Education

Exam grades: how examiners decide your final mark

Rive News Research Desk

Published October 29, 2026 · 9 min read

By Rive News / Research desk | Rive News / Published October 29, 2026 at 9:00 AM PKT

Key takeaways

A grade starts with a mark scheme, not a fixed percentage: examiners award marks for evidence in your answer.
Grade boundaries are set after marking, using statistics and senior examiners' judgment of real scripts.
Moderation checks that different markers apply the same standard, so the same work earns the same grade.

Students write exams in a classroom, focused on their tests. Photograph: exam classroom

Image Credits: Pexels

An exam grade is the final label on a chain of decisions. It starts with a question paper, passes through a mark scheme and one or more markers, and ends with a boundary decision that turns a raw mark into a grade.

This article explains that chain, including mark schemes, moderation, absolute and relative grading, and grade boundaries. It focuses on large public exams, where the process is documented, and notes where classroom grading differs.

The mark scheme comes first

Before any script is marked, examiners write a mark scheme. It states what earns each mark: a correct fact, a valid method step, a developed argument, or a piece of evidence used well. For essay questions, the scheme usually describes levels of quality rather than a single right answer.

The scheme is tested before live marking begins. Senior examiners mark sample scripts, compare results, and refine wording where markers disagree. This step matters because a vague scheme produces inconsistent marks, however careful the markers are.

In most large exams, marking is anonymous. The marker sees the script, not the student's name or school. That rule exists so the grade reflects the work, not the writer's reputation.

How a raw mark becomes a grade

A raw mark is a count of marks earned on a paper. A grade is a band. The line between bands is called a grade boundary, and it is not fixed in advance.

This surprises many students. It is natural to assume that 70 percent always means the same grade. In public exams, the percentage needed for a grade moves from year to year and between subjects, because papers differ in difficulty.

OCR, one of England's exam boards, explains the approach as "comparable outcomes." Exam boards look at how the cohort performed in earlier assessments, build a prediction of what grades that cohort should achieve, and then set boundaries so results stay comparable with previous years.

If a paper was harder than intended, the boundary drops. If it was easier, the boundary rises.

Senior examiners also read real scripts near each proposed boundary and compare them with scripts from previous years at the same boundary. If the quality matches, the boundary stands.

If it does not, the boundary moves. Statistics inform the decision, but human judgment of actual work is part of it.

Who sets the boundaries

For each subject, an awarding committee is convened. Evidence given to a UK parliamentary committee describes the committee as a chair of examiners, a chief examiner, and principal examiners responsible for individual papers. The committee's task is to recommend boundaries that keep standards comparable from one year to the next.

The committee weighs two kinds of evidence. Quantitative evidence includes the mean mark on each paper and the cohort's prior attainment profile, for example GCSE results for A-level cohorts. Qualitative evidence is expert judgment of the quality of work in scripts near the boundary.

Once recommended, boundaries are inspected inside the exam board and signed off by its responsible officer. In England, the regulator Ofqual oversees the process. Boundaries are published on or shortly before results day, never in advance.

Moderation: keeping markers consistent

When many markers grade the same exam, consistency is a real risk. Two markers can read the same essay and award different marks, especially on open questions. Moderation is the system that reduces that risk.

The University of Toronto Mississauga describes a moderated grading session in which all graders meet to create or refine a rubric, mark sample papers, compare grades, and discuss criteria until they reach a consensus standard. Only when general agreement is reached does the group proceed to live marking.

Moderation does not end there. During marking, graders stay in touch about borderline cases.

After marking, a second marker or supervisor may re-mark a sample of scripts to check that standards held. Where a marker's grades drift from the agreed standard, scripts can be re-marked.

For coursework marked inside a school, moderation works slightly differently. The school marks the work, and the exam board samples it. If the sample shows the school marked generously or harshly compared with the national standard, the whole cohort's coursework marks can be adjusted.

Absolute grading versus grading on a curve

Two broad systems decide grades, and they produce different results. In absolute grading, grades are assigned against fixed criteria set by the teacher or board.

If every student meets the standard for an A, every student gets an A. If nobody meets it, nobody does.

