3.1 Decision rules describe how measurement uncertainty is accounted for when a statement of conformity is made alongside a reported result. Eurofins | BLC will, when required, apply the below decision rule as standard unless the client requests an alternative which can then be discussed further.
3.2 A decision rule is only required when a statement of conformity is reported with a measured value.
3.3 If a statement of conformity is not requested (e.g., the testing is carried out “For Information Only”) then no decision rule is required and the measured value will be reported, along with the measurement uncertainty value.
3.4 When a statement of conformity is required against a specification or standard, and the defined specification states how to account for the measurement uncertainty, then the test result will be reported in accordance with the requirements of the selected specification or standard.
3.5 For test methods where a subjective assessment of a property has been made, a simple acceptance decision rule has been used based upon the conditions of the testing being within the criteria specified within the test method.
3.6 If a statement of conformity is requested to a defined specification or standard, and the standard does not account for measurement uncertainty, then Eurofins | BLC will apply a “simple acceptance shared risk decision rule” or a “non-binary simple acceptance decision rule” to all compliance, pass/fail statements. In this case, uncertainty values for all test methods will be stated on the test report, along with the results and one of the following statements where applicable.
(a) “The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor of k =2, providing a level of confidence of approximately 95%.
Unless otherwise specified all compliance and pass/fail statements are binary simple acceptance based on the tolerance interval and, except for graded methods, a test uncertainty ratio (TUR) greater than 4:1. For graded methods, the TUR will drop to as low as 0.5:1 when the tolerance limits are within a grade of the upper scale limit. The uncertainty budgets are stated for each test method, these are for reference, and should be considered when results are on or close to specification limits / requirements. In such cases, it should be noted that the risk of false acceptance or rejection may be as high as 50%, for further information please refer to ILAC G8.”
(b) “A non-binary simple acceptance decision rule based on guard bands has been used as the decision rule. The guard band is equal to the expanded standard deviation stated in the test result table. When the difference between the test result and the requirement is less than or equal to the expanded uncertainty of measurement, then a risk of false acceptance or false rejection is possible. The risk of false acceptance or false acceptance is 2.5% based on a conformance probability of 97.5%.”
3.7 It should be noted that, for some test methods, an uncertainty of measurement value may not be available, this will be due to the nature of the test method and reported result.