Report verdicts look like judgments about quality, but beneath each pass, fail, or hold sits a statistical and procedural argument about what was examined and what it represents. A lot verdict claims something about thousands of units from examining dozens; a sample verdict claims something about buildability from examining one or two. Buyers who grasp the sampling logic behind the words respond with the right intensity: neither treating a lot pass as a purity certificate nor treating a sample hold as a disaster. This article unpacks that logic for the final part of the report-reading series.
Key takeaways
- Every verdict answers two questions at once: what was found in the examined units and what that finding implies for the unexamined remainder.
- Acceptance sampling provides the statistical link between examined units and lot decisions, defining how many to draw and when to reject. [1]
- Acceptance-test practice shows verdicts must record conditions and criteria, because the same finding means different things under different agreed rules. [2]
- Lot verdicts depend on lot discipline: mixed revisions or mixed production dates in one lot invalidate the reasoning behind the numbers.
- Reading the verdict tables before the verdict line turns the report from an opinion into a reproducible argument.
Busy readers open the report at the verdict line and stop there, which wastes the pages that give the verdict its authority. The finding tables, the sampling statement, and the criterion references explain exactly what the inspector saw, how many units it came from, and which rule converted those observations into the headline word. Without that context, pass and fail are just moods.
The two verdict types in these reports also work differently and are often confused. Sample-stage verdicts judge whether a documented configuration is approved for production, while lot-stage verdicts judge whether a defined population of finished goods may ship. The sampling reasoning behind each is distinct, and applying lot expectations to a sample verdict, or the reverse, produces the wrong response.
This closing part of the series teaches the reading order professionals use. It explains the sampling ideas in plain terms, shows how lot discipline underwrites lot verdicts, and gives a routine for converting any verdict into correct next steps.
What the verdict line assumes without saying
Every verdict silently assumes that the examined units were drawn fairly from the population the verdict speaks about, that the criteria applied were the agreed ones, and that the conditions of examination are on record. When those assumptions hold, the verdict carries real authority; when any of them fails quietly, the verdict is decoration. Professional readers therefore verify the assumptions before acting on the headline.
The report supports this verification with specific pages: the sampling statement names the population and the draw, the criterion column names the rule behind each finding, and the condition notes record the environment and equipment. Readers who check these three elements before the verdict line rarely misread reports.
This discipline matters most under schedule pressure, when the temptation is to act on the headline alone. A pass reached against the wrong revision or from a non-random draw is worse than no verdict at all, because it authorizes shipment with false confidence.
Reading any verdict in five checks
- Identify which population the verdict coversConfirm whether the verdict speaks about a sample configuration approval or about a defined lot of finished goods, since the two imply entirely different next steps.
- Check the sampling statementRead how many units were examined, how they were drawn, and from which segregated population, because verdicts without a stated basis cannot be reproduced or defended.
- Walk the criterion columnFor each finding, confirm the criterion cited was the agreed specification and note any finding judged against an unstated rule for follow-up.
- Separate configuration findings from lot findingsTreat design and material findings as approval questions for the configuration and treat variation findings as acceptance questions for the lot, routing each to its proper remedy.
- Convert the verdict into assigned actionsTranslate pass, fail, or hold into named owners, evidence requirements, and dates before closing the report, so the file drives work instead of archiving opinion.
Sample-stage versus lot-stage verdicts
| Aspect | Sample-stage verdict | Lot-stage verdict | Reader mistake to avoid |
|---|---|---|---|
| Population | One configuration approval | Defined finished-goods lot | Treating sample pass as lot proof |
| Basis | Full examination of few units | Statistical draw from the lot | Demanding lot statistics from one unit |
| Pass means | Configuration approved for production | This lot may ship | Reading lot pass as design purity |
| Fail means | Configuration needs redesign | This lot is rejected | Scrapping design over one bad lot |
| Hold means | Approval pending corrections | Decision pending re-draw or evidence | Ignoring hold pending items |
Keep this distinction visible whenever reports circulate internally, because most verdict misreadings confuse one column with the other.
The sampling ideas behind the tables
Two established practices supply the reasoning that report tables encode, and buyers need their working logic rather than their mathematics.
How examined units speak about unexamined ones
Acceptance sampling defines the relationship between sample size, defect counts, and lot decisions through published tables and operating characteristics. Examining more units tightens the discrimination between good and bad lots, while the accept-reject numbers translate observed defects into a ship-or-reject decision with known risks to both sides. [1]
For report readers, the consequence is concrete: the sampling statement is not boilerplate but the load-bearing part of a lot verdict. Verdicts reached from properly segregated lots and stated draws deserve decisive action, while verdicts with vague populations deserve challenge before anyone acts.
Why conditions and criteria belong in the verdict
Acceptance-test tradition records the conditions of examination alongside the findings, because environment, equipment, and procedure shape what the numbers mean. A function pass at room temperature, a shade reading under specified lighting, and a safety run on stated firmware are meaningful precisely because their conditions are recorded. [2]
Readers should therefore treat condition notes as part of the verdict rather than as decoration. Findings without recorded conditions cannot be compared across rounds, and configuration approvals without stated firmware, material, or revision references cannot anchor a bulk run.
A worked reading of a mixed report
Consider a report whose headline reads hold: dimensions pass, functions pass, shade sits marginally outside tolerance, and one laboratory sheet is pending. The untrained response debates whether this is really a pass or really a fail, while the trained response never joins that debate. It splits the report into a configuration question and a lot-readiness question, then acts on each.
The configuration answer is that approval is pending two items: a shade correction with re-measurement evidence and a current laboratory sheet for the exact build. Neither item restarts the program, and both have known remedies with known proof. The bulk answer is that no lot verdict exists yet, because no lot has been drawn, so production release waits on the configuration pass.
Notice how naturally sample evaluation reports support this reading when approached tables-first. The finding rows already separate dimensions from documents, the criterion column already names the rule behind each row, and the hold line simply summarizes what the tables show. Teams that read in this order close mixed reports in days; teams that argue about the headline first close them in weeks, if at all.
Sources
- Wikipedia: Acceptance sampling —— Overview of acceptance sampling including AQL concepts used in product inspection.
- Wikipedia: Factory acceptance test —— Overview of factory acceptance testing performed before shipment or installation.
Frequently asked questions
Can a sample-stage pass guarantee bulk quality?
No, and no competent report claims that it does. A sample pass approves the documented configuration as buildable; bulk quality depends on whether the factory repeats that configuration across volume with controlled processes. The sample verdict earns the program a production run with proper lot controls, not an exemption from them.
What invalidates a lot verdict after it is issued?
Discoveries about the population it claimed to cover: mixed revisions in one lot, units added after the draw, non-random selection, or criteria different from the agreed specification. Any of these breaks the link between examined units and the population, and the honest remedy is segregating proper lots and re-drawing rather than defending the compromised verdict.
How should a fail on one characteristic be weighed against passes elsewhere?
By the classification agreed before examination. A critical or safety fail rejects regardless of other passes, while an isolated minor finding may convert to a hold with defined correction. The report criterion column shows which case applies; weighing findings by gut feel instead reopens decisions the classification already settled.
What belongs in the file when the verdict closes?
The report, the sampling statement with the draw record, the criterion references, the closing evidence for every hold item, and the frozen configuration revisions. That bundle lets any later reader reproduce the decision and gives lot inspections an authoritative baseline. Files missing any element invite relitigation months later when memories have faded.