Why teachers don’t know best

The blind leading the blind by  Pieter Bruegel the Elder, 1568It is not surprising that teachers get impatient when others tell them how to do their job: “we are the experts”, they complain, “not you”. What should surprise the rest of us is how wrong they are: most teachers know little about teaching as a technical discipline.

This post responds to a comment by someone nicknamed subminiature, who argued on the Radio Times website that teachers knew what they were doing and should just be left to get on with the job. In this lengthy response, I argue the opposite: through no fault of their own, teachers do not have the skill-set that is required to improve the chronic under-performance of our education service. This will only be achieved by the implementation of education technology, backed by sound pedagogy. It is not surprising that teachers are not technology-experts: what is surprising is that they are not experts in pedagogy either. Expecting teachers to lead the sort of transformative development that is required in education is about as sensible as expecting a group of horse-drawn carriage drivers to design the first steam engine. Yet that is precisely the assumption on which government policy has been based over the last 15 years. A policy based on teachers sharing ed-tech best practice is analogous to Breugel’s allegory of the blind leading the blind.

(11,200 words).

My conversation with subminiature

Stephen Fry and the French

The British comedian Stephen Fry recently pronounced that the French are “a better educated race” than the English. I made a comment on the Radio Times website suggesting that there might be something in what he was saying. In response to this comment, subminiature responded that teachers were tired of being told what to do by Joe Public, who always considered himself an expert just because he once went to school too. I copy his comment in full below.

Everyone who went to school and even a number home taught believe they know what is best in the classroom. Their simplistic view of teachers teaching and pupils learning breaks down when you see how the system enables and blocks learning for a given pupil. The changes imposed upon teachers by those who have no training or experience in teaching are the biggest problem in education.
There is no great magic to better education. The research has been done, but largely ignored in flavour of research that reenforced a preconceived view even if the evidence points the other way.
First funding for all pupils should be similar across the country. Some counties get significantly less per pupil than others.
Reduce class sizes.
Make teaching a profession with a governing body.
Government may monitor that standards are maintained and that schools are run correctly and require improvements but the way and how education and care of the pupil is undertaken day to day is down to the ‘profession’.
The first task of education is to teach children how to learn and to enjoy learning for learning sake. That seems to have been lost, if in fact it ever was a priority in English schools.

I take this comment as the text for this article:

  • because subminiature articulates what is, on the face of it, a reasonable position and one that I do not doubt is widely held among teachers;
  • because I believe that there is such an strong body of evidence to suggest that this position is fundamentally wrong.

The points at issue

First, let me list the points in subminiature’s comment that I agree with:

  • that teaching practice should as far as possible be based on evidence-led research;
  • that the formula used to fund education should at least be transparent and based on a capitation grant that provides the same basic funding for every student, even though a certain degree of variability (e.g. in respect of the proportion of students on free school meals and other measures of deprivation or need) is also sensible;
  • that government should not interfere with the day-to-day teaching and care of pupils, leaving these operational matters to teachers, given also that teachers are held to account for delivering satisfactory and consistent outcomes;
  • that in principle, teachers should be regarded as a genuine profession, though I think there are practical difficulties in achieving this in practice.

The points that I disagree with are that:

  • “there is no great magic to better education”, which I take to mean “teachers know what works”;
  • “the research has been done”, which I take to mean that the research evidence for the knowledge alluded to above is secure;
  • it is a priority to “reduce class sizes”;
  • “the first task of education is to teach children how to learn…”
  • “…and to enjoy learning for learning sake”.

I shall address the points of disagreement in reverse order

Our points of disagreement

Learning for learning’s sake

Now, don’t get me wrong. I think that learning for learning’s sake is great. In his Nichomachean Ethics, Aristotle called it “contemplation” and rated it the highest of all human activities, on the grounds that it had no instrumental (therefore superior) benefit. [Comment by Dylan Wiliam] The reward was entirely intrinsic. It is surely good that teachers, through the passion that they show for their subjects, should imbue their students with the same sense of intrinsic wonder, curiosity and delight in the subject that they feel themselves. But this should be their ultimate task, not their first:

  • teachers are likely to alienate their students if they welcome them by condemning the motivation with which they start the course;
  • it is unrealistic to expect people to understand the intrinsic joy of knowledge until they have a deep experience of that knowledge, and this will be something that a minority of students may attain at the very end of their course and not something that can be expected of the majority of students at the beginning;
  • even at the end of the course, it is not for teachers, as suppliers of a service, to pass judgement on the motivation of their customers, any more than it is for a plumber to pass judgement on whether their customer is right to order a Jacuzzi for his bathroom;[Comment by Richard]
  • we should not complain too much about materialistic motivations when we all depend on the fruits of materialism for the delivery of what we consider as our basic human rights, not to mention our pleasures, which most of us regard as pretty good too;
  • the assumption of superiority by the academic can quickly turn into arrogance; and the doctrine of intrinsic good into a rejection of accountability—as Frederick Nietzsche argued, there is no greater ego trip than standing in a pulpit.

While teachers should try, preferably by example, to convey a sense of wonder, curiosity and delight in learning for its own sake, they should not regard this as their primary goal and should certainly not be arrogant enough to deprecate students who want to master their subjects merely because they want to get a good job.

Learning to learn

This is a cornerstone of the pedagogical doctrine that has underlain much of the thinking during the Becta years on education technology. The internet, so the theory goes, has given students the tools to enable them to become independent learners. Technology has also driven the requirement for them to do this as the skills that learners require are changing so fast that anything that they learn at school will be obsolete in a few years time. So teachers not only can but also need to prepare students to be “lifelong learners”. What matters will be what they learn after they leave school, not what they learn in the classroom.

I believe that this attitude has been profoundly damaging to teaching and learning over the last decade—but before making my arguments, I need again to add a caveat. Just as I think that it is highly desirable that students should acquire a sense of wonder, curiosity and delight in the world around them, so too I think it is great when these attitudes lead people to carry on learning all through their lives. This might almost be called the mark of an educated person. The problem lies not in the state of mind achieved, but how we expect students to get there.[Comment by Dylan Wiliam]

The paradox of learning

One of the more popular aphorisms of Aristotle at the moment (perhaps because it contains fewer than 140 characters) is “For the things we have to learn before we can do them, we learn by doing them”. It is a paradox, of course, a “Catch 22”. We can only learn to do something by doing it and we cannot do it until we have learnt how. But like most paradoxes, it is not a contradiction in terms. What appears to be a contradiction is in fact resolvable, and in this case by resorting to practice. If we repeatedly try to do something, our attempts while in a uneducated state might sometimes work and sometimes not, but given the ability to discriminate between our relative successes and failures, we can repeat and incrementally improve upon our successes.

This leaves us with two “takeaways” from Aristotle’s aphorism:

  • we learn by doing;
  • doing is not always easy, especially at first.

If we “learn by doing”, then the best way of learning to ride a bike is simply to ride a bike. How then are we meant to “learn to learn”? The answer, of course, is simply to learn. What insight, therefore, is contained in the advice that you shouldn’t just “learn” but that you should go one better and “learn to learn”? The answer is “none at all”. The best way of “learning to learn” is just to do some plain vanilla learning.

The argument against dependency

The “learning to learn” enthusiasts will accuse me at this point of playing sophistical word games. The point, they will say, is that learner should be enabled to learn by himself and not in a state of dependence on the teacher.

The perception that students become over-dependent on teachers tends to be conflated with the perception that that learning is seen too much in terms of passively receiving information. The dependent relationship is supposed to be one in which the student sits in the classroom or lecture hall and takes notes, ready to regurgitate those same notes in the test. The independent learner, by contrast, is meant to go out into the world and find things out through active experimentation, learning skills and attitudes and not mere facts.

Of course I agree that an education in which the teacher merely hands out information to students is very limited. And so, across the ages, do all the other educationalists who have stressed the fundamental importance of active approaches to plain vanilla “learning”.

Socrates argued for a form of teaching which he likened to midwifery, in which no information at all was conveyed to the student by the teacher. The role of the teacher, he argued, was to be a critical interlocutor who, through a process of constantly challenging the student’s superficial assumptions, drew out the knowledge that was already inside the student.

William Cory, Assistant Master of Eton, wrote in 1861: “You go to school at the age of twelve or thirteen and for the next four or five years you are engaged not so much in acquiring knowledge as in making mental efforts under criticism”.[1].

