Record a supplier table that omits expected observations

Establish what was actually expected

Record it separately so the register does not look complete simply because every printed row was extracted.

Our public material-register example shows a synthetic report beside downloadable output. Use it to inspect how values and source references travel together. The self-service ParseMill trial is for invoice PDFs; supplier test reports and engineering review need a separate fit discussion.

Separate three questions: was the observation expected, is a corresponding result present, and has its technical adequacy been assessed? A present row can still lack a usable result. An absent row does not establish that testing failed, passed or was never performed.

Build an expected-to-reported map

Give each expectation a stable reference and record sample, property, requested context and scope basis. Map supplied rows to those references using reviewed identifiers. Microsoft: Left anti join describes a left anti join that finds left-table rows without matches. With expected records on the left, it can help list unmatched expectations after the matching keys have been checked.

Inspect unmatched supplied rows as well. They may represent extra tests, different sample names or a mapping error. A successful join proves that keys matched under the chosen rule; it does not establish that the result satisfies the intended requirement.

Reconcile a four-observation fixture

Synthetic example: the approved request expects length and mass observations for samples S1 and S2. The supplier report contains three corresponding rows.

There are four expected observations, three matched report rows and one absent row. Only two rows contain numerical observations. Calling the package 75 percent complete from row matching alone would obscure the explicitly unreported mass result; state the separate counts instead.

Ask for the missing relationship or evidence

A focused clarification identifies the request reference, sample and property: ask whether the omitted observation was performed and where its result is documented. For an explicitly unreported value, ask what that label means in the source and whether further evidence is available. Avoid suggesting that the supplier should enter a zero merely to fill the row.

W3C: PROV Data Model provides a model for relating records to the activities and sources behind them. Keep the scope record, supplied report and clarification as separate evidence items. A later explanation should supplement the history rather than overwrite the original absence.

Verify closure against the current scope

Close an omission only when the authorized reviewer accepts new evidence, approves a documented scope change or records another explicit disposition. Keep remaining blockers assigned. If the supplier issues a revised report, remap the affected observations and preserve the earlier version that prompted the question.

Use this expected-versus-reported matrix to define a supplier-report pilot before work begins. Keep the expectation list independent from extraction, and agree how the source register would be reconciled to it. The self-service invoice trial does not manage laboratory test expectations.