ABSTRACT
In Chapters 5 to 7 we explained various statistical methods for estimating parameters and performing hypothesis tests, for example, to decide whether soil at a Superfund site has been adequately cleaned up. In all of these chapters we concentrated on specifying the confidence level for a confidence interval, or specifying the Type I (false positive) error rate for a hypothesis test (e.g., the probability of declaring a site is contaminated when in fact it is clean; see Tables 7.1 and 7.2 in Chapter 7). In this chapter, we will talk about how you can determine the sample size you will need for a study. If the study involves estimating a parameter, you will need to decide how wide you are willing to have your confidence intervals. If the study involves testing a hypothesis, you will need to decide how you want to balance the Type I and Type II error rates based on detecting a specified amount of change. The tools we talk about in this chapter are used in Steps 6 and 7 of the DQO process (see Chapter 2) to specify limits on the decision error rates and in the iterative step of optimizing the design.