John Stuart Mill argued the same point at much the same time. Instead of being told the right answer, students should be forced to argue their case. The role of the teacher in these circumstances was not to be the fount of all knowledge but rather to play the devil’s advocate, even if that meant the teacher arguing strenuously for a position that he/she thought was wrong[2]:

He who knows only his own side of the case, knows little of that. His reasons may be good, and no one may have been able to refute them. But if he is equally unable to refute the reasons on the opposite side; if he does not so much as know what they are, he has no ground for preferring either opinion. The rational position for him would be suspension of judgment, and unless he contents himself with that, he is either led by authority, or adopts, like the generality of the world, the side to which he feels most inclination. Nor is it enough that he should hear the arguments of adversaries from his own teachers, presented as they state them, and accompanied by what they offer as refutations. This is not the way to do justice to the arguments, or bring them into real contact with his own mind. He must be able to hear them from persons who actually believe them; who defend them in earnest, and do their very utmost for them…Where this advantage can no longer be had, I confess I should like to see the teachers of mankind endeavouring to provide a substitute for it; some contrivance for making the difficulties of the question as present to the learner’s consciousness, as if they were pressed upon him by a dissentient champion, eager for his conversion.

Educationalists have long recognised that education is not merely about the transfer of information. They have always recognised education is an active process requiring “learning by doing” or in the case of abstract learning, “learning by conversation”. At the same time this realisation does not remove the student’s dependency on the teacher—on the contrary, it is precisely this requirement for activity, debate and criticism which establishes the students dependence on the teacher.

This point does not seem to be understood by many teachers and educationalists. Large numbers of teachers continue to place their confidence in the transmission of information by lecturing; those who oppose such expositive pedagogies argue that students can only become active do-ers if the teacher retreats from the stage and becomes only a “guide on the side”. Both are wrong. What traditional approaches to pedagogy teach is that the student can be both an active do-er and at the same time can be—and in the early stages of learning needs to be—dependent on the teacher for that guided activity.

At this point we come back to the paradoxical part of Aristotle’s aphorism. Even though we learn by doing, it does not follow that we can necessarily “do” straight away. If you want to learn to do a triple somersault, you would be wise not to throw yourself of the high diving board on your first attempt. It would be better to start with a single back-flip off the side of the pool, progressing through a graded sequence of dives and exercises, the nature of which the student diver is almost certainly ignorant. They may know that they want to do a triple somersault of the high board but they do not know how they can get from here to there and for that, they need direction from a teacher.

In the same way, you do not necessarily learn to be an independent learner by straight away trying to learn independently—you start by learning under the direction of a teacher, progressing to make independent forays under supervision, short at first but increasingly long as you grow in confidence. An infant grows into an independent adult because of early parental support and not despite it The reason why such dependency is necessary has little to do with knowledge transfer and much to do with the need for criticism and assignment management. Premature independence does not encourage students to be more active because at that stage in their development they probably do not even have a clear conception of what it is that they are meant to do, let alone have the ability to do it. The role of the good teacher is to encourage the development of independence incrementally, as the student shows that they are able to make use of such opportunities.

The doctrine that we should “learn to learn” is not just meaningless—it is also positively damaging in its implications of premature independence.

Independent learning in practice

In practice the theory of independent learning has been closely connected with theories, popular over the last 15 years, about how we should use technology in education. It was proposed that the internet provides the tools through which the students can break free from their dependency on their teachers at a much earlier stage than was possible previously. This in turn has been behind the concept of ICT in which the best way of using technology to improve teaching and learning was to teach students a set of digital literacies, which were supposed to free them from the restrictive dependence on the teacher.

Once they have mastered these basic digital literacies, students are meant to be able to learn either from online learning resources or from the use of technology as an empowering tool. At school level, the Khan Academy provides an example of a large collection of online videos, all of which focus on the transfer of information. At HE, most Open Education Resources and MOOCs use the same predominantly expositive approach.

The  two ways in which ICT was perceived to support learning—first as a medium for the dissemination of information and second as an enabler of interactivity—came together in the expectation that it could provide a platform for “internet research”, which supposes an active approach to the collection of information.

Professor Stephen Heppell was described by Microsoft as “Europe’s leading online education expert” and by the UK Department for Education as “the most influential academic of recent years in the field of technology and education”. When in 2008 he appeared on Radio 4’s Today programme to give his backing to the recently published Rose Review, the interviewer, Evan Davis, expressed his surprise at the assumption that primary school children would be sufficiently mature to do this sort of research [at 01:30]:

EVAN DAVIS: Some of it [the Rose Review] sounded very adult, I have to say. It was trying to give youngsters “research skills” [laughs] rather than facts about History, for example. Well maybe you need just a little bit of the basic outline of the facts before you start on the more advanced stuff.
STEPHEN HEPPELL: Well you need both and the report is very sensible about valuing the subject teaching alongside the project-based work that we know is invigorating for children as learners.

Professor Heppell does not acknowledge that there is any problem with pushing students prematurely into independent research but parries the criticism by suggesting a dual approach between subject teaching and project-based work. While the former presumably takes care of the facts, the main benefit of the latter is that it is “invigorating” (which I guess is much the same as saying that it is motivating). He goes on to give an example of how this works in practice: [at 03:03]

I was in a school in West London just this last week where the children are out collecting 100 faces: a 1-year-old, a 2-year-old, a 3-year-old, a 97-, 98-, 99-year-old, and they’ve got a standard question for each one: “what was the first thing you had for breakfast this morning?”. And then they’re finding 99 other places in the world where they’ve done the same thing. And its a really stunning piece of research. And they are swapping backwards and forwards: “what did everybody have for breakfast?” “what were all the 10-year-olds doing?” “why can’t we find anyone in their ’80s in West Africa?” Really interesting stuff.

If this was “research”, then what exactly were its “really interesting” conclusions? that a regular full English breakfast led to a long life, perhaps? Who was drawing these conclusions? What skills of data analysis and research methodology were the students displaying? This is the sort of international project that may sound good on radio and which may need a globe-trotting consultant like Professor Heppell to co-ordinate; but it is not clear that any learning is occurring at all. In fact, no-one seems to claim that it is delivering learning—that is for dull old subject teachers to worry about. This is all about sexy project work which earns its keep by being “ambitious” and “invigorating”.

In its 2001 report ICT in Schools, Ofsted gave a prescient warning about the educational benefits of so-called “internet research” (paragraph 35 on page 10).

For too many pupils, the location of information remains an end in itself, and they present information unprocessed. In order to progress, pupils require much more sensitivity, determination and understanding to handle large volumes of potentially relevant information, as well as strategies for focusing on the most useful material for the purpose in hand. Too often the use of ICT involves unnecessary extra work or unproductive waiting, such as when finding and downloading a map from the Internet or printing graphic images.

Ofsted’s more dispassionate assessment suggests that internet research delivered very few educational benefits at all. Download speeds may have improved since 2001 but the problem of information processing remains. Unfortunately this was the last clearly expressed warning that Ofsted gave. After that, the bandwagon dedicated to hyping up the benefits of dodgy ICT-related pedagogies became too strong. Ofsted was told to get its tanks off Becta’s lawn and mainly confined itself to the inspection of ICT as a subject. Encouraged by Becta and the advocates of independent learning, “internet research” became one of the favourite homeworks to be set by lazy teachers. Instead of setting a challenging activity that required forethought, design and feedback, all the teacher had to do as they high-tailed it out of the classroom, was to say, “see what you can find out about x for next time”. And the marking was easy too: what came back was invariably pages and pages of well crafted, beautifully presented prose. Children who spent their time collecting cut-and-paste, unprocessed information, increasingly saw education as the collection of boring facts, and failed to learn the distinction between research and plagiarism.

So how have things moved on since 2001 and 2008?

When Michael Gove was asked by Peter Gothard of Computing Magazine about the potential for online learning in schools, the Secretary of State replied that:

It is not for me or for government to wade in to dictate how schools and teachers should use [digital resources] within the curriculum. I do think that it is important for teachers to have the skills to use technology well, and for schools to learn from innovators in this field. That is why the Department for Education is supporting the National Teaching Schools’ new technology group to identify and share good practice.

I cannot find any direct evidence of the NTAB’s activities – but those involved in leading the NTAB have been active building a Google website which is advertised as a collection of “resources for teachers by teachers”. I copy below a random selection of links from the “Resources for self study” section, taken from the first four topics listed for KS3:

The offering comprises painfully uninspiring, expositive resources—video lectures and documents—from which any half-competent teacher would run a mile before recommending to a student. The last video ends with the great strap-line, “If you have any questions, ask a librarian”. It is at least disappointing to those of us who wish to promote education technology as an exciting medium through which teaching and learning in schools can be improved that the initiatives that seem to achieve most recognition by the DfE are so heavily based on the ineffective use of the internet for the dissemination of factual information.

