Workshop Statistics: Discovery with Data, Fourth Edition

TOPIC 16

Confidence Intervals: Proportions

The generation of children being raised in the 2000s has been dubbed Generation M, because of the extent to which various forms of media (hence the M) permeate their lives. The Kaiser Family Foundation commissioned an extensive survey to investigate this phenomenon. In this topic, you will learn more about how to use sample results to estimate population values. For example, what proportion of all American teens have a television in their room? Is this proportion higher for boys or for girls? Is the proportion higher or lower for other forms of media, such as CD players and video game players and computers? How can we answer these questions without asking all American teens?

Overview

In the last unit, you explored how sample statistics vary from sample to sample. You studied this phenomenon empirically through simulations and theoretically with the Central Limit Theorem. You learned that this variation has a predictable longterm pattern. This pattern enables you to make probability statements about sample statistics, provided you know the value of the population parameter. These probability statements allow you to turn the tables and address the much more common goals of estimating and making decisions about an unknown population parameter based on an observed sample statistic. These are the goals of statistical inference.

There are two major techniques for classical statistical inference: confidence intervals and tests of significance. ...