Laboratory Measurement Uncertainty: Resolution vs Calibration

A digital readout can show many decimal places without establishing how well the result is known. In laboratory records, resolution, calibration and measurement uncertainty answer different questions. Keeping them separate makes analytical reports easier to assess and prevents a precise-looking number from carrying more meaning than the evidence supports.

This guide explains measurement terminology and documentation. It is not a preparation protocol, a certificate of analysis, or evidence about any material supplied by Peptides Centre Canada.

Start with the quantity being measured

Before interpreting an instrument reading, identify the quantity and the conditions that define it. A balance indication, an analytical concentration and a chromatographic area percentage are not interchangeable results. Record the method, units, sample identity and reporting basis alongside the number. A result without that context is difficult to compare with another laboratory’s result.

For the document-level questions, start with our guide to reading a peptide certificate of analysis. That guide covers what a report identifies; the distinctions below help interpret the measurement information within it.

Resolution: what change can the instrument distinguish?

In the International Vocabulary of Metrology, resolution concerns a detectable change in an instrument’s indication. It can depend on the measured value and on effects such as noise. A small display increment does not, by itself, establish a small uncertainty for the final result. See JCGM VIM, resolution.

For example, an illustrative display reading of 10.000 does not explain whether a correction was applied, whether repeated readings agree, or which conditions the calibration covered. Adding zeros when copying that reading into a spreadsheet supplies none of the missing evidence.

Calibration: how is the indication related to a reference?

Calibration establishes a relationship between reference values and instrument indications, with their uncertainties, under specified conditions. That information is then used to obtain a measurement result. Calibration is not the same operation as adjusting an instrument, and neither term should be used as a synonym for verification. See JCGM VIM, calibration.

When reviewing a calibration record, check that the instrument identifier matches the equipment used and that the documented range and conditions are relevant. A label saying “calibrated” is not a substitute for the record, its results and any applicable corrections. Whether the equipment is suitable for a particular method requires a separate assessment.

If a reference material supplies an input value, check which property is certified and what uncertainty and traceability accompany it. Our guide to certified reference materials and research reagents explains how to read that scope without extending certification to unlisted properties.

Uncertainty: how well is the result known?

Measurement uncertainty describes the dispersion of values attributed to the quantity being measured, using the available information. It can include several components, including contributions associated with corrections and reference values. It is not simply the number of decimal places, nor is it an admission that a mistake occurred. See JCGM VIM, measurement uncertainty.

A useful report makes clear which result the uncertainty belongs to and how it is expressed. An uncertainty quoted for a calibration is not automatically the complete uncertainty of every subsequent laboratory measurement.

Type A and Type B describe evaluation methods

A Type A evaluation uses statistical analysis of measurement data. Examples include suitable analyses of repeated observations or fitted calibration data. Repetition can provide information about variability, but it does not by itself account for every possible influence on a result. NIST’s Type A guidance describes these statistical evaluations.

A Type B evaluation uses other relevant information, such as calibration reports, instrument specifications or prior knowledge. The assumptions used to interpret that information matter. “Type B” does not mean inferior evidence, and “Type A” does not guarantee that a measurement is fit for its intended purpose. See NIST’s Type B guidance.

Combined and expanded uncertainty are not the same

Combining contributions requires a measurement model. The influence of each input and any relevant correlations must be considered; uncertainty percentages should not be added without a justified model. NIST explains the first-order propagation approach and its assumptions in Combining uncertainty components.

Expanded uncertainty, commonly written U, is obtained by multiplying the combined standard uncertainty by a coverage factor k. A statement such as “±0.20” is therefore incomplete if its units, meaning and coverage information are missing. Under the conditions described by NIST, k = 2 corresponds to approximately 95% coverage; that interpretation is not universal for every distribution or data set. See NIST’s expanded-uncertainty guidance.

Traceability belongs to a measurement result

Metrological traceability connects a result to a stated reference through a documented calibration chain, with each link contributing uncertainty. NIST also cautions that traceability alone does not establish fitness for purpose: the uncertainty still needs to be suitable for the intended application. See NIST’s traceability policy and FAQ.

This is different from tracking the movement of a vial or matching a report to a lot. Both records can matter, but a lot number does not establish metrological traceability. Our receipt-to-result lot-traceability guide explains the material-record side of that distinction.

A practical report-review checklist

Use these questions to identify missing information, not as an accreditation checklist or a pass/fail rule:

• Result: What quantity, units and reporting basis are stated?
• Identity: Which sample, lot and instrument records connect to the result?
• Method: Which procedure and relevant conditions were used?
• Calibration: Is the applicable record available, and were relevant corrections considered?
• Uncertainty: Is the quoted value standard or expanded, and what coverage information accompanies it?
• Scope: Which contributions are included, excluded or not evaluated?
• Decision: If a conformity statement is made, what specification and decision rule were applied?

If an answer is absent, ask the issuing laboratory what the report supports. Do not fill the gap by assuming that a fine display resolution, a calibration sticker or a lot identifier proves the missing property.

What this means for laboratory calculations

An arithmetic result cannot be more informative than its inputs and assumptions. Keep source values and units in the working record, distinguish measured inputs from nominal labels, and avoid presenting extra calculated digits as newly acquired measurement evidence. A qualified laboratory should select the measurement model and evaluate suitability for its intended analytical use.

Prepared by Peptides Centre Canada. Sources checked September 23, 2026. This is general measurement-literacy content, not independent testing, accreditation, a product-quality guarantee, or human or veterinary use guidance. The cited organizations have not endorsed this website.