Sample size problems with

The sample size problem can be understood best with a simple example. Let us suppose we are going to test some chemical, call it X, using lab rats. We intend only a simple experiment, with two dose groups. We have available a supply of young male rats, and from that supply we randomly select 50 animals for the control group and 50 (unlucky) animals for the one test group. 

This brief summary of the state of current instrumentation for the total automation of pesticide analyses will hopefully entice the reader to apply these principles of mechanized unit operations to his own analytical problems. There are enough creative scientists in this field, each trying to lighten his workload, that it seems safe to make this prediction by 1984 we ll have a pesticide analyzer which will take our unmeasured, untreated sample into one end of the instrument and give us final answers, printed in correct concentration units, twice as fast as today, with one tenth the sample size, and with half the CV s. If an analyst makes up his mind to do so, he can likely automate anything. 

As Whitehead (1997) points out, this can even occur with fixed sample size studies of survival analysis. For example, consider a fixed sample size trial with a recruitment over one year. We may have determined to analyse the results once the last patient recruited has been followed up for one year. This means, though, that patients recruited earlier in the trial will have been followed up longer those recruited at the beginning will, in fact, have been followed up for two years. Often this extra information will also be included in the analysis, which thus reflects a mixture of follow-up times from one to two years. However, once this analysis is complete it will still be possible to obtain further data and in a year s time an analysis of patients with 2-3 years follow up could be carried out. I think it is fair to say, however, that it is more likely to be a problem which makes itself known in a sequential rather than a fixed trial framework. Nevertheless, it is a potential feature of all forecasting systems that they are hostages to the future further information which embarrasses us can always arise. 

The number of samples required to estimate the sample mean to some degree of precision or uncertainty is an important aspect of sampling. Increasing the sample size or the number of units tested to obtain a mean increases the precision or reduces the uncertainty. However the cost of this extra work increases linearly with sample size, while the precision increases at a much slower rate with the square root of the number of samples. Sample size problems are approached on the basis of the total uncertainty in the mean. There are two criteria that must be specified to calculate a sample size n. 

The fracture mechanics approach is potentially a more flexible tool than the stress-life approach as it allows the progression of cracking to failure to be modelled and can be transferred to different sample geometries. Problems with the traditional fracture mechanics approach include selection of initial crack size and crack path, selection of appropriate failure criteria, load history, and creep effects. Also the fracture mechanics approach does not accurately represent the accumulation and progression of damage observed experimentally in many cases. However, recent modifications to the standard fracture mechanics method have seen many of these limitations tackled. 

Chi-Square Distribution For some industrial applications, produrt uniformity is of primary importance. The sample standard deviation. s is most often used to characterize uniformity. In dealing with this problem, the chi-square distribution can be used where = (.s /G ) (df). The chi-square distribution is a family of distributions which are defined by the degrees of freedom associated with the sample variance. For most applications, df is equal to the sample size minus 1. 

The other major problem concerned with sampling is that of the sample size. The size of the sample taken from a heterogeneous material is determined by the variation in particle size, and the precision needed in the results of the analysis. 

A problem long appreciated in economic evaluations, but whose seriousness has perhaps been underestimated (Sturm et al, 1999), is that a sample size sufficient to power a clinical evaluation may be too small for an economic evaluation. This is mainly because the economic criterion variable (cost or cost-effectiveness) shows a tendency to be highly skewed. (One common source of such a skew is that a small proportion of people in a sample make high use of costly in-patient services.) This often means that a trade-off has to be made between a sample large enough for a fully powered economic evaluation, and an affordable research study. Questions also need to be asked about what constitutes a meaningful cost or cost-effectiveness difference, and whether the precision (type I error) of a cost test could be lower than with an effectiveness test (O Brien et al, 1994). 

The value of spruce-oil chemistry in sorting out problems of hybridization and introgression—major factors in Picea taxonomy—was succinctly summarized by von Rudloff who defined three situations (1) Terpene variation is limited such that it is not possible to use these characters in studies of introgression this is the case in eastern North America where the ranges of black spruce and red spruce overlap. (2) Sufficient variation in terpene profiles exists for the compounds to be useful markers in systematic studies as seen in white spruce. Brewer s spruce, and Sitka spruce. (3) Tree-to-tree variation in terpene content is so variable that use in che-mosystematic studies is precluded, or at least requires very large sample sizes for statistical reliability, as seen with Engelmann s spruce. 

On-line dialysis also separates the analyte from tissue matrix based upon molecular size, but in this case, the sample extract is passed over a membrane filter through which the analyte (and other low molecular weight compounds) is diffused into a second solvent on the other side of the membrane filter. Usually, the second solvent is then concentrated on to an SPE column to minimize the dilution effect that is caused by the dialysis process. Agasoester used on-line dialysis to separate oxytetracycline from muscle, liver, milk, and egg tissue matrix components. A problem encountered with on-line dialysis is the inability of analyte molecules that are bound to proteins in the sample extract to pass through the membrane filter. Problems with membrane clogging are reduced with on-line dialysis compared with ultrafiltration because no external force is being applied to bring the analyte across the membrane filter. 

Solid-phase sorbents are also used in a technique known as matrix solid-phase dispersion (MSPD). MSPD is a patented process first reported in 1989 for conducting the simultaneous disruption and extraction of solid and semi-solid samples. The technique is rapid and requires low volumes (ca. 10 mL) of solvents. One problem that has hindered further progress in pesticide residues analysis is the high ratio of sorbent to sample, typically 0.5-2 g of sorbent per 0.5 g of sample. This limits the sample size and creates problems with representative sub-sampling. It permits complete fractionation of the sample matrix components and also the ability to elute selectively a single compound or class of compounds from the same sample. Excellent reviews of the practical and theoretical aspects of MSPD " and applications in food analysis were presented by Barker.Torres et reported the use of MSPD for the... 

Trace analysis is particularly attractive for SFE-HPLC since quantitative transfer of all analytes extracted to the chromatographic system becomes possible. At present, on-line SFE-HPLC appears to be feasible for qualitative analysis only quantitation is difficult due to possible pump and detector precision problems. Sample size restrictions also appear to be another significant barrier to using on-line SFE-HPLC for quantitative analysis of real samples. On-line SFE-HPLC has therefore not proven to be a very popular hyphenated sample preparatory/separation technique. Although online SFE-HPLC has not been quantitatively feasible, SFE is quite useful for quantitative determination of those analytes that must be analysed by off-line HPLC, and should not be ruled out when considering sample preparatory techniques. In most cases, all of the disadvantages mentioned with the on-line technique (Table 7.15) are eliminated. On- and off-line SFE-HPLC were reviewed [24,128]. 

With improving detector performance, the smaller can be the sample size and, consequently, the more rapid the sample pretreatment. However, as shown repeatedly in quantitative analysis, small sample sizes (several mg) face homogeneity problems and set a... 

---