✦ScienexySCIENCE KNOWLEDGE LIBRARY
reference · SCI-REF-003-078

Absolute, Relative and Percentage Error

Absolute, relative and percentage error are numerical ways of describing the difference between a measured value and a reference value. Absolute error expresses the magnitude of the difference in the same unit as the measured quantity. Relative error compares that magnitude with the reference value, while percentage error expresses the relative error as a percentage. These measures are useful for comparing measurement deviations, evaluating experimental results and communicating the scale of an error.

Signed error: e = xₘ − xᵣ
Absolute error: Eₐ = |xₘ − xᵣ|
Relative error: Eᵣ = |xₘ − xᵣ| / |xᵣ|, for xᵣ ≠ 0
Percentage error: Eₚ = [|xₘ − xᵣ| / |xᵣ|] × 100%, for xᵣ ≠ 0
If signed relative deviation is required: eᵣ = (xₘ − xᵣ) / xᵣ, for xᵣ ≠ 0.

Explanation

A measured value may differ from a reference value by a positive or negative amount. The signed difference indicates the direction of the deviation, while absolute error expresses only its magnitude. Relative and percentage errors describe the size of the deviation compared with the reference value.

Signed error is calculated by subtracting the reference value from the measured value. A positive signed error indicates that the measured value is greater than the reference; a negative signed error indicates that it is smaller. The sign is useful when investigating systematic bias because it shows the direction of the deviation.

Absolute error is the magnitude of the signed error. It is expressed in the same unit as the measured quantity. For example, if a reference length is 10.00 cm and a measurement is 10.15 cm, the signed error is +0.15 cm and the absolute error is 0.15 cm.

Relative error compares the absolute error with the magnitude of the reference value. It is dimensionless because the numerator and denominator have the same unit. It provides a scale-independent way to describe the deviation, making it easier to compare measurements involving different units or magnitudes.

Percentage error is relative error multiplied by 100%. It is commonly used in school experiments, laboratory reports and practical measurement comparisons. The reference value must be nonzero for the standard relative and percentage error formulas. When the reference value is zero or extremely close to zero, these measures are undefined or unstable, and an alternative measure should be selected.

These formulas are most straightforward when the reference value is sufficiently reliable. If the reference itself has significant uncertainty, the calculated difference should not be interpreted as an exact error. A proper comparison should consider the uncertainty of both values and whether they refer to the same measurand under comparable conditions.

Absolute, relative and percentage error should also be distinguished from measurement uncertainty. Error is a difference from a reference value. Uncertainty characterizes the dispersion of values attributed to the measurand. A percentage error calculated from one experiment does not automatically represent the measurement uncertainty or the probability that a result is correct.

When a result is derived from multiple measured quantities, the uncertainty of the result may depend on the uncertainties of all inputs and their correlations. In such cases, a single percentage error calculation against an assumed reference may be insufficient to describe the reliability of the result.

Variables

SymbolMeaningUnit

xₘ

Measured quantity value

Unit of the measurand

xᵣ

Reference quantity value

Unit of the measurand

e

Signed measurement error

Unit of the measurand

Eₐ

Absolute error magnitude

Unit of the measurand

Eᵣ

Relative error magnitude

Dimensionless

Eₚ

Percentage error

%

Example or application

A worked example has not yet been added.

Definition

Signed Error

The measured value minus the reference value, retaining the sign of the difference.

Absolute Error

The absolute magnitude of the difference between the measured value and the reference value.

Relative Error

The absolute error divided by the magnitude of the nonzero reference value.

Percentage Error

The relative error multiplied by 100%.

Purpose

The purpose of these error measures is to quantify the difference between a measurement and a reference value in ways that communicate the magnitude, direction and relative significance of the deviation.

Importance

These measures help students and researchers evaluate experimental results, compare measurement methods, identify potential bias and communicate deviations clearly. They are especially useful when the same type of quantity is measured at different scales or with different units.

Interpretation Of Error Measures

MeasureInformationUnit

Signed error

Indicates both the magnitude and direction of the difference.

Same as the measurand

Absolute error

Indicates the magnitude of the difference without direction.

Same as the measurand

Relative error

Indicates the error magnitude in relation to the reference value.

Dimensionless

Percentage error

Expresses the relative error as a percentage.

