What works is a lot better than what doesn’t

photo-main

Teachers often talk about the vital nature of their work and the fact that for the young people we teach there are no second chances. I’ve heard teaching compared to air traffic control and the risks in the classroom compared to the risk involved in miscalculating the landing of a plane. These kinds of comparison are made to alert us to the importance of what we do, more clearly they’re over dramatic and, in a very real way, untrue. I don’t want to make out that what we do is unimportant but if we teach algebra badly no one dies. But what if it were true? What if education really was life or death? What might we do differently?

In their important new book, Make It Stick: The Science of Successful Learning, Brown, Roediger and McDaniel point out that almost everything teachers and students believe about learning is untrue and based on unsubstantiated theories, folklore and intuition. We do what we’ve always done. We do what everyone else does. We do what sounds right. And we do it based on very little in the way of empirical evidence. But what if we had a gun to our heads? What if we were staking our lives and the lives of our students on the efficacy of  choices we make? Would that change our approach or shake our faith? Would we be willing to bet our lives that the strategies and techniques we advocated and practised were the best ones?

There’s very little I’d want to bet my life on, and nothing that I would place such faith on that wasn’t based on the most impeccable of evidence. I might be in a minority. Clearly faith plays an important role in the lives of millions of people around the world. Maybe that’s you. Maybe you are so sure you’re right that you’d take that bet. If so, I envy you your certainty.

But I would also worry. I’d worry because I know what it feels to believe I’m right and I experienced the shock of being presented with irrefutable evidence that I’m wrong. More than once. And it always blindsides me. Take this example: I was due to fly home from Japan at 2 pm. I get anxious about being late and wanted to make sure I’d arrive at the airport in good time so I got up at 8.30, showered and ate a leisurely breakfast before packing up and wondering about what to do in the intervening hours. Knowing how pensive I get while waiting I decided it would be best to get underway and get the train to the airport. I dragged my backs to the station and trundled off kicking myself for being such a poor traveller. Instead of spending my last hours doing something fun I’d be hanging around in Osaka airport terminal for a few extra hours. When on the train, I dug out my ticket to double-check which terminal I was leaving from and was staggered to discover that my flight was due to depart not at 2pm but at 11am! It was now 10.30. I can’t tell you how sick I felt; I broke out in a cold sweat and sat there staring in blank incomprehension. How could I have been so wrong? I had known, been so certain I hadn’t even felt the need to check. I might not have bet my life on the fact I knew the time of my flight, but I’d certainly have bet several hundred pounds on it! And I’d have lost. In fact that’s exactly what I had to pay to book a new flight the next day.

The point of that sad story is that we often place unwarranted faith in what we believe. We can be so certain that we feel no need to check; we just know we’re right. But to err is human and we are all wrong all the time about things both trivial and fundamental. consider this little gem that reemerged on Twitter today:

BmQzlmaIEAA2Z9o

It seems so obviously true, doesn’t it? partly because it bears out and validates our experience as teachers: we get to know our subjects so much better because we teach them, so it follows that the best way to retain new information is to teach it to someone else. And look: there’s some statistics so it must be true! Well, unfortunately not. Darren Kuronatwa kindly got in touch to send me Multimodal Learning Through Media:  What the Research Says which dismisses the claim in short order:

If most educators stopped to consider the percentages, they would ask serious questions about the citation. They would inquire about the suspicious rounding of the percentages to multiples of ten, and the unlikelihood that learners would remember 90 percent of anything, regardless of the learning approach.

But it’s not just the statistics that are dodgy, it’s the dubious idea that a one-size-fits-all, magic bullet approach to learning will always work in every context for every student. The paper goes on to demonstrate that contrary to what we may believe. They conclude that the truth is, unsurprisingly a little more complex:

The percentages related to the cone of learning were a simplistic attempt to explain very complex phenomenon. The reality is that the most effective designs for learning adapt to include a variety of media, combinations of modalities, levels of interactivity, learner characteristics, and pedagogy based on a complex set of circumstances.

