Reliability-measurement uncertainty

All these methods pretend to represent the intuitive way an expert deals with uncertainty. Whether this is true remains an open question. No method has yet been evaluated thoroughly. Modelling uncertainty to obtain a reasonable reliability measure for the conclusions remains one of the major unsolved issues in expert system technology. Therefore, it is important that in the expert system a mechanism is provided to define its boundaries, within which it is reasonably safe to accept the conclusions of the expert system. 

The choice of method depends on the purpose for which the analysis is being performed. The customer requesting the analysis may specify the method to be used. Even in this situation, it is the responsibility of the laboratory to demonstrate that the method is capable of producing results that are reliable. When no method is specified the points to consider have already been identified in Section 4.2. The acceptable level of measurement uncertainty specified or implied will, to a certain extent, set the precision and bias levels. All of the topics covered in the following sections may be crucial, depending on the purpose of the analysis, and should appear on your list. 

It is important to have some knowledge of the reliability of all measurement results. Measurement uncertainty is the parameter used to describe the range within which the true value (or right answer) for a particular measurement is expected to lie. Evaluating measurement uncertainty involves a number of distinct steps, which are described in this chapter. 

Do you normally give a measurement uncertainty/reliability when reporting results to your customer ... 

The maximum acceptable measurement reliability (also known as measurement uncertainty) for each analytical result. 

Figure 24.9a shows a plot of measured total carbon (CO plus CO2, mole percent) versus equivalence ratio. The solid line was calculated assuming chemical equilibrium at the measured temperatures. The data points represent the measured CO and CO2 mole fractions (dry basis) using the fast extractive-sampling system. Horizontal bars represent the uncertainty in (f> due to reading and calibration errors vertical bars represent the uncertainty in the CO and CO2 mole-fraction sum due to line strength and absorption measurement uncertainty. The data are consistent to within 4% of the equilibrium predictions at all values of (p, indicating reliable operation of the system. 

The purpose of an analytical method is the deliverance of a qualitative and/or quantitative result with an acceptable uncertainty level. Therefore, theoretically, validation boils down to measuring uncertainty . In practice, method validation is done by evaluating a series of method performance characteristics, such as precision, trueness, selectivity/specificity, linearity, operating range, recovery, LOD, limit of quantification (LOQ), sensitivity, ruggedness/robustness, and applicability. Calibration and traceability have been mentioned also as performance characteristics of a method [2, 4]. To these performance parameters, MU can be added, although MU is a key indicator for both fitness for purpose of a method and constant reliability of analytical results achieved in a laboratory (IQC). MU is a comprehensive parameter covering all sources of error and thus more than method validation alone. 

The standard recognizes the factors that determine the correctness and reliability of test results human factors, accommodation and environment, methods, equipment, sampling, and the handling of test items. In this list, measurement traceability is mentioned, but in fact metrological traceability, with measurement uncertainty and method validation, are really subsumed in methods. (subsection 5.4). The effect of each of these factors on measurement uncertainty will differ considerably among kinds of tests. 

An intriguing aspect of these measurements is that the values of D determined from NMR and from sorption kinetics differ by several orders of magnitude. For example, for methane on (Ca,Na)-A the value of the diffusion coefficient determined by NMR is 2 x 10 5 cm2 sec-, and the value determined for sorption rates only 5 x 10"10 cm2 sec-1. The values from NMR are always larger and are similar to those measured in bulk liquids. The discrepancy, which is, of course, far greater than the uncertainty of either method, remained unexplained for several years, until careful studies (267,295,296) showed that the actual sorption rates are not determined by intracrystalline diffusion, but by diffusion outside the zeolite particles, by surface barriers, and/or by the rate of dissipation of the heat of sorption. NMR-derived results are therefore vindicated. Large diffusion coefficients (of the order of 10-6 cm2 sec-1) can be reliably measured by sorption kinetics... 

In metrology in general, the closer the similarity between two specimens, the smaller the relative uncertainty of the measured difference between them and the easier it is to make a reliable measurement. Thus, by the use of suitable standards, measurements in the field can become highly reliable and far less demanding and costly. 

If there is no reason to choose one specific measure, the separate (reliable) measures are combined [10] into a single assigned value, being the inverse-variance weighted (for uncertainty) average of the independent measures of analyte concentration. 

The concept of traceability depends on a chain of standards (artefacts or measurements) linked back to the appropriate international primary standard series of calibrations (intercomparisons between two standards in the chain). A measurement result obtained through calibration against one of these standards will itself be traceable. The uncertainty of calibration and the measurement result will depend on the uncertainties of the values assigned to the standards in the chain and the measurement procedure used. Unless the measurement uncertainty of each transferred value is reliably known there is no way to estimate the accuracy of the standard being calibrated and hence the accuracy... 

This paper has examined the role of calibration and evaluation of measurement uncertainty in clinical laboratories arising from the request for traceability assurance. To produce results which are accurate and reliable within the stated uncertainty, all uncertainties of the quality measurement process and the traceability chain should be demonstrated. Also, the quality of a spectrophotometric result depends critically on RMs and photometric systems whose traceability have been properly demonstrated. 

