What Is a Grade Boundary?

Why the number cannot be decided in advance

Two papers written to the same specification are never identical in difficulty. Question wording, the combination of topics, how much reading each question demands — all of it moves a little from one series to the next, and no amount of care in writing the paper removes the variation entirely.

If boundaries were published in advance, that variation would land on the students. A slightly awkward paper would depress a whole cohort's grades, and an accessible one would inflate them, so a grade 6 would mean something different every June. Setting the boundary afterwards puts the variation where it belongs: on the number, not on the candidates.

The decision is made at an awarding meeting once marking is complete. Senior examiners look at scripts sitting around the candidate marks under consideration alongside statistical evidence about the cohort, and place the boundary where the work matches the standard from previous years. The principle behind that statistical evidence is usually summarised as comparable outcomes: a cohort of similar ability should get broadly similar results, so a shift in results has to be explained by something other than how the paper turned out.

What moves, what does not, and what is not comparable

The grade is set on the total across all the papers in the qualification, not on each paper individually. A weak Paper 1 followed by a strong Paper 3 is arithmetic rather than a failed component, and boards that publish per-paper figures publish them for information — the grade comes from the sum.

Boundaries are not comparable across anything. Different tiers have different boundaries for the grades they share. Different boards have different boundaries for the same subject, so an Edexcel figure tells you nothing about AQA. And boundaries in reformed GCSEs in England are given in raw marks, not in the uniform marks that older modular qualifications used, so old advice about UMS does not transfer.

Using last year's boundaries without being misled by them

Published boundaries from previous series are genuinely useful for setting a target, and genuinely dangerous when treated as a promise. They tell you roughly what a grade has cost recently. They cannot tell you what it will cost this time, because this time's paper has not been marked.

The sensible use is to take the recent figures for your board and tier, look at the range across the last few series rather than a single year, and aim comfortably above the top of it. A student who trains to hit last year's grade 7 boundary exactly has left themselves no margin for the one thing nobody can predict.