How to give effective feedback on mock exams and assessments

Marking mocks is one of the more thankless jobs in teaching. You sit with a stack of scripts on a Sunday afternoon, write thoughtful comments, hand them back on Monday, and watch most students flip to the grade, glance at the comments, and put the paper away. Whatever you wrote often makes no measurable difference to the next piece of work.

The research on feedback is more interesting than that depressing pattern suggests. Black and Wiliam's Inside the Black Box (1998) and the EEF's feedback guidance reports agree on the central point: Feedback only changes learning if students actually do something with it. The most beautifully written comment is wasted if it does not trigger a behaviour change. The most effective feedback is often the simplest, provided you build in time and structure for students to respond.

This guide is aimed at teachers facing two or three mock cycles a year, plus regular assessed work in between. It covers what good feedback looks like at this scale, how to cut workload without cutting impact, and what to do in the lesson after mocks come back.

A few well-documented patterns explain why the standard approach often fails.

The grade-first problem. As soon as a grade appears on a script, students focus on it almost exclusively. Butler's 1988 study showed that students given grades alongside written feedback performed no better on a subsequent task than students given grades only, and worse than students given written feedback only. The grade swallows the comment.

The vague-comment problem. 'Add more detail' or 'develop your argument further' rarely tells a student what to do differently. Comments general enough to apply to half the class are usually too general to change any individual student's next piece of work.

The no-response problem. If students do not have a structured chance to act on feedback, the loop never closes. The next lesson moves on to new content and the feedback becomes a sunk cost. One of the biggest levers in feedback design is building in dedicated time for students to respond, not writing better comments.

The highest-leverage change to most feedback routines is whole-class feedback. Instead of writing the same comment thirty times in thirty different scripts, you read through the class set once, identify the patterns, and respond to them in a single lesson.

The process: As you mark, keep a single sheet open with four columns: Misconceptions, recurring errors, strong examples, and individual issues. As you read each script, add tally marks or anonymised quotes. By the time you have read the class set, you have a structured picture of what the class needs.

In the feedback lesson, you teach to the patterns. Project the recurring errors and re-teach the underlying point. Project strong examples and ask students to identify what makes them work. Provide a model answer for the trickiest question and ask students to compare it against their own. Whole-class feedback typically takes thirty to forty minutes for a class set, compared to two or three hours of individual marking, and the impact is often higher because the response is built into the lesson.

Whole-class feedback handles common patterns. Individual feedback handles cases outside them. Some students need a comment specific to their script, and folding their issue into a whole-class lesson would dilute both.

A workable rule of thumb: Use individual written feedback on perhaps three or four scripts per class set. A high-achieving student with a specific stretch point, a student whose script reveals an issue not seen elsewhere, and one or two students whose engagement or confidence would benefit from personal recognition. For everyone else, the whole-class response does the job.

When you do write individual comments, keep them short, action-focused, and tied to a specific next step. 'Your point about market saturation is well argued. Next time, link it explicitly to the diagram in question 4(b).' That sentence is more useful than a paragraph of general encouragement, and it gives the student something concrete to do.

The clearest signal feedback is working is that student behaviour changes between one piece of work and the next. The clearest signal it is not is that the same errors keep showing up. Designing feedback around the response rather than the comment is what closes the gap.

A simple structure for most subjects. Hand back scripts without grades attached, ideally with feedback codes or annotations rather than full comments. Project the model answer and the key feedback points. Give students fifteen to twenty minutes to do something specific: Re-write a paragraph using the model, redo a maths question with the corrected method, identify three lines from the source missed in analysis. Then, and only then, share grades.

This sequencing matters. If students see grades first, the response time becomes performative. If they see grades after the response, the lesson has done its work. Some teachers withhold grades until the next assessment, on the basis that grades on a single piece of work cause more anxiety than they generate useful information.

Writing the same comment in thirty scripts is a tax that produces little extra learning. A feedback code lets you mark a recurring issue with a single letter or symbol, while the meaning is held in a key students refer to.

A typical English code set might include S for spelling, EX for needs an example, AN for needs analysis, Q for embed the quotation more smoothly. A maths code might include UM for unit missing, WS for show working steps, CC for calculator misuse. Each code is paired with a brief instruction in a printed key the student keeps in their folder.

The pay-off is twofold. Marking gets faster, often cutting time by half on routine assessments. And students start internalising the categories. After a term, they begin to spot 'this paragraph needs an EX' without you writing it, which is exactly the metacognitive habit you want. Departments adopting a shared code set find consistency improves across teachers too.

Mocks deserve a slightly different approach than ordinary assessed work. Stakes feel higher, scripts are longer, and there is a window of weeks or months before the real exam in which behaviour change is most valuable. The temptation to write detailed comments is at its strongest here, and it is largely the wrong instinct.

A workable structure. After marking, produce a one-page sheet for the class showing the mark scheme breakdown, the most commonly missed marks, the questions where the class scored well, and three or four specific improvement targets. Hand back scripts with code annotations but no narrative comments. Provide the one-page sheet for context.

Use the feedback lesson to walk through one or two of the worst-performing questions in detail. Project a model answer. Ask students to compare it with their own. Then run a re-do, either of the same question with the model hidden, or a similar question from a different paper. The re-do is what converts feedback into changed behaviour. Without it, the lesson becomes a post-mortem rather than forward-looking.

For high-stakes assessments like the final mock before the real exam, a five-minute individual conversation can be more valuable than written comments, especially for students approaching grade boundaries. The conversation is direct, two-way, and lets the student ask the question they were too unsure to write down.

The practical issue is time. A class of thirty at five minutes each is two and a half hours. Two approaches help. First, spread conferences across two or three lessons of independent practice, conferring with a few students per lesson while the rest work. Second, prioritise. Students near grade boundaries, students whose performance dropped unexpectedly, and students needing a confidence boost are highest-value. Students whose work is steady may need less of your time.

Keep conferences structured. Start with the student's own reflection: What went well and what would you change? Follow with two specific points from your marking. Agree one concrete action for the next piece of work. Close with a check on confidence. Four to five minutes produces a clearer next step than a written comment.

Students often need to be taught how to respond to feedback, particularly if earlier years gave them feedback they were not expected to do anything with. The first time you build in a structured response lesson, expect some confusion about what to do.

A scaffold helps. Provide a written prompt sheet: Identify the three feedback points that came up most often in your script. Pick one and re-write the relevant paragraph or redo the relevant question. Compare your re-done version with your original and note what you changed. Set a target for the next piece of work that addresses one of the feedback points.

After two or three rounds, the prompt sheet becomes less necessary as students internalise the routine. Departments adopting a shared response template find this transfer happens faster, because students see the same structure in multiple subjects. The goal is for students to do this without prompting by the time they leave Year 11.

Sustainable feedback routines last; unsustainable ones collapse around half-term. The best feedback system is the one you can keep running in week six of a long term, not the one you can manage in week one.

A few honest observations. Whole-class feedback, once mastered, saves significant time without losing impact. Feedback codes save more time than most teachers expect, and the gains compound over a term. Detailed individual written comments are usually the lowest-impact and highest-cost option, despite often being the default. Conferences are time-expensive but high-impact for targeted students, so use them surgically.

If you are spending more than five hours marking a single class set of mocks, the system is probably wrong somewhere. The fix is usually structural (shift to whole-class feedback, introduce codes, narrow individual comments) rather than working faster. Departments that audit marking workload usually find significant savings without compromising outcomes.