✦ScienexySCIENCE KNOWLEDGE LIBRARY
reference · SCI-REF-003-075

Sources and Mechanisms of Measurement Errors

Measurement errors arise from interactions among instruments, measurement methods, operators, environmental conditions, reference standards, samples and data-processing procedures. Understanding these sources makes it possible to identify error mechanisms, determine which effects influence a result and design suitable controls. This topic examines common physical, instrumental, environmental, observational and computational causes of measurement deviations.

General measurement model: Y = f(X₁, X₂, ..., Xₙ)
For small input variations, the first-order sensitivity coefficient is cᵢ = ∂f/∂Xᵢ.
A simplified independent-input uncertainty model is u_c²(Y) = Σ[cᵢ²u²(Xᵢ)].
For correlated inputs, covariance terms must also be included:
u_c²(Y) = Σ[cᵢ²u²(Xᵢ)] + 2Σᵢ<ⱼ[cᵢcⱼu(Xᵢ, Xⱼ)].

Explanation

A measurement result is produced by a complete measurement system rather than by an instrument alone. The result depends on the measurand, measurement principle, instrument, calibration, operator, environment, sample condition, data acquisition and mathematical model. A weakness in any of these components can influence the final result.

Instrument-related errors arise from the construction, adjustment, calibration and operating characteristics of measuring equipment. Examples include zero offset, scale-factor error, nonlinearity, hysteresis, backlash, friction, resolution limitations, electrical noise, drift and sensor ageing. An instrument may also respond differently under different loads, frequencies, temperatures or measurement ranges.

Environmental errors arise when external conditions influence the measuring instrument, sample or measurement process. Temperature can cause thermal expansion, alter electrical resistance or change sensor sensitivity. Humidity can affect materials and insulation. Vibration can disturb dimensional measurements, while electromagnetic interference can affect electrical signals. Pressure, airflow, lighting and acoustic noise can also influence particular measurement methods.

Observational errors arise from how a person reads, positions, operates or interprets a measuring system. Parallax occurs when an analogue scale is viewed from an unsuitable angle. Incorrect alignment can affect length and angle measurements. Reaction time can affect manual timing. Inconsistent contact force can influence measurements made using mechanical instruments.

Method-related errors occur when the chosen procedure does not correctly represent the measurand or when necessary corrections are omitted. Measuring the diameter of a non-circular object at only one location may not represent the intended dimension. Measuring a moving object with a method that assumes it is stationary can also produce a method-dependent result. The measurement procedure must define the quantity sufficiently clearly.

Sample-related effects occur when the object or material changes during measurement. Examples include thermal expansion, evaporation, moisture absorption, deformation under contact pressure, chemical reaction, contamination and ageing. In biological or environmental measurements, the sample may also vary naturally with location or time.

Reference and calibration-related effects arise from the uncertainty or limitations of reference standards, calibration procedures, reference materials and traceability chains. A calibration result is not automatically exact. The reference, calibration method, environmental conditions and instrument stability all contribute to the final measurement uncertainty.

Data-processing errors arise during conversion, calculation, software processing, filtering, rounding, storage or transmission. An incorrect unit, formula, calibration coefficient or software setting can systematically affect a large set of results. A transcription mistake can affect an individual result. Automated processing therefore requires validation, version control and independent checks appropriate to the risk.

The influence of a source depends on the measurement task. Temperature may be negligible in a rough classroom measurement but significant in precision dimensional metrology. Reaction time may be important when timing a short event manually but negligible when measuring a long duration with an automated clock. Error investigation should therefore be based on the actual measurement model and required performance.

