方法标准验证实验数据中离群值的识别

对方法标准验证实验中测量数据进行合格性审核,对于后续方法精密度计算是一个重要环节。文献中识别离群值的Grubbs法、Dixon法等经典方法有时不能满足要求。探讨了用稳健统计法识别离群值的可行性。基于2套文献数据和XRF方法标准验证实验精密度测量数据,对Grubbs法、Dixon法、Mandel h检验法、质控指标法和稳健统计法(四分位法、迭代法、合格数据范围判定法)进行了比较。结果表明:稳健统计法可有效识别离群值。但四分位法存在过度"检出"现象。综合考虑多种方法识别结果有利于提高离群值判定结论的可靠性。对于个别难以判断的情形,可借助质控指标、技术要求以及数据是否剔除对实验室间标准偏差的影响进行取舍。     

基于稳健统计方法对实验室比对中土壤汞的质量控制指标研究

利用稳健四分位间距法和迭代法2种稳健统计方法,通过对全国31个省、自治区、直辖市范围内共334家环境监测单位开展土壤中汞的实验室间比对,对比对测定结果和质控数据进行统计分析,系统性地研究了土壤中汞的质量控制指标。建议实际监测工作中土壤汞的实验室间相对标准偏差范围为7%～19%,相对误差控制指标为±8%,低浓度水平下可适当放宽至±10%,加标回收率控制范围为81%～109%,为日常监测开展土壤汞的质量控制工作提供了评价依据,具有广泛的应用价值和较好的指导作用。     

环境检测领域能力验证工作的组织及评价方法

介绍了环境检测领域能力验证的基本情况和统计方法，以及当前国内外能力验证计划提供者采用的数据处理方法，分析了该领域能力验证的发展趋势。指出环境检测领域化学分析方面的能力验证计划一般采用参加者结果的中位值或稳健均值作为指定值，采用标准化四分位距或稳健标准偏差作为能力评定标准差，以 z比分数法评价参加者的能力。     

实验室能力验证中的分割水平检测样品与稳健统计技术

论述了分割水平检测样品和稳健统计技术在对实验室能力验证活动中的应用,并对导致CNAL T040水中金属元素检测能力验证计划中部分结果异常的原因进行了分析。     

环境监测实验室水中砷、汞监测能力考核结果评价

报道了全国环境监测系统136个实验室地表水中As、Hg监测能力考核的结果。采用四分位稳健统计法处理各实验室的测定数据,砷和汞结果均满意的实验室为101家。分析了各实验室质量管理体系运行中存在的问题,质控环节的问题是质量体系运行中较为突出的问题。提出了综合数据结果和质量体系运行情况进行结果评价的方法。     

用稳健统计法处理方法标准验证实验中方法检出限数据

方法检出限是分析方法基本性能参数之一。根据《环境监测分析方法标准制修订技术导则》(HJ 168—2010)的要求,此参数需要6家以上实验室基于特定样品测试数据的标准偏差得到。目前,标准文本大多采用各家实验室方法检出限中最大值。研究提出用稳健统计法处理方法验证数据,无需识别和删去离群值,可将离群值的影响降到最小。基于4组公开发表数据和2组实测数据,比较了不同方法计算结果。结果表明:参与统计的3组数据29个检测项目中,常规方法确认的方法检出限被判为离群值占27.6%。取最大值存在数值偏高的风险。提出了用曼德尔k检验法识别1组方法检出限数据中的离群值,该方法同样可以用于实验室数据与标准文本中方法检出限符合性判断。     

应用稳健马氏距离评价实验室间比对能力考核

基于稳健马氏距离的多元统计算法，综合评价了全国环境监测系统96个实验室4种有机氯农药监测能力考核的结果，从中筛选出25个考核结果存在异常的实验室。将多元统计结果与稳健Z比分数结果进行比较，发现基于稳健马氏距离筛选出的异常实验室其稳健Z比分数结果也存在一定的问题。同时简单介绍了稳健马氏距离及其相关的异常值筛选的算法原理，并详细介绍了其基于R语言的实现过程。     

2种统计方法在一次实验室间比对中的应用比较

比较了稳健统计法和误差系数法在水中苯系物实验室间比对结果中的应用，分别采用两种方法对比对结果进行了分析，探讨了两种统计方法的差异和适用范围。结果显示，两种方法对此次比对结果的分析结论基本一致，在实际应用中应跟据实验室间比对的目的和样品特点选择合适的统计方法。     

氟化物和总硬度分析项目相关性的检验

目前,通过中国实验室国家认可委员会的实验室,按照ISO/IEC17025:2005《校准和检测实验室能力通用要求》,都要制定并实施质量控制程序以监控检测和校准的有效性,所得数据的记录方式应便于可发现其发展趋势,如可行,应采用统计技术对结果进行审查。文章利用认可准则规定的五种方法中的一种方法,就分析一个物品不同特性结果的相关性,即利用某地下水水质的监测数据对氟化物、总硬度进行一元线性回归,检验其相关性是负相关,并建立了回归方程,绘制回归直线,充分验证了监测数据的有效性和可靠性,从而证明实验室利用的质量监控方法有效。     

能力验证是环境监测质量保证的重要环节

能力验证是近几年来开展的一项新的技术活动,是环境监测质量保证的重要环节.有组织、有计划的开展环境监测系统实验室的能力验证来确定实验室的能力并确保实验室维持较高的检测水平,也是保证环境监测数据具有科学性、准确性、可比性的必要措施.本文对此进行了论述.     

