Probability method for evaluating measurement errors with the distribution of the measured quantity taken into consideration

Conclusions

• 1. The distribution of the random measurement error depends in a general case on the distribution of the measured quantity. This relationship becomes essential and should be taken into account when the root-mean-square error of the measuring device and the root-mean-square deviation of the measured quantity are of the same order. In order to account for the above relationship for normal distribution laws of the measured quantity and the measuring device error, it is necessary to apply to the measurement results correction m_δ/y according to (10), and to determine the root-mean-square measurement error from (8).

• 2. If the root-mean-square deviation of the measured quantity is of a higher order, the effect of its distribution becomes unessential. In such a case, the distribution of the measurement error virtually coincides with that of the measuring device error, and it is no longer necessary to account for the distribution of the measured quantity.