Variables

SymbolMeaningUnit

Y

Output quantity or measurement result

Unit of the output quantity

Xᵢ

ith input quantity in the measurement model

Unit appropriate to that input quantity

f

Mathematical measurement model relating inputs to the output

Depends on the model

cᵢ

Sensitivity coefficient showing how the output responds to a change in input Xᵢ

Output unit divided by input unit

u(Xᵢ)

Standard uncertainty associated with input quantity Xᵢ

Same unit as Xᵢ

u_c(Y)

Combined standard uncertainty associated with output quantity Y

Same unit as Y

u(Xᵢ, Xⱼ)

Covariance between input quantities Xᵢ and Xⱼ

Product of the units of Xᵢ and Xⱼ

Example or application

A worked example has not yet been added.

Definition

A source of measurement error is an instrument, environmental, procedural, observational, sample-related, reference-related or data-processing influence that contributes to the difference between a measured quantity value and a reference quantity value.

Purpose

The purpose of studying measurement-error sources and mechanisms is to understand how the measurement system produces deviations, identify influential factors and determine where controls, corrections, calibration or improved procedures are needed.

Importance

Identifying sources is necessary for a meaningful uncertainty evaluation and for improving measurement reliability. Without understanding the mechanism, an investigator may apply an ineffective correction, overlook an important influence or incorrectly attribute a systematic effect to random variation.

Classification Of Sources

SourceMechanisms

Instrument-related

MechanismExplanation

Zero offset

The instrument produces a nonzero indication at the reference zero input.

Sensitivity or scale-factor error

The change in indication per unit change in input differs from the intended response.

Nonlinearity

The instrument response does not follow the assumed linear relationship over its operating range.

Hysteresis

The indication depends on the direction or history of the input change.

Backlash and friction

Mechanical clearance or resistance affects the indicated position or movement.

Drift

The instrument response changes with time even when the input remains unchanged.

Limited resolution

Small changes in the measurand cannot be distinguished by the instrument indication.

Environmental

MechanismExplanation

Temperature variation

Changes instrument dimensions, material properties, electrical resistance and sensor response.

Humidity

Can affect moisture-sensitive samples, insulation and material dimensions.

Vibration

Can disturb readings in dimensional, mass, pressure and optical measurements.

Electromagnetic interference

Can introduce unwanted signals into electronic measurement circuits.

Air pressure and airflow

Can influence pressure, flow, density, buoyancy and thermal measurements.

Illumination and acoustic conditions

Can influence visual observations and sound-related measurements.

Operator and observation

MechanismExplanation

Parallax

The apparent position of an indicator changes with viewing angle.

Misalignment

The instrument axis or reference direction is not correctly aligned with the measurand.

Reaction time

A human response delay influences manual timing.

Inconsistent contact force

Different applied forces can deform the sample or alter instrument contact.

Reading and recording mistakes

An incorrect scale interpretation, digit or unit is entered into the record.

Measurement method

MechanismExplanation

Incorrect measurement model

The assumed relationship between measured inputs and the desired quantity is incomplete or unsuitable.

Incomplete sampling

The selected locations or times do not adequately represent the quantity being studied.

Uncorrected influence

A known physical effect is omitted from the measurement procedure.

Improper instrument positioning

The setup does not satisfy the geometry or operating conditions required by the method.

Sample and measurand

MechanismExplanation

Thermal expansion

The sample changes dimensions with temperature.

Deformation

The sample changes shape under measurement force or pressure.

Evaporation or moisture exchange

The sample mass or composition changes during measurement.

Contamination

Foreign material changes the property being measured.

Natural variation

The measured quantity varies spatially or temporally in the sample itself.

Reference and calibration

MechanismExplanation

Reference-standard uncertainty

The value assigned to the reference standard is not perfectly known.

Calibration procedure

The calibration method introduces uncertainty or does not adequately cover the operating range.

Traceability-chain limitations

Uncertainty is accumulated or evaluated through successive comparisons with higher-level references.

Data processing and software

MechanismExplanation

Incorrect unit conversion

A conversion factor or unit relationship is applied incorrectly.

Incorrect mathematical model

The calculation does not correctly represent the physical relationship.

Rounding and numerical precision

Premature rounding or unsuitable numerical representation changes calculated results.

Software configuration

Incorrect settings, calibration coefficients, algorithms or data selections affect the output.

