# 2.6 Three Sampling Approaches

The total error associated with an estimate may be considered in the following simple equation that relates the true but unknown value of the parameter of interest (in this case the mean concentration) to the estimate of that parameter:

true mean concentration = estimate of that concentration ± total error

This equation emphasizes several important concepts:

• There is a true mean concentration in any volume of soil.
• Any type of sampling and analysis is capable of providing only estimates of the true mean concentration.
• The best estimate is the one with the least total error.

These concepts provide a basis from which to compare different methods for estimating the mean through different sampling approaches: discrete, composite, and ISM sampling. Any sampling design must include consideration of these questions:

• Which parameter of the population is being estimated (e.g., mean, maximum, proportion)?
• To what soil volume does that estimate apply?
• How will total error be controlled, measured, and assessed?

Although these questions should be addressed at the beginning of any sampling effort, they commonly are not. One of the strengths of the ISM process is that it necessitates a thoughtful consideration of these topics as well as an assessment of the strengths and weaknesses of the various sampling approaches prior to sampling.

The characteristics of these three sampling approaches relevant to providing an estimate of the mean are discussed in the following sections.