✦ScienexySCIENCE KNOWLEDGE LIBRARY
reference · SCI-REF-003-009

Measurement Process and Measurement Models

A measurement process is the organized set of operations used to obtain a measurement result for a defined measurand. A measurement model is the mathematical or physical representation connecting the measurand to relevant input quantities, instrument indications, corrections and influence factors. Together, they provide the foundation for reliable, interpretable and uncertainty-aware measurement.

General measurement model: y = f(x₁, x₂, ..., xₙ). For a rectangular object's density: ρ = m/(lwh). For independent inputs and a first-order approximation: u²(y) ≈ Σᵢ[cᵢ²u²(xᵢ)], where cᵢ = ∂f/∂xᵢ. Correlation terms must be added when inputs are correlated.

Explanation

A measurement process is more than the act of using an instrument. It is a structured sequence of operations designed to obtain a value that can reasonably be attributed to a specified measurand. It includes defining the quantity, choosing a method, selecting instruments, preparing the measurement system, controlling conditions, collecting data, applying corrections, evaluating uncertainty and reporting the result.

The first essential stage is the specification of the measurand. The measurand must be described clearly enough to identify the quantity intended to be measured. For example, measuring the temperature of a metal block requires specifying whether the target is surface temperature, internal temperature or a representative temperature under an assumed uniform condition. Without this information, different measurement procedures may produce values that are not directly comparable.

The next stage is selecting a measurement method and procedure. A measurement method describes the general logical approach used to determine the measurand. A measurement procedure provides detailed instructions for carrying out the measurement, including instrument configuration, sample preparation, measurement sequence, environmental requirements and data recording.

Instrument selection depends on the measurand, measurement range, required resolution, expected uncertainty, operating environment and practical constraints. A suitable instrument must be capable of measuring the required quantity without unacceptable influence on the system. Instrument performance characteristics such as sensitivity, stability, linearity, repeatability and response time may be relevant.

The measurement environment must also be considered. Temperature, humidity, pressure, vibration, electromagnetic interference, illumination and other influences can affect the instrument or measurand. In high-precision work, environmental conditions may need to be controlled, monitored or included in the measurement model.

During data acquisition, indications and relevant conditions should be recorded systematically. Repeated measurements may be required to evaluate variability. Where appropriate, measurements should be taken at different positions, times or operating conditions to characterize the system adequately.

The raw instrument indication may require corrections. These can include calibration corrections, zero corrections, environmental corrections, geometric corrections or corrections for known influence quantities. A correction should be applied only when its definition, applicability and uncertainty are understood. Incorrect or unjustified corrections can introduce additional errors.

A measurement model describes the relationship between the measurand and the quantities that influence its determination. In general, the model can be written as y = f(x₁, x₂, ..., xₙ), where y is the output quantity and xᵢ are input quantities. Some inputs may be obtained directly from instruments, while others may come from reference data, constants, calibration certificates or environmental measurements.

For example, density can be calculated using mass and volume. If volume is determined from geometric measurements, the complete model may depend on length, width and height. For a rectangular solid, ρ = m/(lwh). The uncertainty of density then depends on the uncertainty of mass and each dimension, as well as any correlations and assumptions about the shape.

Measurement models can be simple or complex. A simple model may involve one ratio or product. A complex model may include nonlinear equations, multiple sensors, environmental corrections, dynamic response, signal processing or numerical simulation. In advanced metrology, a model may also include time dependence, spatial variation and correlations between input quantities.

The measurement model is central to uncertainty evaluation. Uncertainty in the output depends on the uncertainties of input quantities and on the sensitivity of the output to those inputs. For a model with independent inputs, a first-order approximation can use sensitivity coefficients and the law of propagation of uncertainty. When inputs are correlated, covariance terms must also be included.

A measurement result should be reported with enough information to support interpretation and comparison. Depending on the application, this may include the measurand definition, value, unit, uncertainty, coverage factor, measurement method, relevant conditions, calibration status and traceability information.

The measurement process should be validated or verified to the extent required by its intended use. Validation evaluates whether the method is suitable for the intended purpose. Verification provides evidence that specified requirements have been fulfilled. These concepts are particularly important in laboratories, industrial measurement systems and regulated applications.

A well-designed measurement process minimizes avoidable errors, makes the model assumptions explicit, supports traceability and provides a defensible estimate of uncertainty. It also allows other users to understand how the result was obtained and whether it is appropriate for a particular scientific or engineering decision.

Variables

SymbolMeaningUnit

y

Output quantity or measurand value.

Depends on the measurand.

xᵢ

Input quantity in the measurement model.

Depends on the input.

f

Mathematical or physical model relating inputs to the output.

Depends on the model.

u(y)

Standard uncertainty associated with the output estimate.

Same as y.

u(xᵢ)

Standard uncertainty associated with input quantity xᵢ.

Same as xᵢ.

cᵢ

Sensitivity coefficient describing how the output changes with input xᵢ.

Unit of y divided by unit of xᵢ.

