能力验证均匀性检验的两种统计分析方法比较及结果合理性判断  

Comparison of Homogeneity Checking Statistics Methods of Profi ciency Testing Items

作　　者：戴雪伟[1] 姜瀛洲[1] 王璐[1] 

机构地区：[1]上海出入境检验检疫局,上海200135

出　　处：《检验检疫学刊》2014年第5期28-31,共4页Journal of Inspection and Quarantine

基　　金：上海出入境检验检疫局科研项目(HK030-2014)

摘　　要：分析了0.3σ方法与F检验两种统计方法各自的优缺点和适用范围。0.3σ方法以能力统计量作为评价的依据,F检验使用正态分布下的显著性偏差作为临界条件。因此,在使用0.3σ方法时,应关注样品的不均匀性和方法的精密度;在使用F检验时,应预先进行数据正态性的验证。Homogeneity checking is of essential importance to proficiency testing programs. The most common-used homogeneity checking statistics methods is, 0.3 ？σ and F-test. Based on different statistical design, this essay introduced the advantages and disadvantages of the two statistics methods and their scopes of application. The 0.3 ？σmethod quotes the standard deviation of PTs, while F-test uses the significance of a normal distribution deviation as the critical condition. Therefore, the heterogeneity of the PTs and the precision of the testing methods should be taken into account in the implementation of the 0.3 ？σ method. For F-tests, normality should be checked in advance.

关 键 词：能力验证样品 均匀性检验 统计方法 显著性偏差 正态性 

分 类 号：O652[理学—分析化学] N32[理学—化学]