Assessment holds up when it tests the stated outcome, marks against explicit criteria, and produces the same decision from two assessors on the same evidence. That last point is the one most providers underdo: if two assessors would grade the same work differently, the assessment is not defensible, however good the questions look. Design for that consistency from the start.
What makes assessment defensible
Assessment is defensible when it does three things: it tests the outcome the course claims, it marks against criteria written down in advance, and it yields the same decision from two competent assessors on the same evidence. Those three together are what a regulator and an external verifier actually check. Miss any one and the result cannot be trusted.
The common failure is an assessment that looks rigorous but tests the wrong thing, or one that leaves too much to the marker's judgement. Both are fixable at the design stage and expensive to fix afterwards.
Match the method to the outcome
Choose the assessment method to fit what the outcome asks for: knowledge, skill or behaviour. A knowledge outcome can be tested with questions; a skill outcome needs the learner to perform the skill; a behaviour outcome needs evidence of the behaviour in a realistic setting.
So read the verb in the outcome and let it choose the method. If the outcome says apply or demonstrate, a recall test cannot assess it, because recall is not application. Matching the method to the outcome is the biggest thing that makes assessment valid.
Write marking criteria before you deliver
Write the marking criteria before the first cohort sits the assessment, not after the scripts come in. Criteria written in advance say what a pass looks like, so the marker knows the standard before they mark and the learner could, in principle, know what is expected. Criteria written after the fact tend to fit the scripts rather than the standard.
Good criteria are specific enough that two markers reading them would apply them the same way. Vague criteria like good understanding push the decision back onto the marker's gut.
Design for consistency between assessors
Design the assessment so different assessors reach the same decision, because consistency is what makes a result mean anything. Explicit criteria are the foundation, and standardisation is how you keep it: assessors meeting to compare their judgements against the same criteria on the same sample of work, so drift is caught and corrected.
Build standardisation into the delivery model, not just the design, and record that assessors did it. This is where internal quality assurance and internal verification meet the assessment: sampling assessors' decisions is how you prove the criteria are being applied consistently, not just written down.
Keep the assessment valid and current
Keep the assessment valid over time by reviewing it against the outcome and the current subject matter, and version it like the rest of the course. An assessment that was valid at launch can drift out of date as the field moves or as learners share past questions.
So build a review point into the assessment and put a version and date on it, so you can show which learners took which version. This also protects integrity: refreshing tasks keeps them from becoming a memory test of last year's paper.
Build the evidence trail
Record the assessment so each result can be traced from outcome to decision, because the record is what an audit inspects. For each learner you want the attempt tied to the outcomes, the result, and the assessor decision with a name and a date against it.
Decide where these records live before delivery, not after. In Malta this sits inside the internal quality assurance system that MFHEA expects; in England it is what an awarding organisation's external verification checks against. Our course development service builds the assessment and the trail together.
Common assessment design faults
Most assessments that fail scrutiny fail for a short list of reasons, and every one of them is avoidable at the design stage. Check your assessment against these before it goes live.
- The method does not match the outcome, so it tests the wrong thing
- Marking criteria are vague, so two assessors would disagree
- No standardisation, so assessors drift apart over time
- The assessment is never reviewed, so it goes stale
- No evidence trail, so results cannot be traced or defended
Common questions
What is valid assessment?
Valid assessment tests the outcome the course actually claims, using a method that matches what the outcome asks for. A knowledge outcome can be tested with questions, but a skill or behaviour outcome needs the learner to perform, so a recall test cannot validly assess an apply or demonstrate outcome. Validity is what every other quality rests on.
What are marking criteria?
Marking criteria are the written statements that say what a pass, and each grade, looks like, set out before the assessment is marked. Good criteria are specific enough that two markers would apply them the same way. Vague criteria like good understanding push the decision onto the marker's judgement.
How do you keep marking consistent?
You keep marking consistent with explicit criteria plus standardisation. Standardisation is assessors meeting to compare their judgements against the same criteria on the same sample of work, so drift is caught and corrected. Sampling assessors' decisions through internal verification then proves the criteria are applied the same way in practice.
What records should assessment produce?
Assessment should produce, for each learner, the attempt tied to the outcomes, the result, and the assessor decision with a name and a date. Decide where these records live before delivery, because reconstructing them afterwards is where providers come unstuck.
Can Apollo design our assessments?
Yes. Apollo Accreditation designs assessments and marking criteria for training providers in the UK and Malta, built to test the outcome, mark to explicit criteria and produce a defensible evidence trail. Book a scoping call and we will map your outcomes to an assessment approach and the records it needs.
Planning a programme, an accreditation route or a quality system?
Book a 30 minute scoping call. We will map the route, the standards that apply and a realistic timeline. No preparation needed.
Book a 30 minute scoping callPrefer to write first? Contact us.