Calibration Fundamentals
Use accuracy, precision, error, correction, repeatability, reproducibility, and uncertainty correctly in measurement records and technical decisions.
When a measurement is reported as 20.00 mm, the number looks precise. It does not, by itself, tell you whether the measurement is close to the true or reference value, how consistently the measurement system produces results, or how much uncertainty surrounds the result.
That distinction matters in dimensional inspection, calibration, manufacturing quality control, laboratory measurement, and process monitoring.
Accuracy, precision, and error describe different aspects of measurement. Confusing them can lead to incorrect conclusions about a part, an instrument, or a manufacturing process.
This guide explains the difference and shows how the concepts connect in practical measurement work.
What is accuracy in measurement?
Accuracy describes the closeness of agreement between a measured quantity value and a true quantity value.
In practical metrology, the true quantity value is generally not known exactly. A reference value or accepted reference value may therefore be used when evaluating measurement results.
For example, suppose a reference standard has an accepted value of 50.00 mm and an instrument produces a result close to that value. The result may be considered accurate to the extent that it agrees closely with the reference value, within the conditions and uncertainty associated with the measurement.
Accuracy is therefore related to closeness to the reference or true value.
What is precision in measurement?
Precision describes the closeness of agreement between repeated measurement results obtained under specified conditions.
A measurement system can produce results that are tightly grouped even when the group is offset from the reference value.
For example, repeated measurements might produce:
- 50.21 mm
- 50.20 mm
- 50.21 mm
- 50.20 mm
These results are closely grouped. That indicates good precision under the stated measurement conditions.
The group may still be biased relative to the reference value. Precision alone does not establish accuracy.
What is measurement error?
Measurement error is the measured quantity value minus a reference quantity value.
The reference value used for this purpose must be appropriate to the measurement being evaluated. In many practical situations, the reference value is an accepted value rather than an exactly known true value.
A simple expression is
Measurement error = measured value - reference value
For example, if the measured value is 50.02 mm and the reference value is 50.00 mm:
Error = 50.02 mm - 50.00 mm = +0.02 mm
The positive sign indicates that the measured value is above the reference value.
Error and measurement uncertainty are different concepts. An error can be described when a reference value is available. Uncertainty expresses the information associated with the dispersion of values that could reasonably be attributed to the quantity being measured.
Accuracy and precision are different.
Accuracy and precision are often used as if they mean the same thing. In metrology, they describe different properties.
| Concept | What it describes | Practical interpretation |
|---|---|---|
| Accuracy | Closeness to a true or reference value | How close the result is to the accepted value |
| Precision | Closeness among repeated results | How closely repeated results agree |
| Error | Difference between a measured value and a reference value | How far the result differs from the reference |
| Uncertainty | Quantifies the dispersion associated with the quantity being measured | How much uncertainty is associated with the measurement result |
Four common measurement situations
High precision and high accuracy
Repeated measurements are close to one another and close to the reference value.
This is generally the situation measurement engineers want when a measurement is used for an important engineering decision.
High precision with poor accuracy
Repeated results are tightly grouped but consistently displaced from the reference value.
This pattern can indicate a systematic effect or bias in the measurement system. Investigation may be needed to determine the source.
Poor precision with reasonable average agreement
Individual results vary considerably, while their average may be close to the reference value.
This situation can still create problems. A measurement system that produces large variation may not provide sufficient information for a particular acceptance decision.
Poor precision and poor accuracy
The results vary widely and do not agree closely with the reference value.
Both the measurement system and the measurement method may require investigation.
Why repeated measurements matter
A single measurement provides limited information about measurement behavior.
Repeated measurements allow engineers to examine variation under specified conditions.
Suppose a component is measured five times and produces nearly identical results. That gives information about repeatability under those conditions.
If another operator measures the same component using the same defined procedure and obtains materially different results, the investigation can examine reproducibility.
This is why measurement-system evaluation goes beyond checking whether an instrument has a current calibration certificate.
How calibration relates to accuracy and precision
Calibration establishes, under specified conditions, a relationship between quantity values provided by measurement standards and corresponding indications of a measuring system.
A calibration certificate can provide evidence about the instrument's performance under the conditions of calibration.
The manufacturing measurement may involve different conditions.
Temperature, operator technique, fixture arrangement, workpiece geometry, measurement method, and other factors can affect the result depending on the application.
Calibration therefore needs to be considered together with the actual measurement task.
A current calibration status does not, by itself, establish that every measurement made with the instrument is suitable for every possible decision.
Measurement error and measurement uncertainty
Error and uncertainty answer different questions.
Error describes a difference between a measured quantity value and a reference quantity value.
Measurement uncertainty describes the dispersion of quantity values that could reasonably be attributed to the measurand based on the available information.
Consider a dimensional measurement reported as
20.00 mm
The displayed value alone does not describe the complete measurement result.
