How to run standardisation meetings, and why regulators expect them

What standardisation is

Standardisation is assessors meeting to compare their judgements on the same piece of work against the same criteria so that marking is consistent. It exists to answer one uncomfortable question: if two of your assessors marked the same evidence, would they agree? If not, a learner's result depends on who happened to mark them, which is the unfairness quality assurance is meant to remove.

Regulators expect standardisation because consistency is the promise behind every qualification. A grade means something only if it means the same thing regardless of the assessor.

How to run one

Run a standardisation meeting by having assessors mark the same sample independently, then compare and resolve the differences against the criteria. The independent marking is the part people skip and the part that matters, because if assessors see each other's decisions first they converge on the loudest voice rather than the criteria. Blind marking exposes the real spread of judgement.

Once the marks are in, work through the differences one at a time. For each disagreement, go back to the criteria and the evidence, decide the correct judgement, and note why. The output is a shared understanding of the criteria that assessors carry into their own marking.

How often to hold them

Hold standardisation regularly and always at the points where inconsistency is most likely, rather than on a fixed calendar alone. Run one at the start of a new programme, when a new assessor joins, when the criteria change, and periodically through delivery so drift does not creep in. A single meeting a year is not standardisation, it is a ritual.

Match the frequency to the risk, the same way you match sampling to risk. High stakes assessment, new teams and new material need standardising more often. Settled delivery needs it less, but never zero, because assessors drift apart quietly.

What to record

Record enough of each standardisation meeting to prove it happened and show what was agreed, because an unrecorded meeting counts for nothing when a verifier asks. The record does not need to be a transcript. It needs the sample, who attended, where judgements differed, how each difference was resolved, and the agreed position assessors will now apply.

• The date, the attendees and the sample of work used
• Where assessors initially disagreed, and by how much
• How each difference was resolved against the criteria
• The agreed interpretation assessors will now apply
• Any action, such as re-marking or feedback to an assessor

New assessors and standardisation

Bring every new assessor through standardisation before you rely on their decisions, because a new assessor is the single biggest source of inconsistency in a team. Do not assume a qualified assessor will mark the way your team does. Have them mark the same sample and see where their judgement sits against the agreed position.

This protects the learner and the assessor alike. It gives the assessor a clear picture of what a pass looks like in your context, and it gives you evidence that you inducted them properly, which an external verifier checks when a new name appears on decisions.

Turning standardisation into evidence

Turn every standardisation meeting into evidence by keeping the record and linking it to the marking that followed. When a verifier asks how you know your assessors mark consistently, a run of records resolving real disagreements against the criteria answers directly.

Standardisation also feeds your wider system. Findings can trigger re-marking, feed an assessor's feedback, and update your quality assurance records. Run alongside internal verification, it is how a provider shows its marking genuinely means what it says.