The issues of method validation and assessment of measurement uncertainty in the determination of potentially toxic trace elements in rice are of permanent interest for the scientific community. In this context, the sources of uncertainty associated with the determination of Cd, Cu, Pb, and Zn have been recently estimated in rice through an interlaboratory comparison [30]. Four Brazilian laboratories participated in the proficiency test. The analytical technique used were FAAS, ET-AAS, and ICP-AES. The rice samples were supplied by the Institute for Reference Materials and Measurements (IRMM), Joint Research Center of the European Commission, within the scope of the interlaboratory comparison International Measurement Evaluation Programme (IMEP) 19 Trace Elements in Rice (see also Chapter 7 in this book). Three out of the four laboratories reported values close to the reference values. It was emphasized that, in order to establish a reliable uncertainty budget, all significant sources of uncertainty should be identified. 

Analytical method attributes measures of the quality, reliability, and uncertainty of the determinations obtained with an analytical method. Typical analytical method attributes are selectivity, sensitivity, detection limits, signal/noise, recovery, accuracy, bias, precision, and validation. Analytical method attributes are sometimes called figures of merit. 

The difference between b and vs is profound yet simple To properly evaluate wall slip, we simply need to use Eq. (4a) instead of Eq. (7a), i.e., to normalize both sides of Eq. (7a) by Q0. Only when the relative change (Q-Qo)/Q0=8fo/D is beyond the experimental uncertainty, can wall slip be reliably measured Theologically. In other words, evidence for wall slip would be questionable if the magnitude of slip correction, measured by 8h/D, is much smaller than unity, and the difference (Q-Q0)/Q0 is experimentally vanishingly small. 

The present computed equilibrium constant for reacfion (5) af 298 K is 8.9 x 10 cm molecule . For 1 atm of N2, the effective second-order ks 4 x 10 cm molecule s [49]. Thus the corresponding first-order dissociation rate constant is k-5 4.5 x 10 s at these conditions, i.e., the lifetime is 6 h. This estimate of k-5 is about two orders of magnitude smaller than that of Allan and Plane [49], but still high enough for appreciable IONO2 dissociafion to take place overnight. The dissociation rate is very sensitive to DFfo(IO-N02) the present error limits correspond to an uncertainty of a factor of 3.5, and so a reliable measurement of k-5 would allow reliable experimental determination of the thermochemistry. 

It should be noted that the basic and necessary parameters that characterize an analytical result are traceability and measurement uncertainty. An anal3tical result without documented traceability and estimated uncertainty is a source of misinformation. These two parameters are the basic requirements of reliable analytical results. 

Zischka, M., Wegscheider, W. Reliability of and measurement uncertainty for the determination of Au, Pd, Pt and Rh by ICP-MS in environmentally relevant samples. In Zereini, F., Alt, E. (eds.) Anthropogenic platinum-group element emissions. Their impact on men and environment, pp. 201-214. Springer, Berlin (2000)... 

The principal parameter used to indicate the rate of drug absorption is Cmax, even though it is also influenced by the extent of absorption the observed fmaX is less reliable. Because of the uncertainty associated with Cmax, it has been suggested (Endrenyi Yan, 1993 Tozer, 1994) that Cmax/AUCo-loq/ where AUCo-loq is the area under the curve from time zero to the LOQ of the acceptable analytical method, may more reliably measure the rate of drug absorption, except when multiexponential decline is extensive. Estimation of the terms should be based on the observed (measured) plasma concentrationtime data and the use of non-compartmental methods rather than compart-mental pharmacokinetic models. MRTs, from time zero to the LOQ of the analytical method, for the test and reference products can be compared, assuming that first-order absorption and disposition of the drug apply (Jackson Chen, 1987). 

This paper discusses a number of practical problems arising from the request for and use of clinical reference materials for the validation of the performance of photometric systems used in national clinical chemistry laboratories. It shows that uncertainties in the measurement step of photometric analysis have largely been ignored. Uncertainties associated with this step can and do contribute significantly to overall analytical uncertainty. Thus, for a knowledge of trueness and measurement uncertainty, an adequate certified reference materials system and an attempt at a traceability chain are of the utmost importance, since the quality of clinical chemistry results depends critically on the use of reliable reference materials and properly validated instruments. 

It is important to note the differences between error and measurement uncertainty. Errors are differences in measurements while uncertainties are a range of measurement. The importance of measured uncertainty values quoted with the results improves the reliability of a result and adds confidence in the decision and reporting of the same. Knowledge of the uncertainty value also give credence to strive to reduce the uncertainty value associated with measurements which will facilitate better and more confident reporting. 

The data from these experiments were analyzed using the statistical methods described in Chapters 5, 6, and 7. For each of the standard arsenic solutions and the deer samples, the average of the three absorbance measurements was calculated. The average absorbance for the replicates is a more reliable measure of the concentration of arsenic than a single measurement. Least-squares analysis of the standard data (see Section 8C) was used to find the best straight line among the points and to calculate the concentrations of the unknown samples along with their statistical uncertainties and confidence limits. 

---