The lesson that starts four minutes late

Look for the fault in this ordinary lesson change. The bell goes, and most of the class arrives more or less on time, while two or three drift in a few minutes later. The teacher waits, because starting twice is worse than starting late. Bags come off, books come out, and someone needs a pen.

By the time the room is ready to learn, four or five minutes have gone.

Did you find one? No child was rude, and no teacher made a mistake. But run the same scene again at the next lesson change, and the one after that, five or six times a day, and the arithmetic turns serious. Twenty-five minutes a day is more than two hours a week, and over a school year those minutes run into weeks of teaching.

Almost no parent ever sees this happening, yet it may be the single biggest difference between one school and another.

The school that decided to count it

Michaela Community School sits in Brent, one of the poorer boroughs of London. Between a third and two fifths of its pupils are eligible for free school meals.^[1] Around half attract the pupil premium, the extra money given to schools for children from disadvantaged homes, and the school does not select by ability.

In 2025, over 80% of all its GCSE grades were 7 to 9. More than four in ten pupils walked away with five or more grade 9s, and every single pupil passed both English and Maths.^[1] Ofsted has rated the school Outstanding twice, in 2017 and again in 2023, describing behaviour there as “exemplary”.^[2]

What would you suspect, reading those figures? The obvious suspicion is that the school must be quietly choosing clever children. There is a measure designed to test precisely that suspicion, and that measure rules it out.

Progress 8 compares every pupil’s GCSE results with the results of pupils across England who had the same score at the end of primary school.^[3] It measures distance travelled rather than where a child started, so a school full of high-attaining eleven-year-olds gets no credit for their high grades at sixteen, because those grades were already expected.

The last year this could be measured was 2024. Michaela’s Progress 8 came out at 2.55, the highest in the country for the third year running.^[1] The Department for Education confirms separately that the school’s admissions policy is non-selective.^[3]

Whatever is producing those results, it happens after the children arrive.

One note about the year. Progress 8 cannot be calculated for 2025, at Michaela or anywhere else, because that cohort never sat their Key Stage 2 tests. Those tests were cancelled during the pandemic, so there is no starting point to measure from.^[3] The 2025 grades above are real, and the proof that they were not simply the result of a clever intake comes from 2024.

Taught rather than punished

Michaela is well known for one visible habit: pupils move between lessons in silent, orderly lines.

Written down, that sounds severe, and the lines attract a lot of argument. The reasoning behind them is duller than the argument suggests. If a class arrives together and settled, teaching starts at once, because nobody is waiting and nobody is repeating the first instruction for the three pupils who missed it. The four minutes at the front of the lesson stop disappearing.

One word in how the school describes this changes the meaning: the behaviours are taught. Pupils are shown what the routine looks like and then practise it. When it slips, they are corrected in the same way they would be corrected on a piece of algebra. The Australian Education Research Organisation reviewed the evidence on classroom routines and found that routines pupils know well “reduce interruptions and increase teaching time”. It also found that explicitly teaching and modelling those routines frees up a pupil’s working memory to spend on the actual lesson.^[4]

Teaching a behaviour is very different from being tough on children. Being tough punishes the absence of a behaviour, whereas teaching puts the behaviour there in the first place.

This has been known for fifty years

None of it originates with any one school. In 1970, an American researcher named Jacob Kounin filmed 49 classrooms for a full day each, and recorded every twelve seconds whether pupils were working.^[5] They were primary classrooms, although what he found has been repeated in secondary classrooms many times since.

He expected to find that the best teachers were the ones who dealt with misbehaviour most effectively. He found something else. What separated the strongest classrooms from the weakest was momentum, meaning starting activities promptly, keeping them moving and finishing them cleanly, together with smoothness, meaning the absence of jolts that break the flow of a lesson. Transitions between activities turned out to be one of the main places where a classroom succeeded or failed.

Since then the finding has been measured many times over. Improving how a classroom is managed carries an effect size of about 0.52,^[6] which is roughly the difference between an average pupil and one at the seventieth percentile.

Inside that research sits a distinction that matters more to a parent than any headline figure.

Researchers separate allocated time, the hours a subject appears on the timetable, from engaged time, the minutes a pupil is genuinely working. Allocated time turns out to predict achievement only weakly, while engaged time predicts it strongly. Taken together, these time measures are described in the literature as the most powerful explanation of differences in achievement once you have accounted for where pupils started.^[7]

Hours in a classroom, on their own, buy remarkably little, while minutes spent actually working buy a great deal.

Present in the room, absent from the lesson

Minutes are not only lost at the door.

A student can be sitting in the right room, at the right desk, with the right book open, and still be somewhere else entirely. Every teacher recognises that student. But how much does the wandering cost? Researchers have measured that cost precisely.

In one study, students who used a laptop for other things during a lecture scored 11% lower on the test afterwards. Students who were merely sitting in view of someone else doing it scored 17% lower, which was worse than the people doing the distracting.^[8] A wandering attention spreads to the desks around it.

A larger study followed 507 students through a course and tracked their non-academic internet use. That use predicted their final exam score even after the researchers accounted for how intelligent and how motivated those students were.^[9] Using the internet for the class itself made no difference either way.

This is engaged time again. The student is present, counted and marked on the register, while very little is going in.

Two limits on what this proves

First, the obvious conclusion about phones does not survive checking. An early English study suggested that schools banning phones saw test scores rise. A Swedish replication found no such effect, and a large 2025 study of over 1,200 teenagers across 30 schools found no difference in English or maths results between schools that ban phones and schools that do not.^[10] School-wide bans do much less than people assume. A phone in front of one working student is a different question, and there the evidence is strong.

Second, one school proves nothing on its own. Michaela offers a narrower range of subjects than most schools, which is a genuine trade-off, and entry to its sixth form is selective. The argument rests on the research rather than on the school, which is simply the clearest existing demonstration that a disadvantaged intake does not decide the outcome.

At home, remove the daily negotiation

The part you can take home is the removal of the daily negotiation, rather than the silence or the lines.

At Michaela, no pupil arrives at a lesson wondering whether today is a working day, or how long they will get before the teacher gives up. The question has already been answered, so no energy goes into asking it.

Revision at home usually goes the other way. Whether work happens, when it starts and how long it lasts are all reopened every evening, and the negotiation itself eats the time. A fixed hour, in a fixed place, with the phone in another room, removes the argument before it starts.

We will not tell you how to parent, and neither does the research above. What the research does say clearly is which variable is doing the work: not the total time set aside, but the minutes genuinely spent working.

GCSE English punishes the difference more than most subjects. Maths gives instant feedback, because a wrong answer refuses to balance, but English gives no such signal. An hour spent reading a set text with a phone face-up produces a warm feeling of having revised, some highlighted pages, and almost no change in what a student can write under timed conditions. Nothing tells the student that the hour was wasted.

That is why our programme is built on rhythm rather than good intentions. Lessons run at the same time each week. Written work is marked in detail against the examiner’s criteria throughout the course, not once a term. Every student knows what they are working on next, because it was decided from their last piece of marking rather than negotiated on the night.

What that rhythm gives a student is engaged time, arranged in advance: the lesson that begins at the bell rather than four minutes after it.

Sources