Compare research peptide suppliersUnited States · English
Peptide Criteria
A practical guide

Measurement uncertainty: comparing close lab results

Learn what to ask before ranking close analytical percentages: measurement definition, method, uncertainty, coverage and decision rules.

Public-source desk researchEditorial owner: Andrei SasContent reviewed

A report showing 99.8% does not automatically establish a better purchase than a report showing 99.7%. Before ranking close numbers, establish what each number measures and how uncertainty is expressed. The extra decimal place can create more confidence than the evidence warrants.

Start with the quantity being measured.

Two percentages can describe different quantities or use different analytical procedures. A chromatographic area percentage, a mass fraction and a stated specification are different kinds of information. Record the method and the quantity before attempting a numerical comparison.

For uncertainty reporting, NIST Technical Note 1297 distinguishes combined standard uncertainty from expanded uncertainty and calls for the coverage factor when expanded uncertainty is reported. That context helps readers understand what an interval represents. A bare plus-or-minus value leaves the interpretation incomplete.

Use an illustrative comparison carefully.

Suppose two hypothetical reports list the same nominal quantity as 99.7% and 99.8%, but neither provides enough information about method comparability or uncertainty. The defensible observation is a difference of 0.1 percentage points in the reported values. The reports alone do not establish that this difference is meaningful for the purchasing requirement.

Do not invent an uncertainty interval to make the comparison work. Do not interpret overlapping intervals as a universal statistical test either. Ask the issuing laboratories to explain the relevant uncertainty and limitations, then let the responsible scientific team decide whether the difference matters.

Request the information behind the number.

Our suggested checklist turns a vague accuracy question into a useful enquiry. It also helps identify cases where the reports cannot reasonably be compared at all. If a requested item is unavailable, keep that gap explicit.

Questions for a close numerical comparison
Ask forWhy it matters
Quantity and unitsEstablishes whether the numbers describe the same thing
Method and reporting basisReveals procedural differences
Uncertainty statementShows the stated measurement limitations
Coverage informationClarifies a reported interval
Decision rule for conformityExplains assessment against a specification limit

Write the purchasing conclusion at the right level.

A purchasing record can say that both suppliers supplied results for review, that method comparability remains unresolved, and that a scientific reviewer must assess the specification. It should not transform the higher displayed percentage into an overall supplier ranking.

Other requirements may be decisive: an identifiable dispatch lot, appropriate documentation, a required assay or an acceptable delivery condition. Keep those requirements beside the analytical result. When the uncertainty information is missing, a clear question is more useful than another decimal place in a comparison table.