How boards keep marking consistent.

Two careful examiners given the same script do not award the same total. Every board knows this, and the machinery they build around it — head examiners, spot evaluation, moderation — exists for that one reason. Most of it is invisible from outside, and one piece of it is worth copying.

The premise: marking varies, and everyone knows it

It is worth being clear that this is not a claim about careless examiners. It is the assumption the entire system is built on. If marking were reproducible, there would be no reason to sample it, no reason to adjust it afterwards, and no reason to put a senior examiner between the marker and the result. All three exist. What makes a handwritten paper hard to mark is the longer account of why.

The head examiner, and the one rule worth stealing

A head examiner is responsible for a group of assistant examiners. The instruction that matters is this: on the first day, the head examiner goes through the first five answer books marked by each examiner, before that examiner is given the rest of their allocation. Where the marking varies from the scheme, it is resolved by discussion, and the remaining scripts are released only once there is no significant variation left.

Read that again as a design. It is not a quality check at the end. It catches drift in the first five scripts of a batch of hundreds, when correcting it is still cheap, and it does so by comparison between markers rather than by inspection of one. There is very little in institutional marking that is this well thought out.

And almost no school or coaching centre reproduces it. That is the honest answer to why two teachers in the same staff room mark the same paper differently: not because either is careless, but because nobody ever compared their first five scripts and had the conversation.

Spot evaluation

Marking done at a designated centre, under supervision, inside a fixed window — as opposed to examiners taking scripts home. The point is that marking becomes observable and keeps to a timetable. A head examiner can actually sample work as it happens, and the discussion that resolves a variation can happen in the room rather than by post.

It also sets a workload. A board treats twenty answer books a day in a main subject, across an eight-hour day, as a full day's work for an appointed examiner marking to a scheme. Twenty-five in other subjects. That is a useful external number the next time somebody is handed a hundred and fifty scripts and a weekend.

Moderation, and why it is not normalisation

Moderation adjusts marks after checking so that different examiners, or different centres, are brought onto a comparable standard. It is a correction applied to the marker's behaviour, not to the student's work.

Normalisation is a different thing that gets called the same thing. It adjusts across different papers or sessions of the same examination, so a candidate who sat a harder shift is not disadvantaged. It belongs to multi-session entrance tests, not to school marking. Conflating the two is the most common mistake in any discussion of marks being "adjusted". Both terms, and the rest of the vocabulary.

What on-screen marking changed, and what it did not

CBSE moved Class 12 answer books to on-screen marking in February 2026; Class 10 stayed on physical scripts. The marking rules did not change. What changed is the examiner's physical relationship to the paper: they cannot hold it, cannot flip freely between pages, cannot tilt a faint pencil mark to the light, and cannot flick back to compare question four with question nineteen as quickly.

Those sound like small things. They are the conditions under which the judgements above are actually made, which is why a change in the medium shows up as a change in the marks even when every written rule has stayed the same.

What an institution can actually copy

Of the three mechanisms, two need a board's scale. Spot evaluation needs a centre and supervision. Moderation needs enough scripts for a statistical adjustment to mean anything.

The head examiner check needs neither. It needs one person to read the first five scripts from each marker and have a conversation before the rest go out. On a staff of four that is twenty scripts and an hour, once per paper, and it is the single highest-return thing an institution can do about marking consistency. The other decisions no board makes for you are worth settling in the same hour.

Common questions

What does a head examiner do?

Supervises a group of assistant examiners: briefs them on the marking scheme, samples the first five answer books each of them marks before releasing the rest of their allocation, resolves variation by discussion, and is the person an examiner refers an unusual answer to. Most of what makes marking consistent across a board happens at this level rather than at the individual examiner's desk.

What is spot evaluation?

Centralised marking. Examiners are called to a designated centre and mark under supervision within a fixed window, instead of taking answer books home. It exists so that marking is observable and finishes on schedule, and so that a head examiner can sample work while it is being done rather than afterwards.

Is moderation the same as normalisation?

No, and the difference matters. Moderation brings different examiners or centres onto a comparable standard for the same paper. Normalisation adjusts across different papers or sessions of the same examination. School marking uses the first; multi-session entrance tests use the second.

Can a school run its own moderation?

Not meaningfully — a statistical adjustment needs more scripts than a single school produces. But it can run the head examiner check, which is the mechanism that actually prevents the variation rather than correcting it afterwards, and which costs about an hour per paper.