Chapter 12 Statistics and data evaluation

Abstract

This chapter focuses on the statistics and data evaluation. The analyst is responsible for the sampling procedure to make sure that samples which are to be analyzed do not differ from the material where they come from. In case of trace elements no contamination must be introduced in the sampling procedure. The analyst also participates in the discussion about the sample mass and what should be calculated from the material to be determined. Estimation is made about the sample size needed to detect differences between populations. The three major random sampling techniques are: systematic sampling; simple random sampling; and stratified random sampling. All three techniques purposes that the samples are representative of the population or compartment to be sampled. Commonly simple random sampling and stratified random sampling are used and considered in the chapter. Stratified sampling is appropriate to the case in which a certain logical or natural grouping may be expected within the population. In the determination statistical methods are used to improve the analytical method. With many methods a calibration curve is required. There are two methods commonly used for testing the homogeneity are the Cochran and the Bartlett tests.