Linking a Test Sample to Its Production Batch

The essential connection

A test result applies first to the sample that was actually examined. To relate that result to a production batch, the documentation must create an unbroken chain from the SKU and submitted-sample identifier to a defined lot. A model name alone is insufficient because the same name may cover different batches, dates, or configurations.

The sample submission record should identify the sample, SKU or configuration, submitting party, submission date, and stated batch number. The batch register should independently define that batch, including its production identifier and quantity. The sampling record should then show when, where, and by whom the sample was taken, how it was selected, and how many units were collected.

Preserving identity after sampling

A sealed reference sample or documented custody record helps show that the item received by the laboratory is the item taken from the recorded batch. Useful links include matching seal numbers, labels, photographs, dispatch records, receipt records, and the sample number printed in the final report. Any mismatch or unexplained relabeling breaks the evidentiary chain and should be resolved before a batch-level conclusion is made.

What the report can establish

The report should be read together with the sampling and batch records, not in isolation. The NIST Engineering Statistics Handbook explains that acceptance sampling uses information from a sample to decide whether a lot is likely to be acceptable; its purpose is not to estimate every aspect of the lot’s quality. The sampling method and decision rule therefore matter when extending a sample result to a batch.

If the only common identifier is a product or model name, the defensible conclusion is limited: the report records the result for the submitted sample. It does not, by itself, establish that the sample came from a particular production batch or that every unit in that batch has the same characteristics.