like with Starbucks’ round tables, we see these things and assume something that isn’t necessarily true. Just because students are busy or engaged or calm, it doesn’t mean that learning is taking place.

Whilst these things are logically desirable, they don’t really have anything to do with progress. And whilst it is certainly okay for schools to ask for these standards in lessons, the sad thing is that careers are often made or broken on the achievement of them, irregardless of progress. I have a friend (not in my school, I should add) who always gets excellent GCSE results. However, this teacher has been placed on capability measures due to failing lesson observations. Meanwhile a colleague of theirs has poor GCSE results yet revels in ‘Outstanding’ observations. I’m sure we all know stories like this.

These poor proxies were highlighted by Professor Coe a couple of years ago, yet still aren’t widely known in schools. I think academic work like this is too important to not be recognised by teachers and school leaders. Organisations such as NTENEEF and ResearchED are working well to reach schools and teachers that are engaged with research. But what about those that aren’t? How do this information reach them?

Maybe Starbucks have got something they can teach us about ubiquity or presence too?

Stop looking for the new Brain Gym. Start looking for the new Monkey Crouch

We all know that Brain Gym is guff. It’s so obvious to us now. In fact, it is such guff that it has even given birth to an edu-snowclone:

X is the new Brain Gym.

What is interesting is how schools swallowed it in the first place. Why did we? I think much of it came from its truthiness, and the problem there is that nobody was really measuring its impact. As far as my brief fledgling complicity goes, I was told it was based in ‘science’ and it looked interesting so why would I question it? Why would I need to measure the impact it had made over a period beyond that moment when everybody was actually engaged in the activity?

And this is the crux of the issue: with much of what we do in the classroom, we respond to the moment. Our education system seems built on what is happening at the moment of teaching, over and above what impact that moment has. From the very start of our careers when we are observed during training, we are encouraged to reflect on the individual lesson. More so than the sequence of lessons. More so than how the learning has developed over the term.

Whilst there is an acknowledgment from some camps – Ofsted include themselves here – in moving away from it, lesson observation feedback (and grading, where it exists) is largely given based on what was seen in the moment – notes are made and I’d wager that any grading is decided on in the majority of cases whilst the observation is still taking place.

But wouldn’t it be more useful for the observer to come back later (a lesson later, a week later… longer than that?), find out what the pupils have actually learned from the lesson – what they have retained and remembered; what they can do now – and give feedback based on that?

With such a culture of immediacy, it is no wonder that we fall foul of so many ineffective teaching practices – as highlighted in the Sutton Trust report published this week. The report, written by Professor Robert Coe et al. from Durham University, offers up some examples of practices that have strong evidence of impact, as well as those strategies which are prevalent in schools but aren’t supported by evidence. In many of the cases of those that aren’t supported, I would suggest that they are approaches that are quite visible or tangible in the moment – praise, grouping, discovery learning, active learning, etc. – and perhaps that is where their appeal lies: we can see them so we are drawn towards them as ‘evidence’. Conversely, the two factors with the strongest evidence are those which might often be invisible or less tangible in the moments of a lesson – their impact is in the long term and so, perhaps, is their visibility:

Screen Shot 2014-10-31 at 22.22.01

An area where a similar dislocation in the judgement of practice has occurred is in the age-old sport of horse racing. Although, ancient as it is, it is a sport that hasn’t really changed an awful lot since Areion of the Black Mane won the 2.30 at Nemea by a nose (no doubt in front of a young John McCrirrick). Indeed, in the past 100 years – when we have made huge strides in almost every arena of civilisation – horse racing times have only improved by a trifling 1%.

But, actually, in the late 19th century, there was development in horse racing that improved times by around 7% in just a few years. Yet it was a development that, when judged in the moment, was actually derided and ridiculed by jockeys and the sports press alike. At this time, jockeys used to race in an upright position, as in the picture below. jockeyonhorse-early However, an American jockey noted for his observant mind, Tod Sloan, took a different approach. A 1900 edition of Vanity Fair tells us that he “studied the problems of wind resistance and adopted a posture of crouching along the neck and shoulders of the horse.” That same publication also tells us that he was ridiculed for this at first – indeed VF‘s founder Jehu Junior mockingly likened his “peculiar seat” to that of “a monkey on a stick”.

In those days, Britain was the centre of horse racing. Having conquered the sport in the US, Sloan came to London in 1897 with his ‘monkey crouch’ style. He was instantly rounded upon by the British press, who took great joy in ridiculing and mocking his methods in both the written word and with satirical cartoons.

But Sloan won races. Lots of them.

And as he won more and more races, slowly people started to copy him. In the few years that followed, as more jockeys began to take up the ‘monkey crouch’, race times improved by much more than they have in the 100 years since. The Daily Mail finally had to concede: “It is useless to deride the style and methods of a jockey who keeps winning.” In 1900, he was asked by the Prince of Wales to ride for the royal stable, wearing the soon-to-be-king’s colours. Today, every jockey rides using the approach that Sloan was ridiculed for at first.

Those in the horse racing world made the mistake that we have in recent years – they judged what they saw in the moment ahead of looking at the results of the method. Instead, as the Sutton Report suggests, we should look at the effectiveness of our approaches and evaluate teaching based on this. Professor Robert Coe suggests caution in judging the moment:

“Given the complexity of teaching, it is surprisingly difficult for anyone watching a teacher to judge how effectively students are learning. We all think we can do it, but the research evidence shows that we can’t. Anyone who wants to judge the quality of teaching needs to be very cautious.”

How long before we start to evaluate more effectively, outside of the moment? Place your bets.