Professor Brian Cox’s educational truthiness

Professor Brian Cox’s Twitter bio reads, “I think the world needs more experts.” I agree with this sentiment. We are cursed by generalists who think they can apply generic skills they learnt in one area to solving problems in a quite different area. However, generic skills of this kind don’t really exist. Instead, you need domain specific knowledge.

That’s why it was odd to see Cox in The Sunday Times come out with a half-baked prescription for education. Cox does not appear to have studied education or the learning sciences. He does teach a highly select group of students at university level, but you can hardly extrapolate from that experience to make claims about schools, so I called him out.

I was on the way to the Hope Show in Derbyshire with my family for the bank holiday and, once Cox had replied to me, my Twitter notifications rocketed. It was mostly fans of Cox calling me names, telling me that he cannot be an amateur because he is a professor or suggesting I was trying to promote my book. I couldn’t keep track of all of it. However, I did agree to detail why Cox is wrong. So here it is.

Firstly, the article says that Cox, “…criticised the effect of schools when they drill teenagers for tests”. I’m not sure whether these are Cox’s exact words but they represent one of the worst cliches in education. For over a hundred years, people have been having a go at drilling, often adding ‘rote knowledge’ or ‘facts’ to the description.

It is based on a naturalistic view of learning that suggests that learning should be fun and relatively effortless. Many educationalists (eg here) have explicitly made the comparison with learning your mother tongue – nobody does this through drilling and rote repetition so why teach school subjects this way?

David C Geary has an answer: evolution has had time to affect our ability to learn to speak and listen. We are primed to learn these abilities. Writing, and all other academic subjects, are relatively young, with mass literacy only being around for about 150 years. Learning these subjects is not natural and is necessarily effortful, requiring a lot of not particularly enjoyable deliberate practice.

Why is this? According to Cognitive Load Theory, learning new academic content in this way requires the use of an extremely limited working memory. It is limited for a reason – if not, our long term memories would be a riot of chaotic change. We can only attend to a few items at a time in working memory. However, we can boost this by drawing on knowledge held in long term memory in a process known as ‘chunking’. If we simply know that 7 x 8 = 56, we don’t have to work this out in working memory and so we free up working memory resources to attend to other aspects of the problem. Drilling and testing are ideal ways of ensuring the storage of this knowledge in long term memory in networks of relationships or ‘schemas’. Drilling and testing therefore serve a critical purpose.

Cox carries on with this theme, creating a false choice between students who ‘develop a real understanding of their studies’ and those who are merely ‘exam-passing machines’. In reality, these two factors are highly correlated. This downer on exams is fashionable, but I worry about the consequences. Exams are not perfect, but they are still the fairest way of selecting students for university and careers. If you replace them with portfolio assessments or extra-curricular checklists, you create a system that is far easier to game by the privileged.

Instead of focusing on exams, students should focus on ‘big questions’ such as ‘how did the universe begin?’ What does Cox imagine this would look like? Are school students simply to speculate and debate in the absence of knowledge? Perhaps they are to look up stuff on the internet and cut and paste it into posters which they don’t understand (because they lack schema in long term memory)? Or maybe watch a video? Or perhaps we explicitly teach them about red shift, cosmic microwave background radiation and the Big Bang, with lots of questioning and practice? If it’s the latter then that’s what we do already and that’s what is assessed in exams.

Cox links exams to a rise in mental health issues. I am not sure about the strength of the evidence for this, particularly since exams have been a constant of teenage life in Britain for many years. It is possible that other factors account for any rise in mental health issues, including greater awareness and a loss of taboo leading to greater diagnosis. I also don’t think it helps when children are given the explicit message that exams will stress them out rather than them being a normal part of life: no big deal. Cox’s comments are an example of such a message.