Mechanism Investigation

Steps

  • Define the measurand and the conditions under which it is measured.

  • Draw or describe the measurement system, including instruments, references, samples and data-processing stages.

  • Identify possible input quantities and physical influences.

  • Determine how each influence can affect the measurement result.

  • Estimate the sensitivity of the result to important inputs.

  • Use calibration, experiments, environmental monitoring or comparison methods to investigate influential effects.

  • Apply justified corrections where appropriate and evaluate the uncertainty remaining after correction.

  • Document assumptions, procedures, instrument status and environmental conditions.

Measurement Model Example

Quantity

Density of a uniform sample

Model

ρ = m/V

Potential Sources

  • Mass indication and balance calibration

  • Volume measurement and dimensional geometry

  • Temperature-dependent sample expansion

  • Air buoyancy where relevant to the weighing method

  • Sample nonuniformity and trapped air

  • Unit conversion and numerical processing

Explanation

The density result depends on both mass and volume. An accurate mass measurement alone does not guarantee an accurate density if the volume measurement or sample model is inadequate.

Worked Examples

TitleProblemSolutionInterpretation

Temperature influence on a length measurement

A metal component is measured at a temperature different from its reference temperature. Explain why this can influence the result.

For a uniform material under a small temperature change, the approximate length change is ΔL = αL₀ΔT, where α is the coefficient of linear thermal expansion, L₀ is the reference length and ΔT is the temperature difference.

The measured length can differ because the sample itself expands or contracts. The effect should be considered when the required measurement accuracy makes it significant.

Identifying a data-processing source

A speed is calculated from a distance of 100 m and a time of 20 s, but the software uses 10 s.

Correct speed = 100 m / 20 s = 5 m/s. Incorrect software result = 100 m / 10 s = 10 m/s.

The error originates in the input data or processing configuration rather than necessarily in the measuring instruments.

Sensitivity to input uncertainty

A quantity is calculated using Y = 2X. Explain how a small uncertainty in X affects Y.

The sensitivity coefficient is ∂Y/∂X = 2. For a small input uncertainty u(X), the corresponding standard uncertainty contribution is approximately 2u(X).

A measurement model can amplify or reduce the influence of input uncertainty.

Common Misconceptions

  • All measurement errors do not originate in the instrument.

  • An environmental influence may affect the sample, instrument or both.

  • Calibration does not remove every possible measurement influence.

  • A high-resolution instrument can still have substantial calibration or environmental errors.

  • A correct calculation cannot compensate for an incorrectly defined measurand.

  • Software-generated results are not automatically free from measurement errors.

  • A correction should be supported by evidence and should not be applied without considering its uncertainty.

Applications

  • Laboratory measurement-system design

  • Dimensional metrology and precision manufacturing

  • Environmental monitoring

  • Sensor development and testing

  • Electrical and electronic measurements

  • Mass, force and pressure measurement

  • Scientific data processing

  • Calibration and quality-assurance laboratories

Related Topics

  • Introduction to Measurement Errors

  • Systematic, Random and Gross Errors

  • Measurement Instrument Characteristics

  • Measurement Uncertainty

  • Calibration and Metrological Traceability

Reference

TitleOrganizationUrl

Guide to the Expression of Uncertainty in Measurement

JCGM

https://www.bipm.org/en/committees/jc/jcgm/publications

International Vocabulary of Metrology (VIM), JCGM 200

Joint Committee for Guides in Metrology

https://www.bipm.org/en/committees/jc/jcgm/publications

NIST/SEMATECH e-Handbook of Statistical Methods

National Institute of Standards and Technology

https://www.itl.nist.gov/div898/handbook/

Related Topic

  • SCI-REF-003-073

  • SCI-REF-003-074

  • SCI-REF-003-076

  • SCI-REF-003-080

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Source verification pending; check an authoritative textbook, university, standards body, or peer-reviewed source.

Suggest a correction or source · More in Units & Measurement