Chapter 10: Statistical Inference: One-Sample Hypothesis Test

Chapter 10: Statistical Inference: One-Sample Hypothesis Test

Looking Ahead: What Is This Chapter About?

Evaluating the effectiveness of a new teaching technology or assessing attitudes towardviolence on TV involves making a decision on the basis of incomplete information.The researcher’s information is usually incomplete because it is impossible orimpractical to observe all the people in the population of interest—for example, allschoolchildren or all TV viewers. Fortunately, there are procedures for making rationaldecisions about populations that use a sample containing only a small portion ofthe elements in the population. These procedures, called statistical inference, are thesubject of this and subsequent chapters.

Several approaches to making decisions about a population use information froma sample, but I will limit my discussion to classical statistical inference, whichevolved from the work of Ronald A. Fisher and, more directly, Jerzy Neyman andEgon Pearson. Two complementary topics fall under classical statistical inference:null hypothesis significance testing, the subject of in this chapter, and confidence intervalestimation, which is described in the next chapter. I will examine hypothesistesting first because the procedure is so widely used in the behavioral sciences,health sciences, and education.

In this chapter you will learn about a new sampling distribution called the t distribution.You also will learn how to use a t statistic to test a hypothesis about themean of a population. You will use the concepts that you learn in this chapterthroughout the remained of the book.

After reading this chapter, you should know the following:

■ The difference between scientific hypotheses and statistical hypotheses

■ The five steps used to test a statistical hypothesis

■ How to use a t statistic to test a statistical hypothesis about a population mean

■ The relative advantages of one- and two-tailed tests

■ The two kinds of errors that can occur in testing a statistical hypothesis

■ How to specify an appropriate sample size, n