Keep a unit record and a separate inspection-event history. Then report two clearly named results: which original units passed the first time, and which were ultimately accepted. The distinction tells a reviewer what rework recovered without hiding the initial result.
Define the batch and inspection basis
ASQ defines first-pass yield around units meeting requirements without rework or retesting. A later pass therefore does not turn an initially failed unit into a first-pass success. ASQ's first-pass yield definition
This original fictional batch contains ten printed paper packets, U1 through U10. The supplied requirements include correct page order and legible printing. Initial inspection records show seven passes and three failures. No actual packets were inspected for this example.
| Units | Initial result | Recorded issues | Later evidence | Final disposition |
|---|---|---|---|---|
| U1–U7, seven units | Pass | None recorded | Initial passes retained | Accepted |
| U8 | Fail | Smudge; wrong page order | Reworked; later reinspection pass recorded | Accepted after rework |
| U9 | Fail | Smudge | Reworked; later reinspection pass recorded | Accepted after rework |
| U10 | Fail | Tear | Scrap disposition supplied | Scrapped |
U10's failure and scrap disposition are supplied sample outcomes. The listed criteria do not specify how a tear is evaluated; before using this inspection plan, clarify the relevant requirement with the actual reviewer. Do not invent a severity class or assume that every tear violates a particular printing requirement.
Keep both rates tied to the same ten units
The initial pass group is U1–U7: seven units out of ten, or 70% first-pass yield. U8 and U9 still belong to the initial failure group after they are repaired.
The final accepted group is U1–U9: nine units out of the original ten, or 90% final acceptance for this example. U10 remains scrapped. Show the difference as “7 initial passes; 2 additional acceptances after documented reinspection,” rather than replacing the first-pass number.
The two reinspection events do not add two new units to the batch. There are ten unit identities throughout. An event count answers how many inspections were recorded; the unit denominator answers how many original packets were being evaluated.
Count issues separately from failed units
The initial failures contain four recorded issues: two on U8, one on U9, and one on U10. They affect three initially failed units. U8 counts once as a failed unit even though it has both a smudge and a page-order issue. ASQ's glossary distinguishes defects from units containing one or more defects; keep the same counting distinction visible in your own records. ASQ's quality glossary, D
Use unit ID, inspection-event ID, event type, outcome, observed issues, requirement reference, and supporting record reference in the native ledger. Keep the first event immutable in the review history. The later event points back to the same unit, with the rework and reinspection evidence separately identified.
For the demonstration, you can label U8's first event DEMO-U8-I1 and its reinspection DEMO-U8-I2. These are original illustrative identifiers, not references to real inspection records. A reviewer inspecting U8 should see both its initial two issues and its later pass, with neither replacing the other.
Require the later result before changing acceptance
“Reworked” describes an action, not the result of inspecting that action. If the pass evidence for a repaired unit is missing, show “Awaiting reinspection evidence.” Keep it out of the final accepted count until the appropriate result is documented. The initial failure remains unchanged either way.
The local HTML review compares initial outcomes with final dispositions within the same ten-unit scope. Inspect U8 to see both events and the two initial issues. A separate missing-evidence demonstration should be labeled as a variant; it must not silently erase the supplied pass record from the main example. Test the local review controls in the intended host before using them with actual inspection records.
Copy this prompt for the inspection-history dashboard
Prepare a complete US English quality dashboard review packet and native inspection-ledger specification. Use ten fictional printed paper packets U1-U10; label all records illustrative, with no actual inspection performed.
Supplied requirements: correct page order and legible printing. Initial outcomes: U1-U7 pass; U8 fails with smudge and wrong order; U9 fails with smudge; U10 fails with tear. U8/U9 are later reworked and have recorded reinspection passes; U10 is scrapped.
U10's recorded outcome is supplied, but the listed criteria do not explain how tears are evaluated. Keep this limitation visible and request the actual requirement; do not invent severity or change the supplied outcome.
First-pass yield7/10=70% remains70% after rework. Final acceptance9/10=90%,including2 documented rework passes. The denominator is the same10units,not12 because of repeated inspections. Four initial recorded issues affect3failedunits; U8's2issues still represent1unit.
Return complete initial/final summaries, a unit table and separate event-history specification. Fields: unitID,eventID,eventtype,outcome,issues,requirementreference and evidence reference. Illustrative U8events may be DEMO-U8-I1 and DEMO-U8-I2. Preserve initialfail/twoissues beside laterpass.
Rework without documented reinspection pass stays awaiting evidence and is not accepted. If illustrating that case, label it a separate variant without altering the main supplied records. Proposed initial/final views preserveunitIDs/scope and reveal eventhistory; implementation/testing pending. No approval/mark-safe controls,arbitrary qualityscore,AQL,control limits,root-cause diagnosis or standards compliance. Keep actual records in the native inspection workflow.
Review the inspection history in NOS
NOS can save and render a prepared HTML review copy of the unit summaries and event history. Use the local example to compare the initial failure with the later evidence. Keep actual inspection decisions and their supporting records in the maintained workflow.
Finish with the 70% first-pass result still visible beside 90% final acceptance, and make every recovered unit traceable to its own reinspection result.