I finish this section by returning to the core point, which is about the mis-selling of independent learning as a means of delivering more active style of learning. It was thought that by reducing dependency on the teacher, the student would become a more active learner. In practice, the application of these theories has had the opposite effect, making the student more dependent on the passive accumulation of irrelevant information from the internet. As Professor Diana Laurillard comments:

Technology opportunists who challenge formal education argue that, with wide access to information and ideas on the web, the learner can pick and choose their education—thereby demonstrating their faith the transmission model of teaching. An academic education is not equivalent to a trip to the public library, digital or otherwise[3]

The theory of independent learning, fuelled by the publication of expositive learning resources, has done nothing but damage to the quality of learning in our formal education system.

That it is a priority to reduce class sizes

The evidence against this proposition is overwhelming.

In 2002, research at Harvard’s Education Dept into education systems around the world found that:

Smaller classes exhibit beneficial effects only in countries with relatively low teacher salaries [Greece and Iceland].

In 2007, Dylan Wiliam of the Institute of Education in London stated at his keynote to the Association for Learning Technology:

Class size reduction—reducing class size by thirty percent—actually gives you a twenty percent increase in the speed of learning. But it costs twenty thousand pounds per classroom, per year… But if you get teachers doing formative assessment in their classrooms, you get a seventy-five percent increase in the speed of learning and it costs about two thousand pounds per classroom.

A shame then that teachers representatives routinely criticise government policy on the grounds that its focus on assessment is a distraction from the serious business of teaching. [Comment by Dylan Wiliam]

In 2012, PISA’s Education Indicators in Focus states that

Reducing class size is not, on its own, a sufficient policy lever to improve the performance of education systems, and is a less efficient measure than increasing the quality of teaching…Between 2000 and 2009, many countries invested additional resources to decrease class size; however, student performance has improved in only a few of them.

Pearson’s 2012 Learning Curve report states:

reducing class size is expensive and has little or no impact on system performance.

I could go on.

That teachers know best

I have already shown that in advocating the importance of “learning to learn” and that class sizes should be reduced, subminiature (who I presume is a teacher) clearly shows that he or she does not know best. But this is just the tip of the iceberg. Teachers routinely make statements (often in highly polemical tone) that show that they understand very little about teaching as a systematic, technical process, or of learning as a neurological process. I list some of them, by way of example, below.

The primacy of personality in the classroom

Teachers commonly view their most important contribution in class as being through personality, relationships and inspiration. So long as the teacher makes sure that children want to learn, then they will get on and do the rest by themselves.

Twitter conversation on inspiration

Most teachers would say something similar—the key contribution of the teacher is inspiration. But what about “expertise”, “subject knowledge”, “criticism”, “analysis”, “progression management”?

Listen to the audience reaction to Professor Mazur’s talk to the Specialist Schools and Academies Trust (go to 18:48 if the bookmark does not take you there straight away). Mazur asks a conference of teachers “what do you need [in education] in addition to the transfer of information?” The audience (composed almost entirely of teachers) answers:

  • challenge;
  • relationships;
  • personality;
  • feedback.

“Relationships” and “personality” are in the same vein as “inspiration” and demonstrate most clearly how teachers think about their role in the classroom. “Challenge” and “feedback” are warmer but still don’t come close to giving the correct answer to what must be one of the most basic questions about pedagogy that you could ask. And that is from a conference hall packed with teachers to which Professor Mazur had already given the answer in the earlier part of his talk.

What is needed in addition to the transfer of information is the “practice” or “activity” that will allow the student to assimilate the information and apply abstract knowledge to a wide variety of concrete contexts. These are the processes that would be delivered by a scientific, activity-based approach to teaching, justified by reference to the evidence of what works. These are the processes that our current model, focusing on teaching as a personal craft and on information retrieval via the internet, is so bad at doing that hardly any of the people who man the current system even understand what it is that is required of them.

I should make clear at this stage that I am not saying that personality, relationships and inspiration do not matter too. They partly matter, as most teachers believe, because motivation is an important prerequisite for learning. But they matter even more in a sense that I do not believe that most teachers understand properly: that imitation and role-modelling are vital to human learning. It is a very tall order to ask teachers to provide compelling role models to their adolescent pupils in an age dominated by superficial but media-packaged celebrities. But that is another problem which I shall address in a future post. The main point for the purpose of this article is that while motivation, inspiration, relationships and personality all matter, they are not sufficient to deliver consistent learning outcomes. For that, you need not just to inspire, you need to teach.

That assessment distracts from teaching

Teachers commonly complain that the government’s emphasis on assessment distracts from teaching. In the same interview on the Today programme, Professor Stephen Heppell made clear his view that assessment is irrelevant to learning in the modern world [go to 03:40]:

HEPPELL: If there’s a worry here—and I think there is a worry—its that there is a bit of a gap opening up between the schools that are being ambitious and ingenious with their learning and the schools that are just doing the predictable old stuff—you might call them educational fundamentalists…
EVAN DAVIS: Which of those two types of school did better in the tests?
HEPPELL: Well, you see, the tests are quite interesting. You sense…that the tide is turning on all this, the tests are ebbing away. But there’s an interesting little point here. In the last century, we built our education system—and certainly our tests—on a model of “have you met this before?” …and yet the world we are in is a world in which we have never met anything before.

Professor Heppell is obviously wrong in at least one respect—the tests have not ebbed away but have rather flooded back with a vengeance, much to the disgust that is evident in his attack on this trend as “unthinkably foolish, unutterably irrelevant“. He is also wrong (though perhaps less obviously) about the reasons that he originally gave for claiming that the tests are no longer required.

It is not true that tests are necessarily built on a model of “have you met this before?” Students can easily be tested on their ability to create things or solve problems that they have not met before, drawing on abstract, transferable skills (see Mazur on transferable skills at 34:46–35:18). And yet critics of Michael Gove, a great number of which are teachers, follow Pressor Heppell in endlessly parroting the same old gross simplification, that assessment is all about “the three Rs—reading, remembering and regurgitating” (as summarised by Dr Brian Lighthill in a letter to Guardian); or that exams are for children “who thrive on memory and recall testing”, as suggested in Fiona Millar’s video giving “Gove ‘F for Fail’ in exam plans”.

Nor is it correct to say, even allowing for hyperbole, that we live in a world in which we have never met anything before. As I argue in the conclusion to Education’s coming revolution, to claim that “21st century skills” were invented a few years ago is about as convincing as claiming that sex started in 1963. If you look at the issue in detail, you see that most of the so called 21st century skills would have been just as familiar to Socrates as they are to us. What changes is the context in which abstract learning is applied and communicated: the fundamental principles change very little. This even applies to the field of Computer Science, which is at the eye of this supposed storm of change. The Royal Society states in its report Shut down or Restart? that the discipline of Computer Science:

has longevity (most of the ideas and concepts that were current 50 or more years ago are still applicable today), and every core principle can be taught or illustrated without relying on the use of a specific technology.

Professor Heppell’s view that assessment is irrelevant is echoed by very many teachers who claim that it is a distraction. 

Twitter conversation on assessment

In this Twitter conversation, Fran does nothing more than restate the commonly held position of teachers that assessment means “time-out” from learning. This ignores the research evidence produced by Professor Dylan Wiliam that the exact opposite is the case: assessment is the best form of re-enforcing and assimilating what we have already been told and is therefore a vital part of learning.

The evidence from Dylan Wiliam is exactly what would be assumed by someone who followed the argument of Professor Mazur. The essential second element of teaching is practice or activity—and assessment is nothing more than a tracked activity. While students are sat in rows listening to the teacher, then it is not really possible to assess them. As soon as they are active—practising, applying, creating, problem solving—then why would you not want to track how well they were doing, if only so that you can give them appropriate feedback? And once they are active and you are tracking the success of their activity, you are running an assessment—an assessment that has nothing at all to do with regurgitating facts.

Assessment is not just important for diagnostic purposes: it also lies at the very heart of good teaching. In fact the whole language used by teachers, suggesting a dichotomy between teaching and assessing, is deeply unhelpful. All assessment should be formative (i.e. promote learning) and all learning should be assessed. Achieving this state of affairs is what according to Dylan Wiliam the research shows would provide the most effective means of improving the quality of teaching and learning in our schools. What a shame, then, that the language commonly used by teachers assumes an antithesis between learning and assessment, rendering such an objective difficult for most practitioners even to conceive.

