土壤重金属分析方法间检测结果一致性判断方法——数据对t值检验法的改进与完善

Consistency Criterion Between the Results of Different Analytical Methods Among Soil Heavy Metals in the Verification Experiments-Improvement and Perfection of t-value Test Method by Data

不同分析方法测定结果间的比较,或者说如何判断两种分析方法测定结果的一致性,是检测实验室经常遇到的实际问题。其应用场景包括方法标准验证实验、新方法与经典方法间的比对,特别是在缺少标准样品、测试数据不满足t值检验的条件下,选用合适的判据尤其重要。以采用单波长激发型能量色散X射线荧光光谱法、电感耦合等离子体质谱法、波长色散X射线荧光光谱法测定土壤和水系沉积物样品中Cd、As、Pb、Cu、Zn、Ni、Cr、Co、V、Mn、Mo、Sb等重金属含量所得数据为例,探讨了《环境监测分析方法标准制订技术导则》(HJ 168—2020)附录B中规定的“数据对t值检验法”的不足及改进办法,即用数据对相对偏差和再现性限合格率作为判据,判断分析方法测定结果的一致性。对一组验证样品分别用两种分析方法进行了测定,在数据对t值检验的基础上,取各元素测定结果的相对偏差与相关限值进行比较,或者取两者的差值与基准分析方法标准规定的再现性限进行比较,利用合格率大小协助判断了两分析方法测定结果的一致性。研究结果表明,此方法符合日常质控技术规范中关于精密度的规定,实用性强,是一条将数理统计原理与专业知识相结合的合理技术路线。

The comparison of measurement results between different analytical methods is used to judge the consistency between the two methods,which is a practical problem often encountered by testing laboratories.Its application scenario includes the comparison between the new method and the classic method in the method standard verification experiment.Especially in the absence of standard samples and do not meet the t-value test data conditions,the selection of appropriate criteria is particularly important.In this paper,the shortcomings and improvement methods of“data-paired t value test method”stipulated in Technical Guideline for the Development of Environmental Monitoring Analytical Method Standards(HJ 168-2020)were proposed based on results of elements such as Cd,As,Pb,Cu,Zn,Ni,Cr,Co,V,in soil and sediment samples determined by inductively coupled plasma mass spectrometry(ICP-MS),single wavelength excitation energy dispersive X-ray fluorescence spectrometry(SW-ED-XRF)and wavelength dispersive X-ray fluorescence spectrometry(WD-XRF).The possibility of judging the consistency of analysis methods by the criteria of relative deviation of data pairs and pass rate of reproducibility limit were explored.The results showed that,for a group of verification samples,two analytical methods were used to measure them respectively.On the basis of the t value test in data-pared,the relative deviation of the determination results of each element was compared with the related limit,and the difference between them was compared with the repeatability limit of reference method or calculated by Horwitz model,which could help to judge the consistency of the results of the two analytical methods according to their pass rate.This improved method is in line with the standard of daily quality control technology about precision,and it has strong practicability.It is a reasonable technical route combining mathematical statistics principle with professional knowledge.