Sampling techniques, error

Errors in the analytical laboratory are basically of two types determinate errors and indeterminate errors. Determinate errors, also called systematic errors, are errors that were known to have occurred, or at least were determined later to have occurred, in the course of the lab work. They may arise from avoidable sources, such as contamination, wrongly calibrated instruments, reagent impurities, instrumental malfunctions, poor sampling techniques, errors in calculations, etc. Results from laboratory work in which avoidable determinate errors are known to have occurred must be rejected or, if the error was a calculation error, recalculated. 

Manual sampling techniques can introduce error by virtue of variations in strength and size of the human hand, from analyst to analyst. As a result, the pulling velocity through the filter may vary considerably. Too rapid a movement of liquid through the filter can compromise the filtration process itself. 

The qualification concept was used by Wargo and Drennen to verify the homogeneity of solid mixtures. Qualitative analytical algorithms based on the Bootstrap Error-adjusted Single-sample Technique (BEST) proved sensitive to variations in sample homogeneity. Plugge and van der Vlies converted NIR spectra into polar coordinates and used them to calculate the Mahalanobis distance in order to identify nonhomogene-ous samples. 

In practice, the application of x-ray measurement techniques to thin films involves some special problems. Typical films are much thinner than the penetration depth of commonly used x-rays, so the diffracted intensity is much lower than that from bulk materials. Thin films are often strongly textured this, on the other hand, results in improved intensity for suitable experimental conditions but complicates the measurement problem. Measurements at other than ambient temperature, not usually attempted with bulk materials, constitutes additional complexity. Since typical strains are on the order of 1 X 10 , measurements of interplanar spacing with a precision of the order of 1 X 10 are needed for reasonably accurate results hence, potential sources of error must be kept to a low level. In particular, the sample displacement error can be a major source of difficulty with a heated sample. The sample surface must remain accurately on the axis of the instrument during heating. 

The literature cited is intended to illuminate the topics covered but in no sense represents an exhaustive review. Furthermore, numerous alternative sampling techniques are omitted in favor of more comprehensive treatments of what I judge to be the most generally useful strategies. The most important omission is probably impactor collection, a technique that may reduce sampling errors with some labile species (I). 

The sources of sampling error are constitutional error, distributional error, and error caused by use of the incorrect sampling technique [GARFIELD, 1989 MINKINNEN, 1987],... 

Indeterminate errors are errors that cannot be eliminated and are inherent in the analytical technique. Determinate errors are errors whose cause and magnitude can be determined, and they include poor sampling technique, decomposition of the column, change in detector response, improper recorder performance, calculation errors, and operator prejudice or error. 

Here, T is the observed line width (Av << F), 7d is the peak-to-valley intensity in the difference spectrum, and To is the peak height of the Raman line. Although this equation is for Lorentzian-shaped bands, the results are approximately the same for Gaussian-shaped bands (the constant 0.385 becomes 0.350). In the case of carbon disulfide-benzene mixtures, the smallest shift observed was -0.06 cm-1, and the associated error was 0.02 cm-1 (77). A convenient rotating system that can be used for (1) difference spectroscopy, (2) normal rotating sample techniques (solid and solution), and (3) automatic scanning of the depolarization ratios as a function of the wave number has been designed (45). 

The accuracy of interference microscopy depends on the calibration standards used. For measurements in the secondary wall, these values are 1.604 for Nb and a range of values for Na depending on the specimen. The variation in Na arises presumably from variation in microfibril angle (Hermans 1946). The need to measure Na for each sample is one disadvantage of this technique that increases the amount of work necessary to perform the measurement. If Na is not measured for each sample, an error of 11% is introduced compared to the usual error of 2% (Donaldson 1985a). Na does not vary between earlywood... 

Since the weighing of such small sample quantities can influence the analytical error, it is desirable to avoid the need to weigh the sample. One channel is occupied by the reference absorption measurement, for example with pure iron or low-alloy steels an iron hollow-cathode lamp is used and the measurement takes place at 372 nm. In this case, only one further element can be determined. The carrying out of the method is simple (Fig. 6). Tests of this compact sample technique with standard materials, of which there are at present only a few containing certified trace elements, show satisfactory agreement (Fig. 7). As regards traces, the chips of standard materials are very homogeneous. 

Random Measurement Error Third, the measurements contain significant random errors. These errors may be due to sampling technique, instrument calibrations, and/or analysis methods. The error-probability-distribution functions are masked by fluctuations in the plant and cost of the measurements. Consequently, it is difficult to know whether, during reconciliation, 5 percent, 10 percent, or even 20 percent adjustments are acceptable to close the constraints. 

Sampling technique Percentage standard deviation a% Estimated maximum sample error (E%) Efficiency (C%)... 

---