Perhaps the most clear cut sign that Cox is just talking off the top of his head is his call for money to be spent on lowering class sizes. It seems like an obvious good thing. It’s truthy. And yet it is very well known in education research that the evidence for such a proposal is mixed, at best. If Cox knew this then he would at least preface his comments with ‘despite the lack of evidence at present, I still think…’. But he uses no such caveat.

Why are smaller classes not such an obvious win? Firstly, you can spend a lot of money reducing classes from 27 to 25 and yet this is unlikely to change anyone’s experience very much. You also need to recruit more teachers and ensure they are of the same quality as the current ones and this is not trivial. In some of the highest performing education systems, schools tend towards the opposite approach, teaching larger classes but giving more spare periods to teachers for planning and collaboration. Lowering class sizes is a classic example of the kind of suggestion you would get from a non-expert operating outside of their field.

Brian Cox is right that we do need more experts, particularly when it comes to ideas for improving education. He should heed his own advice.


44 thoughts on “Professor Brian Cox’s educational truthiness

  1. Brave again Greg. Brian Cox, like inclusion and differentiation are so popular, questioning will lead to critical defense.
    Bet your twitter feed will cop a pounding.

  2. Pingback: Argumentum ad Celebritatem – another update – Pocket Quintilian

  3. I’m reminded again of one of my favourite quotes from one of my favourite authors, David Lodge. From “Nice Work”:

    “How did you get to where you are?”
    “How did you become a university lecturer? By doing better than other people in exams, right?”
    “Actually, I’m opposed to competitive examinations,” said Robyn.
    “Yes, you would be,” said Wilcox. “Having done all right out of them, you can afford to be.”

  4. No matter how much we wish primitive and traditional educational system will survive either in form or in aspect. The bulk of work in education is recent development ie as near as 19th century..the basic classroom structure, dominant role of teacher and formality are hallmark of this eduction. Brain Cox presented his wish list and it should be honoured.

  5. I would disagree that we ‘ pick up’ language effortlessly. There’s an element of casually observed truth here; it is ‘truthy’, and while it seems we are ‘wired for language’, I’ve also noticed my children have lots of continuous trial and error, drilling if you like, as they learn words, grammar, sentence construction and expression by being children. One doesn’t become eloquent overnight, or even over years…it can take decades.

    Secondly, on the Big Bang, etc. Here’s a great example of real content where students can be introduced to the real hard work of the maths and theory that underlie this, and also to critically think out the limits of knowledge in theory building. I wonder if Cox, who while brilliant, is a bit one-eyed, and never talks about the discussions in cosmology regarding the Big Bang and the continuous ‘saves’ as the theory morphs. Now that would be educative as presumably disinterested scholars debate using observation, analysis, lots of maths, and good hard thinking.

  6. 1. The initial opposition to ‘drill’ was because in many schools that was all there was. Paper was so expensive there were few textbooks and no exercise books, so rote learning was about all that was left. Add multiple-age classes into the mix and many children learned little and understood less.

    2. See my critique of Geary’s model here

    3. Is Cox suggesting abandoning exams?

    4. Reducing class size is a suggestion you’d expect from someone who has extensive experience of working with groups of varying sizes. The calibre of teachers is a *different* issue.

      • Stan says:

        To be fair – he showed it works for K-3 nothing else. Also he doesn’t mention opportunity cost but compares it to doing nothing.

        I am intrigued about the point about kids in some countries being socialized to work better in large classes. Is this a euphemism for better behavior management strategies in these countries?

        Wouldn’t that be the right intervention to compare smaller classes sizes to?

        (I am hoping that the internet visibility into other countries cultural contributions has destroyed the myth that we always create better people from a creative or cultural point of view.)

      • Well, it seems we agree on something! This was actually the one stricture of his (Cox, that is) in that article that I agreed with, although his proposed solution was unrealistic. Most of the evidence I’ve seen presented in the past to downplay the importance of class sizes has relied quite heavily on the South Koreas and Singapores and the like, where there are obviously other factors.