The argument against knowledge

Allied to the argument against assessment is the argument against knowledge, which is popularly deployed against Michael Gove’s new curriculum. Back again to Professor Heppell, writing in the Financial Times on 17 July 2013, giving an account of a speech by Kenneth Chen, the undersecretary of state for education in Hong Kong (who uses language peculiarly reminiscent of Heppell himself, perhaps because, although not mentioned in the article, Heppell was working for him at the time as a consultant—a rather curious case of self-referencing):

He [Chen] urged educationalists to “move away from a focus on content knowledge” and to embrace the concept of “learning to learn” that he had placed at the heart of education reforms. Technology was crucial to achieving that ambition.

Heppell continues to criticise the new “fact-based English history curriculum”:

Focusing on content is a common economic mistake. In the 20th century, in parallel with the expensive misunderstandings of the dotcom bubble, education-based companies thought their market would be content delivery. Surely, content was king and delivery equalled dollars? The answers were it wasn’t and it didn’t.
In a world awash with content, much of it free, ingenuity and creativity were increasingly scarce and valued commodities.
However, will English schoolchildren, newly returned to rote learning, sitting down to an exam paper and hoping there are no surprises, be ready for the continuing uncertainty and the constant surprises that characterise our current economic circumstances?

This passage is full of unjustified assumptions and non-sequiturs:

  • why does Heppell assume that the new English exam system will focus on rote learning, just because the curriculum specifies the range of topics that should be taught?
  • why does Heppell conflate content-meaning-information, with the content-meaning-digital-resources—resources that can include tools, games and simulations just as well as expositive texts?
  • why does Heppell assume that knowledge plays no part in the ability to deal with “continuing uncertainty and constant surprises” or (as proposed by Bloom’s taxonomy) in the development of “ingenuity and creativity”?

At the heart of Professor Heppell’s position are two beliefs:

  • that information accessible on the internet is—to all intents and purposes—just as useful as knowledge that is directly accessible to us in our own memories;
  • that ingenuity and creativity can be plausibly developed in the absence of such knowledge.

Both of these points are false—but I will leave them to a future post on the role of knowledge in learning for a detailed rebuttal.

That teachers should be facilitators

Government ICT Skills champion and previous Dungeons and Dragons inventor Ian Livingstone appeared on a video debate at Computing magazine, on which I was also present. Arguing that the lack of skilled Computer Science teachers should not be regarded as a problem when introducing a rigorous new computer Science curriculum, Mr Livingstone echoed a common perception among teachers by arguing that:

If ICT teachers don’t feel qualified today to serve a rigorous and robust curriculum, let them be facilitators.

According to Mr Livingstone, you do not need teachers who know much about their subject, any more than you needed bread to feed the population of Paris in the late eighteenth century. You can just let them eat cake. It is not that Ian Livingstone should be blamed too much for adopting such an absurd position—he is merely providing a window on the general consensus among teachers and educationalists. He picked up his views from the wider community, in which it is believed that teachers need be no more than a “guide on the side”.

Following my write up of the Computing magazine round-table, I received an email from Terry Freedman, a well respected blogger on ICT teaching, who says:

Totally agree with your comments about “let them be facilitators”… Like you (I think) I get tired of all the hype which bears no evidence of having been thought through, let alone acknowledgement of any research…

After recommending a similar article that he had written about facilitators, Terry continued:

I do sometimes feel like the king in a Kahlil Gibran story: someone poisons the water supply, and everyone in the kingdom wakes up one morning totally insane—except for the king, who had not drunk any water during the night. All his subjects thought HE had become insane—so he drank the water, and became as mad as they were, upon which they rejoiced that their ruler had regained his sanity!

Et cetera, et cetera, et cetera

I could go on all day listing silly theories that are widely believed by teachers. In fact, in his recent book Teacher Proof, Tom Bennett has done just that. I will review the conclusions of this book in more detail in my next blog but, for the time being, its significance here lies in its survey of multiple lunatic theories of education, most of which have achieved a considerable following in the teaching community. These (some of which I have already addressed) include:

  • multiple intelligences;
  • Neuro-linguistic programming;
  • Brain Gym (TM) ;
  • group work;
  • emotional intelligence;
  • twenty-first century skills;
  • digital natives;
  • the three-part lesson;
  • learning styles;
  • gamification;
  • learning to learn;
  • divergent thinking (making some use, I suspect, of my January 2012 post on Sir Ken Robinson).

I do not agree with all of Tom Bennett’s conclusions (more of which in my next post) but the point that he makes very persuasively in this middle section of the book is that this long list of educational theories, despite being promoted on the basis of non-existent evidence and transparently nonsensical justifications) have been lapped up by gullible practitioners and officials.

Variability of outcomes

So much for the theories and theorists that are popular among teachers today: what about their performance? As the proof of the pudding is in the eating, so too are outcomes the ultimate criteria against which expertise should be judged. The problem here is in deciding what constitutes a good outcome.

This is where we come back to Stephen Fry’s perception that the French are a better educated race than the English. If he is right, then the outcome of English education must be judged to be worse than the outcome of French education and, while there are very many factors that might be responsible for such a difference, it becomes rather harder for teachers to rule out of order a suggestion that the style of education in England (which places very much less emphasis on assessment than the French system) might have something to do with the outcomes.

If teachers wish to argue that the difference can be explained by cultural differences, then they will also need to explain why, in spite of much rhetoric and hundreds of millions of pounds spent on outreach programmes, the private sector and selective grammar schools are increasingly monopolising the places at the most selective UK universities: 64% of private school children get these places, 58% of grammar school children, and only 20% of comprehensive school children. The previous government became increasingly hostile in their criticisms of Russell Group universities for not doing more to encourage comprehensive school children to apply to university. The Russell Group universities replied (though maybe in more diplomatic language than this) that there was no way that you could make a silk purse out of a sow’s ear: the fault lay with the comprehensives. Although some of the discrepancy between comprehensives on the one hand and private and grammar schools on the other can of course be explained by the amount of money being spent on each child’s education or the selectivity of their intakes, it does not explain the wide and ever widening divergence in their success rates. Maybe a better explanation is the lower quality of teaching in a comprehensive sector, overly impressed by theories of independent learning? [Comment by Dylan Wiliam]

But the strongest evidence comes from the same talk already cited by Dylan Wiliam, who claims that the research shows that:

There’s a four-fold difference in the speed of learning created by the most and the least effective teachers. And it’s not class size, it’s not between class grouping, it’s not within class grouping – it’s the quality of the teacher…

This is like saying that there is a four-fold difference in the death rates between two surgeons in the health service: a statistic that would surely hit the headlines and cause a national scandal. The best teachers may well be very good indeed but the wide variation in performance shows that this expertise is a matter of the personal craft of the individual teacher and is not embedded in a body of established, professional knowledge. If you go to a doctor, you feel entitled to expect that he or she will be good. If you go to a teacher, most of us know from personal experience (and Dylan Wiliam’s evidence confirms) that we are rolling up our sleeve and thrusting our hand deep into the lucky dip. In these circumstances, the establishment of a new professional body for teachers would amount to little more than an empty gesture.

That the research has been done

Research has been done, of course, and much of it, as I show above, flatly contradicts what subminiature thinks has been securely established. But what has not been done is the research—research which is so conclusive that we can regard what works in education as being a question that is done and dusted.

Most of the research I reference above is at a very high level—it doesn’t help very much when it comes to exactly how teachers should apply teaching strategies in the classroom.

When it comes to focused research on pedagogy, it has been clear for some time that the quality of most educational research is of poor quality and largely irrelevant.

In 1996, Professor David Hargreaves, Professor of Education at the University of Cambridge, stated in the Teacher Training Agency’s annual lecture that there was a considerable amount of:

frankly second-rate educational research which does not make a serious contribution to fundamental theory or knowledge; which is irrelevant to practice; which is uncoordinated with any preceding or follow-up research; and which clutters up academic journals that virtually nobody reads.