ρ

Mass density.

kg/m³.

m

Mass.

kg.

l, w, h

Length, width and height of a rectangular object.

m.

Example or application

A worked example has not yet been added.

Definition

A measurement process is a set of operations performed to determine the value of a measurand. A measurement model is the mathematical or physical relationship among the quantities involved in the measurement, including the measurand, input quantities and relevant influence factors.

Purpose

To explain how measurement operations are organized and how measurement models connect instrument indications and input quantities to a defensible measurement result.

Importance

A defined process and valid model are essential for reliable measurements, uncertainty evaluation, calibration, traceability, method validation and reproducible scientific work.

Measurement Stages

StageDescription

Measurand definition

Identify and specify the exact quantity intended to be measured.

Method selection

Choose an appropriate measurement principle and general procedure.

Instrument selection

Select measuring equipment suitable for the range, resolution and uncertainty requirements.

Measurement preparation

Prepare the sample, instrument and environment according to the procedure.

Data acquisition

Collect instrument indications and relevant experimental observations.

Corrections and processing

Apply justified corrections and calculate the measurement result.

Uncertainty evaluation

Identify and evaluate important uncertainty contributions.

Validation and review

Check that the procedure and model are suitable and that the result is consistent with the available evidence.

Reporting

Report the result, unit, uncertainty and relevant supporting information.

Measurement Model Types

TypeDescriptionExample

Direct-response model

Relates an instrument indication to the measurand through a calibrated response.

A calibrated thermometer's indication related to temperature.

Algebraic model

Uses mathematical operations such as addition, multiplication, division or powers.

Density calculated from mass and volume.

Geometric model

Determines a quantity from dimensions and geometric relationships.

Volume of a cylinder calculated from radius and height.

Physical model

Uses a physical law or theory to connect measured inputs with the measurand.

Resistance determined from voltage and current under appropriate conditions.

Statistical model

Uses statistical relationships to estimate quantities from observations.

Estimating a population mean from sampled measurements.

Dynamic model

Accounts for time-dependent system behavior and instrument response.

Determining the true temperature of a changing system using a sensor response model.

Measurement Quality Factors

  • Clarity of the measurand specification.

  • Suitability of the measurement method.

  • Instrument calibration and performance.

  • Environmental stability and influence control.

  • Sample preparation and measurement geometry.

  • Operator competence and procedural consistency.

  • Data processing and correction validity.

  • Measurement uncertainty and input correlations.

  • Traceability and documentation.

  • Method verification and validation.

Worked Examples

TitleGivenModelCalculationAnswerInterpretationMethod

Measurement model for density of a rectangular block

  • Mass = 0.600 kg

  • Length = 0.100 m

  • Width = 0.050 m

  • Height = 0.020 m

ρ = m/(lwh)

Volume = 0.100 × 0.050 × 0.020 = 0.0001 m³. Density = 0.600/0.0001 = 6000 kg/m³.

6000 kg/m³

The density is calculated from four input measurements or values. Their uncertainties contribute to the uncertainty of the density result.

A temperature measurement process

A defensible temperature result requires more than reading a display; it requires a defined measurand and a suitable measurement process.

  • Define the measurand as the temperature at a specified location in a sample.

  • Select a thermometer with an appropriate range and uncertainty.

  • Check calibration status and instrument condition.

  • Place the sensor at the specified location.

  • Allow sufficient time for the sensor and sample to approach thermal equilibrium.

  • Record the indication and relevant environmental conditions.

  • Apply justified corrections and evaluate uncertainty.

  • Report the temperature with its unit and uncertainty.

Limitations

  • A measurement model is only as reliable as its assumptions and input data.

  • Simplified models may omit important influence quantities.

  • Nonlinear models may require more advanced uncertainty evaluation.

  • Correlations between input quantities can make simple uncertainty calculations invalid.

  • Dynamic systems may require time-dependent models rather than static equations.

  • A well-documented procedure cannot compensate for an unsuitable measurement principle.

Common Misconceptions

  • The measurement process is not limited to data acquisition.

  • A measurement model is not necessarily an ideal theoretical law; it must represent the actual measurement procedure.

  • A numerical result does not become reliable merely because a computer calculated it.

  • All important input quantities and corrections must be considered in the model.

  • Measurement uncertainty is associated with the result and is not simply the instrument's display resolution.

Related Concepts

  • Measurement

  • Measurand

  • Direct and Indirect Measurement

  • Measurement Uncertainty

  • Calibration

  • Method Validation

  • Measurement Traceability

Reference

  • Evaluation of Measurement Data — Guide to the Expression of Uncertainty in Measurement (GUM), JCGM 100:2008.

  • International Vocabulary of Metrology (VIM), JCGM 200:2012.

  • The International System of Units (SI Brochure), 9th edition, BIPM.

Related Topic

  • SCI-REF-003-001

  • SCI-REF-003-008

  • SCI-REF-003-010

  • SCI-REF-003-073

  • SCI-REF-003-081

  • SCI-REF-003-086

  • SCI-REF-003-093

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