        During my stint training English teachers in China I was amazed to hear that some of them were regularly teaching classes of 70 or 80 (!). And (so they told me) it wasn’t as if they never had classroom management issues, but they were mitigated to a large extent by the Confucian norms of respect for authority and loss of face if parents found out about misbehaviour.

        In my own classroom experience I’ve found that class sizes do matter, very much so. But perhaps not chiefly for pure academic reasons.

      • You’ve brought the evolutionary biology issue up before but you’ve never explained it. You just reference a book that most of us don’t have a copy of. Try explaining your critique.

        I would point readers to my interview with Geary on this blog where he clears up a lot of questions.

      • Tom Burkard says:

        Just as signficantly, the paper only covers nations that are predominantly white and wealthy, and ones where progressive ideology has long dominated K-3. In schools where learning is personalised, it’s pretty obvious that smaller classes will produce better results. There’s also the minor matter of researcher bias–as one wag once commented about research in the social sciences, “Tell me the conclusions you’d like, and I’ll provide you with the references”.

        In fairness, schools that use intensive synthetic phonics do use quite a bit of small-group and individual instruction in the first year or two of formal instruction, but this is only for short periods each day and is often delivered by TAs. Hence it does not impact on class size or add significantly to workload, and in event it reduces the need for differentiation in other subjects and in subsequent years.

    • Stan says:

      On point 4 – no it is a perfectly valid alternative to compare paying for smaller class sizes to and to worry that the cost of lowering class sizes goes up worse than linearly if you want to keep the same quality of teacher. That is because like any supply curve as you try to find more the cost per unit goes up in the short term.

      • Stan says:

        Can you see how these 3 factors are not independent? That is you cannot reduce class size without increasing cost or reducing teacher quality or both.

        The class size – cost dependence seems obvious to me.

        The size-quality interdependence is based on the assumption that you have selected the best available teachers today. So if you need more tomorrow you will be adding less able teachers. The effect might be small if you are just reducing class size by 5%. But then so would the effect of smaller class sizes.

        The quality-cost dependence also comes from the idea that people considering their career options will consider the financial return. So to be more selective in who can become a teacher paying more may be necessary to increase the pool of people considering it.

        So the complaint is not that these are not different but that they are not open to changing one without impacting the others.

  7. Stan says:

    I can’t understand how anyone with a claim to an opinion worth sharing can propose smaller class sizes without a mention of opportunity cost.

    All else being equal the cost goes up in proportion as size goes down. Half the size, double the cost. So comparing to the status quo is just pointless. You have to compare to what else you might do with that money to improve outcomes. Perhaps it is feeding kids. Perhaps it is paying for a better quality of teacher. (Arguments about paying teachers more must be arguments about getting better teachers right ?)

    • Tara Houle says:

      re: class size. Tell me about it. With the SCOC ruling in BC, we’ve seen an additional $3m/year for the next few years just to satisfy that argument. It’s been a nightmare! We now have schools turning kids away because class sizes are full. We’ve had daycares being turned out of schools because of limited space. We’ve had Superintendents filling in for teachers because we don’t have enuf teachers hired.

      For what?

      All this has done, is place even more strain on an already bloated, inefficient school system with even more parents turning to either the private school or homeschooling. More money for the teacher unions, negative impact for the taxpayer and for the community.

      • chrismwparsons says:

        So if money can be better spent than on reducing class sizes, how should that be spent…?

      • Stan says:


        Some might be –
        providing text books,
        providing separate remedial classes
        providing after hours tutoring as is done in Ontario high schools.
        providing school meals
        hiring more subject matter expert teachers
        providing training for existing teachers
        updating the curriculum or other guiding documents to incorporate the core ideas from Greg’s book.
        – any other improvement you could make that costs money.

        The point is before doing anything options should be compared and evaluated given there is a limited amount of resources to be allocated even if you also increase from the amount from what you have today.