Sediment grain size estimation using airborne remote sensing, field sampling, and robust statistic

Remote sensing has been used since the 1980s to study parameters in relation with coastal zones. It was not until the beginning of the twenty-first century that it started to acquire imagery with good temporal and spectral resolution. This has encouraged the development of reliable imagery acquisition systems that consider remote sensing as a water management tool. Nevertheless, the spatial resolution that it provides is not adapted to carry out coastal studies. This article introduces a new methodology for estimating the most fundamental physical property of intertidal sediment, the grain size, in coastal zones. The study combines hyperspectral information (CASI-2 flight), robust statistic, and simultaneous field work (chemical and radiometric sampling), performed over Santander Bay, Spain. Field data acquisition was used to build a spectral library in order to study different atmospheric correction algorithms for CASI-2 data and to develop algorithms to estimate grain size in an estuary. Two robust estimation techniques (MVE and MCD multivariate M-estimators of location and scale) were applied to CASI-2 imagery, and the results showed that robust adjustments give acceptable and meaningful algorithms. These adjustments have given the following R(2) estimated results: 0.93 in the case of sandy loam contribution, 0.94 for the silty loam, and 0.67 for clay loam. The robust statistic is a powerful tool for large dataset.     

How well should laboratories perform in proficiency tests

The long-term performance of laboratories is examined using results from three proficiency testing programmes in the water and environmental monitoring sector. A medium term proficiency test pass rate of greater than approximately 75% is suggested as a minimum target for routine analysis to demonstrate fitness for purpose. This pass rate is proposed as a practical yardstick to be used by QA managers and accreditation bodies. This recommendation is contingent on a laboratory's continued and largely uninterrupted participation in a PT programme.     

采用能力验证数据评估环境监测结果测量不确定度

测量不确定度评估是实验室检测能力的体现,能力验证是实验室质量控制的有效方法,对环境监测领域实验室采用能力验证数据进行测量结果不确定度评估方法进行了研究。依据Nordtest准则,根据实验室内再现性标准差和测量偏倚,评估了重铬酸钾法测定水中质量浓度为100 mg/L的化学需氧量测量结果的相对不确定度为6.00%。该评估方法避免了ISO GUM评定方法自下而上不确定度评估过程的繁琐,还充分考虑了实验室内外误差的来源,能够促进环境监测结果不确定度评定的一致性。     

A proficiency testing scheme for aromatic hydrocarbons in air by the manual thermal desorption-GC method: a comparison of laboratory performance with the uncertainty requirements of the European Union Ambient Air Directive

The Workplace Analysis Scheme for Proficiency (WASP) is a proficiency testing scheme for the analysis of occupational hygiene and environmental air samples and is operated in the UK by the Health and Safety Laboratory. Since 1997, WASP has offered samples of benzene, toluene and m-xylene, at environmental levels on Tenax, and has about 35 laboratories participating, mostly from industry, local government and consultancy organisations in the UK. The results reported cover the first 10 rounds of the environmental analytes (1997-1999) and demonstrate the important role of proficiency testing in assessing the quality of laboratory performance. Estimates are obtained for within-laboratory precision and the total variability at each analyte level. The estimates of within-laboratory precision suggest that laboratories have more difficulty analysing toluene and m-xylene than benzene. Linear relationships for the reproducibility relative standard deviation (RSDT) with loading level are evident for the analytes at occupational levels. At environmental levels, the relationship between loading level and reproducibility is much less well defined. The standard deviation for the proficiency testing assessment for all three analytes at the environmental level is 14%, as derived from the benzene data. Expanded uncertainty estimates (k = 1.96), for the analysis of samples since the scheme started, are obtained from the average total variance, and are 27% for benzene, 39% for toluene and 36% for m-xylene. Although the linear trend of performance against round number was not significant at the 95% level of confidence (p = 0.23 for benzene, p = 0.3 for toluene and p = 0.32 for m-xylene), there was a general improvement in RSDT from 26-34% to about 8-13% 10 rounds later. Currently, for a laboratory to meet one of the data quality objectives in the Ambient Air Directive (indicative measurement of benzene, expanded uncertainty +/- 30% or less), it would have to achieve a level of analytical performance to satisfy the category 1 (best performance) limit of better than +/- 8.8%. In the last proficiency testing round, discussed in this paper, only 58% of laboratories obtained performance scores that indicated that they were able to consistently achieve this level of performance.     

第三方环境检测机构发展现状与展望

介绍了我国第三方环境检测行业的现状,指出了存在的问题,分析了我国第三方环境检测行业发展趋势,并对完善我国第三方环境检测行业发展提出了健全机制,建立资质、能力验证、市场监管,全面的实验室比对,限制服务范围,创建高端品牌化第三方环境检测实验室的建议.     

Analysis of two-way layout of count data with negative binomial variation

A number of methods has been proposed for dealing with single-factor or factorial experiments when the requirements for performing the normal theory analysis of variance procedure are not satisfied. This paper suggests the use of the likelihood ratio statistic for testing the main effects and the interaction between the factors in two-way layout of count data following negative binomial distributions with a common dispersion parameter. The likelihood ratio statistic for testing the equality of the dispersion parameters of several groups of count data is also derived. The methods is illustrated by an example concerning the study of spatial and temporal variation of bacterial counts.