There was a predictably hostile reaction to this statement from the research community. Professors Michael Bassey and Hilary Constable retorted with an analysis of the titles of 12,000 research papers submitted to the Education Panel of the Higher Education Funding council’s 1996 Research Assessment Exercise, showing by an analysis of the titles of these papers that research was relevant to education. However, Bassey and Constable’s paper was in turn refuted by the Tooley Report (James Tooley and Doug Darby, published by Ofsted in 1999). This analysed first 264 articles, which represented all the articles published between 1994 and 1996 in four of the most important peer-review journals for education research; and then a sub-sample of 41 articles, which were analysed against a set of 30 questions, with the answer to each question being agreed by two researchers coming from contrasting wings of political opinion. If found the following:

  • while it found that almost all the research reviewed was relevant to education in some form, it did find that only 10% of all research addressed general questions of teaching and learning (other topics were gender and race issues, curriculum issues, governance etc.);
  • it found that only 15% of all research presented quantitative evidence (as opposed to qualitative, “non-empirical”—i.e. argument only, or “reflexive”—i.e. personal opinion, none of which can provide reliable, repeatable, empirical proof);
  • it found that only 36% of all research conformed to basic standards of good practice.

The reason why almost two thirds of all research failed to meet Tooley and Darby’s criteria for good practice were as follows.

  • partisanship (normally about politics, gender issues or pedagogies such as constructivism), the problem in these cases being “not that the author holds these views but that he introduces them as if they are the only possible interpretation of events, without seeking to defend them or subject them to critical scrutiny”. (p39)
  • failure to control or even describe the methods used for sample selection;
  • the failure to guard against subjective conclusions in the collection of qualitative evidence from unstructured interviews;
  • failure to “triangulate” (i.e. to check evidence derived from a single source against other sources of evidence);
  • tendency to quote secondary sources and the works of perceived “great figures” without subjecting those sources to criticism or going back to primary sources, leading to “a game of ‘academic Chinese whispers'”;
  • contradictory arguments or non sequiturs;
  • failure to consider more than one interpretation of the facts or to acknowledge controversy around positions presented as universally acknowledged.

The problem that this research reveals is even worse that the headline figures might suggest, for a number of reasons:

  • if you combine the figures above and ask what proportion of all education research is likely to (a) use qualitative evidence to (b) address issues of pedagogy while (c) conforming to good practice, you will come up with the staggeringly low figure of 0.5% (see this Excel spreadsheet for the calculation)[Comment by Steve];
  • the fact that so much unsatisfactory research was being published in the UK’s top academic journals showed that the procedure for peer review was totally dysfunctional;
  • there was almost no attempt to criticise other research papers, so poor research went unchallenged and there was no process whereby a consensual position was developed by a process of cumulative argument;
  • no examples at all were found of researchers trying to repeat the findings of the small amount of quantitative research that had been done, a matter of some importance as the repeatability of quantitative research is an essential criterion by which the robustness of the evidence can be tested.

In their conclusion, Tooley and  Darby endorsed the earlier comments of Hargreaves, implicitly and in substance if not quite explicitly or in tone:

The picture emerged of researchers largely doing their work in a vacuum, unnoticed and unheeded by anyone else…
The tentative suggestion here is that the individual shortcomings in particular facets of the research articles noted above are, in general, serious enough to raise grave misgivings about the quality of the research surveyed.

Nothing much appears to have changed since 1999. In March 2013, Ben Goldacre, author of the 2008 book Bad Science, was asked by Michael Gove to publish a paper on how to improve the quality of evidence in education. Tom Bennett published Teacher Proof, his castigating review of education research in the same month.

Even the academic researchers are striking a more cautious note regarding the success of their work to date . The recent report from the London Knowledge Lab, Decoding Learning, sounded a cautionary note, even though it was produced by a group of academic researchers who might have been expected to blow the trumpet for education research.

Many research studies have addressed the impact of particular technological innovations, and many meta-analytic reviews have aggregated these findings. Typically, these synthesising reviews do find some evidence of positive impact. However, there are two important complicating factors that limit the strength of the claims that can be made.
Firstly, the evidence is drawn from a huge variety of learning contexts: the wide range of teacher experience and learner ability means that too often the impact identified is relatively modest in scale. Secondly, these findings are invariably drawn from evidence about how technology supports existing teaching and learning practices, rather than transforming those practices. What is clear is that no technology has an impact on learning in its own right; rather, its impact depends upon the way in which it is used.

The conclusion from the London Knowledge Lab is supported by Michael Fullan and Katelyn Donnelly in their recent report, Alive in the Swamp, which summarises the state of current research on ed-tech and pedagogy on page 11 as follows:

If we consider digital innovations, the field is currently characterised by either weak or undeveloped pedagogy, or strong technology and pedagogy confined to a small number of schools; that is, the best examples tend to be small-scale exceptions that are not representative of the main body of schools. Additionally, there is in general a lack of strong efficacy evidence demonstrating the impact of the digital innovations on student learning. Robust academic meta-analysis research, such as that by Steven Higgins et al., shows a current lack of causal links between the use of technology and student attainment.

The issue identified by Alive in the Swamp is that there is a strong dependency between education technology and pedagogy: the first is weak very largely because so is the second.

At a general level, the same point is supported by Chris Wormald, Permanent Under Secretary at the DfE, who told the Commons Select Committee on 23 January:

The evidence base in education – not just in the department but more generally across the UK – is not as good as it should be. It is not as good as the evidence that is available to my colleagues in the Department for Health, for example, or in a variety of other government departments – so there are a lot of questions in education where the current evidence base does not provide you with a completely clear answer.

So subminiature’s assertion that “the research has been done” is startlingly at odds with a decade-old consensus among academics and civil servants.

Summary of our points of disagreement

I have explained at length why I think that subminiature is wrong to suggest that the teacher’s first task is to generate a joy in learning for learning’s sake and the ability to learn to learn. In both cases, these are skills and attitudes that develop through the experience of learning and it is unrealistic and unhelpful to try and teach them outside that context. What is more, the second of these formulas implicitly emphasizes a strategy of independent learning, which is not an effective way of teaching higher order skills.

The assertion that reducing class sizes flies in the face of all the research, which is what subminiature claims is informing his or her position.

The assertion that we have clear research evidence on teaching practice is also untrue: there is a widely held consensus that when it comes to statistically significant, quantitative data about different teaching strategies, the quality and relevance of the great majority of educational research is extremely poor.

Finally, I have rebutted subminiature’s position that teachers know best by listing several theories to which teachers subscribe widely but which quickly fall over on examination.


So far, this article has swung the hammer at a wide variety of targets, set up for me by subminiature. But in supporting the charge made in the title, that teachers do not know best, it is not my intention to pin the blame for the current situation on over 400,000 teachers. If there is a problem at such a large scale, then the cause is clearly systemic. Teachers are rather the victims of the current situation than its cause.

We will all have had experience of some teachers who are very good and others who are pretty poor. What the evidence shows about teachers in general, however, is that they are not familiar with academic research. This is at least partly because most academic research is neither helpful nor reliable. Teaching, as it is practiced today, is very largely a matter of personal craft, and teachers themselves value inspirational relationships more than anything else.

A system which places such a premium on the personality of its staff is a high-stress environment in which to work. This point was understood very well by Kim Taylor, from whose 1971 book I quoted extensively in Education’s coming revolution:

The modern teacher, as he emerges in conferences and in articles, is expected to achieve prodigies of co-ordination, busking his restive audience like a one-man band[4].
Often the proliferation of opportunities is merely an embarrassment, an added strain, a further widening of the gap between what imagination conceives and frailty allows. Period after period, the schoolmaster stands exposed and alone on stage, forced to put on too many performances, the only source of an often unpalatable curriculum, concocting what to do next to keep thirty or more adolescents busy[5].

Not only is the model of teaching as personal craft stressful for teachers—it does not deliver the goods either. Our current education service achieves remarkably inconsistent outcomes, to an extent that would not be tolerated from a commercial business or even from public service like the NHS with which the adult electorate had regular, direct contact. And there is an increasing body of evidence to show that, in absolute terms, the British state education service performs badly when compared with the education services of other countries and sectors.

There are many and complex reasons for this under-performance, of course—but one of the most critical is the lack of a commonly understood body of knowledge about what constitutes good teaching; a body of knowledge that can be applied, like a surgical procedure, independently of the personality of the practitioner applying it.

The reason why teachers don’t have this sort of knowledge is not because they are lazy or stupid: it is because this sort of expert, professional knowledge simply does not exist outside the repertoires of the best and most experienced teachers. We do not have a clear, evidence-based understanding of pedagogy. Without such a body of knowledge, the achievement of consistent outcomes will always be elusive; the attempt to replicate good practice from one classroom to another will always be uncertain, and any claim to professional status will only ever amount to empty posturing.