%

Calculation Procedure

  • Identify the measured value and the reference value.

  • Confirm that both values represent the same physical quantity under comparable conditions.

  • Convert both values to the same unit.

  • Subtract the reference value from the measured value to obtain signed error.

  • Take the absolute value when the error magnitude is required.

  • Divide by the magnitude of the nonzero reference value to obtain relative error.

  • Multiply the relative error by 100 to obtain percentage error.

  • Report a suitable number of digits and consider the uncertainty of the reference.

Special Cases

CaseExplanation

Reference value equals zero

Relative and percentage error are undefined because division by zero is not possible. Report the absolute difference or use a suitable application-specific metric.

Reference value is close to zero

Relative error can become extremely large and unstable. Absolute error or an alternative normalized measure may be more informative.

Negative reference value

For error magnitude, use the absolute value of the reference in the denominator. Preserve the sign separately if the direction of the deviation matters.

Reference value has uncertainty

The measured-reference difference is not an exact statement of the measurement error. Consider uncertainty in both values and their relationship.

Reference and measured values use different units

Convert them to a common unit before subtraction or division.

Worked Examples

TitleProblemSolutionInterpretation

Absolute and percentage error in length

A reference length is 25.00 cm, while the measured length is 24.80 cm. Calculate signed error, absolute error and percentage error.

Signed error = 24.80 − 25.00 = −0.20 cm. Absolute error = 0.20 cm. Relative error = 0.20 / 25.00 = 0.008. Percentage error = 0.008 × 100% = 0.8%.

The measurement is 0.20 cm below the reference, corresponding to a deviation magnitude of 0.8%.

Comparing errors for different quantities

Measurement A differs from its reference by 2 mm when the reference is 100 mm. Measurement B differs by 0.5 m when the reference is 50 m. Compare their percentage errors.

A: (2/100) × 100% = 2%. B: (0.5/50) × 100% = 1%.

Although the absolute errors have different magnitudes and units, Measurement A has the larger relative deviation.

Negative signed error

A reference voltage is 12.00 V and the measured voltage is 11.70 V.

Signed error = 11.70 − 12.00 = −0.30 V. Absolute error = 0.30 V. Percentage error = (0.30/12.00) × 100% = 2.5%.

The negative signed error indicates that the measured voltage is lower than the reference.

Zero reference value

A sensor should indicate 0.00 V under a specified zero-input condition but indicates 0.04 V.

Absolute difference = |0.04 − 0.00| = 0.04 V. Relative and percentage error using the standard reference-based formulas are undefined.

The absolute difference is useful for describing the zero offset. A different normalization may be used only if it has a clearly defined basis.

Error And Uncertainty Distinction

Error

A difference between a measured value and a reference value.

Uncertainty

A parameter describing the dispersion of values attributed to the measurand.

Explanation

A calculated percentage error is not the same as percentage uncertainty. Percentage uncertainty is commonly formed by dividing an uncertainty by the magnitude of the measured estimate, where meaningful. It describes the relative scale of uncertainty, not a known deviation from a reference.

Common Misconceptions

  • Absolute error and relative error are not the same quantity.

  • Absolute error has units, while relative error is dimensionless.

  • Percentage error is not calculated by dividing by the measured value unless a particular convention explicitly defines that denominator.

  • The sign of signed error should not be discarded when the direction of bias matters.

  • Relative error is undefined when the reference value is zero.

  • A small percentage error does not automatically establish that a measurement is sufficiently reliable for every application.

  • Percentage error does not replace a complete uncertainty evaluation.

Applications

  • School and university laboratory experiments

  • Instrument calibration comparisons

  • Engineering measurements

  • Experimental physics and chemistry

  • Sensor testing

  • Manufacturing quality control

  • Numerical model validation

  • Scientific reporting

Related Topics

  • Introduction to Measurement Errors

  • Accuracy, Precision, Trueness and Repeatability

  • Significant Figures and Rules of Numerical Reporting

  • Measurement Uncertainty

  • Error Detection, Correction and Reduction Techniques

Reference

TitleOrganizationUrl

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/

Guide to the Expression of Uncertainty in Measurement

JCGM

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

Related Topic

  • SCI-REF-003-073

  • SCI-REF-003-076

  • SCI-REF-003-079

  • 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