Use the audit to protect learning from activity metrics
This audit decides whether a proposed practice loop is ready, needs revision, or should stop. It does not award points for hours, exercises, or output. A hard failure—no target behavior, no retest, inaccessible task, unsafe feedback, or workload that requires unpaid overextension—cannot be averaged away. Gather the draft exercise, rubric, feedback plan, transfer task, learner context, and time allocation before answering.
The independent synthesis in How People Learn II supports attention to prior knowledge, context, feedback, and metacognition. Those elements guide the audit, but they do not certify one local exercise. Record evidence and uncertainty beside each check. Approval should expire when the role, task, learner needs, rubric, or work environment changes.
Evidence: National Academies of Sciences, Engineering, and Medicine
Checks 1–3: target and task validity
Check 1 asks whether the target is an observable decision or behavior rather than an identity or broad aspiration. Check 2 asks whether the exercise actually elicits that target without unrelated production work obscuring it. Check 3 asks whether difficulty is challenging but feasible and whether the task resembles the conditions where the skill matters. Write one reason the exercise could give a false impression of competence.
Test the design with a counterexample. If a learner can complete the task successfully without using the intended skill, the exercise fails Check 2. If only an expert can understand the instructions, it may fail Check 3. CMU’s active-learning strategies offer useful task forms, but selecting a technique is not enough; the form must make the specific reasoning visible.
1. Target is expressed as observable behavior.
2. Task completion requires the target behavior.
3. Difficulty and conditions represent the intended use.
Evidence: National Academies of Sciences, Engineering, and Medicine; Carnegie Mellon University Eberly Center
Checks 4–6: criteria, baseline, and access
Check 4 requires a small rubric with behavioral anchors available before judgment. Check 5 preserves an uncoached baseline plus task conditions, time, and tools. Check 6 separates access needs from the substantive criterion and offers an appropriate way to engage with materials and demonstrate performance. Removing support that enables access is not a legitimate test of independence.
CAST’s guidelines support multiple means of engagement, representation, and action or expression. Apply that lens by asking whether the format introduces vision, hearing, motor, language, timing, or participation barriers irrelevant to the target. Keep legitimate assistive technology recorded and available across comparisons. If a mode itself is the skill—for example, spoken facilitation—state that boundary rather than pretending the criterion is format-neutral.
4. Rubric anchors point to visible evidence.
5. Baseline and conditions are preserved before feedback.
6. Access support is appropriate and consistently recorded.
Evidence: Carnegie Mellon University Eberly Center; CAST
Checks 7–10: feedback and correction integrity
Check 7 verifies that the feedback source has enough perspective or expertise for the target. Check 8 requires comments to identify a trace and consequence rather than preference alone. Check 9 ensures the learner selects and performs the correction. Check 10 uses a parallel immediate task so a rewritten answer or memorized example is not misread as learning.
Inspect power and timing. The learner needs a safe way to question feedback, and the comment must arrive before the next relevant attempt. An expert silently replacing the work may improve the artifact while weakening the learning signal. A peer consensus can still be wrong. Preserve disputed feedback and state why one correction was chosen. The goal is an interpretable cycle, not agreement at any cost.
7. Feedback source fits the knowledge and perspective required.
8. Comment links criterion, trace, and consequence.
9. Learner owns the correction decision and execution.
10. Immediate retest is parallel, not an answer replay.
Evidence: National Academies of Sciences, Engineering, and Medicine; Carnegie Mellon University Eberly Center; Scrum Guides
Checks 11–14: retention, transfer, and sustainability
Check 11 schedules a delayed attempt after the immediate rehearsal. Check 12 changes at least one surface condition while holding the target stable. Check 13 records prompts, assistance, contradictory cases, and the decision that follows. Check 14 confirms practice time is inside a sustainable workload and will not displace recovery, essential work, or participant safeguards.
The Scrum Guide’s inspection-and-adaptation pattern is relevant to the final decision: continue, raise difficulty, change feedback, redesign the task, or close the target. It does not establish the right delay or transfer standard. Choose timing proportionate to the work and label it provisional. If the learner succeeds only with a prompt, record partial support rather than converting it to a pass or fail.
11. Delayed retest has a date and owner.
12. Transfer task varies context without changing the criterion.
13. Assistance and contrary evidence remain in the log.
14. Practice fits paid or protected capacity and recovery needs.
Evidence: National Academies of Sciences, Engineering, and Medicine; CAST; Scrum Guides
Close failures with an owner and dependent recheck
For each failed item, name the repair, responsible person, evidence required, and date. Re-run dependent checks: changing the target invalidates the task and rubric; changing the format may require a new access review; changing the feedback source may alter safety and timing. Archive the old audit so later reviewers can see why the plan changed. Set a full review no later than 2027-02-10.
Passing the fourteen checks means the loop is sufficiently specified to try; it does not demonstrate retained skill or predict job performance. Begin with a limited cycle and preserve an unsuccessful result. If transfer evidence remains absent, do not compensate by counting more content. Redesign the observation or conclude that a workflow aid, collaboration pattern, or structural change is more appropriate than further individual practice.
Evidence: National Academies of Sciences, Engineering, and Medicine; CAST
Sources and further reading
These references informed this article. A source supports a claim; it does not imply endorsement of TenMultigure or any future product reference.
- How People Learn II: Learners, Contexts, and CulturesNational Academies of Sciences, Engineering, and Medicine · Accessed August 10, 2026
Provides the evidence context behind checks for prior knowledge, feedback, metacognition, delay, and transfer, while the audit preserves uncertainty about any individual result.
- Concrete Strategies for Active LearningCarnegie Mellon University Eberly Center · Accessed August 10, 2026
Contributes active-learning, worked-example, peer, and self-assessment practices used to test whether an exercise makes the selected reasoning visible.
- CAST Universal Design for Learning GuidelinesCAST · Accessed August 10, 2026
Supports the explicit access review and appropriate ways of engaging with or demonstrating a capability without lowering the substantive behavioral criterion.
- The Scrum GuideScrum Guides · Accessed August 10, 2026
Informs the final inspect-and-adjust decision after a cycle; the audit does not equate practice planning with Scrum or infer delivery performance.
Reviewed by TenMultigure Editorial Team. See an error or a source that has changed? Tell the editorial team.
Review method: AI-assisted desk research with editorial checks. Reviewed ; next scheduled review . Rebuilt TM-215 as a fourteen-check audit covering task validity, accessible baseline, feedback integrity, learner-owned correction, delayed transfer, sustainable workload, and dependent rechecks.