In general, multimodal learning has been shown to be more effective than traditional, unimodal learning. Adding visuals to verbal (text and/or auditory) learning can result in significant gains in basic and higher-order learning. The meta-analytic findings in this report provide insights into when interactivity augments multimodal learning of moderately to complex topics, and when it is advantageous for students to work individually when learning or building automaticity with basic skills

Maybe it might be true that teaching others could be an effective way to learn, but what’s the evidence beyond our intuition? This kind of belief is an emotional one and, that being the case is like pretty much everything else that gets believed. But that doesn’t excuse people making stuff up to support what they believe.

Screen Shot 2014-05-03 at 00.10.50

So if we can’t rely on theory, lore and intuition, what can we rely on? There are some compelling voices in education that are happy to tell us that research in education is a chimera. This popular tract by Gert Biesta claims that “what works” won’t work. Biesta is against using research to inform education policy because we don’t know everything, what we know might not be quite right every time,  and because it can be misapplied. Instead he wants us to embrace a ‘value based’ model based on what we as educators believe to be in the best interests of our children. Fortunately for Biesta, no one is holding a gun to his head and no one is likely to die if he’s mistaken. But this kind of attitude is, I think, much more likely to be misunderstood and misapplied than giving due consideration to admittedly incomplete picture that research might give us. It’s likely to result in people tweeting the ‘cone of learning’ as if it’s a fact and, metaphorically, lots of people missing flights.

Brown, Roediger and McDaniel are at pains to point out that what cognitive psychology tells about what works is counter-intuitive, and therefore easy for us to ignore. But inconveniently it’s based on well designed, consistent and repeatable trials. In other words, it ‘works’. These are just a few of the ideas they present which, practising what they preach, are spaced and interleaved throughout the book:

  • We’re poor judges of how we learn best
  • Rereading and massed practice are the most popular but among the least effective ways to learn
  • Spacing and interleaving feel unproductive but are a much more effective way to learn
  • Allowing yourself to forget before attempting to retrieve will boost your ability to ‘store’ information
  • Making it harder to learn  is more effective than making it easy
  • Every time we learn something new, we change the architecture of our brains

If you’ve been following my blog over the last year or so, these ideas won’t be new to you, but this book presents an evidence base which confirms and goes beyond the faith I’ve placed in cognitive scientists like Willingham and Bjork (Both of whom endorse this book incidentally.)

So Biesta may well be right: we certainly don’t know everything and what we do know will inevitably be ballsed-up by enthusiastic, well-intentioned folk like me. But the irrefutable evidence that we are so often wrong should at least give us reason to question our intuition. If the cognitive scientists are right, we could make a profound difference to how well our students learn. If tall their empirical evidence turns out to be wrong, no one’s died. It may not be worth betting your life on, but it outweighs the risk of going with a hunch.

Related posts

Deliberately difficult – why it’s better to make learning harder
Everything we’ve been told about teaching is wrong and what to do about it
The Cult of Outstanding: the problem with ‘outstanding’ lessons

 

19 Responses to What works is a lot better than what doesn’t

  1. […] Teachers often talk about the vital nature of their work and the fact that for the young people we teach there are no second chances. I’ve heard teaching compared to air traffic control and the risks in the classroom compared to the risk involved in miscalculating the landing of a plane. These kinds of comparison  […]

  2. ijstock says:

    Thought-provoking as always – thanks.

    But there are bigger problems with evidence – namely evidence for *what*?

    You can’t identify what evidence suggests works if you can’t define what ‘works’ is. Even Hattie was forced to concede that he was looking at attainment rather than learning. I certainly don’t think that learning is the same as mere retention (it must surely also include evaluation and application, perhaps even modification) – and even if it were, what do you make of the person who drags something that they misunderstood at school out of their mind thirty years later and finds that it suddenly all makes sense (that would be me and Economics, then…) Was that ‘learning’ effective or not?

    Until we can define ‘learn’ all research that claims to tell us what works is fundamentally flawed – and I don’t think we’ll ever get to that point because there is no single such thing as learning.

