土壤中铅含量的测定能力验证结果分析

Analysis of proficiency testing results for determination of lead in soil

组织了土壤中铅含量的测定能力验证计划,有120家实验室参加,其中有115家提交了结果。综合考虑能力验证目的、测试方法特性及满意率情况等因素,采用稳健算法A计算结果稳健均值和稳健标准差,再以稳健Z比分数来评估实验室能力。该次能力验证满意率为88.7%,基本能够反映参与实验室的测试水准。对不同测试方法的统计结果分析表明,两种测试方法得到的稳健均值和结果满意率均很接近,方法差异并不显著。从测试数据的尧敦图可以看出,少数实验室存在较大随机性或者系统性偏差。最后,结合实验室原始数据分析了可疑和离群结果可能存在的具体原因,为检验机构提高土壤中铅含量检测能力提供参考依据。

In the proficiency testing scheme for the detection of lead in soil,the testing results were submitted by 115 of 120 registered labs.Considering the purpose of proficiency testing,the characteristics of testing methods and the satisfaction rate,robust statistical Z-Score of the assigned value and standard deviation by robust algorithm A were used to evaluate lab capability.The satisfaction rate of this proficiency testing was 88.7%,which could basically reflect the test level of participatory labs.The robust mean values and satisfaction rates of the two different detection methods were both nearly approaching,and the difference between the two methods was not significant.In addition,from the Youden plot of sample pairs,it could be seen that a few labs had large random or systematic deviation.Finally,combined with the original data of faulty labs,the particular problems that causes suspicious and outlier results were explored,which provides a reference for the testing agency to improve the performance of determining lead in soil.