        Given to make a big difference in class sizes has a huge cost and means you must have available a pool of available teachers of similar quality to those there today it is hard to see it being obviously the first way to spend more or retarget existing funds.

  8. chrismwparsons says:

    I think Professor Cox’s comments about the overall impact of certain styles of education – ‘drilling’ (or maximising knowledge retention through explicit/direct intruction in the context of a high stakes exam sytem if you like) – is a reasonable ‘big-picture’ philosophy of education perspective, which almost all of us are ‘amateurs’ at being able to talk about, but which all of us have a dutiful perspective on as members of the society which the education system serves. And let’s be honest, his role in trying to engage a huge range of people in gaining some kind of nourishing value-added to their lives by an intellectual interest in going past what’s in front of their noses, means that his perspective is probably wider than that of many.

    Just because fruit is the best way of indulging in the right combination of certain desireable nutrients (let’s say certain sugars), this doesn’t mean that we overall thrive on a diet composed only of fruit.

  9. I think that not having an exam (a single major time-limited assessment) doesn’t need to entail ending drilling (a learning technique). It might be possible to encourage students to enjoy drilling if they see how it leads to their own growth and improvement at skills and recall. I’m sceptical of the benefit of exams. There is a lot of extra stress and pressure, and often some amount of luck, in having one big, time-critical test as a long-lasting measure of a person’s capabilities.

      • I haven’t looked at empirical data on alternatives, but this is probably (at least in part) a philosophical question. What is fair and what is gaming the system? How much should we correct for imbalances in performance that result from imbalances in privilieges? I don’t think these questions have easy answers, but I agree that the privileged shouldn’t be excessively advantaged.

        A test environment might be the only way to fairly determine skill levels and information recall. Still, I think having a single major test (an exam) to determine a final grade can be incredibly stressful, exaggerates randomness in performance, and doesn’t account for or encourage ongoing growth.

        Encouraging students to retake small, specific tests until they demonstrate competency could reduce stress, remain a fair measure of their ability, encourage a growth mindset, and improve learning outcomes. Lots of small tests on specific information or skills can also more readily produce a detailed account of a student’s competencies.

        At the moment, transcripts and rankings provide a pretty nebulous evaluation of a student’s capabilities. A more fine-grain record of attained student competencies could be used for university admissions and by employers for hiring. Admissions and graduations based directly on required competencies could also be more resistant to grade inflation.

        A minimum or required set of information learned and standards for skills could be used for university admission. This might provide a fair and less stressful process for admission. I think giving students this detailed list of competency requirements also gives them a better goal to aim for than an overall grade or percentile ranking.

        I suspect that competition in schools and admissions often results in pathological outcomes. Competition can result in some students being stressed beyond their limits trying to achieve and many others giving up. Limiting the use of relative grading and ranking schemes and placing an upper bound on the expectations on students might also be effective in reducing stress and dissatisfaction in school.

        Competency requirements could still allow different levels of admission without explicit student ranking. Students that meet a higher set of competency requirements could be accepted automatically and receive advanced credit. If too many applicants only achieve a minimum level of requirements, they could be entered into a lottery for admission. (Is a lottery fair?)

        Replacing exams with many small tests would increase the amount of assessment work substantially. This could be dealt with by using the student body to do the majority of the assessment. Students could cycle between being marking tests, receiving instruction or feedback on their graded tests, and taking the next tests. A teacher could take a role instructing the class, providing individual in-person feedback and suggestions for study, and overseeing the student testing and marking.

        How much does this conflict with your ideas about education?

      • I think students marking tests would be very dangerous if these tests are high stakes, as you suggest they would be. There would be a motivation to cheat. You are also suggesting replacing one exam on which students’ futures depend with lots of mini exams, which hardly seems like a way of reducing an emphasis on exams. To use lots of little tests, you would also have to chop the curriculum up into lots of little bits. Teachers are then likely to drill each bit until the test is passed and then forget that and move on to the next one. How will you know if any student has ever grasped the entirety of the subject?

