Learning delivery workshop / 6 MIN READ
Let a learner review completed material without issuing a new result
Build a cmi5 launch-mode acceptance sheet that checks allowed result effects, not just whether the course opens.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources checked 2026-11-07
Reopening completed learning can be a review activity rather than a new assessed attempt. The launch mode must preserve that purpose so exploratory clicks do not silently create a fresh success or failure claim against an existing assignment.
What you will leave with: Build a cmi5 launch-mode acceptance sheet that checks allowed result effects, not just whether the course opens.
Use this when
- A cmi5 course supports browsing or reviewing material.
- Revisiting content unexpectedly changes reported learning outcomes.
Not deleting failed attempts, concealing criticism or selecting the best grade.
Read the mode supplied for the launch
The cmi5 Quartz specification’s launch-mode section, checked on 7 November 2026, distinguishes Normal, Browse and Review. In Browse and Review, satisfaction-related data must not be recorded; initialized and terminated remain required cmi5-defined statements. This lesson applies to Quartz with its referenced xAPI 1.0.3 contract, not every LMS preview feature.
Ask the integration owner to identify the actual launch mode in each intended journey. A button labeled review is only interface wording until the underlying launch context supports that purpose. Record the learner’s intended action and the mode supplied to the assignable unit. Keep the prior result as a baseline so the acceptance test can detect an unintended outcome change rather than merely confirming that the page displayed.
Create contrasting mode fixtures
Prepare synthetic cases for a normal assessed launch, a browse launch and review of already completed material. Use the same distinctive interaction where appropriate so differences in result effects can be attributed to the mode rather than a different exercise. State what the learner may explore and which prior outcome should remain intact.
Do not assume that suppressing visible feedback is sufficient. The course may still emit a result behind the interface. Conversely, an absence of a new pass or fail does not mean the launch produced no records at all. Have the implementation owner inspect the allowed lifecycle evidence and outcome effects under the profile. Keep that inspection separate from ordinary event retry or duplicate-delivery diagnosis.
Check both the content behavior and the reporting effect
Follow the learner through opening, interacting and leaving in each mode. Does review let the person revisit the material as promised? Does the normal path still produce the intended assessed outcome? A repair that suppresses results for every launch would protect review only by breaking ordinary assessment.
Compare the baseline outcome with the post-session record and any newly issued claims. If review changes satisfaction improperly, identify whether the wrong mode was supplied or the content ignored the correct mode. Do not erase the new record as the first response or rewrite a failure into a pass. Use isolated test records to locate the defect, then ask the responsible owner for the supported correction and a separate plan for any real affected history.
Explain the boundary before someone reopens the course
Make the learner-facing choice clear: revisiting material is different from undertaking a newly assessed enrollment or attempt. If the service offers only one of those routes, state the limitation instead of presenting a misleading review button. Do not imply that review grants renewed certification, extends an expired qualification or satisfies a new assignment.
Retain the mode sheet with the course revision, integration version and profile scope. Include a normal-path regression test whenever review behavior changes. Recheck the journey from the transcript as well as from the course catalog if both exist, because their labels may conceal different launch handling. Acceptance means each supported mode has the intended learning-record effect while preserving the prior result where required, not that all interactions are unrecorded or that any past outcome can be hidden by reopening content differently.
Make the reasoning concrete.
A fictional brand-guidelines course offers a review link after completion. A synthetic learner revisits a question and selects a wrong option; the content mistakenly sends a new failure claim as though the review were a normal assessment.
Delete the inconvenient result
Opening: Remove the new record and leave the launch path unchanged.
Hypothesis: The summary looks correct again.
Risk: The underlying mode defect can recur for the next reviewer.
Test the mode contract
Opening: Compare supplied launch mode, content behavior and resulting claims in isolated normal and review fixtures.
Hypothesis: The team can correct the cause while preserving legitimate assessment behavior.
Risk: Any real affected records require a separate authorized remedy.
The team fixes the test integration and verifies both review protection and normal assessment, without deleting or rewriting production learning evidence.
Take the questions into your next brief.
Use these prompts in your own notes. Nothing is submitted from this page.
Purpose
Is this journey browsing, review or a normal assessed launch?
Mode
What mode does the integration actually supply?
Baseline
Which prior outcome should remain unchanged in this fixture?
Effect
What lifecycle and outcome evidence appears after the session?
Regression
Does the normal assessed path still report its intended result?
Watch for the wrong conclusion.
- Trusting a review button label without checking launch mode.
- Assuming review means no records of any kind.
- Suppressing every result or deleting evidence instead of fixing the mode mismatch.
Your next action
Compare one synthetic normal launch with one review launch, preserving the prior result as the review baseline and inspecting their different outcome effects.
Sources & scope
Workflows, worksheets and fictional examples are original editorial analysis, not supplier implementation instructions or measured business results. Sources support only the specifically attributed platform facts. Supplier controls, contracts and local requirements can change; verify the actual placement and current provider guidance before implementation.
- AICC: cmi5 Quartz specification, launchMode →Primary technical specification · Supports: Supports Normal/Browse/Review satisfaction and lifecycle boundaries under Quartz; the mode acceptance sheet is editorial.