Marking and Feedback That Actually Helps Pupils Improve (Not Just Your Workload)

There’s a particular kind of tired that comes from a Sunday evening spent triple-marking thirty exercise books, writing a comment, leaving space for a pupil response, then writing a second comment underneath their response, in a different colour pen, because that’s what the policy says. If you’ve done it, you’ll know it rarely felt like the thing that was actually going to move a Year 8 class forward in fractions or a Year 10 group forward in their essay structure. It felt like proof. Proof for someone else that you’d done your job.

That feeling wasn’t imagined, and it wasn’t just you being inefficient. It was, for a long stretch of the last fifteen years, genuinely what a lot of schools expected, and it grew from a fairly thin evidence base. Here’s what the research actually says works, where deep and triple marking came from, and how to give feedback that helps pupils improve without quietly wrecking your evenings.

How triple marking became “what good teachers do”

Deep marking, triple marking, dialogic marking, the purple pen of progress: different schools gave it different names, but the shape was similar. A teacher marks the work in detail, often with a written comment against a success criterion. The pupil responds, sometimes correcting an error, sometimes answering a question the teacher has posed. The teacher then marks the response too, closing what got called the “feedback loop.”

None of this was ever a national requirement. But through the early 2010s, a mix of pressures pushed it from “something a few outstanding lessons did well” into “what every book scrutiny expected to see.” Senior leaders under pressure to demonstrate progress to inspectors pointed to visible, dated, colour-coded evidence in books as the easiest thing to show a visitor in thirty seconds. Some multi-academy trusts wrote it into whole-school policy as a way of standardising practice across dozens of schools at once. Individual examples of thorough marking, praised in specific lesson observations, got generalised into “this is what all marking should look like, in every subject, every week.” Once a few schools were doing it and being held up as exemplars, the practice spread through book scrutinies, learning walks and mock inspections, largely detached from any evidence that it actually helped pupils learn more than simpler approaches did.

By 2015, the government’s own workload surveys were showing that marking was one of the biggest single drivers of teachers’ unmanageable hours, and headteachers and unions were both raising it as a retention issue. In response, the Department for Education set up the Independent Teacher Workload Review Group, chaired by Dawn Copping, specifically to look at marking. Its 2016 report, Eliminating unnecessary workload around marking, is blunt about what it found: extensive, deep, or triple marking had little to no evidence behind it as a driver of pupil progress, and in many schools it had become, in the group’s words, a “time-wasting burden for teachers that has to stop.”

The report is also direct about where responsibility sat. It found that heavy marking regimes were frequently being driven not by pupils’ learning needs but by a wish to satisfy “other, mainly adult, audiences” — senior leaders, governors, inspectors — rather than to move a pupil’s understanding on. Crucially, the report traced this back to a misreading of what Ofsted actually required. Ofsted’s own position, restated clearly in the review, is that inspectors do not expect to see any specific frequency, type or volume of marking; how work is checked and responded to is for the school to decide through its own assessment policy, and inspectors are meant to judge only whether that policy is being followed, not whether it matches some unwritten gold standard of purple-penned dialogue.

That distinction matters enormously, because it means a great deal of the marking load teachers carried, and in some schools still carry, was never actually imposed from outside. It was a defensive practice, schools protecting themselves against an inspection expectation that didn’t exist in the form everyone assumed it did.

What the evidence says actually helps pupils improve

If triple marking wasn’t the answer, what is? The most useful single source here is the Education Endowment Foundation’s guidance report, Teacher Feedback to Improve Pupil Learning, published in 2021 after a systematic review of the international evidence combined with input from practising teachers and school leaders. It’s worth reading in full if you get the chance, because it deliberately avoids prescribing a single method, and instead sets out principles that hold regardless of whether the feedback ends up written, verbal, whole-class or one-to-one.

