Get the Offline ArchiveFree lifetime updates

Assessment Methods

5 min read

Judge an apprentice mainly by watching them do real work, supported by a teaching test, a problem-solving test and peer review, and put at least two assessors on every serious decision, with one of them outside the master-apprentice pair. A system that lets the master alone pass or fail the apprentice can be abused in both directions, so give every failure written reasons, a date to try again and a way to ask a second panel to look.

Part of Skill Progression Framework

How apprentices are tested: the four methods in the carrier article, who should judge, and how to stop assessment itself becoming a tool of favour or control.

Four Methods, One Of Them Primary

The carrier article lists four ways to assess an apprentice:

  • Practical demonstration. The apprentice performs the skill while the master observes. This is the primary method.
  • Teaching test. The apprentice teaches a technique to someone else. Clear teaching shows deep understanding.
  • Problem-solving test. An unusual scenario, such as “this board is warped, how do you work with it?”, tests adaptability and not only procedure-following.
  • Peer review. Other practitioners evaluate the work, and multiple perspectives catch blind spots.

The corpus’s Assessment Systems article agrees that watching someone do the task is the most direct measure, and the Oral Examinations article describes how a conversation alone can examine understanding with nothing to write on. The Competency Testing article adds that the candidate should be able to perform the task, explain it at any step and say what could go wrong, because a candidate who can only do it has skill without understanding, and one who can only explain has theory without practice.

Who Watches Matters As Much As What Is Watched

The Competency Testing article is firm on three points:

  • The panel needs at least one person of master-level competence, since a panel of non-experts cannot judge high-level performance.
  • The evaluator should not be the apprentice’s primary trainer. The trainer wants the trainee to pass, and is too familiar with their particular strengths to be objective. The trainer can attend and give context but not decide.
  • The panel should be at least two people for any non-trivial certification, since a single evaluator brings too much personal bias.

When only one person in the community has mastery, the same article offers three options: invite an evaluator from a neighbouring community, let the sole master assess but require written reasoning that others can scrutinise, or convene a panel of senior practitioners from other fields to judge the observable outcome.

Assessment Can Be Abused In Both Directions

The corpus recognises that a master can exploit an apprentice as cheap labour without teaching, and that a trade can restrict its own numbers. Assessment is where both happen quietly:

  • A master who never passes the apprentice keeps them as cheap labour. The Labour apprenticeship systems article says an apprentice doing the same simple tasks for two years is a sign that something is wrong.
  • A master who passes everyone hands out credentials that carry no meaning.
  • A trade that fails outsiders and passes its own protects a monopoly.
  • A hostile or incompetent assessor can damage a good apprentice.

The Competency Testing article suggests a check on the trainer: if several graduates of the same trainer keep failing at the same skill, the trainer’s method needs correcting. The same logic can be applied to assessors by comparing pass rates for insiders and outsiders.

What You Need

  • A panel of at least two, including one from outside the pair.
  • A written standard for each milestone (Milestone Definitions).
  • A form to record the task, conditions, observers, result and reasons.
  • An appeal route, which can be the community’s dispute process (Dispute Resolution).
  • A record in the registry (Record Keeping).

Method: Running An Assessment

Step 1: Announce the task and standard beforehand. Surprise tests measure nerves.

Step 2: Observe the performance without coaching. The panel writes notes separately.

Step 3: Ask questions during and after: why this step, what could go wrong here, what would you do if.

Step 4: Panel members decide separately, then compare. If they disagree, add a third assessor and record both views.

Step 5: Record the result with reasons. A pass names what was shown; a fail names what was missing and when the candidate may try again.

Step 6: Treat safety-critical steps as all or nothing. A candidate who skips a safety step fails the milestone however good the rest.

Step 7: Review the pattern yearly: pass rates by trainer, by assessor and by family background.

Common Mistakes

  • Letting the master test alone.
  • Giving a failure with no reason. The apprentice cannot improve and cannot appeal.
  • Using the same panel of friends every time.
  • Testing only what is easy to test.
  • Coaching during the performance.

What’s Next