Show what the judgment rests on
L-18’s assignment says, “Maya will send the revised schedule by Friday.” It identifies an owner, task and deadline. It says nothing about handling a blocked task. Under the exercise rubric, the first three elements are present and the escalation element is not evidenced. The missing element does not prove L-18 never escalates problems in real work.
Write the criteria before reviewing the cohort. For each one, describe what evidence would count and what would remain uncertain. Avoid a vague “good communication” score whose meaning changes between reviewers.
Preserve the submitted work. A suggested summary should point back to the actual words or observed behavior. If a reviewer changes the judgment, keep the corrected judgment and its reason; do not rewrite the participant’s source to match the category.
Use several sources without blending their meanings
A participant may love the course while struggling with the assignment. Another may dislike the format but demonstrate the skill. Keep experience feedback, learning evidence and later practice as separate findings that can inform different decisions.
Open comments can help explain difficulties. “The example assumes I manage a team” may suggest a relevance problem for individual contributors. Treat that as a theme to investigate with other evidence, not proof that the entire course is ineffective.
For multi-language work, retain the original and review the interpretation. The same criterion should not become stricter because one participant writes less fluently in the delivery language. Decide whether language proficiency itself is part of the intended skill before allowing it to affect judgment.
Make review useful for the next attempt
Ask two reviewers to independently assess a small fictional set, including an incomplete answer and a borderline example. Discuss disagreements and improve the criterion wording. Agreement is a useful check, not proof that the rubric measures the right construct.
If configured AI analysis prepares a score, require the criterion, cited evidence and uncertainty alongside it. Test invented quotations, unsupported positive judgments and missed qualifications. Route unresolved cases to a named reviewer rather than accepting every output because it arrived quickly.
Return feedback that helps someone act: “Add who to contact when the schedule is blocked” is more useful than “3 out of 4.” Track the next attempt as new evidence. Over time, this creates a reviewable learning history instead of a pile of marked files.
Work through the Bridge example
This is a fictional teaching example. Adapt the fields and rules to the question your own training program needs to answer.
| Criterion | Evidence in the assignment | Useful feedback |
|---|---|---|
| Task and owner | Maya sends the revised schedule | Present under the exercise rule |
| Deadline | By Friday | Present; clarify time if needed |
| Escalation | Not described | Explain what happens if the task is blocked |
| Workplace use | No workplace observation supplied | Do not infer from this exercise |
Build this part of your plan in more detail
These lessons address the next practical questions in this module. Choose the detail your workflow needs, then return to complete the exercise.
- How to Analyze Multilingual Feedback and Evaluate Software
Review feedback without losing meaning across languages.
- How to Analyze Open-Ended Survey Responses
Build and review a codebook, then report themes with a clear denominator.
More practical questions in this module (6)
- How to Identify the Common Barriers Your Clients Keep Hitting
Count recurring difficulties and overlap without inferring causes or blaming participants.
- How Do You Assess Learning From Student Reflections?
Review reasoning and reported application with a rubric, without treating a reflection as verified workplace performance.
- How to Document Mentoring Sessions Without Inventing Emotion
For mentoring or coaching, distinguish what was said, observed and interpreted in a session note.
- How to Review Participant Support Needs Mid-Program
Review a current support need, confirm its context and track whether an agreed action was delivered.
- How to Turn a Job Description into a Requirements Checklist
For employment placement, agree the role requirements and accepted evidence before reviewing people.
- How to Score Candidate–Role Matches with a Clear Rubric
For employment placement, review candidate–role evidence with explicit anchors, missing states and human decisions.
Add this part to your plan
On workbook page 3, assess the handoff example. Add a source passage, interpretation, uncertainty and next action for each criterion.
Download workbook (fillable PDF)Compare with a suggested answer
The review recognizes what is explicit, marks escalation as not evidenced and requests a revised attempt. It reports a practice-task result rather than a claim about real-world leadership performance.
Self-check: Could the learner or another reviewer challenge a judgment using the same source?
Questions you may have
Can AI grade every assignment automatically?
It can prepare judgments for a defined task, but reliability must be checked against the rubric and representative examples. Decide which judgments require human approval.
Should feedback and scores share one average?
Only if the combination has a justified meaning. Satisfaction, demonstrated skill and later practice usually answer different questions.
Apply the method to your work
Use the five-part plan to assess the sources, analysis and permissions your team needs. The solution page shows where a connected platform can support that workflow.
Explore the training and programs solution →