Design Learning That Produces Evidence
Why this matters
Attendance is not competence. Completion is not capability. A quiz score proves recall; a certificate proves presence. Neither proves that the person can decide better when the tool is in front of them and the stakes are real. If the learning experience ends with a score, the evidence is about the session, not about the work.
What you will be able to decide after this
- Whether a learning activity measures judgment or just attendance.
- What the real artifact for the activity is — the observable work that counts as evidence.
- Whether the activity can be passed without the capability actually being there.
Core lesson
Effective enablement requires people to produce artifacts, show their reasoning, verify outputs, and revise when the first version is weak. The same evidence standards from Core Judgment apply here: an output is only evidence when someone can look at it and tell whether it was done with judgment.
Design every significant learning experience so that the output is observable work, not a quiz score or a certificate:
- Produce a real artifact. A decision, a map, a contract, a verified output, a post-mortem — the same kind of thing the person will produce in the work, not a worksheet about it.
- Show the reasoning. The artifact carries the trail: the constraints, the verification performed, the uncertainty that remains.
- Verify the output. The participant checks their own work against the standards, and the checking is visible.
- Revise when the first version is weak. The second version is part of the evidence. A design that only accepts the first pass measures confidence, not judgment.
The test that matters: "Could someone pass this without demonstrating actual judgment?" If the answer is yes, the design is still weak — it can be gamed, which means it measures compliance, not capability.
Worked example
A one-hour "AI Fundamentals" workshop used to end with a multiple-choice quiz and a certificate. Redesigned: the quiz is gone. The participant brings a real piece of their own work — a document, a draft, a decision they actually made — and the session's artifact is a one-page constraint and verification record for it: the goal, the constraints set, the verification performed, what still cannot be verified, and the decision about whether the output is usable. The facilitator checks the record against the standards; a weak first version gets revised in the session, and the revision is part of what is assessed. Someone can now fail by writing a form-filling record with no actual checking — and the design catches that, because the verification step is performed, not described. Attendance is still taken. It is no longer the evidence.
Practice
Take one existing or planned learning activity you know. Write what it currently measures — attendance, recall, or judgment. Then write the artifact that would make it measure judgment, in one sentence.
Apply — produce the artifact
Take one existing or planned learning activity. Redesign it so that participants must produce a real artifact — decision, map, contract, verified output, or post-mortem — and perform verification. Remove any element that only measures attendance or recall.
Verify
Ask: "Could someone pass this without demonstrating actual judgment?" If the answer is yes, the design is still weak.
Sources
This module is original practice guidance based on the authoring standard and does not depend on a specific external factual claim. Editorial review is still required.