24/7 emergencies 814 777 6921812 230 6079WhatsApp 811 786 0768
Electrónica MonterreyElectrónica MonterreyLaboratorios Creativeit Talk to an engineer
Menu

What is the GUM uncertainty guide?

The GUM is an international document that sets out general rules for evaluating and expressing measurement uncertainty, with a method for combining contributions and stating a result with its expanded uncertainty.

GUM is the English abbreviation for the Guide to the Expression of Uncertainty in Measurement. It is the reference document that unifies the way uncertainty is estimated and communicated in laboratories around the world. Its aim is that two laboratories measuring the same thing can express their uncertainties in a comparable way. This summary is general; for details, consult the current edition of the guide and its supplements.

Main elements

  • Measurement model: the mathematical relationship between the measured quantity and the input quantities.
  • Type A evaluation: uncertainty estimated by statistical methods from series of observations.
  • Type B evaluation: uncertainty estimated from other information, such as certificates, specifications, experience or prior data.
  • Standard uncertainty: each contribution expressed as a standard deviation.
  • Sensitivity coefficients: these indicate how much each input quantity affects the result.
  • Combined uncertainty: the result of combining the contributions, normally by a weighted sum of squares.
  • Expanded uncertainty: the combined uncertainty multiplied by a coverage factor, for a stated level of confidence.

A key idea

The GUM treats uncertainty as a quantification of what is known, not only of random errors. Even systematic effects, once corrected, leave an uncertainty associated with the correction. And all available information, statistical or not, is handled with the same probabilistic logic.

Supplements and limits of the method

The method in the guide assumes a reasonably linear model and a distribution of the result close to normal. When the model is strongly nonlinear or the distributions are asymmetric, complementary methods are used, such as propagation of distributions by Monte Carlo simulation, described in supplements to the guide.

What the engineer sees in practice

Calibration certificates report an expanded uncertainty with its coverage factor, following this approach. Knowing how it was built lets you judge whether it is reasonable, whether it is dominated by a single factor and what can be improved.

We do real tests, on video! Watch them on our YouTube channel →

Do you have a problem nobody has been able to measure?

Talk directly to the engineer in charge, with no middlemen.

Talk to an engineer