Defining Populations and Selecting Samples

Defining Populations and Selecting Samples

Chapter 1 provided an introduction to the field of biostatistics. We discussed applications of statistics, study designs, as well as descriptive statistics, or exploratory data analysis, and inferential statistics, or confirmatory data analysis. Now we will consider in more detail an aspect of inferential statistics—sample selection—that relates directly to our ability to make inferences about a population.

In this chapter, we define the terms population and sample and present several methods for selecting samples. We present a rationale for selecting samples and give examples of several types of samples: simple random, convenience, systematic, stratified random, and cluster. In addition, we discuss bootstrap sampling because of its similarity to simple random sampling. Bootstrap sampling is a procedure for generating bootstrap estimates of parameters, as we will demonstrate in later chapters. Detailed instructions for selecting simple random and bootstrap samples will be provided. The chapter concludes with a discussion of an important property of random sampling, namely, unbiasedness.