Use lessons 01–02 or equivalent skills before this exercise. You can revisit them from the lesson list.
A browser and text editor. Live source discovery needs browsing; all supplied exercises also work offline.
Start with the idea
Match each important statement to evidence before drafting.
Write the statement, its source and whether the source supports it. Narrow claims that overreach. Remove measured-benefit claims when there was no measurement.
Claim ledger: A table linking each statement to its evidence and review decision.
See the example and check your understanding
A made-up teaching example
Fourteen of twenty attendees answered the survey. Eleven respondents rated the session useful. That supports a statement about respondents, not all attendees or a 25 percent time saving.
Ask yourself
Can respondent feedback establish a measured processing-time reduction?
Show the explanation
No. Report how many people answered. Remove the unmeasured benefit and explain that some attendees did not reply.
Start with these checks
- Use the correct total when calculating a rate.
- Separate feedback from measured workplace results.
- Remove claims with no measurement.
Use the different practice case below and check every requirement in its full instructions.
Full learning objectives
- Separate evidence, inference and unknowns.
- Verify citation support at passage level.
- Repair unsupported numerical and causal claims.
Captions appear here during playback.
Construct a claim ledger before drafting
Match each important statement to evidence before drafting.
Video playback is unavailable. Read the descriptive transcript below, go directly to practice, or open the video file.
Source-based teaching example · Fictional data · Reviewed 2026-10-03. Microsoft Mark · locally synthesized. On a small screen, use fullscreen for diagram details or read the transcript.
Read the descriptive transcript
0:00 · What you will learn
Match each important statement to evidence before drafting.
What appears on screen
Match each important statement to evidence before drafting.
0:06 · The idea
Write the statement, its source and whether the source supports it. Narrow claims that overreach. Remove measured-benefit claims when there was no measurement.
What appears on screen
Write the statement, its source and whether the source supports it. Narrow claims that overreach. Remove measured-benefit claims when there was no measurement.
0:19 · One useful term
Claim ledger. A table linking each statement to its evidence and review decision.
What appears on screen
A table linking each statement to its evidence and review decision.
0:27 · A worked example
Fourteen of twenty attendees answered the survey. Eleven respondents rated the session useful. That supports a statement about respondents, not all attendees or a 25 percent time saving.
What appears on screen
Fourteen of twenty attendees answered the survey. Eleven respondents rated the session useful. That supports a statement about respondents, not all attendees or a 25 percent time saving.
0:42 · Your turn to think
Can respondent feedback establish a measured processing-time reduction? Pause and decide before continuing.
What appears on screen
Can respondent feedback establish a measured processing-time reduction? Pause the video if you need more time.
1:00 · Why this answer works
No. Report how many people answered. Remove the unmeasured benefit and explain that some attendees did not reply.
What appears on screen
No. Report how many people answered. Remove the unmeasured benefit and explain that some attendees did not reply.
1:11 · Check 1
Use the correct total when calculating a rate.
What appears on screen
Use the correct total when calculating a rate.
1:17 · Check 2
Separate feedback from measured workplace results.
What appears on screen
Separate feedback from measured workplace results.
1:23 · Check 3
Remove claims with no measurement.
What appears on screen
Remove claims with no measurement.
1:28 · Now try a different case
Open the lesson and try its different practice case. Use every check in the full task. Watching alone does not mark the lesson complete.
What appears on screen
Open the lesson Try the new case Check and revise The full task and its requirements are in the linked lesson.
Watching explains the example. Complete the lesson practice to check what you can do. Explore all visual lessons →
The method
A claim ledger is a small table: proposed claim, source and passage, evidence type, support status, and correction. Split compound sentences when different parts need different support. “The pilot was popular and reduced delays” contains two claims; evidence for popularity does not establish a delay reduction.
Ask the model to extract candidate claims, then audit the result yourself. A reference may exist yet fail to support the sentence. Do not equate two assistants agreeing with independent verification: they can share missing evidence. Preserve calculated values with the arithmetic and denominator; qualify inference instead of presenting it as an observed fact.
Choose your model, harness and offline route
Text-only assistant
Use supplied text. Ask for a plan or draft; do not claim browsing, code execution or file access that was unavailable.
Tool-enabled application or harness
Record which tools actually run, allowed files and actions, approval boundaries and observable results. Tools are optional unless the task needs them.
Local or cloud deployment
Inspect inference, retrieval, connectors, logs and storage separately. Use only authorized environments for real data; training fixtures are fictional.
