Interlaboratory comparisons and improvements of methods for acid number determination in used motor oils

The results of two interlaboratory comparisons of acid number determinations in used motor oils are discussed. It is shown that the comparability of the measurement results is not as good as that required by known standards for petroleum products. The problem is motor oil contaminants which accumulated during use, and which are the source of a matrix effect in the acid number determination. The standard methods’ drawbacks are analyzed and some improvements are proposed. Repeatability and accuracy of the improved methods are evaluated. Received: 11 December 2001 Accepted: 15 February 2002     

服装主观热感觉的统计分析

本文用多因子方差分析模型和多重比较分析了服装穿着时的主观热感觉与服装品种、环境条件、穿着者的活动状态与籍贯之间的关系。通过主成分回归建立了用生理指标预测主观热感觉的模型，从而得出一种定量地评价服装属性的方法。     

大规模界约束优化的子空间截断牛顿法

给出了大规模界约束优化的一个子空间截断牛顿法。利用截断牛顿法修正非有效约束所对应的变量，用投影梯度法修正有效约束所对应的变量，文中证明了方法的整体收敛性，并对方法进行了数值试验，且与子空间有限内存拟牛顿法进行了数值比较。     

Statistical treatment of proficiency testing data

 There are three stages to evaluating a laboratory's results in an interlaboratory proficiency test: establishing the correct result for the test item, determining an evaluation statistic for the particular result, and establishing an acceptable range. There are a wide variety of procedures for accomplishing these three stages and a correspondingly wide variety of statistical techniques in use. Currently in North America the largest number of laboratory proficiency test programs are in the clinical laboratory field, followed by programs for environmental laboratories that test drinking water and waste water. Proficiency testing in both of these fields is under the jurisdiction of the federal government and other regulatory and accreditation agencies. Many of the statistical procedures are specified in the regulations, to assure comparability of different programs and a fair evaluation of performance. In this article statistical procedures recommended in International Organization for Standardization Guide 43, Part 1, are discussed and compared with current practices in North America. Received: 22 April 1998 · Accepted: 12 May 1998     

Robust estimates of the theoretical standard deviation to be used in interlaboratory precision experiments

Interlaboratory experiments often contain results that strongly deviate from other results obtained in the same laboratory under repeatability conditions, or laboratory means that strongly deviate from other laboratory means. In ISO 5725-2 [1] and IUPAC [2], the basic standards for the organisation and analysis of interlaboratory experiments for the determination of precision of a measurement method, outlier tests are performed in order to detect such situations and to finally decide whether these values are retained in the analysis or discarded as outliers. This outlier treatment, which always has a subjective aspect, becomes unnecessary by using robust estimates of the repeatability and reproducibility standard deviation. This paper proposes to use Rousseeuw’s Q _n as an extremely robust and efficient estimate of the standard deviation. Two examples show the performance of the new method.     

Visualisation of interlaboratory comparison results in PomPlots

In this work a novel graphical method is applied to the presentation of intercomparison results. This is demonstrated with the results of a recent intercomparison in measuring the ¹³⁷Cs, ⁴⁰K, and ⁹⁰Sr activity concentration in milk powder. The “PomPlot”, an intuitive graphical method, is used for producing a summary overview of the participants’ results of a common measurand. The “PomPlot” displays (relative) deviations of individual results from the reference value on the horizontal axis and (relative) uncertainties on the vertical axis.     

Monotonicity of certain integrals involving gamma distributions and their applications in multiple comparisons

Barlow and Gupta (1969) and Alam (1970) studied the monotonicity of two integrals, involving gamma distributions, that arise in certain ranking and selection problems. In this paper, we shall unify their results by studying the monotonicity of two generalized versions of integrals considered by them. We will also provide applications of derived results in study of certain multiple comparison procedures.     