      • Thanks for your thoughts, Greg. I’m not an education expert, but I do enjoy thinking about alternative approaches to education. It’s nice to hear what problems these ideas might have and think about possible solutions. Some responses to your last comments:

        1. These [small] tests would be high stakes:

        If the assessment is broken into lots of small, specific tests (perhaps dozens for each subject) that students are required to take multiple times, the stakes for each individual test are substantially lower than a single big exam.

        2. Students marking tests would be very dangerous; there would be a motivation to cheat.

        I am confident there are ways to make cheating very difficult, even with students doing the marking. For example, completed tests could be anonymised and randomly assigned to students in another class to mark. The student that marks a test could be recorded. Between submission and marking, the tests could be copied and the teacher could sometimes review the tests and marks and give the students marks on their marking. Given that there would be lots of tests that occur over a long period of time, cheating to much effect would require a sustained effort over a long period of time. I think this system would be challenging to cheat.

        3. Replacing one exam with lots of mini exams does not reduce the emphasis on exams:

        In a superficial sense it doesn’t, but the testing regime is qualitatively very different. If a students is able to take the many small tests as many times as it takes to succeed, that is very different from taking a big test once to assign the student their eternal grade. My understanding of the research (you referenced) is that repeated testing in intervals improves retention. My suggestion of repeating small tests is exactly that. A single final exam doesn’t provide this advantage of repeated recall.

        4. Teachers are likely to drill each bit until the test is passed and then forget that and move on to the next one.

        Ideally, I think the system would be designed to allow students to learn and take most tests at their own pace, apart from periodic retesting of recall. Teachers could hopefully focus on giving individual students feedback on the pace of their progress and provide support to students that have difficulty with any particular skill or topic of information.

        5. How will you know if any student has ever grasped the entirety of the subject?:

        I’m guessing this is pointing to big-picture understanding of the content (grasping the subject is more than recalling each individual part). I think that tests could be designed to assess their understanding and ability to articulate the connections between concepts.

        If you are referring to demonstrating recall and competency at the end of a course of study, there could be a final period of testing and assessment that covers the full suite of content. I don’t immediately see why this couldn’t keep the format of having small tests and retesting, rather than resorting to big one-off exams.

      • Stan says:

        If the multiple tests verses single test aims just to reduce the chance of a bad day hurting someone’s future prospects this can be mitigated with a simple repeat the test option. Say bad days make a serious change to 10% of the students then a redo with the same issue would reduce the number to 1%.

        Offering 2-4 redo opportunities to a small fraction of students would be way more efficient than running standardized tests multiple times for all students.

        If small tests are just to reduce stress because the test is easier then it has not achieved its aim. The point of the test is to assess the student’s learning in an area that may or may not be easily learned. If the normal proportion of students can do well in the current exam system then it is not the exam that is at fault for those who don’t.

      • Stan says:

        Even if you can find a way to consider using students as free labour ethical you should expect to get what you pay for.

  10. Tempe Laver says:

    My kids love drilling. They loved it when they learned their times tables, they loved singing songs over & over and hearing the same books when they were little. Fairly recently my eldest’s science teacher asked them to go home a rote learn the periodic table. Again, she enjoyed every minute and felt that she’d really accomplished something. Her little sister automatically joined in. Eldest reported to me that it was the ONE thing that all the students had gone home and done. This suggests to me that children actually enjoy memorising stuff. The same yr 9 daughter, when asked how her study for her maths exam was going said she would do it all easily but the hard part was memorising all the procedures & formulas. Both my anecdotes suggest that the common narrative we hear regarding memorisation may not be true. Memorising/drilling can be fun and it can also be hard work, rather than boring and easy.

