Tools & Microfiber2 min read

Measurement Resolution vs Uncertainty: Why Extra Digits Do Not Prove a Coating Difference

Distinguish instrument resolution from measurement uncertainty and avoid ranking coatings by differences the test cannot reliably resolve.

Updated · NANOLAB Knowledge

An instrument can display a small numerical change without establishing that the change is meaningful. Resolution concerns the ability to distinguish changes; measurement uncertainty describes the dispersion associated with the reported result. Displayed decimal places are not a substitute for either assessment.

Define the difference you need to detect

Suppose an illustrative comparison gives readings of 80.1 and 80.3 in the same units. If the measurement process varies by an amount comparable to that difference, declaring a clear winner would be premature. The required confidence depends on the instrument, method, surface and sampling design.

Before purchasing equipment, state the smallest difference that would change the decision. A device suitable for broad incoming checks may be unsuitable for distinguishing subtle changes between two coating trials.

Include the whole measurement process

Instrument specifications are only part of the picture. Positioning, surface curvature, sample variation, environmental conditions and operator technique can contribute. Repeating a reading without moving the probe may underestimate the variation of the actual inspection process.

Use suitable reference checks and a documented procedure. Where a formal uncertainty statement is required, ask a competent laboratory to define and evaluate the relevant components rather than copying a display increment into the uncertainty field.

Report a result that matches the evidence

Keep raw readings, units and method information. Round the final result sensibly in relation to the uncertainty. Do not average many digits into a claim of accuracy that the method cannot support.

For a purchasing trial, an unresolved small difference can be reported as inconclusive under the chosen method. That is more useful than inventing a ranking. Other relevant outcomes—application effort, compatible surfaces and observed defects—can still support the decision, provided they are evaluated independently and not used to disguise an unsupported measurement claim.

Related guidance

Technical references: NIST: Measurement-system resolution; NIST: Measurement uncertainty.

For product-specific application instructions, consult NANOLAB technical documentation or ask about your surface and intended use.