Even though we do not have that knowledge, that clearly defined pedagogy, I would contend that we do have some idea of what it will look like. I have quoted with approval from Professors Dylan Wiliam and Eric Mazur. I do not contend that they have produced completely convincing evidence that does more than (in the case of Professor Wiliam at least) suggest a useful direction of travel. This direction of travel, I suggest, is likely to have two central pillars: formative assessment and progression management.

As I suggest that we might see the basic outline of this pedagogy, even if it is only through a glass darkly, we also understand why such a pedagogy cannot be consistently applied under present conditions and why we have never managed to develop a clearer, evidence-based view of what it really looks like in practice. The fact is that in current circumstances, good pedagogy is simply too difficult.

If you have not yet done so, please listen to the parts of Professor Mazur’s talk where:

On the surface, the example that he gives of this particular type of formative assessment (or “learning activity”, as I would prefer to call it) is simple. It can also be used repeatedly, in any subject and for any topic. But if it seems to you an easy matter to implement such a technique, think again. You need some infrastructure, of course: student response systems, appropriate software and wi-fi for communications; you may also want some tracking software to record the answers given by individual students—but none of that should be too difficult. What is really clever is the question that Professor Mazur asks. It is a concrete application of the abstract principle that Mazur is teaching, whose difficulty is carefully calculated to ensure that a little more than half the class gets it wrong (in fact, he miscalculates with his audience of teachers, but more of that another time). It is also a question that will motivate you to discover the correct answer and which, when you to get it for the first time, will give you a real “Ah ha!” insight into the principle being taught. That is a difficult question to devise. And if “peer mentoring” is a technique which you are going to use regularly in your class, you will need hundreds of such questions, and ways of selecting which questions are just right, not just for your current topic but also for your current class.

So good formative assessments, or productive learning activities, are hard to design. And that is only the first half—and much the easier half—of good pedagogy. The really difficult part will be adaptive progression management. You need to be able to profile the particular cognitive abilities of each individual student and route them to the particular learning activity that will best address their particular individual requirement, grouping them as with other students appropriately for each activity. That’s like managing the air traffic control at Heathrow airport on a busy day, single handed. Good pedagogy is simply out of reach for any mortal human teacher. And so mortal human teachers fall back on a sort of fire-and-forget strategy that we are used to: try to motivate the student, get them on your side, give them some information, and hope that they make sense of it themselves. And it is that same fire-and-forget strategy that people have attempted to transfer to the digital world by stressing independent online learning. What was originally borne out of necessity has been elevated into a virtue.

So teachers make the best of the difficult job that they have been given. Some of them manage fairly well and some manage less well—but there is no shame in the fact that none of them are ever really on top of the job. To re-quote another of my favourite passages from Kim Taylor’s 1970 Resources for Learning, first used at the end of my Education’s coming revolution:

The present gap between research and daily application is such that teachers generally turn for help to those in the same boat with them, all awash in a vast sea…and we suspect that the academics, secure from the daily fret of wind and wave, have forgotten what it is like to feel a little seasick all the time.

Teachers are expert “copers”—but that experience does not give them expertise in solving the systemic challenges that they face. That would be like expecting a collection of horse-drawn carriage drivers to design the first steam engine. And it is a similar kind of paradigm-shifting technological advance that is now required, if we are to address the chronic under-performance of our education system.

While it is beyond the capacity of any mortal human teacher to deliver personalised, adaptive, activity-based teaching programmes, such a task is not beyond the capacity of digital software systems. Software runs business accounting systems, supermarket logistics systems, military defense systems, even the air traffic control system at Heathrow on a busy day. But no-one has designed software systems that deliver personalised, adaptive, activity-based learning programmes because no-one yet knows what such a pedagogy looks like. Without the software systems that would make such a feat possible, no-one has ever been able to implement such an approach.

In his book Teacher Proof, already cited, Tom Bennett lists “Techno, techno, techno, TECHNO: digital natives in flipped classrooms” as one of the examples of barmy pseudo-science with which teachers have been too commonly duped. As practised to date, I agree with him. But that is not to say that technology does not have a huge potential to improve education, even if its implementation has not been successful so far. If Tom Bennett had been writing his book in 1900, he might well have listed manned flight as one of the barmy ideas of pseudo-science. Empirical evidence is never very helpful when it comes to innovation.

The skills that teachers have developed to allow them to cope in current circumstances are not the skills that will allow them to design the education technology systems of tomorrow. And yet government continues to look to teachers and educationalists to lead innovation in this area—and teachers and educationalists and their representatives still guard jealously the right to be regarded as the sole experts in a field in which they have no expertise at all. The result is exactly the sort of pseudo-science that Tom Bennett attacks, led by the charasmatic gurus who preach, who never doubt the truth of their creed, who never respond to critical questions, and who never admit the almost total lack of evidence to back up their theories. As far as education technology and pedagogy is concerned, we live still in the age of alchemists and leech doctors, not the age of rational science.

In my next post, I will look at the important question raised by Tom Bennett’s Teacher Proof as to whether a technological, scientific approach to education is even possible. Having (spoiler!) concluded that it is, I will deal in more detail with some of the pedagogical issues, such as assessment, that I have touched on in this article; I will try and explain what I suspect that good pedagogy will look like and I will look at the relationship, touched on by Alive in the Swamp, between pedagogy and education technology. And in case that some readers may think that it is preposterous that a single blogger like me should dare to criticise the collective wisdom of 400,000 teachers, then I shall clarify the widely misunderstood dictum about the “wisdom of the crowd” which is so often by the lemmings to justify their lack of independent thought.

I hope that by Christmas I will have made sufficient progress through these preliminary topics to address the key question raised by this article: given that we are not condemned to shun counter-consensual innovation in deference to the supposed wisdom of the crowd, and given that a technological approach to education is both possible and conceivable in outline, how in practical policy terms can the current impasse in ed-tech and pedagogical innovation be broken?


To link to a section in this article, right click on the appropriate section and choose “Copy link address”.

1. My conversation with subminiature
1.1. Stephen Fry and the French
1.2. The points at issue

2. Our points of disagreement
2.1. Learning for learning’s sake
2.2. Learning to learn
….2.2.1. The paradox of learning
….2.2.2. The argument against dependency
….2.2.3. Independent learning in practice
2.3. That it is a priority to reduce class sizes
2.4. That teachers know best
….2.4.1. The primacy of personality in the classroom
….2.4.2. That assessment distracts from teaching
….2.4.3. The argument against knowledge
….2.4.4. That teachers should be facilitators
….2.4.5. The importance of learning styles
….2.4.6. Variability of outcomes
2.5. That the research has been done
2.6. Summary of our points of disagreement

3. Conclusion


[1] Quoted in Resources for Learning, by L C Taylor, Penguin Education, 1971, p. 22. L C Taylor takes the quotation from Anthony Chevenix-Trench, the Public Schools, in Peter Blander (ed.), Looking Forward to the Seventies. Smythe, 1968, p. 76.

[2] Mill, J.S., On Liberty, 1860, http://www.constitution.org/jsm/liberty.htm.

[3] Teaching as a Design Science, Diana Laurillard, Routledge, 2012, p. 4.

[4] L C Taylor, Resouces for Learning, Penguin Education, Hammondsworth, 1971, p.178.

[5] L C Taylor, op. cit., p.237.
[6] I added the word (necessarily) into the title in response to a Tweet suggestion by Dylan Wiliam.

21 thoughts on “Why teachers don’t know best

  1. An excellent, thoughtful post. Especially nice to see Aristotle’s Nicomachean Ethics used in educational thought. If more people understood that teaching is more phronesis than techne or episteme, then I think we would understand better how to improve education. [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3624-a ]

    Some specific points

    Learning how to learn
    Paul Black and I spent four years of our lives working on a research project called “Learning how to learn” and in the end started calling it “the swamp”—whenever we tried to define what learning how to learn might actually be, it blew up in our faces (or more precisely fell to pieces). I have put the boot into “21st century skills” in a forthcoming pamphlet entitled “Curriculum Redesign” to be published by the SSAT in the autumn term. [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3624-b ]

    Success in higher education
    Anna Vignoles’ research shows that the differences between students from state and private sectors attending highly selective universities are almost entirely explained by A-level grades.
    That said, the PISA evidence show clearly that the quality of teaching in the independent sectors is actually lower than the state sector—controlling for social class, private schools achieve in classes of 13 what state schools achieve in classes of 25. [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3624-c ]

    Benefits of testing
    For those who want to know more about the benefits of testing, you could look at:
    Willingham, D. T. (2009). Why don’t students like school: A cognitive scientist answers questions about how the mind works and what it means for your classroom. San Francisco, CA: Jossey-Bass.
    or my own summary of the evidence of the benefits of high-stakes testing:
    Wiliam, D. (2010). Standardized testing and school accountability. Educational Psychologist, 45(2), 107-122. [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3624-d ]

    • Dylan,

      Many thanks for your kind comments. I shall certainly look at the articles on testing that you cite – certainly before I write my post on assessment – and also the PISA evidence. Is that summarised in one of the PISA pamphlets?

