EBSILON®Professional Online Documentation
The specifications about the uncertainties of the measurement values are based on assumptions and "experience values", which need not always be correct in real cases. A well-founded quality control of these assumptions in the sense of mathematical statistics is possible only by processing of additional redundant measurements and should be an obvious thing for the acceptance measurements. If the Eq. (140) is not fulfilled, then the measurement and the results obtained from it are to be rejected owing to broad errors.
EBSILON®Professional makes it possible to include redundant measurement values in the calculation. For these measurement points the flag FVAL is to be set to "only for validation". These measurement values are highlighted visually through colour flags. Measurement points, which are used in the simulation as well as in the validation mode, are highlighted by means of black flags.
For the complete evaluation after the calculation, a parameter "Chi²-test ratio" is available under "Extras"à"Model Options"à"Results", which shows the standardized error square sum relative to the F-distribution according to Eq. (140).
This value can be used as a characteristic value for the complete evaluation:
If Chi^2 > 1, then a broad error is present.