The measurement method, calibration information, environmental conditions, repeatability, reference standards, and other uncertainty contributions may affect the interpretation of the result.
For measurements close to an acceptance limit, these considerations become particularly relevant.
What causes measurement error?
Measurement error can have many sources. The relevant sources depend on the measurement method and application.
- Instrument effects: Characteristics of the measuring instrument can influence the result.
- Environmental effects: Temperature and other environmental conditions can affect some measurement processes.
- Operator effects: Technique and interpretation can contribute to measurement variation.
- Workpiece effects: Geometry, surface condition, and material behavior can affect some measurements.
- Method effects: The measurement procedure itself can introduce sources of variation or systematic effects.
- Fixture effects: How a component is positioned or constrained can influence some dimensional measurements.
These sources should be evaluated according to the actual measurement process. A generic list cannot determine the uncertainty of a particular measurement.
How accuracy and precision affect manufacturing inspection
Consider a production dimension with an upper and lower specification limit.
An inspection system produces repeated results that are tightly grouped near one limit.
The measurement system may have good repeatability, but engineers still need to determine how the results relate to the reference value, measurement uncertainty, and the acceptance rule being applied.
Now consider a measurement system with large variation.
Two operators may obtain different results from the same component. The inspection decision can then depend on measurement variation rather than the actual difference between the component and the requirement.
This is why measurement-system performance matters to manufacturing quality.
Accuracy, precision, and process control
Metrology becomes more useful when measurements are studied as a sequence rather than isolated values.
Suppose a critical dimension is recorded during production.
The individual measurements remain within specification, but the results gradually move in one direction.
The pattern may indicate a change in the manufacturing process. Tool condition, machine behavior, temperature, material, or another process variable may be relevant.
The measurement trend does not prove the cause. It provides evidence that can guide an investigation.
This creates a practical chain:
Measurement → Trend → Investigation → Process cause → Corrective action → Verification
The quality of this chain depends on the quality of the measurement data and the understanding of the measurement system.
Accuracy vs. precision vs. error in simple terms
A practical way to remember the difference is
- Accuracy: closeness to the reference or true value.
- Precision: closeness among repeated measurements.
- Error: difference between a measured value and a reference value.
- Uncertainty: quantified information about the dispersion associated with the measurement result.
These concepts are connected, but they should not be treated as interchangeable.
Accuracy and precision in calibration work
Calibration laboratories need to consider the measurement method, reference standards, environmental conditions, measurement equipment, and uncertainty associated with the calibration result.
Traceability also needs careful interpretation. Metrological traceability applies to a measurement result and involves an unbroken documented chain of calibrations, with each contribution to measurement uncertainty taken into account.
Traceability alone does not establish fitness for purpose.
The uncertainty associated with a measurement needs to be suitable for the intended measurement requirement and decision.
What engineers should check when a measurement looks wrong
When a measurement appears unusual, avoid assuming immediately that the component is defective or the instrument has failed.
Check the measurement process systematically.
- Review the measurement method.
- Check the instrument's calibration status.
- Repeat the measurement where appropriate.
- Check whether the result changes with the operator or method.
- Review environmental conditions relevant to the measurement.
- Compare the result with previous measurements.
- Review the workpiece and its setup.
- Consider measurement uncertainty when the decision is close to a specification limit.
- Investigate the manufacturing process if the measurement pattern indicates a possible process change.
This approach helps separate a product problem from a measurement problem.
Why the distinction matters
Accuracy, precision, and error are basic measurement concepts, but they have practical consequences.
An engineer evaluating a measuring instrument needs to know whether the instrument produces results close to a reference value.
A quality engineer studying repeated measurements needs to understand variation and measurement-system behavior.
A manufacturing engineer studying a dimensional trend needs reliable measurements before drawing conclusions about process drift.
The same numerical result can therefore have different meaning depending on how it was obtained, what reference was used, and what decision it supports.
Accuracy vs. precision vs. error: the practical takeaway
A measurement result is more useful when its context is understood.
Accuracy concerns closeness to a true or reference value. Precision concerns agreement between repeated results. Error describes the difference between a measured value and a reference value. Uncertainty describes the dispersion associated with the measurement result based on the available information.
In manufacturing, these concepts help engineers evaluate whether measurement data can support inspection and process decisions.
For your next measurement-system review, start with the actual decision the measurement must support. Then examine the reference, repeatability, reproducibility, uncertainty, method, and operating conditions that affect that decision.
That approach gives the number a proper engineering context.
References
- National Institute of Standards and Technology (NIST), International Vocabulary of Metrology (VIM) terminology and metrology resources.
- NIST, Engineering Statistics Handbook, measurement process characterization, and measurement systems.
- NIST, measurement uncertainty, and metrological traceability resources.
Comments
Post a Comment