A re-examination of platinum-group element concentrations in the environmental certified reference material BCR-723

A detailed statistical examination of replicated data used to certify platinum-group elements (PGEs) in environmental reference material BCR-723 is presented. Certification of Pt, Pd, and Rh concentrations in BCR-723 was based on 16, eight, and nine accepted data sets, respectively. Each accepted data set contained six replicated measurements for each PGE, and the statistical properties of these concentration data were examined, i.e. 96 for Pt, 48 for Pd, and 54 for Rh. This level of investigation has received limited attention but is critical in furthering our understanding of PGE variability and representativeness. Concentrations of Pt, Pd, and Rh were shown to differ significantly between accepted data sets. Palladium and Pt differed in their quantification between detection techniques. Additionally, Pd and Pt concentrations varied significantly between laboratories using a similar definitive method (inductively coupled plasma-isotope dilution mass spectrometry). The distribution of Pd concentrations was found to be bimodal, with a secondary population exhibiting a contamination signal of about 15%. The secondary population, not previously reported in BCR-723, is likely a measurement artifact and not due to a nugget effect. Comparisons of BCR-723 with other environmental media from Europe, i.e. airborne particulate matter, tunnel dust, and road-deposited sediment, indicated that Pd is uncommonly low in BCR-723 (6.0?ng?g^?1) and is generally not representative in terms of its distribution relative to Pt and Rh. Serious consideration should be given to developing a new PGE certified environmental reference material.     

In-house validation for analytical methods and quality control for risk evaluation of chlordecone in food

Chlordecone was used until 1993 as a pesticide in the banana plantation of Martinique and Guadeloupe (French Antilles) against the root borer. This organochlorine pesticide was lipophilic, remnant, and toxic for human beings with both acute and chronic effects. Chlordecone was strongly absorbed and stored in soil and weakly decomposed in environment. Surveys conducted in 2001 revealed its presence in soil, rivers, and domestic food products. Local food (fruits and vegetables, cattle, poultry, and fish) was growing on soils, widely contaminated by chlordecone, used in the past as banana plantations. In 2003, French Administration asked for a risk evaluation for the Antilles population. The French Agency for Food Safety, proposed a Provisional Tolerable Daily Intake of 0.0005?mg?kg^?1?b.w.?day^?1, and an Acute Reference Dose of 0.01?mg?kg^?1?b.w.?day^?1, based on a toxicological risk assessment. The French National Reference Laboratory for pesticides has carried out two analytical methods, one for food of animal origin and another for food of fruit and vegetable origin. These methods were validated in the reference laboratory and dispatched to 13 laboratories for a proficiency test before the launch of two studies on Martinique and Guadeloupe food. About 900 samples from Martinique were sent to the network of laboratories for analysis of chlordecone. Performance parameters obtained through the proficiency test were briefly reported. The quality control test proposed in this study was discussed to shed light on the true variability achievable by intra-laboratory and inter-laboratory analysis. The limits of conventional quality-control procedures were discussed, and a process was proposed in order to get better confidence in analytical results.     

Convergence properties and practical estimation of the probability of rank reversal in pairwise comparisons for multi-criteria decision making problems

In this paper, we address the impact of uncertainty introduced when the experts complete pairwise comparison matrices, in the context of multi-criteria decision making. We first discuss how uncertainty can be quantified and modeled and then show how the probability of rank reversal scales with the number of experts. We consider the impact of various aspects which may affect the estimation of probability of rank reversal in the context of pairwise comparisons, such as the uncertainty level, alternative preference scales and different weight estimation methods. We also consider the case where the comparisons are carried out in a fuzzy manner. It is shown that in most circumstances, augmenting the size of the expert group beyond 15 produces a small change in the probability of rank reversal. We next address the issue of how this probability can be estimated in practice, from information gathered simply from the comparison matrices of a single expert group. We propose and validate a scheme which yields an estimate for the probability of rank reversal and test the applicability of this scheme under various conditions. The framework discussed in the paper can allow decision makers to correctly choose the number of experts participating in a pairwise comparison and obtain an estimate of the credibility of the outcome.