TABLE OF CONTENTS In previous chapters we detailed correct sampling principles, analyses, tools, and techniques. We also illustrated traps and pitfalls that result in poor sampling. We did this by presenting a structure for the different sampling errors, which is useful for assessing particular situations. We now know that sampling is made up of components and that these components contribute to the overall sampling error. We have seen ways to improve many sampling situations and reduce the sampling error.
Now we need to understand how to apply these ideas as a strategy, to go beyond particular problems that are brought to our attention, to address all the components of sampling. In this chapter, we propose an approach to sampling that uses the previous theoretical information in a practical way. The strategy has three parts: audit, assessment, and action.
When we learn of a sampling problem, we often focus on an obvious culprit. The particular issue may be a procedure, a collection device, or handling, for example. With this approach, we do not look beyond what we know to be wrong. We are content to solve the "visible" problem at hand. Since we don't see the whole picture, we are oblivious to other sampling issues that may arise in the future. We thus recommend that the first step be a sampling audit, allowing us to look at all the sampling components. An audit provides a firm basis for a detailed assessment and, if...
