2.5.2 Compositional Heterogeneity

Fundamental error is controlled by collecting samples of sufficient mass.Before sampling errors can be discussed, the Gy theory concepts of constitutional and distributional heterogeneity must be introduced. Constitutional (or compositional) heterogeneity is a measure of the differences in composition between individual fragments or particles of the population being sampled with respect to a given parameter of interest. It refers to the fact that soil is made of many different types of particles that interact with contaminants in different ways. CH is a direct cause of a sampling error termed “fundamental error” (FE). A way to control FE is to have large enough samples (or subsamples) so that the probability is high that the composition of the sample will match the composition of the population. Figure 2-10 presents a population with two samples of different masses. Although both samples were collected from the same population, they are not equally representative of the parent population. The larger of the two samples (Sample B) better represents the composition of the population and reduces FE relative to the smaller sample (Sample A). Also, the larger the particles, the larger the sample mass must be to minimize FE. More illustrations of this concept can be found in Hyperlink 12.

Illustration of the effects of sample mass on representativeness of the population.

Figure 2-10. Illustration of the effects of sample mass on representativeness of the population.
Source: USEPA 2002e.