The transformative effect of weekly knowledge tests

  • Colin McKenzie is a science teacher who is starting a tutoring agency in Lancashire
  • Prof Tom Burkard is the principal author of the Sound Foundation decoding and spelling programmes and has recently served on Ofsted’s Maths Advisory Group


Tests—no other word strikes such terror in the teaching profession. Even schools like Michaela avoid using the ‘T’ word. At Ofsted, we weren’t afraid to insist that pupils had to achieve automaticity in each topic before moving on to more complex material, but we still got it in the neck when we suggested that ‘quizzes’ might be used for assessment. Of course, nothing short of a timed test can establish whether a child has automatic recall of any knowledge, but we knew that the mere mention of such a procedure was enough to induce mental illness in our progressive critics. Even PTE has presented tests as a necessary evil, suggesting that they can be administered without pupils even being aware that they’re being tested.

First, the trigger warning: read no further if the thought of children being tested every week and actually enjoying it offends your deepest and most cherished beliefs. We have now tested hundreds, if not thousands, of children – rigorously and regularly – with none of the adverse effects so often reported. Of course, we’re not talking about any old tests; clearly, it helps if pupils understand what they’ll be tested on, which is why an increasing number of schools now use knowledge books which clearly define what pupils need to know at each stage. Nor are we talking about tests to see how clever kids are. This isn’t the place for a discussion of a knowledge-rich curriculum, but we know from experience that the vast majority of pupils seldom climb very far up Bloom’s Pyramid before their GCSEs. In fact, they never climb anywhere if they’ve missed out on the relevant knowledge and understanding for a given subject.

Our work in literacy and maths confirmed that a secure knowledge of basic facts forms the very bedrock upon which any pyramid of higher skills can be built. For example, before the phonics revolution, many pupils failed to develop automatic recall of grapheme-phoneme correspondences. Inevitably, these pupils were unable to read well enough to understand the simplest written information, and they invariably ended up in the lowest sets where the only thing they learned was that school was a humiliating waste of their time. Similarly, extensive research in the cognitive sciences has demonstrated that the automatic recall of number bonds – and the well-practised use of simple traditional algorithms – frees pupils’ working memory to engage in higher-order tasks.

Working on the same principle, we set out to produce a bank of essential science facts – separated into individual topics – that pupils would learn and be tested on, until they could recall them automatically. And the crucial word here is ‘tested’. We had no desire to produce yet another expensive handout to be left discarded on desks at the end of a lesson or be stashed and then ignored forever in a school bag.

Our pupils were simply going to have to learn these facts and we’d check their knowledge by testing them every single week. By then, we had few hesitations about regular testing. Our GCSE results were already improving dramatically with a strong focus on pupils sitting practice tests under exam conditions. There can be no better prophylactic against exam stress than knowing enough to pass the exam, along with lots of practice taking tests.

First, our teachers produced a knowledge book of facts and each week our pupils were given a section to learn. The information was set out in simple question and answer form, so if they had learned the answers, they’d have no problems. In the last lesson of each week, pupils were tested on exactly those facts. They were shown how to test each other ‘in pairs’ and we enlisted the help of parents too. But the main motivator was the test itself.

The competition to come top in each group was friendly but fierce, with some students even demanding to be tested on extra material from previous topics – or even topics we hadn’t yet covered in class – just to get one-up on their peers. Ironically, competition also produced collaborative efforts beyond the wildest dreams of our more progressive teachers, with pupils spontaneously pairing up to test each other, or to coach the weaker pupils. Initially, those weaker pupils were motivated by beating their own ‘personal best’ scores, but it has to be said, they didn’t stay much weaker for long. Those that made the effort made the most progress and it was plain for all to see. We were really beginning to ‘close the gap’.

So, did any of these children approach our tests in trepidation? Were they stressed or fearing failure? I can honestly report – not a single one. It’s true that at first, some didn’t really engage. Some of our lads even made a point of not trying to learn the facts, but once the reward of a little effort was plain to see – as knowledge scores were published and test results got better by the week, the ball really started to roll. We had never seen anything like it. Pupils who had previously dallied on corridors, deliberately late for lessons, were lined up early and quickly testing each other on those corridors before their tests. If any teacher decided not to administer a test on the given day, there could be uproar. Our children loved those tests. For once, they knew what they had to learn, and they knew that they were learning it.

Revision for our weekly tests replaced almost all homework. New ‘higher stakes’ termly tests were also introduced, with a ‘Knowledge’ paper consisting only of knowledge book content. We still maintained a SAT style paper to test ‘higher order’ skills – and found that performance on this paper improved too. Because the memorisation of knowledge was largely accomplished outside class, during lessons we were able to relate the knowledge to previous learning and enable our pupils to develop a deeper understanding of the material we covered. This in turn made it easier for them to understand the significance of new knowledge and integrate it into ever-expanding schemata. Three years after our first experiments with a knowledge-based approach to teaching and – more importantly – to testing, our GCSE results were well above target and massively above anything previously achieved. Other factors certainly contributed towards this success, but there can be no denying that what we’d instigated was a complete culture change.

Nothing that we did was entirely original—we’re old enough to have attended schools that used routine testing. Distant drums told us that we were not alone—and many more schools have since discovered the transformative effects of weekly knowledge tests, but they have better things to do than cope with the Twitter storm that inevitably follows disclosure. We don’t Tweet, so we don’t care. Sadly, we’re not the only ones involved, and we have to consider our younger colleagues who have their careers to think about.

This is not to say that what we’ve done isn’t research-based. Take the ‘forward effect’–researchers have found that tests can “can increase long-term retention of subsequently studied new information” even when it’s not related to the previously studied material. Pastötter and Bäuml (2014) conclude that “The forward effect of testing has been shown to be a robust and replicable phenomenon in laboratory studies of memory research.”

For so long as there have been schools, teachers have known about the forward effect. If pupils know they’re going to be tested, they pay a lot closer attention to what their teacher says. They actually think about what they’ve just read (aka ‘metacognitive monitoring’) and try to remember it. Even though the profession is now largely ignorant of the forward effect, this is only because most teachers use tests rarely, or use them for all the wrong reasons, with poorly-prepared children who have little chance of succeeding.

So even if Pastötter and Bäuml and their ilk haven’t discovered anything new, we owe them a massive debt for dressing up common sense in respectable scientific clothing.

The views expressed here do not necessarily reflect those of PTE or its employees.