      The comparison between quality of teaching in private and state schools sounds like an interesting challenge to me. I guess these conclusions very much depend on how reliably you can control for social class. I hear it said sometimes that the upper classes are not – necessarily – great intellectuals!

      Thanks again, and also for the interesting research-update service you provide on Twitter.


    • With respect to phronesis, techne and episteme, this sounds like another blog post…But let me explore a little in a comment first. I am not sure I would put all my chips on phronesis – but I think teasing apart these different accounts of virtue would be a very useful exercise.

      As I remember, Aristotle defines these terms along the following lines, in order of ascending importance:

      * techne: technical skill, the basis of expertise;
      * phronesis: prudence: the attitudes and understanding that aims at practical benefits (which might – I can’t remember – include the sort of worldly cleverness shown by bank robbers);
      * sophia: wisdom about the eternal verities, implying virtue, which includes episteme, i.e. knowledge. This references Socrates’ counter-intuitive (but I think rather good) theory that wickedness boils down to ignorance of what virtue is.

      So I think the intrinsic delight in truth – what I call “contemplation” in the blog, falls into the “sophia” box and phronesis is the sort of practical virtue, also often called wisdom (incorrectly, according to Aristotle), that might allow teachers to react with understanding to the sort of unexpected contingencies that are always being thrown up in classrooms and which allows you to “do the right thing”, in the current political jargon. Techne might tend towards “procedural knowledge” – a bit more like painting by numbers.

      Aristotle’s taxonomy raises an interesting question in my mind about the rather lowly position of “understanding” in Bloom’s taxonomy. Aristotle puts deep knowledge at the top of the tree, with the practical, prudential benefits of that knowledge (application, evaluation, creation) lower down.

      I think that the “capital T Teacher” needs all three of these – but my argument, rather against your position, Dylan. More phronesis would be good – but I wonder whether it is a realistic aim. I am suggesting that we need to do rather more to codify the procedural knowledge that will allow beginning and imperfect teachers to do an ordinary job of work. “Teacher” (capital T) is how people addressed Jesus when they were asking for some profound insight into the human condition – and we can’t all attain that depth of insight, certainly during the first twenty years of our careers. As I argue in “Education’s coming revolution” at https://edtechnow.net/2013/01/26/industrial-revolution/, I think the chronic problem facing education lies in the under-supply of teachers in what is still run as a craft-based service. Crispin.

  2. Reblogged this on Tick Tock Maths and commented:
    This is a really interesting post. I’m not sure I agree with all of it. In particular [blockquote]”even at the end of the course, it is not for teachers, as suppliers of a service, to pass judgement on the motivation of their customers, any more than it is for a plumber to pass judgement on whether their customer is right to order a Jacuzzi for his bathroom”[/blockquote] is horrible. I hate the idea of students as customers as it implies a very one directional relationship. I strongly believe that teaching is more than filling student’s heads with knowledge. It’s a collaborative process on the part of both the student and teacher. [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3625 ]

    It’s a really interesting read, however, and I don’t want to add too many of my own thoughts as I’m sure you’ll have your own reaction reading it.

    • Hi Richard, Thanks for raising the “customer” issue, which I acknolwedge is a thorny one!

      In your comments about directionality, I think you are conflating the transmission of factual knowledge with the provision of a service. I think the latter *is* one directional, though the style of the service is very conversational. A psychoanalyst provides a service, for example, although it is the patient that does 99% of the talking. Directionality in the first sense does not imply it in the second.

      The “customer” analogy is a complex one, I agree – and I admit to being a little sloppy in my use of the word in this passage. I think of the “customer” as the student in 20 years time, or the idealised, rational student. In practice, this mythical figure could be represented by a collection of people, including the student and the student’s parents, when engaged in rational deliberation, and the agents of government, who represent the taxpayer who is paying. I stand by the main point, though, which is that however difficult it is to define the customer – who exactly is sitting on the other side of the table – it is not the teacher, who is sitting on *this* side of the table. The teacher is still a supplier and is therefore not the person to define what the ultimate aims of the service should be. That would be to make the teacher judge and jury – or perhaps high priest – which I also find a horrible idea, though one that many teachers seem to subscribe to. Crispin.

      • I agree with your last point there. Also: I quickly edited myself when I realised that later on you talk about teaching being that two way relationship.

        Some very good points raised by you here, and I think it’s something that should be required reading, even as just a jumping off point for people’s own thoughts

        • Also: I find comments about teacher quality interesting. As I do the idea of a ‘four fold’ increase in learning. I’m not sure how this was measured, measuring educational outcomes is VERY tricky, but as a teacher I find two things:

          1) There is a difference between embedding long lasting knowledge and coaching well for tests.
          2) On some days I’m a much better teacher than others. My ‘teacher quality’ is in no way a static stat. I am a better teacher now than I was last year. I have been better or worse depending on the culture of the school I am in.

          • Hi Richard, I have not looked into the detail of Dylan’s research either. I think it would be worth doing, though I included the conclusions as I respect him as academic, my time is limited, and the finding corresponds with my own experience, checked against other anecdotal evidence. Tooley would not have been impressed, I grant you.

            The difficulty of measuring outcomes is a key point, I agree, and I will address it in my next post.

            Variability of quality of a teacher’s work from day to day is important too, I agree, and I think this re-enforces the fundamental point I am making that education should do more to address the question of consistency and this means placing less emphasis on the teacher in the front line firing on all cylinders, day in, day out. I remember an INSET day when I was an NQT in north London in which the question of discipline was raised and the answer from the platform was, “you’ve just got to interest the pants off them”. Laying aside the appropriateness of the metaphor, I remember thinking, “gee, no stress then!”

            I think Dylan Wiliam would be the first to agree that teachers improve with experience – and that is why the wastage rate from the profession is so damaging and why the effect of schemes like Teach First will only ever scratch the surface of the problem. This fundamental problem of teacher supply forms the basis of my argument, which I made in “Education’s coming revolution” at https://edtechnow.net/2013/01/26/industrial-revolution/, based heavily on the argument made by Kim Taylor, Director of Resources for Learning at the Nuffield Institute in the 1970s.


            • Thanks for the reply, Crispin. You make some excellent points here.

              ‘ I think this re-enforces the fundamental point I am making that education should do more to address the question of consistency and this means placing less emphasis on the teacher in the front line firing on all cylinders, day in, day out’
              Is exactly right, I think.

  3. Pingback: Al wat lectuur: genezend mp3′tje, high tech spieken tegengaan, hoe leren we? | X, Y of Einstein?

  4. Really interesting article with lots of cause for reflection and thought.

    One point worth checking is the figure of 0.5% you calculate in relation to education research – I think you’ve assumed the 3 factors you look at are independent (and can thus be multiplied) whereas I suspect they are not.

    [ Refers to text at https://edtechnow.net/2013/08/27/blind/#_cref-3636 ]

    • Hi Steve,

      Thanks for picking this up. I admit that I somewhat glossed over this issue, though I did nod my head at it by including the words “is likely to…” in the sentence that makes the assertion.

      In short, if you have three probabilities, A (10%), B (15%) and C (36%), then it is correct to say that the probability of A *and* B *and* C will be resolved by multiplication. In practice, it is a little more complicated because, as you say, those articles that were about teaching and learning might have been more or less likely than the others to use quantitative research, and more or less likely than the others to conform to good practice. These possibilities could be allowed for if the data had originally been given in tabulated form: e.g. showing how each article rated against these three variables – but as this would have split the original sub-sample of 41 articles into 8 piles and the numbers would have started to get pretty low.

      The basis point is, I think, still valid – that we should really be interested in articles that meet all three criteria and this is going to be a very small number indeed. In theory, it could have been possible that *all* the articles on teaching and learning used quantitative research *and* conformed to good practice. However, this is not only statistically very unlikely, but the text of the Tooley report makes clear that it is not the case (by giving examples of pedagogical partisanship, particularly in researcher’s unexamined advocacy of constructivism and by noting that this partisanship was expressed by following non-empirical – i.e. non-quantitative – research techniques, and by giving a number of common ways in which qualitative research was commonly flawed).

