Proficiency Testing statistics

Table 14 can be regarded as providing a reasonable overall picture, even if the results cannot applied to any particular case. However, if the underlying principle is accepted, it becomes clear that improvements in a single stage, for example the reduction of instrument variation, has a negligible beneficial effect (if this variation was not outside the normal range ). Even if the contribution of repeatability is re-duced to zero, the cumulative uncertainty is reduced by 10% only, i.e. from 2.2 to y(0.0)2 (0.8)2 (1.0)2 + (1.5)2 = 2.0. This statistical view of errors should help to avoid some unnecessary efforts to improve, e.g., calibration. Additionally, this broad view on all sources of error may help to detect the most important ones. Consequently, without participation in proficiency tests, any method validation will remain incomplete. 

This chapter deals with handling the data generated by analytical methods. The first section describes the key statistical parameters used to summarize and describe data sets. These parameters are important, as they are essential for many of the quality assurance activities described in this book. It is impossible to carry out effective method validation, evaluate measurement uncertainty, construct and interpret control charts or evaluate the data from proficiency testing schemes without some knowledge of basic statistics. This chapter also describes the use of control charts in monitoring the performance of measurements over a period of time. Finally, the concept of measurement uncertainty is introduced. The importance of evaluating uncertainty is explained and a systematic approach to evaluating uncertainty is described. 

All proficiency testing schemes should have a statistical protocol which states clearly how the data will be processed and how laboratory performance will be evaluated. This protocol should also describe how the assigned value for any parameter in a test sample is estimated. This is an important consideration, as the performance of individual laboratories is gauged by comparison with the assigned value. 

Participation in proficiency testing schemes is an ongoing activity. It is therefore useful to monitor performance over a period of time and to look for trends. Performance over time can be demonstrated statistically by using measures such as RSZ and SSZ (see Section 7.3.4) but as mentioned previously, these can be misleading. It is better to monitor performance scores by plotting them on a... 

Statistical Methods for Use in Proficiency Testing by Interlaboratory Comparisons , ISO 13528 2005, International Organization for Standardization (ISO), Geneva, Switzerland, 2005. 

As shown above, these include a laboratory to be third-party assessed to international accreditation standards, to demonstrate that it is in statistical control by using appropriate internal quality control procedures, to participate in proficiency testing schemes which provide an objective means of assessing and documenting the reliability of the data it is producing and to use methods of analysis that are fit-for-purpose . These requirements are summarised below and then described in greater detail later in this chapter. 

In ISO 13528 details on possible statistical methods for the evaluation of proficiency tests are given. 

ISO 13528 2005 - Statistical methods for the use in proficiency testing by interlaboratory comparisons. 

Two approaches attempt to solve this problem. The one is a transformation of the data to logarithms prior to the statistical calculations corresponding to a logarithmic normal distribution. The other is a modification of the z-scores with correction factors. This method was introduced first in a German standard for proficiency testing (DIN 38402 - 45), which in the meantime partially was transferred into ISO/TS 20612. 

Traditionally, the education that chemists and chemistry laboratory technicians receive in colleges and universities does not prepare them adequately for some important aspects of the real world of work in their chosen field. Today s industrial laboratory analyst is deeply involved with such job issues as quality control, quality assurance, ISO 9000, standard operating procedures, calibration, standard reference materials, statistical control, control charts, proficiency testing, validation, system suitability, chain of custody, good laboratory practices, protocol, and audits. Yet, most of these terms are foreign to the college graduate and the new employee. 

Wang, W, Zheng, J, Tholen, D W, Cao, Z, and Lu, X (2005), A statistical strategy for discouraging collusion in split-level proficiency testing schemes. Accreditation and Quality Assurance, 10 (4), 140-43. 

Proficiency Testing of Analytical Laboratories Organisational and Statistical Assessment , Analyst Cambridge, 1992, 117, 97. 

Participation in interlaboratory comparisons and proficiency-testing programs provides additional information especially pertinent to controlling interlaboratory variation. Aliquots of homogeneous samples containing the analytes of interest are drawn and distributed to each participating laboratory. The participants results are used to calculate overall and method-specific statistics, such as means, medians, and standard devia-... 

Quality control of laboratories depends on the availability of CRMs, round-robin studies, intercomparisons and proficiency tests between methods and between laboratories. Of special importance is a full knowledge of the complex analytical process and the painstaking pursuit of the true value by defining all sources of errors and the application of an adequate error source budget. The application of Poisson and Bayesian statistics could have some advantage. 

Within two weeks of the study closing date, we issue an interim report, the purpose of which is to provide rapid feedback to participants. At the conclusion of each study, a detailed final report is prepared and issued to participants. This report contains a full description of the study together with statistical analysis and graphical presentation of the results. The report is prepared in a standardised format consistent with ISO and ILAC guidelines for proficiency test reports. 

Consensus value can be defined as the mean of participants results on a test material distributed in a proficiency testing scheme, after outliers have been handled either by elimination or by the use of robust statistics. 

If an ALMERA member wants to keep the evaluation result of his/her participation anonymous, he/she will have the option to only take part in the IAEA world-wide open proficiency test. In this case, his/her results will not be included in the ALMERA report. In addition, the statistical approach used to evaluate the analytical results of the ALMERA network proficiency test will be adapted in the future to take the reporting time into consideration. 

The National Soil Conservation Administration (SCA) has extended an invitation to the I.O.N.S. Corporation to participate in a round robin proficiency testing activity. This is an activity in which a laboratory demonstrates its proficiency in performing certain functions in a statistically accurate manner and, thereby, establishes credibility with its clients. Other laboratories also will be participating. Our results will be compared to that of other laboratories as well as to the results expected by the SCA. The SCA will then grade our proficiency. 

New. A new microscale titration experiment is included, provided by Professor John Richardson from Shippensburg State University, for the analysis of hard-water samples (Experiment 18). The tools an4 techniques used for that experiment could be used to design similar experiments for other titrations if desired. (If your in-stractor tries this with you, I may include your experiment in the next edition ) Two team experiments are added (Experiments 39 and 40) to illustrate the principles presented in Chapter 4 on statistical validation. One is on method validation and quality control, in which different members of teams perform different parts of the validation for a chosen experiment. The other is on proficiency testing, in which you calculate the z-values for all the student results of one or more class experiments and you compare your z-value to see how well you have performed. 

---