engleski

Statistics in Metrology

The two scientific disciplines metrology and statistics are closely related. While the possibility of measurement is a prerequisite for the collection of data in any statistical study it is also true that statistical reasoning, methods and tools are essential for rational quantification of measurement results. The purpose of each measurement is to determine the value of the measured quantity. Since measurements are not perfect due to the effects of random influences and the limited possibilities of correcting systematic effects, a measurement result is only an approximation or an estimation of the value of the measured quantity. In order to deal with this issue, unknown variables, which must be estimated, have been included into the mathematical measurement model. The lack of knowledge about input values, changes in the results in terms of the repeatability and reproducibility as well as the uncertainty associated with the mathematical model, all contribute to measurement uncertainty. This paper presents examples of the application of statistics in the area of measurement system quality assessment, process capability assessment as well as in the evaluation of measurement uncertainty and inter-laboratory comparisons of results.

Statistics; Metrology; Measurement Uncertainty; Comparison Measurements

nije evidentirano