“精密度法则”是分析化学的基础技术

The “Precision Rule” Is the Basic Technology of Analytical Chemistry

摘要: 实践证明，“精密度法则”，即“保持各个因素对同一测定系列各个样品影响的一致性”，是分析化学的基础技术：1) 可保证光度法“校准曲线”的高度直线性。对它几何图形的解析，导出了光度法的“两标样测定法”。它与分析化学“校准曲线”的共同特点，是它们的直线方程都是“两点式”而不是“点斜式”。光度法的“两标样测定法”同样适用于分析化学，只是把其信息量吸光度A替换成共同的信息量I。实例证明了这种替换的正确性，它展现了分析化学的基本原理：物质信息量的改变量正比于其物质含量的改变量；2) 保证了分析化学的样本测得值在极差R区间内的正态分布，其图解展现了“小样本”最大残差eR处于0.5R~R之间。作者参照国标正态样本离群值判断，在大量n > 5的“小样本”中寻找eR，使其超出(X'±eR)的测得值即应剔除统计离群值，最终确定eR = 0.65R，超出(X'±eR)的测得值即统计离群值应予剔除，确保了测定结果的可靠性；3) 保证了样本测量的重复性，在其A类标准不确定度公式的基础上，建立了样本测定结果的不确定度数学模型。表明“小样本”均值的不确定度是样本极差及样本容量的函数，而且应小于0.5R。既符合了JJF1059.1-2012《测量不确定度评定与表示》要求，简便且切合实际。

Abstract: Practice has proved that the “precision rule”, that is “to maintain infection consistency of each factor for each sample in the same measurement series”, is the basic technology of analytical chemistry: 1) It can ensure the high straightness of the photometric calibration curve. The analysis of its geometric figure has led to the “two standard sample determination method” of photometry. The common characteristic of analytical chemistry “calibration curves” is that their straight line equations are “two point equations” rather than “point oblique equations”. The “two standard sample determination method” of photometry is also applicable to analytical chemistry, but its information content absorbance A is replaced by the common information content I. The example proves the correctness of this replacement, which shows the basic principle of analytical chemistry: The amount of change in material information is proportional to the amount of change in its material content; 2) It ensures the normal distribution of the measured values of analytical chemistry samples within the range R, and its diagram shows that the maximum residual eR of “small samples” is between 0.5R and R. The author refers to the national standard normal sample outlier to judge, look for eR in a large number of “small samples” with n > 5, and eliminate statistical outlier if the measured value exceeds (X'±eR). The final determination of eR = 0.65R. The measured value exceeding (X'±eR), i.e. statistical outlier, shall be eliminated, which ensures the reliability of the measurement results; 3) To ensure the repeatability of sample measurement, the mathematical model for the uncertainty of sample measurement results was established based on its Class A standard uncertainty formula. The uncertainty of the mean of “small samples” is a function of sample range and sample size, and should be less than 0.5R. It meets the requirements of JJF1059.1-2012 “Evaluation and Expression of Uncertainty in Measurement” and is concise and practical.