    Brian Cox has a right to say what he thinks about education just as Greg has a right to rebut his arguments. What grates on me when celebrities/tv personalities offer up opinions, it’s invariably printed in the press or shown on tv. Many parents are ignorant in terms of the debates in education on pedagogy and curriculum. They hear a famous guy telling them that tests and drilling are bad and that wonder and awe are great and their kids creative spirits are being stifled or crushed by traditional, outmoded methods. We can be blinded by the celeb. status and likeability and repeat this doctrine because, of course, it sounds wonderful to our ears, especially if it helps to explain/excuse why our child is doing so poorly at school. We want students to think science is great and see the big picture. Well yeah, but how do you get there? They don’t know what they don’t know. Also, how often do we see the counter view presented? It doesn’t make for sexy tv. Instead, images of children wondering through woods & creeks or starring up at stars contemplating the universe and life’s bigger questions, do.

    I think it’s fair to point to the difference between a high school science teachers experience versus a university science professor. Cox’s students have chosen to study science, presumably because they have a propensity for the subject, if not a love of it. No doubt they have quite a bit of knowledge of science due to their schooling and perhaps even read outside of school. How many students in an average high school class room chose to be there? It is not until they have learnt some scientific knowledge and begin to understand what science is that they might become more engaged in it and become fascinated by it. It is often the case that the more you know the more you want to know. Knowing a lot creates wonder & awe.

  11. Tempe Laver says:

    Regarding testing as opposed to portfolios, thankfully Queensland held an inquiry into this very question some years ago as most of our senior assessment was via assignments. Shortly that will change so that at least half will be exam based. The inquiry found that students were gaming the system with
    tutoring companies or parents writing the essays and that writing long essays was not the best way to ascertain how well a student was doing in maths & science.

  12. John Pierry says:

    “You’ve brought the evolutionary biology issue up before but you’ve never explained it. You just reference a book that most of us don’t have a copy of.”

    Actually, I have explained it – see my comments about the “incrementalist” or “accretion” model. And in another comment to one of your recent posts I linked a very nicely worded article by other experts talking about the same thing.

    “I would point readers to my interview with Geary on this blog where he clears up a lot of questions.”

    Nup – he just puts forward the “evolution hasn’t had time to catch up” theory as if we have to accept it at face value. No explanation behind the theory to back it up.

    It reminds me of the arguments from only less than a century ago, when continental drift theory was dismissed out of hand: oh come on, land masses are fixed in place, isn’t it obvious, that’s the end of it sort of thing.

    And as I have pointed out, this “great leaps” model of evolution is rapidly being superseded by the models I have outlined: as Tomlinson points out (with lots of references to various evolutionary scientists), much of our cognitive ability – including mechanisms that (combined) allow for abstract thought – reach back before we became humans.

    I’m not saying I have all the answers. I’m not even claiming to be an expert. I’m just saying that to position this model of evolutionary theory as being the only one is a mistake, particularly as the flaws are increasingly being pointed out.

    • chrismwparsons says:

      Hello John – could you drop a couple of links to things here? I’m interested, but can’t track the references you made in your last comment.

    • Stan says:

      John do you mean this link you provided:

      He points out in two places that any other result would be counter to Darwin’s theories. It really doesn’t sound as though it is any sort of change in the theory of evolution being reported.

      There is pretty good data on the rate of genetic adaptation and everyone agrees that it is genetic adaptation that leads to varying features and capabilities in different species. No one is arguing that any capability doesn’t rely on genes that evolved. So this doesn’t look like the issue with continental drift where people didn’t believe it could have happened.

      If you want to show that written language could be learned in the same way we learn speech you would have to show why all the people on the phonics side of the whole-language vs phonics debate have it wrong when it looks like the data clearly points the other way.

  13. Pingback: It is not a question of priorities – Filling the pail

  14. Pingback: Scientists should be disabused of their pseudoscientific beliefs about learning – Filling the pail

Leave a Reply

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

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

Google photo

You are commenting using your Google 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 )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.