scope:

This International Standard describes a statistical, so-called single-measurement, model (SD model) for using single measurements for production quality control purposes at a producer's manufacturing site, even if a standard test method specifies replicate measurements. The statistical model is also applicable to test methods which call for repeated operations. A general approach to the precision statement in test method standards which produce numerical results is also described. The statistical model is only applicable to test methods which give results that follow a normal (i.e. Gaussian) distribution.

The principle of the statistical model is based on the determination of upper and lower production limits, taking into account the accuracy and the reproducibility of the test method, the latter being added to the production limits in order to define an area outside the production limits in which a test result can fall owing to the nature of the production process and/or the test method.

This International Standard is designed for project managers, heads of laboratories and production managers. However, the support of a statistician is highly recommended, and sometimes indispensable, for providing the necessary technical backup and statistical analysis.

If this International Standard is used in combination with a test method standard, this needs to be clearly indicated in all relevant documents, e.g. product specifications, production process specifications and contracts.