Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL

5.1 Populations and Samples > 5.1 Populations and Samples - Pg. 142

142 Chapter 5 Probabilities and data are closely related, because we use probabilities to anticipate what the data might look like and we use data to estimate probabilities. For example, life insurance companies use mortality data to estimate the probability that a healthy 21-year-old woman will die within a year and use this estimated probability to price a 1-year life insurance policy for a healthy 21-year-old woman. In this chapter, we look at some general principles for deciding whether empirical data are likely to yield reliable information. 5.1 Populations and Samples The starting point is to distinguish between a population and a sample. To illustrate, we look at a study of the adult daughters of women who had immigrated to California from another country [1]. The population consists of 1,111,533 immigrant women who gave birth to daughters in California between 1982 and 2007. A sample might be 30 women selected from this population. We often use samples because it is impractical (or impossible) to look at the entire population. If we burn every lightbulb to see how long bulbs last, we have a large electricity bill and a lot of burnt-out lightbulbs. Many studies are not destructive but simply too expensive to apply to the entire population. Instead, we sample. A lightbulb