The report’s core message is that feedback only works when it sits on top of good teaching, not instead of it. Its recommendations, in broad terms, are:

• Lay the foundations for effective feedback through high-quality initial instruction and careful formative assessment, so you actually know what a pupil has and hasn’t understood before you comment on it.
• Deliver appropriately timed feedback that focuses on moving learning forward, rather than simply flagging what went wrong.
• Plan for how pupils will receive and act on feedback, because feedback that nobody reads, understands or has time to apply doesn’t do anything.
• Carefully consider how to give written feedback, choosing methods that are proportionate to what they add.
• Carefully consider how to give purposeful verbal feedback, which the evidence treats as equally legitimate to written feedback, not a lesser substitute for it.
• Design a school feedback policy that supports and exemplifies these principles, while leaving decisions about specific methods and timing to teachers’ own professional judgement.

Two things stand out from that list. First, specificity matters more than volume: feedback that tells a pupil exactly what to do next, tied to a clear learning goal, consistently outperforms generic praise or generic correction, however much of it there is. Second, the EEF is explicit that a policy should never mandate a single method of feedback delivery across every subject and every lesson, because what works for correcting a maths calculation is not the same as what works for developing an extended piece of historical writing.

It’s also worth knowing what feedback can achieve when it’s done well. The EEF’s wider Teaching and Learning Toolkit, which reviews evidence across dozens of classroom strategies, rates feedback as one of the highest-impact approaches available to teachers, associated with an average of around six additional months’ progress for pupils compared with those who don’t receive it, and at very low cost. That’s a genuinely strong effect. The catch, which the guidance report spends most of its length unpacking, is that the size of that effect depends entirely on the quality and specificity of the feedback, not on how many hours were spent producing it.

Whole-class feedback: the lower-workload alternative with evidence behind it

The practice that’s grown fastest in schools trying to follow both the DfE workload report and the EEF guidance is whole-class feedback, sometimes called WCF. Instead of writing an individual comment in every book, the teacher reads through a set of work, notes the common errors, misconceptions and standout examples of good practice on a single sheet or slide, and then delivers that feedback to the whole class at the start of the next lesson. Pupils then respond directly in their books: correcting a specific error, attempting a follow-up question, or redrafting a section, guided by what’s been shared.

Done well, this does several things that individual deep marking often doesn’t. It lets a teacher read a full set of books once, at a reasonable pace, rather than writing bespoke commentary thirty times over. It means every pupil gets feedback on the actual patterns in their own work, because the teacher has genuinely read it, rather than a rotating set of generic stock phrases. And because the feedback is delivered and then acted on in lesson time, pupils get a structured chance to do something with it immediately, which is exactly the “plan for how pupils receive and use feedback” principle the EEF guidance emphasises. Several secondary schools that have documented their move to whole-class feedback report a substantial cut in weekly marking time, without a corresponding drop in the quality of what pupils actually receive, and in some cases with clearer, more actionable comments than they were managing under a triple-marking regime.

None of this means whole-class feedback is automatically better in every context. A one-word spelling correction on a piece of extended writing might genuinely need an individual comment; a common misconception about equivalent fractions across an entire class is exactly what whole-class feedback is built for. The point isn’t to replace one rigid rule with another. It’s to match the method to what will actually help pupils improve, and to stop assuming that “individual and lengthy” automatically means “more effective.”

Verbal feedback has a real place, not a lesser one

Verbal feedback sometimes gets treated in school policies as the poor relation of written comments, something you do when you haven’t had time to “properly” mark. The evidence doesn’t support that hierarchy. Live, in-the-moment feedback, given while you’re circulating during independent work, has a real advantage that no written comment can match: it arrives while the pupil is still doing the task, which means they can act on it immediately rather than three days later when the exercise book comes back and the thinking behind their original answer has gone cold.

The EEF guidance treats verbal feedback as sitting on exactly the same footing as written feedback, subject to the same underlying principles. That means it needs to be specific rather than a passing “good work,” it needs to give the pupil something concrete to do next, and ideally it needs some form of record, even a quick tick, initial or annotation in the book, so there’s a trace that it happened and what it covered, both for the pupil’s own reference and for anyone checking books later. A quiet word to a pupil that says exactly what to change in their next paragraph, followed by thirty seconds watching them make that change, can do more for their learning than a lengthy written comment they skim and forget. What matters is the same thing that matters for written feedback: is it specific, is it timely, and does the pupil actually get a chance to use it.