Higher-capability or frontier model
Evaluate the exact task and configuration. Capability does not supply missing evidence or grant permission. No provider ranking or guarantee is implied.
Offline alternative
Use the supplied inputs and inspect the worked example manually. Builder exercises use local deterministic Python. No paid account or live model call is required.
Record the exact product/model label when available, tools actually used, inputs, permitted actions, observed output and checks. Live accounts or APIs may have external costs; they are optional and are not used by this page.
Read the general workflow guide →Worked example
Supplied input
Fictional P1: 14 of 20 attendees returned feedback; 11 feedback respondents rated the session useful. No processing-time measurement exists. Draft: “Most attendees loved the training, which cut processing time by 25%.”
A checked result
Supported: 11 of 14 respondents rated it useful. Not established: sentiment of all attendees, “loved,” or a 25% reduction. Ledger: usefulness → P1 → narrower wording; time reduction → no measurement → remove.
Why this works
The correction keeps the actual denominator and avoids turning subjective feedback into operational impact.
Try a different case
Independent fixture: 8 of 12 respondents chose option B; 25 people were invited. Draft: “Two thirds of invitees prefer B, proving it is faster.” Build a ledger and a corrected sentence.
Download this task and checklist →Interactive practice / Fictional fixture
Repair the claim ledger
Use the supplied independent case above. Check the bounded decisions, then review your written artifact against the full acceptance criteria.
Everything runs in this browser. No AI model is called and no work is uploaded. Bounded checks cover specific facts and rules. Your writing is not graded: you review it yourself. Practice never marks a lesson complete automatically.
Attach each claim to evidence; preserve the response denominator and a limitation.
Try the task, run the checks, then revise. Checks are teaching feedback, not an instructor grade.
Observed browser simulation
This trace comes from the bounded local checker. It is not a Python run, model result or production tool execution.
Self-review
Locked. Write every written part, then run the checks. A few words is not an attempt.
This is self-review, not grading. Read our answers, then tick what your own work does. Ticks are optional. They stay on this page and are not saved or sent anywhere. A different answer can be just as good if it meets the lesson criteria.
Your claim ledger and corrected sentence
A strong answer
Claim 1: two thirds prefer B. Evidence: 8 of 12 respondents chose B. Status: supported for respondents only. Claim 2: this covers invitees. Evidence: 25 invited, 12 answered, 13 unknown. Status: not supported. Claim 3: B takes less time. Evidence: none; speed was not measured. Status: remove. Corrected sentence: Eight of the twelve people who responded chose option B. Twenty-five were invited, so the other thirteen are unknown. Speed was not measured.
A flawed answer
Claim: two thirds of invitees prefer B. Evidence: survey. Status: supported. Corrected sentence: Most invitees prefer B, which suggests it is faster.
What is wrong with it: The fraction is right but the population is wrong: 8 of 12 respondents is not two thirds of 25 invitees. “Suggests it is faster” softens the speed claim without any evidence for it.
Keep or restore your work
Drafts are not saved automatically. Save only fictional practice data. A browser draft may be lost if storage is cleared; download a copy to keep it.
Your steps
- List each factual or causal claim separately.
- Recalculate the stated fraction and identify its population.
- Write a correction and a question needed to assess speed.
Check your work
- Report 8/12 respondents, not 8/25 invitees or all invitees.
- Remove the unmeasured speed claim.
- Retain nonresponse as a limitation.
After your attempt: review notes and a common mistake
“Eight of twelve respondents chose B; speed was not measured” preserves evidence. Two thirds is correct for respondents but attached to the wrong population in the draft.
Different wording or a different justified decision can pass. Check your work against every requirement; these notes are not an automated grade.
Explain your approach
What additional study would support a speed claim?
Keep your artifact, one correction or decision, its evidence and one remaining limitation. Nothing on this page sends or uploads your work.
Check your understanding
Choose an answer to see feedback. This is a knowledge check, not instructor assessment.
Review your independent practice
Complete the supplied task before recording progress. Use the acceptance criteria and reveal the review notes after your attempt. Keep your artifact yourself; nothing is uploaded.
Not yet self-reviewed. Completion stays in this browser and does not issue a badge.
Optional deeper study
Explore the related general-course lesson → This opens the existing canonical lesson; it is optional depth, not another copy of this focused task.
Primary references
Revision 2026-10-03. Fictional examples, not observed provider results. Knowledge checks and completion records are self-study aids.