Measurement Uncertainty

Scientific Snapshot

Discipline: Analytical Chemistry and Metrology
Difficulty: Intermediate–Advanced
Course position: Lesson 6 of 10
Core concepts: uncertainty sources, Type A evaluation, Type B evaluation, combined uncertainty, expanded uncertainty, traceability.

Learning Objectives

After completing this monograph, readers should be able to:

• Define measurement uncertainty.
• Distinguish uncertainty from error.
• Identify common analytical uncertainty sources.
• Explain Type A and Type B evaluations.
• Describe combined and expanded uncertainty.
• Explain how uncertainty supports decision-making near specifications.

Executive Summary

Every analytical result contains uncertainty.

That uncertainty reflects the combined influence of sampling, preparation, calibration, standards, instrumentation, environmental conditions, calculations, and normal variability.

Measurement uncertainty does not mean that a result is unreliable. It provides a transparent estimate of the range reasonably associated with the measured value.

Uncertainty Versus Error

Error is the difference between a measured value and the true value.

Uncertainty describes the range of plausible values associated with the result.

Known errors should be corrected where possible. Residual uncertainty remains.

Common Sources

• sampling variability,
• balance calibration,
• volumetric equipment,
• pipette performance,
• reference-standard value,
• detector repeatability,
• calibration model,
• temperature,
• integration,
• analyst execution.

Type A Evaluation

Type A evaluation uses statistical analysis of repeated measurements.

Examples include:

• replicate injections,
• repeated preparations,
• precision studies,
• repeated weighing.

Type B Evaluation

Type B evaluation uses other evidence.

Examples include:

• calibration certificates,
• manufacturer specifications,
• certified-reference documentation,
• historical data,
• published information.

Standard Uncertainty

Each contributor is expressed as a standard uncertainty, usually analogous to one standard deviation.

Combined Standard Uncertainty

Independent contributors are commonly combined using a root-sum-of-squares approach.

The resulting combined uncertainty reflects the total modeled contribution.

Expanded Uncertainty

Expanded uncertainty is calculated by multiplying combined standard uncertainty by a coverage factor.

A coverage factor near two is often used to communicate an interval associated with approximately 95% confidence under suitable assumptions.

Uncertainty Budgets

An uncertainty budget lists:

• source,
• value,
• distribution,
• divisor,
• standard uncertainty,
• sensitivity coefficient,
• contribution.

It helps identify the dominant sources.

Traceability

Traceability links a result through calibrations and standards to recognized references.

Each step contributes uncertainty.

Decision Rules

Uncertainty matters when results approach a specification limit.

A decision rule defines how uncertainty is considered when determining conformity.

Practical HPLC Example

An assay uncertainty budget may include:

• reference-standard assay,
• balance,
• volumetric dilution,
• calibration curve,
• injection precision,
• integration,
• sample preparation.

Practical LC-MS Example

Mass uncertainty may involve:

• instrument calibration,
• resolving power,
• centroiding,
• charge-state assignment,
• deconvolution,
• reference-ion accuracy.

Science Makes Sense

A measurement is not a perfect dot on a number line.

It is more like a carefully estimated zone around that dot. The better the measurement system is understood, the more scientifically meaningful that zone becomes.

Common Misconceptions

“Uncertainty means the laboratory made a mistake.”

No. Uncertainty exists even after known mistakes are corrected.

“Precision is the same as uncertainty.”

Precision is one contributor, not the entire uncertainty.

“Reporting uncertainty weakens confidence.”

Transparent uncertainty usually strengthens credibility.

Laboratory Best Practices

• Define the measurand clearly.
• Map the analytical process.
• Identify significant contributors.
• Use both experimental and documented evidence.
• Avoid double-counting sources.
• Review budgets after method changes.
• Apply documented decision rules.
• Report uncertainty with appropriate significant figures.

Frequently Asked Questions

What is a measurand?

The specific quantity intended to be measured.

What is Type A uncertainty?

Uncertainty estimated from repeated observations.

What is Type B uncertainty?

Uncertainty estimated from specifications, certificates, or other evidence.

Why use a coverage factor?

To convert combined standard uncertainty into a broader reporting interval.

When is uncertainty most important?

When results are close to decision or specification limits.

Key Takeaways

• Every analytical result has uncertainty.
• Uncertainty differs from error.
• Type A and Type B evaluations use different evidence.
• Combined uncertainty integrates significant contributors.
• Expanded uncertainty communicates a broader interval.
• Decision rules should explain how uncertainty affects conformity.

Suggested Figures

1. Error versus uncertainty.
2. Type A versus Type B evaluation.
3. Analytical uncertainty source map.
4. Uncertainty budget.
5. Traceability chain.
6. Decision near a specification limit.

Knowledge Check

1. How does uncertainty differ from error?
2. What distinguishes Type A from Type B evaluation?
3. What is combined standard uncertainty?
4. Why is traceability important?
5. How can uncertainty affect a pass-fail decision?

References

1. JCGM 100. Guide to the Expression of Uncertainty in Measurement.
2. Eurachem. Quantifying Uncertainty in Analytical Measurement.
3. ISO/IEC 17025.
4. Harris DC. Quantitative Chemical Analysis.

Editorial Note

Version 1.0 establishes the metrological confidence framework for advanced analytical interpretation.