Why more marking doesn’t equal more learning

It’s worth sitting with why the “more marking, more progress” assumption took hold, because the logic behind it feels intuitive even once you know the evidence doesn’t back it up. Marking is visible. It produces a physical artefact — a book full of comments — that can be shown to a line manager, a governor, a parent at parents’ evening. Progress in a pupil’s actual understanding is much harder to see directly, so schools reached for the thing they could see instead, and treated it as a proxy.

But a proxy is not the same as the thing itself, and the 2016 DfE review was explicit that quantity and quality had been conflated. Spending longer on a comment doesn’t make it more useful to the pupil reading it; if anything, very long comments are more likely to bury the one thing a pupil actually needs to act on under several things that don’t matter as much. There’s also a workload economics problem specific to deep marking: if the teacher is doing substantially more cognitive work on a piece than the pupil originally put into it, the incentives get skewed. Pupils learn that mistakes get quietly fixed for them, or that the “real” work happens in the second, teacher-guided pass, which can undercut the very independence and resilience that good feedback is supposed to build.

None of this is an argument for marking less carefully. It’s an argument for marking differently: reading pupils’ work closely enough to spot the pattern that matters, saying that clearly, and then getting out of the way so pupils have to do the thinking themselves. That’s harder to standardise into a tick-box policy than “write a comment in green pen every lesson,” which is probably part of why it took a national workload crisis to shift practice at scale.

Following your school’s marking policy without the guilt

Most schools have moved a long way from triple marking since 2016, but plenty of teachers still carry some residual guilt about not doing “enough,” particularly if they trained during the deep-marking years or moved from a school where it was still the norm. A few practical ways to reconcile a sensible policy with a clear conscience:

• Read your actual policy, not the version in your head. Many marking policies were rewritten after 2016 to explicitly allow whole-class feedback, verbal feedback and simple codes, but the memory of the old expectations lingers longer than the policy itself. Check what’s actually written down before assuming you owe more.
• Use a consistent shorthand for routine errors. A shared set of codes for spelling, punctuation or a recurring misconception lets pupils correct their own work with your guidance, which is often more useful for their learning than you writing the correction out in full.
• Batch your close reading, then deliver feedback live. Skim a set of books to identify the two or three things worth addressing, then spend the follow-up lesson on those points with the whole class, rather than writing the same comment thirty separate times.
• Keep a light evidence trail for verbal feedback. A quick initial, tick or one-line note in the margin is usually enough to show feedback happened and roughly what it covered, without needing a full written commentary to accompany it.
• Push back, calmly, with the evidence. If a line manager or book scrutiny asks for something beyond what your policy requires, both the EEF guidance and the DfE workload review are freely available, citable documents. It’s entirely reasonable to reference them in a conversation about what’s proportionate.
• Protect the time you save for the feedback that actually needs it. Whole-class and verbal feedback should free up time for the pieces of work — the mock exam, the extended essay, the piece going towards a moderated assessment — where detailed individual comment genuinely earns its keep.

One practical drag on all of this is how long it takes to create fresh checking material in the first place, whether that’s a quick low-stakes quiz to establish what’s actually gone in before you plan your next lesson, or a set of retrieval questions to structure a whole-class feedback session around. Having a ready-made bank of curriculum-aligned quizzes and checks, like the ones built into Learnaroo Hub, can take that particular piece of the workload off your plate, leaving more of your time for the feedback conversation itself rather than the business of assembling questions to have that conversation about.

The uncomfortable truth about the deep-marking years is that a huge amount of good, well-intentioned teacher time went into something the evidence never really supported, largely because it was easier to point at than to argue against. The upside is that the correction is now well established, evidenced from two directions at once: a government workload review that named the problem plainly, and an EEF guidance report that sets out what to do instead. You don’t need to choose between helping pupils improve and protecting your evenings. The evidence suggests that, done well, the first genuinely makes the second easier, not harder.