Keep report, reproduction and correction separate
Synthetic guide G-341 for v5 tells readers to choose field A, but a controlled v5 trial shows the required output uses field B. Report F-341 names the step and version; F-342 only says “does not work”. The specific report plus matched trial supports a correction from A to B. The vague report still needs scope evidence and must not be silently recorded as another reproduced defect.
| Record | Evidence | Conclusion | Action |
|---|---|---|---|
| F-341 | Names G-341 step and v5 | Testable report | Run matched trial |
| V-341 | v5 trial requires B not A | Specific instruction wrong | Correct tested instruction |
| F-342 | No step or version | Unresolved applicability | Request specific evidence |
Correct the demonstrated error
- Capture guide and step reference
- Match the stated version
- Reproduce and test correction
- Publish corrected applicable step
A correction for v5 cannot silently replace v4 instructions without checking their separate applicability.
How many reported feedback items are supported as demonstrated errors?
Show the worked answer
One: F-341, supported by V-341. F-342 remains an unqualified report; the correction must not pretend it was reproduced.
Retain the specific report and matched trial with the correction. Other versions and vague reports require their own evidence checks.