Competency-Based Assessment: How CBT Learners Are Assessed

Competency-based assessment judges whether a learner can perform a task to the standard the workplace requires, and records the result as competent or not yet competent against written performance criteria. It replaces the end-of-course examination, which ranks learners against each other, with a body of evidence, gathered in more than one way, that shows what each learner can actually do. Done well, it is the part of competency-based training that gives a certificate its meaning.

What is competency-based assessment?

In competency-based training, the outcomes are defined first. Each unit of a programme is drawn from an occupational standard and states what a competent person must be able to do, under what conditions, and to what standard. Those statements are the performance criteria.

Competency-based assessment is the process of collecting evidence about a learner’s performance and comparing it with those criteria. The judgement is binary at the level of each criterion: met or not met. A learner is competent in a unit when all of its criteria are met. A learner who has not yet met them all is not yet competent, is told precisely which criteria remain, and is given further learning and another opportunity to be assessed.

Three features follow, and set it apart from most people’s experience of school.

• It is criterion-referenced. The standard is fixed in advance and the learner is measured against it, not against the rest of the class.
• It is evidence-based. The judgement rests on what the assessor has seen, examined or verified, not on an impression.
• It is continuous and repeatable. Assessment happens when the learner is ready, and a learner who is not yet competent can be reassessed rather than failed.

How does it differ from traditional examinations?

The differences are easiest to see side by side.

Ghana’s move from a multiplicity of testing regimes to standards-driven assessment on the Ghana TVET Qualifications Framework is, at heart, a move from the left column to the right. It is also the most demanding part of the transition, which is why our TVET assessment reform work so often starts here.

What counts as evidence of competence?

Because the judgement is evidence-based, the quality of an assessment depends on the quality and range of the evidence. Assessors typically draw on six types.

1. Direct observation of performance. The assessor watches the learner carry out the task in a workshop, on site or in a simulated environment, and records what is done against a checklist. This is the strongest evidence for practical competence.
2. Product evidence. The assessor examines something the learner has made or completed, such as a welded joint, a garment, a wired panel, a set of accounts or a ploughed plot, against the finish and specification the standard requires.
3. Questioning. Oral or written questions check the underpinning knowledge: why a step is done in that order, what would happen if a condition changed, which safety rule applies. Questioning also confirms that observed performance was understood, not copied.
4. Portfolio evidence. A collection of work samples, photographs, logs, drawings and reflections built up over time, each linked to the criteria it supports.
5. Workplace evidence. Records from an attachment or apprenticeship: job cards, supervisor sign-offs, customer feedback, production records.
6. Third-party reports. A structured statement from a supervisor or employer who has seen the learner perform, used to corroborate other evidence rather than to replace it.

A sound judgement almost always rests on more than one type. Observation plus questioning is the usual minimum for a practical unit; portfolio and workplace evidence broaden it and show that competence holds up over time and in real conditions.

What makes an assessment sound?

Four principles are widely used to test whether an assessment can be trusted. They are often shortened to VRFF.

Validity. The assessment measures what the standard says it should. If the criterion is “installs the fitting to manufacturer’s specification”, a multiple-choice test about fittings is not valid evidence; watching the installation is.

Reliability. Two trained assessors, or the same assessor on two days, would reach the same judgement from the same evidence. Reliability comes from clear criteria, good instruments and assessor training, and it is what verification is designed to protect.

Fairness. Every learner knows in advance what will be assessed and how, has an equal chance to demonstrate competence, and can appeal a decision. Fairness includes reasonable adjustments for learners with disabilities, provided the standard itself is not lowered.

Flexibility. Evidence can be gathered in different ways, at different times and in different settings to suit the learner and the context, as long as validity and reliability are preserved. Flexibility is what makes recognition of prior learning possible.

When an assessment system is weak, it is usually one of these that has slipped, and naming which one is the quickest way to fix it.

Which tools do assessors use?

Good assessment is planned, not improvised. Three tools do most of the work.

Assessment plans. For each unit, a plan sets out which criteria will be assessed, by which methods, where, when, with what resources, and who will assess. It is agreed with the learner beforehand, which is itself a fairness requirement.

Observation checklists. A list of the observable actions and outcomes that correspond to the performance criteria, with space to record what was seen and whether each item was met. Checklists make observation consistent across assessors and are the backbone of reliability.

Rubrics and marking guidelines. Where the standard involves quality of outcome rather than a simple yes or no, a rubric describes what “meets the standard” looks like for each criterion, so that assessors judge finish, accuracy or tolerance in the same way. In Eljen’s methodology these instruments and marking guidelines are developed in Phase 2: Programme Development, alongside the unit specifications and learning materials, and validated with industry in Phase 3: Stakeholder Validation, so that they reflect real workplace expectations rather than a facilitator’s preference. They are also part of what CTVET typically expects to see during programme accreditation.

Who does what: assessor and verifier

Two roles keep the system honest.

The assessor plans the assessment with the learner, gathers and examines the evidence, makes the competent or not yet competent judgement for each criterion, gives feedback and keeps the records. Assessors need to be competent in the trade themselves and trained in assessment practice.

The internal verifier, sometimes called a moderator, checks the assessor rather than the learner. They sample assessment records and evidence, observe assessors at work, compare decisions across assessors and cohorts, and confirm that the instruments and judgements meet the standard. Where they find drift, they act before it reaches a certificate.

External verification is the same check carried out on behalf of the awarding or regulatory body, sampling across institutions so that a certificate means the same thing wherever it was earned. Internal and external verification together form the feedback loop that the quality-assurance system of an institution depends on.

What is recognition of prior learning?

Many skilled people in Ghana and across Africa learned their trade through apprenticeship, informal work or on-the-job experience, and hold no certificate. Recognition of prior learning (RPL) applies competency-based assessment to competence that already exists.

The candidate is assessed against the same occupational standard as a taught learner, using the same principles and the same types of evidence: a portfolio of past work, third-party reports from employers, observation of performance and questioning. Units in which the candidate is competent are credited; the candidate then trains only for the units that remain. RPL is only possible because the standard, not the course, is the reference point. It is one of the strongest arguments for building qualifications on occupational standards generated with industry.

Common mistakes to avoid

Institutions moving to competency-based assessment tend to make the same errors.

1. Assessing knowledge only. A written test is given, a mark is recorded, and the unit is signed off. Nothing has been observed. The certificate now claims a competence that was never demonstrated.
2. One-shot examinations. Assessment is scheduled once, at the end, for everyone, regardless of readiness. Learners who are competent but nervous fail; learners who are not yet competent have nowhere to go.
3. No reassessment route. A “not yet competent” result is treated as a fail with no feedback on which criteria were missed and no second attempt. This removes the whole point of the approach.
4. Instruments that do not match the criteria. Checklists written from memory of the course rather than from the unit specification, so that what is assessed is not what the standard requires.
5. Assessors marking their own teaching without verification. Without an internal verifier, standards drift unnoticed, and every assessor’s judgement becomes a private one.
6. Records that cannot be traced. Judgements are recorded, but the evidence behind them is not kept or cannot be linked to a learner and a criterion, so a verifier or accreditation reviewer has nothing to examine.
7. Converting to marks. Criterion results are quietly turned into percentages “for the parents” or “for the transcript”, reintroducing ranking through the back door.

Each of these is fixable, and most are a matter of training and instruments rather than resources; see our competency-based assessment services. If your institution is reforming its assessment, or preparing assessors and verifiers for the first time, talk to us about where to start.