A count of inspection findings is not a count of affected garments. Before interpreting a lingerie inspection summary, identify the inspection unit, keep each finding attached to that unit, and separate the two totals. Otherwise, several issues on one garment can be mistaken for several affected garments.
This guide concerns the meaning of reported counts. For what to examine on a sample, start with our quality-control guide. Neither guide selects an acceptance sampling plan or determines whether a shipment should be accepted.
Define the unit before counting the findings
The Two-Tone Garter Thigh-High Stockings listing describes connected garter-style construction with an integrated waist detail. For an inspection of that selected construction, agree whether the report counts complete garments or separately identified components. Do not count the complete garment in one column and its leg sections as equivalent units in another.
Give each inspected unit a stable ID. If a finding concerns a particular section, retain that location under the same unit ID. Keep the product version, colour and size beside the inspection record. These identifiers define the scope of the review; they do not establish any result for the listed product.
Keep a finding ledger and a unit summary
| Finding ledger | What to record |
|---|---|
| Unit and location | One unit ID and the exact section being reviewed |
| Finding identity | One finding ID, even if several photographs show it |
| Requirement and evidence | The agreed requirement, observation and supporting reference |
| Classification | Confirmed nonconformity, pending review or other agreed status |
Then create a separate summary showing units inspected, distinct units classified as nonconforming, and the number of confirmed nonconformities. A second photograph of the same issue is supporting evidence, not automatically a second finding. Conversely, two different agreed requirement failures on one unit need not be merged merely because they appear in one photograph.
A small example: three findings, two affected units
The following numbers are invented solely to explain counting; they are not factory results, recommended sample sizes or acceptance limits. Suppose ten complete garments are inspected. Unit G01 has two confirmed nonconformities and G02 has one. The other eight have none recorded against the agreed requirements.
- Confirmed nonconformities: three.
- Distinct nonconforming units: two.
- Units inspected: ten.
- Observed proportion of nonconforming units in this sample: two divided by ten, or 20%.
Calling this “30% defective” would substitute the finding count for the affected-unit count. The 20% describes only this invented inspected sample. It is not an AQL, a shipment acceptance decision or an established defect rate for an entire production lot.
Keep classification separate from arithmetic
The NIST explanation of counts distinguishes individual nonconformities from nonconforming units and stresses the inspection-unit definition. Use that distinction to ask clearer questions, not to assume that every cosmetic observation makes a product unacceptable.
Ask the inspector to identify the agreed classification rules and unresolved findings. Severity and acceptance criteria require their own agreement. Sample selection, lot definition and whether every selected unit was examined also affect how a report can be interpreted. A count alone does not answer those questions.
Request a corrected summary, not a guessed percentage
If a report says only “three defects,” request the unit IDs, finding references, number inspected and classification basis. If components and complete garments were mixed, ask the issuer to reconcile them without deleting the original record. Retain the corrected version and its explanation.
For an OEM or private-label discussion, send the exact product reference and the inspection scope you want reviewed. Ask our team to clarify the required inspection evidence; do not treat this illustrative ledger as a confirmed inspection service or an agreed shipment-release procedure.