    Therefore we are left with doing what seems to be working at the time. Since we can never know what use it will ever be, there’s no point in judging it on anything other than its own intrinsic value – which if essentially a value judgment. The OFT recently described education as a ‘post consumption good’ – which kind of makes the point – you only know what use it was *after* you got it (if then).

    And when it comes to ‘evidence’, I would agree that being seduced by glitzy pseudo-aphorisms and using them to teach is unadvisable (and who do that most of all, but those who would tell us how to teach…?) – but I would argue that thirty years of watching what seems to work with the specific young people in front of one makes a not such a bad sample size when it comes to evidence for one’s decisions.

    • ChrisN says:

      Until we can define ‘health’ all research that claims to tell us what works is fundamentally flawed – and I don’t think we’ll ever get to that point because there is no single such thing as health.

      Does that sentence make sense to you?

      Surely the point of research in both health and education is not so much to discover global principles that can be applied universally (although some may exist), but to find the best ways of dealing with specific, different sub-issues. No one who understands research doubts that research findings will have varying applicability, depending on the content being learned, the age of the learner, etc.

      “I would argue that thirty years of watching what seems to work with the specific young people in front of one makes a not such a bad sample size when it comes to evidence for one’s decisions.”.

      This is true when it comes to techniques for which the teacher receives immediate feedback . However, the longer-term effectiveness of techniques is much more difficult for a classroom teacher to evaluate. (Thinking Fast and Slow by Kahneman explains this well.) Teachers rarely teach children for longer than one year. Yet some types of learning (for example, learning to read), extend over a much longer period than a year. There is a well-known phenomenon called “third grade slump” which affects children taught to read using whole language and whole word methods. Yet many of their earlier teachers remain convinced that these methods work well – because they have never had the opportunity to get the long term feedback.

      A further problem is the serious, and often ignored effect of confirmation bias. And finally, while what one observes may indeed ‘work’; how is one to know that there is not some other method(s) that it has never even occurred to one to try, that will work even better, for more students, more often?

      • ijstock says:

        You may be right – but very often research presents itself as in search of – or even in possession of – universals.

        What you say about short and long term learning may also be true if you are concerned with very specific skills such as learning. I don’t actually think this is a good example though, because there are so many factors that determine how well a person is ultimately able to read – and not all of them are connected to formal teaching. We need to get over the professional hubris that considers the teacher to be the sole educational influence on children.

        But if you are concerned with the overall unified effect of education on some one’s life, it is almost impossible to say ‘what worked’ because the matter is just so complex, and depends entirely on what you consider the ‘objective’ of the exercise was in the first place. Hence my example about learning that I assumed had failed and for which I had no use – only to find much later on that neither was in fact the case.

        Confirmation bias is indeed a problem – against which the only defence is an awareness of it – and even that isn’t a complete vaccine by any means. But we all can ultimately only know what we know, and in the real classroom that’s all we have to go on. And indeed all we need to go on. Given that no one knows the future, we can only do what seems right at the time.

        In that context, I find a lot of educational research to be a lot less use than working knowledge derived on the job. Otherwise, you end up paralysed by the constant effort of trying to second-guess the future.

  3. […] Teachers often talk about the vital nature of their work and the fact that for the young people we teach there are no second chances.  […]

  4. Nigel W says:

    Thanks David…

    “I don’t know what I should do, I only know what I can do”
    (James T. Kirk Star Trek: Into Darkness)

    Teaching is an art. There are skills to be learned, the artistry is identifying when to apply each skill. Whilst all children share a common physiology, their similarities are far outweighed by their differences.

    Whilst I genuinely enjoy discussing and exploring teaching and learning, I can’t help wonder if all the effort going into trying to prove or disprove this theory or has much genuine value. It would if children were the same. But they’re not. The Year 11 class I taught this year are so completely different to the one previous. Their experiences, their aspirations and so on. To try to suggest that “They will all learn best by this method” just seems ridiculous. To suggest that one could apply a retention rate to it, even more so.