      So let us say that the dependency between these three criteria might involve a variation of 2 – i.e. articles on teaching and learning were up to *twice* (or half) as likely to conform to good practice and up to *twice* (or half) as likely to be based on quantitative methods than the others. The correct figure would then lie somewhere in the range of 0.125% to 2% (0.5*0.5*0.5 or 0.5*2*2) and this means that the number is still so small that it really makes no difference: for all the additional reasons that I list, we are still looking suggests a complete and systemic failure of academic research to address issues of pedagogy.

      Another further consideration that I forget to mention in the main article is that this is not a figure for articles that are good, useful or valid – these are questions which are too heavily based on subjective judgement for the research to consider properly). It just a figure for the proportion of articles that make it through the pre-qualification round, as it were – the articles in respect of which it is even worth asking those questions.


      • Crispin,

        Thanks for taking the time to reply to my comment.

        It is key that the 3 events are independent if you wish to multiply their probabilities (which they are almost certainly not in this case) but, as you state, even if we had done an article by article analysis of all research papers the figure would still be far too low.

  5. Pingback: Diigo links 08/28/2013 | DrAlb

  6. Crispin, Your choice of Professor Mazur’s talk was a brilliant choice to make your point on the importance of activity based learning. The methodology of a flipped classroom and peer instruction was a bonus. My own definition for education is to teach people to think. The how of that was exemplified by Mazur’s “test”. Great article and I am looking forward to the next installment.
    You may recall the Knewton article I sent you on Inferred Student Data which is about your point of “you will need hundreds of such questions, and ways of selecting which questions are just right, not just for your current topic but also for your current class.” Any “software systems that delivers personalized, adaptive, activity-based learning programmes” will require I think those questions first, if you believe the Knewton’s approach using algorithmic assessment of inferred student data is on the right path to an adaptive learning system.
    5) Inferred Student Data: Exactly what concepts does a student know, at exactly what percentile of proficiency? Was an incorrect answer due to a lack of proficiency, or forgetfulness, or distraction, or a poorly worded question, or something else altogether? What is the probability that a student will pass next week’s quiz, and what can she do right this moment to increase it?

    Inferred student data are the most difficult type of data to generate — and the kind Knewton is focused on producing at scale. Doing so requires low-cost algorithmic assessment norming at scale. Without normed items, you don’t have inferred student data; you only have crude guesswork at best. You also need sophisticated database architecture and tagging infrastructure, complex taxonomic systems, and groundbreaking machine learning algorithms. To build it, you need teams of teachers, course designers, technologists, and data scientists. Then you need a lot of content and an even bigger number of engaged students and instructors interacting with that content. No one would build this system to get inferred student data for just one application — it would be much too expensive. Knewton can only accomplish it by amortizing, over every app our platform supports, the cost of creating these capabilities. To our knowledge, we’re the only ones out there doing it.

    Educators are sometimes skeptical of adaptive apps because almost all of them go straight from gathering user interaction data to making recommendations, using simple rules engines with no inferred content data or inferred student data. (It is precisely because we envisioned a world in which everyone would try to build these apps that we created Knewton — so that app makers could all build them on top of low cost, yet highly accurate inferred content data and inferred student data.)

    Thanks frank

    • Hi Frank,

      Thanks very much for this comment. It is a bit like having a keen pupil at the back of the class who is two years ahead of what you are trying to explain in this morning’s lesson… Yes, you jump to exactly where I am going with all of this.

      The problem is that many teachers, academics and officials in the UK are likely to recoil in horror at all this talk of “norming-at-scale”. On the basis of some post-modernist clap-trap, many teachers doubt the possibility of making any claims to objective truth at all, let alone representing characteristics displayed by flesh-and-blood human beings in numeric form – so there is a visceral, cultural reaction against this sort of thing. But as you (and indeed Socrates) point out, the proof is in the pudding: the truth of a proposition is shown by its predictive reliability. I predict that these technical approaches are exactly how we will in due course be able to make reliable predictions about future student performance, based on the inferences that we make about their current capabilities from their past performance.

      Mazur shows the power of the atomic learning activity. As I suggest in my conclusion, learning analytics that allow us to profile students’ competencies and react by managing adaptive progression is how those atomic learning activities will be strung together. And just as words have (often ambivalent) meanings, what really count are the sentences.

      My only reservation about Knewton is that I do not think that any of this can be done in a single, closed system. Innovation at the instructional level will be seriously impeded if we allow proprietorial silos to become established at the platform level. Added to which, the ability to “norm-at-scale” means that it is essential that the classroom management layer (or what Dylan calls in his talk the “aggregation” layer) is decoupled from the instructional layer so that we can concentrate all relevant data, from wherever it originates, in one place. Which is why open data standards are absolutely essential. But if I am going to keep my day-job and make this argument thoroughly at the same time, I think it is going to take me the best part of a year to get to the point where the reasoning can be explained to those readers who do not already understand what we are talking about.

      BTW – I also have a half-finished and pretty elaborate blog post to finish, which I started in the Spring in respond to your challenge regarding a use case to illustrate my “learning activity platform” concept. I will get there soon!

      Thanks again for your interesting comments.


  7. Long read!

    Especially the comments.

    I do find the attack, on teachers not knowing the research, followed by somewhere between 98-99.75% of the research being junk, weakened the attack on teachers not knowing the research.

    Personally, the fact that since schools started teaching IT we have had a significant decrease in people wanting to study IT and a massive drop in the quality of applicants is all the evidence that I need that schools can not teach, also the course work was dire.

    I would ideally like to ban schools from teaching any IT until such time that there was established scientific evidence that they could teach IT. (Given the poor quality of the research it might take a very long time before schools provided the evidence and were allowed to teach IT again. Yay!) However, this is just self interest, getting schools to teach all core subjects well would be of greater benefit to humanity.

    Most modern kids take to IT like ducks to water and are much much more competent than their teachers. It would be more productive to get the kids to teach the teachers, and some schools have sent their kids out to teach IT to adults, it was helpful to both sides. The kids were amazed at how useless the adults were and learnt tact and patience. The adults were impressed at the kids knowledge and helpfulness.

    My suspicion is that with a limited number of quality teachers, the focus needs to be on core subjects, and those that the teachers actually understand.

    For those pupils that are tortured by accademic learning they need to be taught as much of the basics as possible with the minimum amount of pain all round and escape to a vocational carrer.

    • Hi Ronald,

      Thanks for the struggling though the post, and for the comment.

      I take your first point. If there is a criticism there, it is for spinning the piece as an attack on teachers which, as you say, it is not. Not only because most of the research is not worth reading but because the research which I believe to be pointing in the right direction, suggests a strategy (adaptive, dialogic, personalised tuition) which I doubt is achievable at scale in the real world. I will have more to say on research in my next post but one (which is taking a long time to write) on the Research Ed 2013 conference.

      So my attack is not so much on teachers themselves as on the teacher-led approach we have taken towards ICT under Becta.

      I agree with your point on core subjects. I remember going to a teachers conference at Cambridge: the Computer Studies man said do Maths at A level (don’t touch Computer Studies), the lawyer said Latin, the Economist also said Maths, and any journalist would say English not Media Studies. But maybe that is taking a university/academic perspective, as you also mention. And the idea of escaping to a vocational career – while not dropping the basics – at adolescence, seems to be the in the spirit of the times.

      The shortage of properly qualified teachers strikes me to lie at the root of all our problems. And my two, interconnected, answers to that are (1) automate what can be automated, balancing budgets away from staffing towards technology, (2) use the same technology to create interlocking professional structures, in which the autonomy of junior teachers is limited by the supervision of senior ones. All of this requires implementations of technology that disrupt the current model of teaching by autonomous craftsman, which is popular with teachers.

      When you have recovered from this read, I would be interested in your thoughts on “Education’s coming revolution” (https://edtechnow.net/2013/01/26/industrial-revolution/), which is based on a previous Director of Nuffield Resources for Learning project.

      Thanks again for the comment,

  8. There is one serious gap in EDRESEARCH I have tried to fill it BUT U R so on the button THE BLIND LEADING THE BLIND only good research can identify correct procedure
    PLEASE try mine 1st stop http://www.system-one-4-every-1.co.uk/
    I HAVE over 700 refs on research proofs



    we can provide good education 4 every child when we teach parents to teach their own
    children to count and read

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s