Hayashi Y, Matsuda R
Kokuritsu Iyakuhin Shokuhin Eisei Kenkyusho Hokoku. 1997(115):206-7.
The precision of instrumental measurements, usually expressed by SD (standard deviation) or RSD (relative standard deviation) is of importance in a society as well as in a world of analytical chemistry. For example, the detection limit and confidence intervals of calibration lines which are based on the precision are inevitable for regulatory science. As is well-known in statistics, however, a small number of repetitive experiments (e.g., 5) lead to a poor reliability of the precision. This presentation shows a method (called FUMI theory) which can provide a precision from a single measurement of noise and signal without any repetitive measurements. The statistical reliability of the precision from the FUMI theory corresponds to that from the repetitive method of 50 experiments.
仪器测量的精度通常用标准差（SD）或相对标准偏差（RSD）来表示，这在社会以及分析化学领域都很重要。例如，基于精度的校准曲线的检测限和置信区间对于监管科学来说是不可避免的。然而，正如统计学中所熟知的那样，少量的重复实验（例如5次）会导致精度的可靠性较差。本报告展示了一种方法（称为FUMI理论），该方法无需任何重复测量，仅通过单次测量噪声和信号就能提供精度。FUMI理论得出的精度的统计可靠性与50次实验的重复方法得出的相当。