    I see it like this. It is my duty to ensure I am as familiar, as practicality will allow, with as diverse range of techniques and strategies as possible. To understand their advantages and disadvantages and then to apply them to each case in order to facilitate the most effective outcome.*

    The kid in front of me does not care what’s at the top of the triangle, and perhaps more importantly, what’s at the top of their triangle is quite possibly different to person who is sat next to them.

    Is it possible that people looking for the “ultimate” method of teaching are those who simply like a world that is either right or wrong? Well it’s not like that is it?

    Lessons cannot be right or wrong, we are dealing with organisms with complexities beyond the classroom. A whole lifetime of experiences will affect the way in which each student responds to stimuli.

    One thing we can say for certain, those wishing to have the spotlight shone on them through trying to prove this theory or that theory will never run out of opportunities for doing so.

    We can’t prove that this method is better that that method. All we can prove is that this method worked well for that particular person on that particular day.

    So, trying to prove this that or the other… is there any point?

    I would say…Know the techniques, know the advantages or disadvantages, know the student.

    *Outcome needs to be agreed by all parties, because Grades are not the only fruit.

  5. […] Teachers often talk about the vital nature of their work and the fact that for the young people we teach there are no second chances. I’ve heard teaching compared to air traffic control and the risks in the classroom compared to the risk involved in miscalculating the landing of a plane. These kinds of comparison  […]

  6. RCA says:

    Thanks very much for your post, I’m curious about this passage?
    “Maybe it might be true that teaching others could be an effective way to learn, but what’s the evidence beyond our intuition? This kind of belief is an emotional one and, that being the case is like pretty much everything else that gets believed. But that doesn’t excuse people making stuff up to support what they believe.”

    This seems to imply that teaching others might be effective for learning but doesn’t actually currently have any evidence behind it or have I misunderstood? If I have inferred your point correctly then you may have overlooked quite a substantial evidence base suggesting that teaching others is an effective way of learning,some of which I know you’re familiar with eg: Karpicke’s work on retrieval. Here’s some of the evidence which would suggest that designing tasks which encourage the students to teach others is an extremely effective activity to include in your pedagogical toolkit:

    1. Teaching others something is a very clear example of retrieval practice in operation, especially when one considers that the ‘teacher’ will almost certainly be taking questions from the ‘student’ as part of the process.
    http://learninglab.psych.purdue.edu/publications/

    2. Teaching others increases a student’s self efficacy which will not only help them in the task at hand but in tackling new concepts/tasks as well. I’m referring specifically to Bandura’s social learning theory here, which again I think you’re aware of as being one of the most evidence supported theories of human learning. I haven’t put a link to this one as there are so many to choose from.

    3. Eric Mazur has well over a decade of data from his undergrad classes at Harvard that Peer Instruction (which has students teaching other students as a core element of its approach) is very successful indeed for learning and identifying misconceptions in the underlying concepts.
    http://research.pomona.edu/kevin-sea/files/2013/07/MazurActiveLearning.pdf
    http://mazur.harvard.edu/research/detailspage.php?rowid=8

    I’ll stop there, but there are very many other evidence supported reasons why teaching others is a very effective way to learn.

  7. Marcus says:

    David, in the book you speak of it mentions: trying to solve a problem before being taught the solution leads to better learning, even when errors are made in the attempt. Isn’t this at odds with your ‘high quality exemplars and knowledge first’ theory? I’m just interested.

  8. […] wrote earlier in the week about why, despite it’s limitations, research is better than a hunch. Since then, I’ve been reading Daniel Willingham’s article on Real Clear Education; he […]

  9. […] wrote earlier in the week about why, despite it’s limitations,research is better than a hunch. Since then, I’ve been reading Daniel Willingham’s article on Real Clear Education; he says […]

  10. […] Didau discusses the ideas of teaching strategies in this post. He says that research shows what the best strategies are compared to what our gut instincts may […]

  11. […] wrote earlier in the week about why, despite it’s limitations, research is better than a hunch. Since then, I’ve been reading Daniel Willingham’s article on Real Clear Education; he says […]

Constructive feedback is always appreciated

%d bloggers like this: