摘要
基于实验室长期积累的质控数据评估测量不确定度的top-down方法具有广泛应用前景,质控图法和稳健统计法是其中常见方法。质控图法要求测量数据中不含离群值,对测量数据时间顺序有明确要求。稳健统计法是指不用识别、剔除离群值,直接应用全部测量数据,将离群值对统计分析结果影响降低到最小的统计分析方法。基于质控图法、稳健统计和实验室长期积累的质控数据对固体样品汞含量热裂解-原子吸收光谱分析方法的测量不确定度进行了评估,对两种方法计算结果、实验室评估的测量不确定度和标准样品标称值的不确定度进行了比较。计算结果表明,两种方法结果基本吻合,基于实验室质控数据期间精密度评估的不确定度明显小于标准样品标称值,结果合理。
There are broad application prospects for evaluation of measurement uncertainty in environmental test laboratory by top-down method based on quality control data accumulated in long-term routine analysis. Quality control charting method and robust statistics are two top-down methods. The quality control charting method needs no outliers and clearly requires time sequence in measured data. Robust statistics is a type of statistical analysis method where it is unnecessary to identify and delete outliers but it can also reduce the effect of outliers on the final results based on all measurement data. In this paper, quality control charting method and robust statistics （iteration method） are used to evaluate measurement uncertainty of mercury in solid samples by pyrolysis atomic absorption spectroph0tometry, after calculating intermediate precision （sR） of two quality data sets collected during last two years. The analysis results of the two methods, the measurement uncertainty evaluated in laboratory and that recommended by the standard reference materials manufacture are compared. The results show that the measurement uncertainty evaluated by two methods is generally in line with each other for solid samples. The values of SR ~ from laboratory data are evidently smaller than that recommended by the manufacture, which is reasonable according to statistics principle.
出处
《中国无机分析化学》
CAS
2015年第2期1-8,共8页
Chinese Journal of Inorganic Analytical Chemistry
基金
环保公益性行业科研专项(201209015)
科技部重大仪器专项(2014YQ060773)资助