In relative grading, often called grading on a curve, students are ranked against each other. The top share gets high grades, the middle gets moderate grades, and the bottom gets low grades, whatever the absolute quality of the work. A survey of more than 8,500 middle and high school students in Virginia, cited in LibreTexts' educational psychology text, supported the finding that teachers commonly mix both approaches in practice.

Relative grading has a fairness problem. An A in a strong class may represent better work than an A in a weak class, yet the label is the same. It can also discourage students from helping each other, since a classmate's improvement can push their own grade down.

Public exams in England use a hybrid. Criteria describe the quality expected, but comparable outcomes keep the share of each grade roughly stable year to year. That is why results rarely swing wildly even when a paper is unusually hard or easy.

Why two similar answers can earn different marks

Students sometimes compare answers with a friend and wonder why the marks differ. Several ordinary reasons explain it. The mark scheme may reward a specific piece of evidence or a method step that one answer included and the other skipped.

Presentation matters too, though less than students often think. An answer that states its reasoning step by step makes it easier for a marker to award each method mark. A correct final answer with no working may earn fewer marks than a wrong final answer reached by sound steps, depending on the scheme.

Level-based essay marking adds another layer. Two essays can make the same broad point, but one develops it with evidence and qualification while the other asserts it. The scheme places them in different levels, and the levels carry different mark ranges.

Classroom grades work differently

Classroom grades are usually set by a single teacher, often combining test scores, homework, and participation. The teacher designs the criteria, so two teachers teaching the same subject can weight the same skills differently.

That flexibility has a benefit. A classroom grade can reflect progress over a term, effort on coursework, and skills that a single exam cannot capture. It also has a cost: grades from different classrooms are harder to compare.

Some schools reduce that cost with common assessments, shared rubrics, and calibration meetings where teachers mark the same sample work and align their standards. The Edvocate's guide to rubrics and standardisation lists common assessments, calibration sessions, benchmark samples, and moderation as the four main tools. Each one trades a little teacher freedom for a little more comparability.

What students can actually control

The most useful fact for students is that boundaries are set after marking. Chasing a rumoured percentage target from last year is a poor strategy, because last year's boundary may not apply to this year's paper.

What students can control is evidence in the answer. Answer every question, even partially, since part marks exist in most schemes.

Show working in maths and science questions. In essays, link each claim to evidence rather than asserting it and moving on.

It also helps to read the mark scheme style for the subject. Past papers and published mark schemes show what examiners reward.

Practising with them is not gaming the system. It is learning the standard the grade will be judged against.

A proctor monitors an exam in a classroom while students take their tests. Photograph: exam classroom

Image Credits: Pexels

Frequently asked questions

Why is a grade 4 not always 40 percent?

Because boundaries move with paper difficulty. A hard paper has a lower boundary for the same grade, and an easy paper has a higher one, under the comparable-outcomes approach used by exam boards.

Who actually decides the grade boundary?

An awarding committee of senior examiners recommends it, weighing statistics and the quality of scripts near the boundary, and the exam board signs it off under the regulator's oversight.

What is moderation in marking?

Moderation is the process where markers align on a rubric, mark sample scripts together, and have their live marking checked, so the same work earns the same grade whoever marks it.

Is grading on a curve fair?

It ranks students against classmates rather than a fixed standard, so the same quality of work can earn different grades in different classes. That is why most public exams mix criteria with cohort-level stability instead.

Can a grade be changed after results day?

Yes, through a review of marking or a clerical re-check, depending on the board's rules. Reviews check whether the mark scheme was applied correctly; they do not re-mark from scratch against a new standard.

The bottom line

An exam grade is not a percentage plucked from a table. It is a raw mark, judged against a mark scheme, checked through moderation, and converted to a grade by boundaries set after marking with statistical predictions and senior examiners' judgment.

That system exists to solve a hard problem: keeping a grade meaningful when papers, cohorts, and markers all change every year. It is imperfect, and borderline cases will always feel harsh. But the chain of scheme, moderation, and boundary-setting is why a grade from one year can be compared, roughly, with a grade from another.

Tags: exam grades, grading, exams, mark schemes

The Rive News research desk writes plain-language explainers from published sources, with no invented claims or unattributed statistics.

Subscribe for one plain-language education explainer each weekday.

Rive News standards: every factual claim in this article traces to a named published source listed above; corrections are noted on the article page.

Taggedexam gradesgradingexamsmark schemes